BREF Mineral Oil and Gas Refineries en
BREF Mineral Oil and Gas Refineries en
BREF Mineral Oil and Gas Refineries en
February 2003
Executive Summary
EXECUTIVE SUMMARY
The BREF (Best Available Techniques reference document) on Mineral Oil and Gas Refineries
reflects an information exchange carried out according to Article 16(2) of Council Directive
96/61/EC. This Executive Summary – which is intended to be read in conjunction with the
BREF preface’s explanations of objectives, usage and legal terms – describes the main findings,
the principal BAT conclusions and the associated emission levels. It can be read and understood
as a stand-alone document but, as a summary, it does not present all the complexities of the full
BREF text. It is therefore not intended as a substitute for the full BREF text as a tool in BAT
decision making. In this information exchange more than 40 people have participated directly.
Oil companies are typically international companies, so people from outside the EU have also
been involved in the process.
Scope
The scope of this BREF for the Mineral Oil and Gas Refinery Industry is based on Section 2.1
of Annex I of the IPPC Directive 96/61/EC, from which it also takes its title. This document
addresses the mineral oil refining industry as well as the natural gas plants. Other related
activities such as exploration, production, transportation or marketing of products are not
included here. All types of mineral oil refineries regardless of capacity and all types of process
activities typically found there are covered in this document. Some activities that are or may be
found in refineries are not covered here because they are covered in other BREFs (e.g. low
olefins and solvent production, generation of power with natural gas). Other activities have not
been fully covered in this document because they are partially covered in other BREFs (e.g.
cooling, storage, waste water and waste gas). Thus, when implementing IPPC permits for a
specific site, other BREFs should also be considered. Soil remediation is not included in this
BREF because it is not a contamination prevention or control technique.
Refinery installations are typically big and fully integrated. Refineries are industrial sites that
manage huge amounts of raw materials and products and they are also intensive consumers of
energy and water. In their storage and refining processes, refineries generate emissions to the
atmosphere, to the water and to the soil, to the extent that environmental management has
become a major factor for refineries. The type and quantity of refinery emissions to the
environment are typically well known. Oxides of carbon, nitrogen and sulphur, particulates
(mainly generated from combustion processes), and volatile organic carbons are the main air
pollutants generated by both sectors. Water is used intensively in a refinery as process water and
for cooling purposes. Its use contaminates the water with oil products. The main water
contaminants are hydrocarbons, sulphides, ammonia and some metals. In the context of the
huge amount of raw material that they process, refineries do not generate substantial quantities
of waste. Currently, waste generated by refineries are dominated by sludges, non-specific
refinery waste (domestic, demolition, etc.), and spent chemicals (e.g. acids, amines, catalysts).
Emissions to air are the main pollutants generated by mineral oil refineries and, to a much lesser
extent, natural gas plants (i.e. number of emission points, tonnes emitted, number of BAT
developed). For every million tonnes of crude oil processed (European refineries range from 0.5
to more than 20 million tonnes), refineries emit from 20000 – 820000 t of carbon dioxide,
60 - 700 t of nitrogen oxides, 10 – 3000 t of particulate matter, 30 – 6000 t of sulphur oxides
and 50 – 6000 t of volatile organic chemicals. They generate, per million tonnes of crude oil
refined, from 0.1 – 5 million tonnes of waste water and from 10 – 2000 tonnes of solid waste.
Those big differences in emissions from European refineries can be partially explained by the
differences in integration and type of refineries (e.g. simple vs. complex). However, the main
differences are related to different environmental legislation schemes in Europe. Main air
emissions from natural gas plants are CO2, NOx, SOx, and VOC. Water and waste are typically
less important than for mineral oil refineries.
Given the progress that refineries have made in the abatement of sulphur emissions to air, the
focus has started to shift towards VOC (including odour), particulates (size and composition)
and NOx, as it has in the environmental debate generally. When the carbon dioxide emissions
debate gathers momentum, it will also strongly affect refineries. Refinery waste water treatment
techniques are mature techniques, and emphasis has now shifted to prevention and reduction.
Reduction of water use and/or the concentration of pollutants in the water can have effects in
reducing the final emission of pollutants.
Techniques applied to
production
Chapter Activity/Process and gases and waste solid TOTAL
section prevention waste gas water waste
1 General
2 Alkylation 3 0 0 0 3
3 Base oil production 14 4 2 1 21
4 Bitumen production 2 5 1 2 10
5 Catalytic cracking 17 13 2 5 37
6 Catalytic reforming 3 3 0 0 6
7 Coking processes 9 19 8 3 39
8 Cooling 3 - - - 3
9 Desalting 13 0 4 1 18
10 Energy system 56 22 2 0 80
11 Etherification 1 0 1 1 3
12 Gas separation processes 3 2 0 0 5
13 Hydrogen-consuming 8 0 0 2 10
processes
14 Hydrogen production 6 0 0 0 6
15 Integrated refinery 33 0 24 6 63
management
16 Isomerisation 3 0 0 0 3
17 Natural gas plants 0 12 5 3 20
18 Polymerisation 1 0 0 2 3
19 Primary distillation units 3 2 3 3 11
20 Product treatments 5 2 4 0 11
21 Storage and handling of 21 19 2 12 54
refinery materials
22 Visbreaking 3 1 1 1 6
23 Waste gas treatments - 76 - 1 77
24 Waste water treatments - - 41 - 41
25 Waste management - - - 58 58
TOTAL 207 180 100 101 588
As can be calculated from the above table, 35 % of the techniques included in Chapter 4 are
techniques dedicated to production and to the prevention of contamination, 31 % are air
abatement techniques and 17 % are for techniques to reduce water pollution and to reduce waste
or to prevent soil contamination. Those figures again reflect the fact that air emissions are the
most important environmental issue in the refinery sector.
This section of the executive summary highlights the most relevant environmental issues and
the key findings that are given in Chapter 5. During the discussion of the information exchanged
by the TWG, many issues were raised and discussed. Only some of them are highlighted in this
summary.
Emission or consumption levels “associated with best available techniques” are presented where
relevant within BAT chapter. BREFs do not set legally binding standards, they are meant to
give information for the guidance of industry, Member States and the public on achievable
emission and consumption levels when using specified techniques. Those levels are neither
emission nor consumption limit values and should not be understood as such. The appropriate
limit values for any specific case will need to be determined taking into account the objectives
of the IPPC Directive and the local considerations.
It was acknowledged that the implementation of BAT in each refinery needs to be addressed in
each case and that multiple technical solutions exist. That is why prevention or control
techniques are given in the BAT as a group of possibilities.
Amongst the many environmental issues addressed in the BREF, the five that are dealt with
below are probably the most important:
• increase the energy efficiency
• reduce the nitrogen oxide emissions
• reduce the sulphur oxide emissions
• reduce the volatile organic compounds emissions
• reduce the contamination of water
BAT under the bubble concept. Five different ranges or values were provided by the TWG for
the concentration bubble approach (three based on different scenarios when implementing BAT)
and two for the load bubble approach (one based on the scenario of implementation of BAT).
BAT related to NOx emissions (~17) typically contain associated emission values.
Emerging techniques
This brief chapter includes the techniques that have not yet been commercially applied and are
still in the research or development phase. However, because of the implications they may have
in the refinery sector, they have been included here to raise awareness during any future revision
of the document.
Concluding remarks
The environmental situation of European refineries varies greatly across the European
Community, so the starting point for each case is very different. Different environmental
perceptions and priorities are also evident.
Level of consensus
The refineries sector is a large and complex one, spread across all Member States except
Luxembourg. This size and complexity are reflected in the number of processes / activities
addressed in the BREF and the number (200+) of BAT it contains. The fact that agreement has
been reached on all but 27 of those 200+ BAT is a measure of the broad commitment of TWG
members to the conclusions reached. Those 27 split views can be summarised and classified in
the following three ways:
• Nineteen relate to the figures in the ranges given in Chapter 5. They represent two points
of view; the first is that control techniques are almost always applicable in all cases and
the second is that control techniques are hardly applicable
• four relate to the structure of the parts of Chapter 5 that relate to SOx and NOx emissions
and are driven by the bubble concept approach
• two relate to the water emissions table; one to the average time period given in the
concentration column and the other to how the metal content should be expressed within
the table
• one is addressed to the introduction to Chapter 5 and relates to the way in which the
upper value of the ranges in Chapter 5 is selected
• only one split view relates fundamentally to a technique; base oil production.
The EC is launching and supporting, through its RTD programmes, a series of projects dealing
with clean technologies, emerging effluent treatment and recycling technologies and
management strategies. Potentially these projects could provide a useful contribution to future
BREF reviews. Readers are therefore invited to inform the EIPPCB of any research results
which are relevant to the scope of this document (see also the preface of this document).
PREFACE
1. Status of this document
Unless otherwise stated, references to “the Directive” in this document means the Council
Directive 96/61/EC on integrated pollution prevention and control. As the Directive applies
without prejudice to Community provisions on health and safety at the workplace, so does this
document.
This document forms part of a series presenting the results of an exchange of information
between EU Member States and industries concerned on best available technique (BAT),
associated monitoring, and developments in them. It is published by the European Commission
pursuant to Article 16(2) of the Directive, and must therefore be taken into account in
accordance with Annex IV of the Directive when determining “best available techniques”.
2. Relevant legal obligations of the IPPC Directive and the definition of BAT
In order to help the reader understand the legal context in which this document has been drafted,
some of the most relevant provisions of the IPPC Directive, including the definition of the term
“best available techniques”, are described in this preface. This description is inevitably
incomplete and is given for information only. It has no legal value and does not in any way alter
or prejudice the actual provisions of the Directive.
The purpose of the Directive is to achieve integrated prevention and control of pollution arising
from the activities listed in its Annex I, leading to a high level of protection of the environment
as a whole. The legal basis of the Directive relates to environmental protection. Its
implementation should also take account of other Community objectives such as the
competitiveness of the Community’s industry thereby contributing to sustainable development.
More specifically, it provides for a permitting system for certain categories of industrial
installations requiring both operators and regulators to take an integrated, overall look at the
polluting and consuming potential of the installation. The overall aim of such an integrated
approach must be to improve the management and control of industrial processes so as to ensure
a high level of protection for the environment as a whole. Central to this approach is the general
principle given in Article 3 that operators should take all appropriate preventative measures
against pollution, in particular through the application of best available techniques enabling
them to improve their environmental performance.
The term “best available techniques” is defined in Article 2(11) of the Directive as “the most
effective and advanced stage in the development of activities and their methods of operation
which indicate the practical suitability of particular techniques for providing in principle the
basis for emission limit values designed to prevent and, where that is not practicable, generally
to reduce emissions and the impact on the environment as a whole.” Article 2(11) goes on to
clarify further this definition as follows:
“techniques” includes both the technology used and the way in which the installation is
designed, built, maintained, operated and decommissioned;
“available” techniques are those developed on a scale which allows implementation in the
relevant industrial sector, under economically and technically viable conditions, taking into
consideration the costs and advantages, whether or not the techniques are used or produced
inside the Member State in question, as long as they are reasonably accessible to the operator;
“best” means most effective in achieving a high general level of protection of the environment
as a whole.
Competent authorities responsible for issuing permits are required to take account of the general
principles set out in Article 3 when determining the conditions of the permit. These conditions
must include emission limit values, supplemented or replaced where appropriate by equivalent
parameters or technical measures. According to Article 9(4) of the Directive, these emission
limit values, equivalent parameters and technical measures must, without prejudice to
compliance with environmental quality standards, be based on the best available techniques,
without prescribing the use of any technique or specific technology, but taking into account the
technical characteristics of the installation concerned, its geographical location and the local
environmental conditions. In all circumstances, the conditions of the permit must include
provisions on the minimisation of long-distance or transboundary pollution and must ensure a
high level of protection for the environment as a whole.
Member States have the obligation, according to Article 11 of the Directive, to ensure that
competent authorities follow or are informed of developments in best available techniques.
Article 16(2) of the Directive requires the Commission to organise “an exchange of information
between Member States and the industries concerned on best available techniques, associated
monitoring and developments in them”, and to publish the results of the exchange.
The purpose of the information exchange is given in recital 25 of the Directive, which states that
“the development and exchange of information at Community level about best available
techniques will help to redress the technological imbalances in the Community, will promote
the world-wide dissemination of limit values and techniques used in the Community and will
help the Member States in the efficient implementation of this Directive.”
The Commission (Environment DG) established an information exchange forum (IEF) to assist
the work under Article 16(2) and a number of technical working groups have been established
under the umbrella of the IEF. Both IEF and the technical working groups include
representation from Member States and industry as required in Article 16(2).
The aim of this series of documents is to reflect accurately the exchange of information which
has taken place as required by Article 16(2) and to provide reference information for the
permitting authority to take into account when determining permit conditions. By providing
relevant information concerning best available techniques, these documents should act as
valuable tools to drive environmental performance.
4. Information sources
This document provides information for two industrial sectors: Mineral Oil refineries and
Natural Gas plants. The rest of this section describes the type of information that is provided in
each section of the document.
The document has been divided into seven chapters plus annexes. Chapters 1 and 2 provide
general information on the mineral oil refinery and natural gas sector and on the production
processes and activities that can be found within the sector. Chapter 3 provides data and
information concerning current emission and consumption levels reflecting the situation in
existing installations at the time of writing. Chapter 4 describes the prevention (including
production techniques) and control techniques that are considered to be most relevant for
determining BAT and BAT-based permit conditions. Each technique within Chapter 4 includes
information on consumption and emission levels considered achievable by using the technique,
some idea of the costs and the cross-media issues associated with the technique, and the extent
to which the technique is applicable to the range of installations requiring IPPC permits, for
example new, existing, large or small installations. Chapter 5 presents the techniques and the
emission and consumption levels that are considered to be compatible with BAT in a general
sense for the mineral oil refineries and natural gas plants as a whole. Chapter 6 describes
briefly techniques that may be applied in both sectors in the future. Chapter 7 summarises the
conclusions and recommendations of the document. In order to help the reader, a Glossary is
also included. Complementing the document, the Annexes contain a summary of the legislation
applicable to the sector as well as summaries of the type of refinery configurations and
descriptions of the refinery products/intermediates generated within refineries.
The purpose of Chapter 5 is thus to provide general indications regarding the emission and
consumption levels that can be considered as an appropriate reference point to assist in the
determination of BAT-based permit conditions or for the establishment of general binding rules
under Article 9(8). It should be stressed, however, that this document does not propose emission
limit values. The determination of appropriate permit conditions will involve taking account of
local, site-specific factors such as the technical characteristics of the installation concerned, its
geographical location and the local environmental conditions. In the case of existing
installations, the economic and technical viability of upgrading them also needs to be taken into
account. Even the single objective of ensuring a high level of protection for the environment as
a whole will often involve making trade-off judgements between different types of
environmental impact, and these judgements will often be influenced by local considerations.
Although an attempt is made to address some of above-mentioned issues, it is not possible for
some of them to be considered fully in this document. The techniques and levels presented in
Chapter 5 will therefore not necessarily be appropriate for all installations. On the other hand,
the obligation to ensure a high level of environmental protection including the minimisation of
long-distance or transboundary pollution implies that permit conditions cannot be set on the
basis of purely local considerations. It is therefore of the utmost importance that the information
contained in this document is fully taken into account by permitting authorities.
Chapters 2, 3, 4 and 6 as well as Section 5.2 have a common internal structure, so as to facilitate
the reading and understanding of the document. Each of those chapters has been structured per
type of process or activity found in the refinery sector and classified in alphabetical order to
make it easier to find them in the document. Those sections do not pretend to be of equal
importance from the environmental impact point of view. Some have more relevance than
others, but this structure is considered a good and clear way to tackle the BAT assessment for
these industrial sectors. Moreover, it is unlikely that a completely new mineral oil refinery
installation will be built in Europe, but it is very likely that new processes will be installed in
existing European refineries.
The first section in each of those chapters is dedicated to a general description of the chapter
issue under consideration in refineries as a whole. Then follow 20 sections covering all the
mineral oil refinery processes and activities and one covering natural gas plants.
Since the best available techniques change over time, this document will be reviewed and
updated as appropriate. All comments and suggestions should be made to the European IPPC
Bureau at the Institute for Prospective Technological Studies at the following address:
EXECUTIVE SUMMARY------------------------------------------------------------------------------------------ I
PREFACE ------------------------------------------------------------------------------------------------------------ IX
SCOPE ------------------------------------------------------------------------------------------------------------- XXV
1 GENERAL INFORMATION ---------------------------------------------------------------------------------- 1
1.1 The purpose of refineries .............................................................................................................1
1.2 Refinery sector in the EU ............................................................................................................1
1.3 European refineries......................................................................................................................6
1.3.1 Technical characteristics of European refineries ....................................................................9
1.3.2 Employment in the European refinery sector .......................................................................13
1.4 Main environmental issues in the refining sector ......................................................................14
1.4.1 Emissions to the atmosphere.................................................................................................15
1.4.2 Emissions to water ................................................................................................................16
1.4.3 Waste generation ..................................................................................................................17
1.4.4 Soil and groundwater contamination ....................................................................................18
1.4.5 Other environmental issues...................................................................................................19
2 APPLIED PROCESSES AND TECHNIQUES ----------------------------------------------------------- 21
2.1 General overview of the refinery processes...............................................................................21
2.2 Alkylation ..................................................................................................................................25
2.3 Base oil production ....................................................................................................................27
2.4 Bitumen production ...................................................................................................................29
2.5 Catalytic cracking ......................................................................................................................31
2.6 Catalytic reforming....................................................................................................................33
2.7 Coking processes .......................................................................................................................35
2.8 Cooling systems.........................................................................................................................38
2.9 Desalting....................................................................................................................................40
2.10 Energy system............................................................................................................................41
2.11 Etherification .............................................................................................................................48
2.12 Gas separation processes ...........................................................................................................51
2.13 Hydrogen consuming processes.................................................................................................53
2.14 Hydrogen production .................................................................................................................60
2.15 Integrated refinery management ................................................................................................63
2.16 Isomerisation .............................................................................................................................66
2.17 Natural gas plants ......................................................................................................................68
2.18 Polymerisation ...........................................................................................................................70
2.19 Primary distillation units............................................................................................................71
2.20 Product treatments .....................................................................................................................74
2.21 Storage and handling of refinery materials ................................................................................76
2.22 Visbreaking................................................................................................................................78
2.23 Techniques for the abatement of emissions ...............................................................................80
3 CURRENT EMISSION AND CONSUMPTION LEVELS -------------------------------------------- 81
3.1 Current emission and consumption levels in refineries as a whole............................................81
3.1.1 Present consumption levels in refineries...............................................................................83
3.1.2 Emissions to air.....................................................................................................................84
3.1.3 Emissions to water ................................................................................................................92
3.1.4 Waste generation ..................................................................................................................94
3.2 Alkylation ..................................................................................................................................96
3.3 Base oil production ....................................................................................................................98
3.4 Bitumen production .................................................................................................................100
3.5 Catalytic cracking ....................................................................................................................101
3.6 Catalytic reforming..................................................................................................................106
3.7 Coking processes .....................................................................................................................108
3.8 Cooling systems.......................................................................................................................110
3.9 Desalting..................................................................................................................................112
3.10 Energy system..........................................................................................................................114
3.10.1 Energy management ...........................................................................................................114
3.10.2 Energy capacity and consumptions.....................................................................................115
Figure 1.1: Percentage of sulphur in the crude used in each European Country ....................................... 3
Figure 1.2: Geographical distribution of the European refineries ............................................................. 8
Figure 1.3: Nelson complexity index of European refineries per country (average 6.1)......................... 12
Figure 1.4: Evolution of the complexity of European refineries ............................................................. 13
Figure 1.5: Employment in European refineries as a function of the complexity of the refinery ........... 14
Figure 1.6: Employment in the European refineries as a function of the crude capacity ........................ 14
Figure 3.1: Example of specific emissions and consumptions found in European refineries ................. 82
Figure 3.2: Specific CO2 emissions from some European refineries as function of the Nelson
complexity index .................................................................................................................. 84
Figure 3.3: Specific NOx emissions from some European refineries as function of the Nelson
complexity index .................................................................................................................. 85
Figure 3.4: SO2 emissions from some European refineries as function of the crude capacity ................ 88
Figure 3.5: Specific SO2 emissions from some European refineries as function of the crude
capacity................................................................................................................................. 88
Figure 3.6: Specific SO2 emissions from some European refineries against the Nelson complexity
index ..................................................................................................................................... 89
Figure 3.7: Energy intensity index vs. Nelson complexity index for some EU+ refineries .................. 115
Figure 3.8: Sulphur and nitrogen contents in HFO (vacuum residues) according to their
geographical origin ............................................................................................................. 116
Figure 3.9: Petroleum coke commercial process................................................................................... 128
Figure 3.10: Historical and distribution trend of calculated annual SO2 combustion bubble for EU
refineries ............................................................................................................................. 131
Table 1.1: Crude oil reserves and consumption per geographical region ................................................ 2
Table 1.2: EU+ capacity for mineral oil refining..................................................................................... 4
Table 1.3: Product demand development Western Europe ...................................................................... 5
Table 1.4: Mineral oil product specifications .......................................................................................... 6
Table 1.5: Percentage of refineries built during different time periods ................................................... 7
Table 1.6: Number of type of processes per country ............................................................................... 9
Table 1.7: Percentage of EU+ refineries that have a certain process (per type of process) ................... 10
Table 1.8: Capacity of the various processes in EU+ refineries ............................................................ 11
Table 1.9: European refineries by configuration.................................................................................... 12
Table 1.10: Main air pollutants emitted by refineries and their main sources ......................................... 15
Table 1.11: Main water pollutants (parameters) generated by refineries................................................. 17
Table 1.12: Main solid wastes generated by refineries ............................................................................ 18
Table 4.1: Information contained in the description of each technique included in Chapter 4 ............160
Table 4.2: Deasphalting-Unit Energy-Reduction Economics...............................................................166
Table 4.3: Emission factors for cracking (units in kg/1000 litres of fresh feed) ..................................182
Table 4.4: Catalytic feed hydrotreatment (typical feeds being atmospheric residue and vacuum
gas oil).................................................................................................................................183
Table 4.5: Economics of SCR applied to FCC units ............................................................................187
Table 4.6: Main cost factors for a Selective Catalytic Reduction (SCR) installation (raw gas) after
the FCC-plant ......................................................................................................................187
Table 4.7: Economics of SNCR applied to FCCs ................................................................................188
Table 4.8: Economic data of third cyclone applied to FCCs ................................................................189
Table 4.9: Economic data of ESP applied to FCCs ..............................................................................191
Table 4.10: Achievable emission levels for wet scrubbers.....................................................................194
Table 4.11: Emission values for petroleum coke production plants (calcination of green coke) ...........206
SCOPE
This document, together with other BREFs in the series, are intended to cover the activity
described in point 1.2 described in Annex 1 of Council Directive 96/61/EC, namely Mineral Oil
and Gas refineries.
The mineral oil and gas industry is organised into four broad sectors: exploration and
production of crude oil and natural gas; transportation; refining; and marketing and distribution.
This document addresses only the refining industry from both sectors (Mineral Oil refineries
and Natural Gas plants); specifically the activities and processes set out in the table below. The
table indicates other reference documents dealing with related issues as shown.
Section no. Activity name used Subactivities or processes included in each Reference to additional
within this within this document category of processes information
document
2 Alkylation HF, H2SO4 alkylation processes
3 Base oil production Deasphalting, aromatic extraction, high pressure
hydrogenation unit, dewaxing, wax processing and
lubricant oil hydrofinishing
4 Bitumen production
5 Catalytic cracking All types of catalytic cracking units. Types
according to feedstock and process conditions
6 Catalytic reforming Continuous, cyclic and semiregenerative
7 Coking processes Delayed and fluid coking and coke calcination
8 Cooling Industrial cooling
systems BREF
9 Desalting
10 Energy system Refinery power plants including all types of Large Combustion Plants
technologies applied and all types of fuels used BREF
within EU refineries
11 Etherification MTBE, ETBE and TAME production
12 Gas separation Separation of light fractions of the crude oil (e.g.
processes refinery fuel gas, LPG)
13 Hydrogen consuming Hydrocracking, hydrorefining, hydrotreatments,
processes hydroconversion, hydroprocessing and
hydrogenation processes
14 Hydrogen production Gasification (coke and heavy oils), steam reforming
and hydrogen purification
15 Integrated refinery Environmental management activities, utility
management management and overall refinery management
(noise, odour, safety, maintenance)
16 Isomerisation C4, C5 and C6 isomerisation
17 Natural gas plants Processes related with the processing of natural gas
18 Polymerisation Polymerisation, dimerisation and condensation
19 Primary distillation Atmospheric and vacuum distillation units
units
20 Product treatments Sweetening and final product treatments
21 Storage and handling Storage, blending, loading and unloading of refinery Storage BREF
of refinery materials materials
22 Visbreaking
23 Techniques for the It merges the contents related to waste gas, waste
abatement of emissions water and solid wastes in Chapters 2
(only in Chapters 2)
23 Waste gas treatments Waste gas and waste
water BREF
24 Waste water treatments Waste gas and waste
water BREF
25 Waste management Waste treatment, waste disposal and incineration
26 Monitoring Monitoring BREF
(only in Chapter 3)
As can be seen in the above table, most of the processes and activities under analysis are
specifically dedicated to mineral oil refineries. The only exception is Section 17, which deals
with the processes that can be found in inland natural gas plants (name used in this document to
identify this industrial sector). Those plants also use some of the end-of-pipe techniques (in
Sections 23 to 25) used in mineral oil refineries and, for this reason, no distinction is made
between the two types of industry.
Some processes or activities that may be present in refineries are not included or are only
partially included in this document, because they are covered in other BREFs. Some examples
are:
• the steam cracker for the production of lower olefins, the production of aromatics (i.e.
BTX), cyclohexane and cumene or alkylation of aromatics in Large Volume Organic
Chemicals BREF
• the energy production techniques when commercial fuel oil, natural gas or gasoil are used
are included in the Large Combustion Plant BREF
• horizontal BREFs as industrial cooling, monitoring, storage and waste water and waste gas
treatments are also applicable to refineries.
TWG has considered that soil remediation techniques are not a part of study within this BREF.
The reason given is that those techniques are not techniques to prevent nor to control emissions.
They are techniques used to clean up the soil when it has been already contaminated.
1 GENERAL INFORMATION
• fuels for cars, trucks, aeroplanes, ships and other forms of transport
• combustion fuels for the generation of heat and power for industry and households
• raw materials for the petrochemical and chemical industries
• speciality products such as lubricating oils, paraffins/waxes and bitumen
• energy as a by-product in the form of heat (steam) and power (electricity).
In order to manufacture these products, these raw materials are processed in a number of
different refining facilities. The combination of these processing units to convert crude oil and
natural gas into products, including its supporting units and facilities, is called a refinery. The
market demand for the type of products, the available crude quality and certain requirements set
by authorities influence the size, configuration and complexity of a refinery. As these factors
vary from location to location no two refineries are identical.
Oil provides 42 % of the European Union's (EU) energy requirement, and 94 % of the fuels
required for transport. The health and viability of the refining industry is of critical strategic
importance to the Union for maintaining a successful and internationally competitive position
for industry as a whole, and for providing competitively priced products to consumers.
The refining industry has suffered from a structural distillation overcapacity most of the time
since the oil crisis in 1973/1974. Only in the early 1980s, early 1990s and last years were
attractive margins achieved due to high oil prices at those times. Moreover, severe competition
costs, environmental compliance of the refining industry and regulatory uncertainty have added
to the decrease in profitability in certain periods. This long-lasting recession led oil and gas
companies to make significant adjustments in upstream and downstream operations such as cuts
in production costs, innovations in technology and organisational restructuring.
However co-operation in the form of partnerships, alliances, joint ventures or mergers between
companies, governments and local communities, contractors and suppliers has led to significant
cost reductions. This trend is also fuelled by increased environmental awareness in society at
large and is supported by the current process of harmonisation of legislation, particularly within
Europe. Examples of new forms of co-operation are the recent (mid-1998) downstream merger
between BP and Mobil which started as a partnership in refining and marketing in 1996 and
secured reportedly significant pre-tax cost savings. Recently a Total-Fina-Elf merger and an
Esso-Mobil merger were concluded. A further recent example is the refining capacity exchange
between Statoil (Mongstad-Norway) and Shell (Sola/Pernis-Netherlands).
Some European refineries have been closed in the last 20 years, but crude oil processing
capacity has increased in recent years, mainly by ‘capacity creep’ (debottlenecking,
improvement in equipment reliability and longer cycles between turnarounds) to cope with the
low but steady increase in product demand of 1 - 2 % per year in Europe. Worldwide ‘capacity
creep’ is estimated to be ‘equivalent to six to ten additional world scale refineries going on
stream every year’.
The crude oil refining industry is complex, with the following issues at stake in Europe:
Feed stock: steady production and a shift towards lighter (North Sea) crudes specifically in the
North of Europe. World reserves of crude oil seems to guarantee the availability of raw
materials for a reasonably long period (~40 years). Table 1.1 shows a summary of the world
reserves and the consumption of crude oil per geographical area during the last decade.
Table 1.1: Crude oil reserves and consumption per geographical region
Source: [247, UBA Austria, 1998], [246, BP-AMOCO, 2001]
The advent of North Sea crude oils and the continuous increase in the production of these light,
low-sulphur crudes is mainly responsible for the lowering of the average sulphur content of the
crude oils processed in European refineries. Since 1985 the average sulphur content has
fluctuated around 1.0 and 1.1 %. However, the difference between type of crude processed in
each European region should be noted, i.e. an average of 1.17 % S in the crude oils processed in
the refineries of North West Europe, an average of 0.91 % S in the Atlantic region, of 1.2 % S
in the Mediterranean and of 0.64 %S in the ‘Others’. The availability of the type of crude oil to
refineries is not the same. In Figure 1.1, the differences in average sulphur content of the crude
feedstock of refineries in different regions/countries are stated.
1.75
1.50
1.25
1.00
% S
0.75
0.50
0.25
0.00
EU
a
n
ce
x
ly
y
vi
ai
U
lu
an
Ita
an
na
Sp
e
m
en
Fr
di
er
B
an
G
1993 1995 1997
Sc
Figure 1.1: Percentage of sulphur in the crude used in each European Country
Source: [261, Canales, 2000]
• location close to oil fields where low-sulphur crudes are produced (freight for crude from
the North Sea to Mediterranean costs up to 1 $/bbl, so low-sulphur crudes from the North
Sea are rarely processed in the Mediterranean area.)
• refinery equipped with insufficient desulphurization / upgrading capacity to process (heavy)
high-sulphur crudes
• exclusion from markets where high-sulphur products can be sold (e.g. bitumen, bunker fuel)
• some other form of specialisation in the use of low-sulphur crudes (e.g. luboil production).
Any refinery that is not influenced by local factors such as those mentioned above will try to
process high-sulphur crudes to the maximum extent, as their prices are considerably lower. (e.g.
in September 1999 Platt’s quoted Arab heavy with sulphur content 2.8 % was 1.1 $/bbl cheaper
than Arab Light with sulphur content 1.8 %). If the flexibility of refineries would be theorically
high, the product slate would be altered to suit market requirements irrespective of the crude
type by the inclusion of conversion units. As consequence of that all refineries would
continuously be running on the cheapest crude available, which probably would be very heavy
and high-sulphur [253, MWV, 2000] [310, Swain, 2000].
Refining distillation capacity decreased substantially in the early 1980s following the oil price
shocks of the 1970s. At the same time, the industry had to invest heavily in conversion capacity
to convert fuel oil and the rebalance demand for lighter transport fuels. Official figures for
nameplate capacity show small further capacity reductions to the end of 1995. Combined with a
slow rise in demand since 1986 this has increased apparent distillation capacity utilisation from
a low point of 60 % in 1981 to an average above 90 % in 1997, higher in northern Europe and
lower in southern Europe.
There are differences in supply and demand balances between different countries, with
Germany in particular having a large supply deficit. Growth of oil products demand in Iberia
has been well above the average for the EU, particularly in transport fuels. However, oil demand
growth in the 1990s shows the Mediterranean region to be in line with the rest of Europe. Even
in the most favourable scenario, European refining capacity will almost certainly exceed
demand for at least the next decade. International trading opportunities will only have a limited
impact on Europe's over-capacity. There is a refining capacity surplus of 70 - 100 million tonnes
per year (equivalent to 9 to 13 refineries) in the EU.
Economics: low refinery margins and low return on investment stimulate refiners either to look
at other products such as electricity, hydrogen and petrochemicals with higher margins or to
shut down (e. g. Sola in Norway).
Political: the political pressures and the economic recession in some regions of the world will
have a significant bearing on European economies and the pace at which upgrading of refineries
or even new refinery projects will be undertaken world-wide.
Product market: growing demand for kerosene and diesel; a stable demand for gasoline; a
declining demand for light heating oil and heavy fuel oil; and an increasing demand for
petrochemicals. Competition is expected from refineries in the Russian Federation and the
Middle East. A picture for the demand in Western Europe is presented in Table 1.3.
Product quality: cleaner fuels. New specifications require reductions of the sulphur content for
all types of automotive fuels, lower aromatics, particularly benzene in gasoline and reduced
polyaromatic hydrocarbons and a higher cetane number in diesel; the phasing out of lead from
gasoline will be further pursued. The trend is that environmental quality requirements will
become more stringent for all refinery products. These requirements will be especially true for
automotive fuels (Auto-Oil I and II) (see Table 1.4).
Product specification
Refinery product Unit Before Auto-Oil I Auto-Oil II
2000 2000 2005
Gasoline
Sulphur ppm 500 max. 150 max. 50 max.
Aromatics % v/v no spec 42 max. 35 max.
Olefins % v/v no spec 18 max. ? max.
Benzene % v/v 5 max. 1.0 max. ? max.
Evaporation at 100°C (summer) % 65/70 max. 46 min. ? min.
Evaporation at 150°C (winter) % -- 75 min. ? min.
RVP, summer kPa 80 60 max. ?
Oxygen % 2.5 max. 2.3 max. ? max.
Diesel
Sulphur ppm 500 max. 350 max. 50 max.
Cetane number 49 min. 51 min. ? min.
Density @ 15°C kg/m3 860 max. 845 max. ? max.
T95 % °C 370 max. 360 max. ? max.
Polycyclic Aromatics % w/w none 11 max. ? max.
Heating gas oil 2008
Sulphur (% w/w) 0.1 max
Inland heavy fuel oil 2003
Sulphur (% w/w) 1
Bunker fuel oil in SOx emission control zones IMO-2003
Sulphur (% w/w) 1.5
Moreover, EU acidification will put additional pressure on the sulphur content of liquid fuels,
and therefore also on the sulphur content of the fuels used in refineries. Meeting these new
specifications will require additional investments, particularly in desulphurisation capacity,
adding more pressure to the restructuring process of the sector mentioned above.
Environmental: emission reduction from refineries is a major issue. As regards the future,
refinery fuel consumption will be higher because of increased conversion (consumption at 49
Mt/yr at current energy efficiency). The composition of the refinery fuel will be adapted and
cannot excluded that the use of the liquid refinery fuel component (estimated at 11 Mt/yr in
1995) may be phased out. Such a reduction of use of liquid refinery fuels, some of them residual
components, leads further to more distillation residue upgrading investments (like coking,
thermocracking or gasification) (on top of the reduced demand for HFO by some 25 Mt/yr by
2010) [118, VROM, 1999].
amounts of cooling water as well as to facilitate the sea transport of raw materials and products.
There are some places in Europe with a high concentration of refineries (e.g. Rotterdam
Netherlands (5); Antwerp Belgium (5) and Sicily Italy (4).
As a result of over-capacity in the European refinery sector, very few new oil refineries have
been built in the last twenty-five years. In fact, only nine percent of the existing refineries have
been built in this period and only two percent in the last ten years, 95 % built before 1981 and
44 % before 1961 (see Table 1.5). Although most refineries will have had upgrades and new
units built since they were first commissioned, their overall structure, and in particular items
like the pattern of sewer systems, will have remained essentially unchanged.
Four on-shore natural gas plants have been identified in Europe. Three have been identified in
the Netherlands and one in Norway.
Table 1.6 shows the number of processes currently operated in the mineral oil refineries in each
country. As can be seen, crude and vacuum distillations, catalytic hydrotreatment and catalytic
reforming are the most common processes, as they are found in the simplest of refineries. It may
be surprising that the number of catalytic hydrotreatment processes is higher than the number of
refineries, but the reason is simply that there is on average more than one catalytic
hydrotreatment in each European refinery. The least common processes in European refineries
are coking and polymerisation/dimerisation.
Some of these refining processes have been analysed in order to discover the type of technique
or techniques used to perform a certain process. For instance, there are currently two
technologies to carry out alkylation, namely sulphuric and hydrofluoric. In this case these two
technologies are competitors but in other cases, such as hydrotreating, one technique does not
exclude another (those cases are marked with an asterisk in Table 1.7). The percentage of
techniques for each process is shown in Table 1.7, based on data in [73, Radler, 1998] reviewed
by TWG members. As can be seen in that table, some techniques are really predominant in
particular processes in European refineries. This category includes delayed coking, visbreaking,
fluid catalytic cracking, hydrofluoric alkylation, C5 and C6 isomerisation, MTBE production and
steam reforming for the production of hydrogen. Other processes where one technique is less
predominant are catalytic reforming, catalytic hydrorefining, catalytic hydrotreating and
hydrogen recovery process.
Table 1.7: Percentage of EU+ refineries that have a certain process (per type of process)
Source: Data from [73, Radler, 1998] reviewed by the TWG.
Note: Total number of EU+ refineries that have a certain process are shown in previous table
Following the analysis of the type of processes, Table 1.8 shows the capacity ranges for the
different processes found in EU+ refineries. As can be seen, the ranges are quite large,
explaining again the diversity of type of refineries found in the EU+. For instance, there exist
huge differences in the capacity of crude units or catalytic hydrotreating. Smaller differences
exist in the coking processes and alkylation processes.
There are various approaches to defining refinery complexity [287, Johnston, 1996]. The Nelson
refinery complexity index was used for the construction of Figure 1.3. In Dutch notes [118,
VROM, 1999] EU refineries are grouped by configuration, as shown in Table 1.9. CONCAWE
[115, CONCAWE, 1999] has grouped refineries by types (that differ in definition from the
configurations of Table 1.9) in Figure 1.4 as they existed in the period 1969 - 1997. Other
definitions are used in the industry, such as FCC equivalent (used by CEC DG Transport reports
on the refining industry) or the Equivalent Distillation Capacity as used by Solomon Associates.
Figure 1.3 tries to illustrate the variety of complexity of refineries within Europe. The Y axis
represents refinery complexity calculated according to the Nelson refinery complexity index
[287, Johnston, 1996]1. The X axis represents the different EU+ countries. If we divide the Y
gap of the EU+ refineries into four categories (less than 3.9, between 3.9 and less than 6.1,
between 6.1 and 8.4 and more than 8.42), we see that 18 % of the refineries belong to the lowest
category, 30 % belong to the second category, 41 % to the third and 11 % to the group of more
complex refineries. Consequently, 7 out of 10 refineries of the EU+ can be considered as
refineries with medium complexity.
1
Nelson developed a system to quantify the relative cost of components that make up a refinery. It is a pure cost
index that provides a relative measure of the construction costs of a particular refinery based on its crude and
upgrading capacity. Nelson assigned a factor of 1 to the distillation unit. All other units are rated in terms of their
cost relative to this unit.
2
Limits of the categories calculated as average ± standard deviation
14
12
Nelson complexity index
10
Netherlands
UK
Austria
1 2 3 4 5 6 7 8 9 10 11 12 13
Italy
France
Norway
Portugal
Spain
Finland
Belgium
Greece
Denmark
Germany
Ireland
Sweden
Switzerland
Figure 1.3: Nelson complexity index of European refineries per country (average 6.1)
COUNTRY No. Base oil and Configuration 1 Configuration 2 Configuration 3 Configuration 4 very
refineries bitumen hydroskimming + catcracker hydrocracker complex refinery with
refineries isomerisation Unit configuration configuration catcracking
Austria 1 1
Belgium 5 1 2 1 1
Denmark 2 2
Finland 2 1 1
France 15 4 10 1
Germany 17 3 2 8 3 1
Greece 4 2 1 1
Ireland 1 1
Italy 17 6 4 4 3
Netherlands 6 1 1 1 2 1
Norway 2 1 1
Portugal 2 1 1
Spain 10 1 2 6 1
Sweden 5 2 2 1
Switzerland 2 1 1
UK 13 3 7 1 2
TOTAL 104 10 26 42 14 12
Figure 1.4 shows the evolution of the complexity of EU+ refineries. It can be seen that the
number of type I refineries has decreased over the years. To make more efficient use of the
crude, increasingly deeper conversion units have been installed in European refineries.
90
80
Number of refineries
70
60
50 Type I
40 Type II+III
30
20
10
0
1969 1974 1978 1981 1984 1987 1990 1993 1997
Type I 34 49 50 38 18 12 9 7 12
TypeII+ III 4 7 61 61 67 67 72 75 76 80
Year
Figure 1.4: Evolution of the complexity of European refineries
Source: [115, CONCAWE, 1999]
Notes:
Type I: Simple or non-conversion refinery. Composed of crude oil distillation, reforming, treatment of
distillate products, including desulphurisation and/or other quality improvement processes (e.g.
isomerisation or speciality manufacturing).
Type II: Mild conversion (Type I plus thermal cracking or visbreaking).
Type III: Complex (Type II plus fluidised catcracking and/or hydrocracking).
Any of above types may have bitumen or luboil production, which clearly increases the relative
complexity. No lubricating or bitumen refinery is included in this figure
It has been estimated that in 1998, there were 55000 direct employees of the refinery operators
and 35000 contract workers (based on the numbers of man-hours worked reported to
CONCAWE for their annual safety statistics report). These figures were calculated by dividing
the reported man-hours by 1840 to give man-years. For the contractors, some of these will be
employed full time at refineries, others will have been working there only during the duration of
the contract. The number of actual people will therefore be larger.
In order to gain a rough idea of the employment generated by a single refinery, the number of
employees (annual average employment) for different EU+ refineries is shown as a function of
the complexity of the refineries (Figure 1.5) and as a function of the crude capacity of the
refineries (Figure 1.6). Indirect or induced employment has also been plotted in both graphs
where the data is available. As can be seen, the employment generated increases as the
complexity or the crude capacity increases.
1000
800
Employment
600
400
200
0
0 2 4 6 8 10 12 14
Nelson complexity index
Figure 1.5: Employment in European refineries as a function of the complexity of the refinery
1000
800
Employment
600
400
200
0
0 5 10 15 20
Crude capacity (Mt/yr)
Figure 1.6: Employment in the European refineries as a function of the crude capacity
Big differences in employment are found for refineries with the same capacity or same
complexity. For instance, employment at refineries with a capacity of 5 million of tonnes per
year ranges from less than 300 to almost 900. Bigger differences can be found in the plot with
the Nelson complexity index.
It is important to know, with respect to the quality and quantity of refinery emissions, that on a
macro scale the crude oils vary only to a certain extent in their composition. Moreover,
refineries are often limited to a comparatively narrow range of crude oil diets. Normally, when
switching from one crude oil to another within this range, large variations in refinery emissions
are not usually expected. Consequently the type and quantity of refinery emissions to the
environment are well known during normal operations. However, from time to time, processing
of crude oils previously unknown to the refinery can have unforeseen impacts on the
performance of refinery processes; leading to an increase in emissions. This is particularly likely
to affect aquatic emissions and to a lesser extent air emissions.
Power plants, boilers, heaters and catalytic cracking are the main sources of emissions of carbon
monoxide and dioxide, nitrogen oxides (NOx), particulates, and sulphur oxides (SOx) to the
atmosphere. Refinery processes require a lot of energy; typically more than 60 % of refinery air
emissions are related to the production of energy for the various processes. Sulphur recovery
units and flares also contribute to those emissions. Catalyst changeovers and cokers release
particulates. Volatile organic compounds (VOCs) are released from storage, product loading
and handling facilities, oil/water separation systems and, as fugitive emissions, from flanges,
valves, seals and drains. Other emissions to the atmosphere are H2S, NH3, BTX, CS2, COS, HF
and metals as constituents of the particulates (V,Ni and others). Table 1.10 shows a very brief
summary of the main pollutants emitted by a refinery, with their main sources.
Table 1.10: Main air pollutants emitted by refineries and their main sources
Carbon dioxide (CO2) is the greenhouse gas which receives most attention in the debate about
global climate change. The main source of CO2 emissions is the production of energy. Since the
first oil crisis in 1973, refineries have made a concerted effort to enhance energy efficiency.
Despite these energy conservation measures, refinery energy demand has increased due to
tightened product specifications and a shift over the years from producing heavy fuel oils to
transport fuels.
Carbon monoxide (CO) always appears as an intermediate product of the combustion processes
and in particular in understoichiometric combustion conditions. However, the relevance of CO
releases from refineries is not very high compared to CO2.
Oxides of nitrogen (NOx), when emitted to the air, can combine with water and form a
component of ‘acid rain’. Further, NOx in combination with volatile organic compounds and
sunlight, can lead to the formation of ground-level ozone. The source of NOx is mainly the
combustion processes; during the combustion of fuel, nitrogen (mainly originating from the
combustion air itself) is transformed to a mixture of NO2 and NO. Combustion conditions play
an important role here. N2O (laughing gas) is a powerful greenhouse gas that contributes to the
stratospheric ozone depletion.
Particulate emissions have become a focus of attention because of their potential adverse health
effects. Particulate emissions are caused by combustion of fuel oils, especially when there is
sub-optimal combustion. Another source is the catcracker.
Sulphur oxides (SOx), when emitted to the air, can combine with water and form a component
of ‘acid rain’. The main source of SOx is the production of energy; during combustion, the
sulphur in the fuel is transformed to a mixture of SO2 and SO3. Another source, typically
smaller, is the flue gas from the sulphur recovery units. There is a direct relation between the
sulphur in the feed to a combustion process and the sulphur oxides in its flue gas. Generally
speaking the sulphur content of the refinery fuel pool is a careful balance between energy
required, the type of crude processed, the emission limits and economic optimisations.
Volatile organic carbons (VOC), as mentioned above, can react with NOx in the presence of
sunlight to form low-level atmospheric ozone. Furthermore, emissions of VOC can give rise to
odour problems, which may result in complaints from nearby residents. The source of VOC
emissions is the evaporation and leakage of hydrocarbon fractions during storage and
distribution. Hydrocarbons may also be emitted during non-optimal combustion conditions, but
these give only a small contribution.
Given the progress that refineries have made in the abatement of sulphur emissions to air, the
focus has started to shift towards VOC (including odour), particulates (size and composition)
and NOx, as it has in the environmental debate generally. When the carbon dioxide emissions
debate gathers momentum, it will also strongly affect refineries.
Water is used intensively in a refinery as process water and for cooling purposes. Its use
contaminates the water with oil products mainly increasing the oxygen demand of the effluent.
Refineries discharge waste water which originates from:
• Process water, steam and wash water. These waters have been in contact with the process
fluids, and apart from oil, will also have taken up hydrogen sulphide (H2S), ammonia (NH3)
and phenols. The more severe the conversion processes, the more H2S and NH3 are taken up
by the process water. The process water is treated in several, well-known steps before
discharge to the environment.
• Cooling water, once-through or circulating systems. This stream is theoretically free of oil.
However, leakage into once-through systems, even at low concentrations, can result in
significant mass losses because of the large volume of water involved.
• Rainwater from process areas. This type of water has not been in contact with the process
fluids, but it comes from rainfall on surfaces which are possibly oil-polluted. It is often
referred to as ‘accidentally oil-contaminated’ water and is typically treated prior to
discharge to the environment.
• Rainwater from non-process areas. This stream is oil-free.
Oil and hydrocarbons are the main pollutants found in waste water generated by refineries.
Other pollutants found in waste water generated by refineries are hydrogen sulphide, ammonia,
phenols, benzene, cyanides and suspended solids containing metals and inorganic compounds
(e.g. halides, sulphates, phosphates, sulphides). Table 1.11 gives a summary of the main water
pollutants as well as their main sources.
Refinery waste water treatment techniques are mature techniques, and emphasis has now shifted
to prevention and reduction. Reduction of water use and/or the concentration of pollutants in the
water, can have effects in reducing the final emission of pollutants.
The amount of waste generated by refineries is small if it is compared to the amount of raw
materials and products that they process. Oil refinery waste normally covers three categories of
materials:
§ sludges, both oily (e.g. tanks bottoms) and non-oily (e.g. from waste water treatment
facilities),
§ other refinery wastes, including miscellaneous liquid, semi-liquid or solid wastes (e.g.
contaminated soil, spent catalysts from conversion processes, oily wastes, incinerator ash,
spent caustic, spent clay, spent chemicals, acid tar) and,
§ non-refining wastes, e.g. domestic, demolition and construction.
Table 1.12 shows a summary of the main types of solid wastes generated in a refinery and their
sources.
Oil retained in sludges or other type of wastes represents a loss of product and, where possible,
efforts are made to recover such oil. Waste disposal depends very much on its composition and
on the local refinery situation. Because of the high operating costs of waste disposal, much
priority has been given to waste minimisation schemes.
Waste generation trends during the last ten years show that oily sludge production is declining,
mainly through housekeeping measures, whereas biological sludge generation has increased as a
result of increased biotreatment of refinery effluent. Spent catalysts production is also
increasing through the installation of new hydrocrackers, hydrotreatment facilities and catalytic
cracker dust collectors. For all these waste categories increased use is made of third party waste
contractors for off-site treatment and disposal.
Most refineries have some areas that are contaminated by historical product losses. Current
refinery practices are designed to prevent spillages and leaks to ground. In the past, the
awareness of the potential risks of these contaminated areas was low. The two main topics here
are prevention of new spills and remediation of historic contamination. As mentioned within the
scope, soil remediation is not included in the scope of this document. Most oil fractions are
biodegradable, given time. The thinking about the cleaning-up of those contaminated areas has
changed over the years. Increased knowledge about soil sciences and the difficulty of soil
remediation on a site that is still in operation, has lead to the pragmatic approach of managing
the risks of these contaminated sites to ensure their fitness for use and ensuring that pollution
does not spread beyond the site. There are a number of ongoing research initiatives for
improving the performance of on-site remediation techniques.
The main sources of contamination of soil and groundwater by oil are typically those places
along the handling and processing train of crude to products where hydrocarbons can be lost to
the ground. These are commonly associated with the storage, transfer, and transport of the
hydrocarbons themselves or of hydrocarbon-containing water. The possibility of contamination
by other substances such as contaminated water, catalysts and wastes also exists.
Besides the environmental issues mentioned in the above sections, especially for refineries
situated near residential areas, nuisance has become an issue of discussion with both local
authorities and with representatives of the local population in so-called neighbourhood councils.
Topics such as noise, light and smoke emission (flaring) and smell which directly impact the
residents receive much more emphasis in these neighbourhood councils than the above-
mentioned ‘major’ emissions which tended to receive more attention historically.
Nuisance by light can be caused through flaring at night by refineries and petrochemical plants
located near densely populated areas.
Since the 1970s refineries have devoted considerable resources to increasing process safety,
both in design and operation by training, procedures and personal protection equipment.
Increased attention, training, safe design and adequate tools and personal protection equipment
has resulted in a steady decrease in the number of unsafe acts, accidents, incidents and near
misses [242, CONCAWE, 1998].
Occupational health is included in the operational safety procedures aimed at protecting workers
from exposure to toxic materials and providing them with all necessary facilities which
contribute to their well-being and their sense of security and safety. Instructions, information
exchange and training of personnel, the provision of personal protection equipment as well as
strict adherence to stringent operational procedures have contributed to a steady decrease in
accidents and health incidents. Typical refinery pollutants and products with a health risk
include hydrogen sulphide, BTEX (of which benzene is the most prominent), ammonia, phenol,
HF, NOx and SOx, for which legally binding Maximum Acceptable Concentrations values
prevail.
The design of the refinery installations and the process control systems needs to include
provisions for a safe shut-down with minimum emissions from the unit involved. During
unplanned operational upsets, these provisions should guarantee that feed supply is terminated
followed by subsequent pre-programmed automated activation of pumps, relief systems,
purging systems, flares and other equipment. Examples of such occurrences are utility failure,
breakdown of equipment, a fire or an explosion. Emergency situations leading to direct spills
occurring in parts of the plant which are neither fully contained nor fully automated, such as
pipeline and tank bottom rupture, have to be addressed with standing emergency procedures.
These are directed to minimisation and containment of the spills, followed by rapid clean-up in
order to minimise the environmental impact.
The major production unit operations and activities typically involved at petroleum refineries
are described briefly, in alphabetical order, in the current chapter. For many of these production
operations, a number of different techniques and or unit operations are used in the industry.
While the major techniques used for each process/activity are described, it is not intended to
discuss and describe all of the different processes currently in use. This chapter specifically
contains the purpose and principle of the process, feed and product streams, a brief process
description of commonly used production processes/activity and operability. This information,
coupled with schematic drawings of the identified processes, provides a concise description of
the process/activity.
The first section is dedicated to a general technical description of refineries and the following 22
sections cover all the processes and activities within the scope of the BREF. Natural gas plants
processes are included in one independent section (Section 17). These sections do not pretend to
be of equal importance from the environmental impact point of view. Some sections have more
relevance than others, but this structure is considered a good way to tackle the BAT assessment
for the mineral oil and gas refinery sectors. No pollution abatement technique is include in this
chapter. However, some of the processes (e.g. hydrotreatments) have an environmental aspect.
Environmental implications of production techniques can be found in Chapter 3. Discussion on
good environmental practices (prevention techniques) of production techniques can be found in
Chapter 4. As consequence of that, no information about emissions can be found in Chapter 2.
Information on, for instance, flares, sulphur recovery units, waste water systems and waste
management is not included in this chapter because they are not considered production
activities. They are security or environmental techniques that are applied in both sectors.
The production of a large number of fuels is by far the most important function of refineries and
will generally determine the overall configuration and operation. Nevertheless some refineries
can produce valuable non-fuel products such as feedstocks for the chemical and petrochemical
industries. Examples are mixed naphtha feed for a steam cracker, recovered propylene, butylene
Mineral Oil and Gas Refineries 21
Chapter 2 Applied processes and techniques
for polymer applications and aromatics manufacture covered under the Large Volume Organic
Chemical BREF. Other speciality products from a refinery include bitumen, lubricating oils,
waxes and coke. In recent years the electricity boards in many countries have been liberalised
allowing refineries to feed surplus electricity generated into the public grid.
Refining crude oil into usable petroleum products can be separated into two phases and a
number of supporting operations. The first phase is desalting of crude oil (Section 2.9) and the
subsequent distillation into its various components or "fractions" (Section 2.19). A further
distillation of the lighter components and naphtha is carried out to recover methane and ethane
for use as refinery fuel, LPG (propane and butane), gasoline blending components and
petrochemical feedstocks. This light product separation is done in every refinery.
The second phase is made up of three different types of "downstream" processes: combining,
breaking and reshaping fractions. These processes change the molecular structure of
hydrocarbon molecules either by breaking them into smaller molecules, joining them to form
larger molecules, or reshaping them into higher quality molecules. The goal of those processes
is to convert some of the distillation fractions into marketable petroleum products (see
Section 10.3) through any combination of downstream processes. Those processes define the
various refinery types, of which the simplest is the ‘Hydroskimming’, which merely
desulphurises (Section 2.13) and catalytically reforms (Section 2.6) selected cuts from the
distillation unit. The amounts of the various products obtained are determined almost entirely by
the crude composition. If the product mix no longer matches the market requirements,
conversion units have to be added to restore the balance.
The market demand has for many years obliged refineries to convert heavier fractions to lighter
fractions with a higher value. These refineries separate the atmospheric residue into vacuum
gasoil and vacuum residue fractions by distillation under high vacuum (Section 2.19), and then
feed one or both of these cuts to the appropriate conversion units. Thus by inclusion of
conversion units, the product slate can be altered to suit market requirements irrespective of the
crude type. The number and the possible combinations of conversion units are large.
The simplest conversion unit is the thermal cracker (Section 2.22) by which the residue is
subjected to such high temperatures that the large hydrocarbon molecules in the residue convert
into smaller ones. Thermal crackers can handle virtually any feed, but produce relatively small
quantities of light products. An improved type of thermal cracker is the coker (Section 2.7), in
which all the residue is converted into distillates and a coke product. In order to increase the
degree of conversion and improve product quality, a number of different catalytic cracking
processes have evolved, of which fluid catalytic cracking (Section 2.5) and hydrocracking
(Section 2.13) are the most prominent. Recently, residue gasification processes (Section 2.14)
have been introduced within refineries, which enable refineries to eliminate heavy residues
completely and to convert them into clean syngas for captive use and production of hydrogen,
steam and electricity via combined cycle techniques.
Supporting operations are those not directly involved in the production of hydrocarbon fuels
but serving in a supporting role. They may include energy generation, waste water treatment,
sulphur recovery, additive production, waste gas treatment, blowdown systems, handling and
blending of products and storage of products.
STABILIZER
ALKYLATION
NAPHTHA
PENTANIZER
RISATION
[2]
[12]
HDS C5 [16]
DEISO-
[12]
[13]
POLYMERIZATION
[18]
DEPENTANIZER
Gasoline
[12]
HDS [11]
[13]
MeOH
LGO Naphtha
Crude DESALTING C 5+ REFORMING pool
HDS
[9] [6]
Oil [13]
HGO H2
HDS
[13]
Kerosene
pool
HC
[13]
VISBREAKING Gasoil
[22]
pool
[5]
Propylene
Butenes
BITUMEN
PLANT Bitumen
[4]
DAO: Deasphalted oil DEA: Diethylamine FCC: fluid catalytic cracker HC: Hydrocracking HDS: Hydro desulphurisation HGO: Heavy gas oil LGO: Light gas oil VGO: Vacuum gas oil
As it has been mentioned in Section 1.3.1, there are various approaches to compare refineries
due to the difference in complexity that they have. This can be done defining refinery
complexity indexes as they have been explained in Section 1.3.1. The Nelson refinery
complexity index was used for the construction of Figure 1.3 [287, Johnston, 1996]. In Dutch
notes [118, VROM, 1999] EU refineries are grouped by configuration, as shown in Table 1.9.
CONCAWE [115, CONCAWE, 1999] has grouped refineries by types (that differ in definition
from the configurations of Table 1.9) in Figure 1.4 as they existed in the period 1969 - 1997.
Other definitions are used in the industry, such as FCC equivalent (used by CEC DG Transport
reports on the refining industry) or the Equivalent Distillation Capacity as used by Solomon
Associates. Some of those parameters will be used in the document.
Table 2.1 summaries the main products obtained from the main refinery processes. As can be
seen, many of the products are obtained from different units and consequently it gives an idea of
the technical complexity and the modes of operation that may be found in a refinery.
PRODUCTS → Section LPG Gasoline Kero/ Heating HFO Base Coke/ Special
REFINERY UNITS ↓ naphtha oil/diesel Oil bitumen
Carbon no range C3-C4 C4-C12 C8-C17 C8-C25 >C8 >C15 >C30
Alkylation 2
Base Oil production 3 wax
Bitumen production 4
Catalytic cracker 5
Residue cracking 5
Catalytic reforming 6 H2
Delayed coking 7
Flexicoker 7 low joule
gas
Gasification 10 syngas
Etherification 11 MTBE
Gas separation 12 Refinery
processes fuel gas
Hydrogen plant 14 H2
Residue 14 H2
Hydroconversion
Hydrocracker 15
Hydrodesulphurisation 15
Isomerisation 17
Crude atmospheric 19
distillation
Vacuum distillation 19
Thermal 22
cracking/Visbreaking
Sulphur recovery unit 23 S
2.2 Alkylation
PURPOSE AND PRINCIPLE
The purpose of alkylation is to yield high-quality motor fuel blending. The term alkylation is
used for the reaction of olefins with isobutane, to form higher molecular-weight isoparaffins
with a high octane number. The process involves low-temperature reaction conditions
conducted in the presence of strong acids.
PROCESS DESCRIPTION
In the alkylation unit HF or H2SO4 is used as a catalyst. When the concentration of acid
becomes low, some of the acid must be removed and replaced with fresh acid.
In the hydrofluoric acid process, the slipstream of acid is redistilled. The concentrated
hydrofluoric acid is recycled and its net consumption is relatively low. In this process (Figure
2.2), the feed enters the reactor and is mixed with the isobutane recycle and HF from the settler.
Both the olefin and isobutane feeds are dehydrated first (not in Figure 2.2), as this is essential to
minimise potential corrosion. The reactor, operating at 25 - 45 °C and 7 - 10 barg, is cooled to
remove the heat generated by the reaction. In the settler, alkylate and excess isobutane are
separated from the HF. The HF is recycled to the reactor, where it is regenerated while the
organic phase consisting of the alkylate and unreacted isobutane flow to the isostripper. There,
the isobutane and some other light components are stripped from the alkylate. This is routed as
product to storage after treating with potassium hydroxide. The decomposition of any organic
fluoride formed is promoted by the high tube wall temperatures. The butane feed (n and iso) is
normally fed to the isostripper. The isostripper overhead, essentially isobutane, is returned to the
reactor. A small overhead slipstream is sent to a depropaniser, where propane is removed. The
bottoms of the depropaniser, isobutane, are routed to the isobutane circulation stream, while the
depropaniser overhead stream, propane, passes through a HF stripper to remove traces of HF
and is sent to storage after final KOH treating. From the unit feed stream, n-butanes are drawn
off as a side-stream from the isostripper, treated with KOH and sent to storage.
FLARE
HF STRIPPER
REACTOR HF REGE-
25-45°C NERATOR
7-10 barg DEPRO-
HF MAKE-UP
FUEL
PROPANE
ALKYLATE
OLEFIN ACID TARS
FEED TO NEUTRALISATION n-BUTANE
AND INCINERATION
SATURATE
FEED (BUTANES)
In the sulphuric acid process, units operate at 4 - 15 °C, which requires chilling. Spent sulphuric
acid generation is substantial and the sulphuric acid removed must be regenerated in a sulphuric
acid plant (not considered as part of the alkylation unit). In this process, the olefin feed and
recycled isobutane are introduced into the stirred, autorefrigerated reactor. Mixers provide
intimate contact between the reactants and the acid catalyst and the reaction heat is removed
from the reactor. The hydrocarbons that are vaporised from the reactor are routed to the
refrigeration compressor where they are compressed, condensed and returned to the reactor. A
depropaniser, which is fed by a slipstream from the refrigeration section, is designed to remove
any propane introduced to the plant with the feeds. The reactor product is sent to the settler
where the hydrocarbons are separated from the acid that is recycled. The hydrocarbons that are
sent to the deisobutaniser, together with make-up isobutane and the isobutane rich overhead, are
recycled to the reactor. The bottoms are then sent to a debutaniser to produce alkylate product.
In the various processes of the base oil production, significant quantities of by-products such as
bitumen, extracts and wax are produced. Based on a total feedstream to the vacuum distillation
unit, on average some 20 - 25 % ends up as final base oil product.
PROCESS DESCRIPTION
A base oil complex consists typically of a vacuum distillation tower (Section 2.19), a
deasphalting unit, an aromatic extraction unit, a dewaxing unit and optional high-pressure
hydrogenation unit and hydrofinishing unit to improve colour and stability, meet product
specifications and remove impurities. Figure 2.3 shows a simplified block diagram of a base oil
manufacturing plant.
TO VACUUM GASOIL
TRANSFORMER OIL
VACUUM RESIDUE
A conventional base oil complex is very labour intensive mainly due to its batch operation, the
many grades of base oil normally produced and the associated intensive product handling
operations.
Deasphalting
Solvent deasphalting produces lubricating oil base stocks by extracting high-boiling lighter
paraffinic and naphthenic hydrocarbons (asphaltenes and resins) from the vacuum residue of the
vacuum distillation unit. This process makes the deasphalted oil extract light and paraffinic and
the bitumen raffinate heavy and aromatic. Propane or propane-butane mixtures are usually used
as solvents. At specific operating conditions, 37 - 40 bar and 40 - 70 °C, lower boiling paraffinic
and naphthenic hydrocarbons are very soluble in propane. At higher temperatures (100 ºC) all
hydrocarbons are almost insoluble in propane. The solvent deasphalting process is a typical
extraction process consisting of an extractor and recovery sections, flash evaporation and
stripping, to separate the propane solvent from the oil and bitumen phase. The deasphalted oil
product stream is run down to intermediate storage; the bitumen product stream can be blended
to heavy fuel, used as feedstock for the coker or used for the bitumen product.
More recently, solvent deasphalting has been adapted for the preparation of catalytic cracking,
hydrocracking, hydrodesulphuriser feeds and hard bitumen (deep deasphalting). For these
purposes, heavier-than-propane solvents (butane to hexane mixture) are used together with
higher operating temperatures. This maximises the yield of valuable deasphalted oil and
minimises the yield of hard bitumen with a softening point usually over 150 ºC.
Aromatic extraction
Aromatic extraction uses solvents to remove aromatics from base oil feedstocks, improving
viscosity, oxidation resistance, colour and gum formation. A number of different solvents can
be used (furfural, N-methyl-2-pyrrolidone (NMP), phenol or liquid sulphur dioxide). These
processes are typical extraction processes consisting of an extractor and recovery sections, flash
evaporation and stripping, to separate the solvent from the oil-rich raffinate and aromatic-rich
extract stream. Typically, feed lube stocks are contacted with the solvent in a packed tower or
rotating disc contactor. Solvents are recovered from the oil stream through distillation and steam
stripping in a fractionator. The raffinate stream is rundown to intermediate storage. The extract,
after solvent recovery, is likely to contain high concentrations of sulphur, aromatics, naphthenes
and other hydrocarbons, and is often fed to the hydrocracker or catcracker unit.
Dewaxing
Dewaxing of lubricating oil base stocks is necessary to ensure that the oil will have the proper
viscosity at lower ambient temperatures. This process is used when paraffinic-rich crudes are
processed. Solvent dewaxing is more prevalent. In these units the high pour point constituents
(mainly paraffins) are removed from the raffinate streams. The oil feed is diluted with solvent to
lower the viscosity, chilled until the wax is crystallized, and then filtered to remove the wax.
Solvents used for the process include propane and mixtures of methyl ethyl ketone (MEK) with
methyl isobutyl ketone (MIBK), toluene or chlorinated hydrocarbons. Solvent is recovered from
the oil and crystallised wax through heating, two-stage flashing, followed by steam stripping.
The wax is removed from the filters, melted and subsequently fed to a solvent recovery unit to
separate the solvent from the wax. The wax is either used as feed to the catalytic cracker or is
de-oiled and sold as industrial wax.
Hydrofinishing
In this unit the colour and colour stability is improved and the organic acid components are
removed. The need for hydrofinishing depends on the crude oil processed and to a certain extent
on the licensor and the design of the preceding units. The design and operation of this unit is
similar to that of a normal hydrotreater unit (Section 2.13).
Normally, a number of different grades of bitumen are produced in campaigns and these are
further modified by blending with other high-boiling components such as vacuum residue,
heavy gas oil or synthetic polymers. In this way a single blowing unit is able to cater for a wide
range of bitumen grades for various applications.
PROCESS DESCRIPTION
The BBU will either operate on a continuous basis or in batch mode depending on the quality of
the vacuum residue feedstock and the required bitumen product specification. Continuous
processes are more spread in refineries. A simplified process flow diagram of the BBU is shown
in Figure 2.4 which represents a typical continuously operated BBU receiving its hot feed
directly from the vacuum distillation unit. Where the bitumen feed is received from storage, an
additional fired heater may be required to preheat the feed to a temperature of about
200 - 250°C, but can be up to 550 ºC. With a batch-operated BBU, a feed buffer vessel is
usually included to store the hot feedstream from the vacuum unit.
WASTE GAS
TO INCINERATOR
STEAM
WATER OXIDIZER
SCRUBBER
VACUUM RESIDUE 300°C SOUR WATER
2.2 barg TO SWS
AIR
SLOP OIL
BITUMEN
STEAM BFW
The residue feedstream is pumped into the top of the oxidation vessel. Operating pressure in the
top of the oxidation vessel is normally around 1 barg and in the bottom around 2 barg,
depending on the height of the vessel. As air is sparged into the base of the vessel, oxidation of
the residue takes place, resulting in heat. The temperature in the oxidation vessel, which
determines to a certain extent the bitumen grade, is normally controlled between 260 – 300 °C.
Different options are applied, taking in colder feed to the oxidation vessel, recirculation of
cooled bitumen product from the bitumen rundown cooler, and in older units even direct water
quenching is applied. The blown bitumen is removed from the bottom of the oxidation vessel
and cooled by raising steam before being sent to storage.
The air rate is normally well in excess of stoichiometric requirements and so a considerable
quantity of oxygen is present in the upper vapour space of the oxidation vessel. To avoid an
explosion in the vapour space, in most units steam is injected at a rate necessary to keep the
oxygen concentration below the lower flammable limit (5 - 6 % vol.). In some units a small
amount of water is also injected into the vapour outlet of the oxidation vessel to reduce the
vapour temperature. This is sometimes considered necessary to prevent afterburning in the
overhead system that could lead to severe coke formation.
The overhead vapours are first passed through a vent gas scrubber to remove oil and other
oxidation products. In most cases gasoil is used as once-through scrubbing liquid. The vent gas
from the scrubber is subsequently cooled to condense light hydrocarbons and sour water,
sometimes in a water spray contact condenser or scrubber. The remaining gas, consisting mainly
of light hydrocarbons, N2, O2, CO2 and SO2, is incinerated at high temperature (~800 °C) to
ensure complete destruction of minor components such as H2S, complex aldehydes, organic
acids and phenolics, which have a highly unpleasant odour.
The majority of the BBUs produce the higher grades of bitumen (roof and pipe coatings) and
normally operate continuously throughout the year. The BBUs which are used to produce road
bitumens operate only when the demand for road asphalt is high.
PROCESS DESCRIPTION
A number of different catalytic cracking designs are currently in use in the world, including
fixed-bed reactors, moving-bed reactors, fluidized-bed reactors and once-through units. The
fluidized- and moving-bed reactors are by far the most prevalent in world refineries.
Fluid catalytic cracking (FCC) units are by far the most common catcracking units. The FCC
unit consists of three distinct sections, the reactor-regenerator section including air blower and
waste heat boiler, the main fractionator section including wet gas compressor and the
unsaturated gas plant section. A simplified flow scheme is shown in Figure 2.5. In the FCC
process, oil and oil vapour preheated to 250 to 425°C is contacted with hot catalyst (zeolite) at
about 680 - 730°C in the riser reactor. To enhance vaporisation and subsequent cracking, the
feed is atomised with steam. The cracking process takes place at temperatures between 500 and
540°C and a pressure of 1.5 - 2.0 barg. Most catalysts used in catalytic cracking are zeolites
(some 15% w/w) supported by amorphous synthetic silica-alumina with metals. The catalyst is
in a fine, granular form which mixes intimately with the vaporised feed. The fluidised catalyst
and the reacted hydrocarbon vapour separate mechanically in a (two-stage) cyclone system and
any oil remaining on the catalyst is removed by steam stripping. The amount of lost catalyst as
fines in the cyclone system is balanced by the addition of fresh catalyst. The catalytic cracking
processes produce coke which collects on the catalyst surface and diminishes its catalytic
properties. The catalyst therefore needs to be regenerated continuously or periodically,
essentially by burning the coke off the catalyst at high temperatures. The method and frequency
with which catalysts are regenerated are a major factor in the design of catalytic cracking units.
The catalyst flows into a separate vessel(s) for either single- or two-stage regeneration by
burning off the coke deposits with air. However, in time the catalyst deactivates gradually and
irreversibly due to high temperature exposure and metal poisoning (mainly vanadium). The hot
regenerated catalyst flows back to the base of the reactor riser, where it is cooled by
vaporisation of the feed and by the heat required for cracking reactions. The cracked oil vapours
are then fed to a fractionation tower where the various desired fractions are separated and
collected. The streams are steam-stripped to remove volatile hydrocarbons prior to cooling and
sending to storage. Slurry oil is taken from the bottom of the column and part of it is cooled
with the reactor feed stream and steam generation and is returned to the column. The recycled
slurry oil is used as wash oil and as a quench for the hot reactor overhead vapours to prevent
aftercracking. The rest of the slurry oil is filtered or decanted to remove catalyst fines, cooled
and sent to storage. The fractionator overhead gas is partly condensed and accumulated in the
overhead drum to separate it into three phases: gas, liquid and sour water streams. The liquid
and gas streams are both fed to the gas plant (Section 2.12) for further separation and the sour
water stream is sent to the sour water stripper for purification (Section 2.23).
FLUE GAS
FRACTIONATOR
CATALYST STEAM
NAPHTHA
REGENERATOR
MAIN
ABSORBER LCO
OR STEAM
STEAM
STEAM STEAM
GENERATION
HCO
FEED
START-UP NAPH. SPLITTER STEAM
HEATER DEBUTANISER GENERATION
AIR 600-700°C
STEAM
STEAM GENERATION
COMPRESSION
FEED FILTER
SLURRY PA
Figure 2.5: Simplified process flow scheme for fluid catalytic cracker
The flow sheet of a residue catalytic cracker (RCC) is basically the same as for a FCC with
the difference that it often has a CO boiler and a catalyst cooler. Sometimes FCC units are
equipped with both a CO boiler and a catalyst cooler and can consequently be used as RCC unit.
Due to the higher coke lay-down on the catalyst as a result of the heavier feeds, the heat balance
around the regenerator may require additional measures as catalyst coolers. As the heavier feed
normally has a higher metal content, especially Ni and V, the catalyst deactivation rate is so fast
that the catalyst needs to be removed continuously, to be replaced by fresh catalyst.
In the moving-bed process, oil is heated to 400 - 700 ºC and is passed under pressure through
the reactor, where it comes into contact with a catalyst flow in the form of beads or pellets. The
cracked products then flow to a fractionating tower where the various compounds are separated.
The catalyst is regenerated in a continuous process. Some units also use steam to strip
remaining hydrocarbons and oxygen from the catalyst before catalyst is fed back to the oil
stream. In recent years moving-bed reactors have largely been replaced by fluidized-bed
reactors.
Some catalytic reformers operate under more severe conditions, resulting in increased aromatics
content in the reformate product. Some catalytic reformers operate for the production of
aromatics as a (chemical) product (see Large Volume Organic Chemical BREF).
PROCESS DESCRIPTION
Feedstocks to catalytic reforming processes are usually hydrotreated first to remove sulphur,
nitrogen and metallic contaminants. The catalysts used in catalytic reforming processes are
usually very expensive (containing Pt) and extra precautions are taken to ensure that catalyst is
not lost. There are several catalytic reforming processes in use today. In general they can be
classified in three categories: continuous, cyclic or semi-regenerative, depending upon the
frequency of catalyst regeneration. The fixed-bed or moving-bed processes are used in a series
of three to six reactors.
FLUE GAS
500-550°C
15-30 barg
REACTOR 1
COMPRESSOR
REACTOR 3
HP
~500°C SEPARATOR LPG
FURNACE
DEBUTANIZER
SPENT CATALYST LP
SEPARATOR
FEED
REFORMATE
Figure 2.6: Simplified process flow scheme for a continuous catalytic reformer
Interheaters are needed between the reactors to provide the heat required for the endothermic
dehydrogenation reaction. As the naphtha feed charge proceeds through the reactors, the
reaction rates decrease and the reheat needed becomes less. Freshly regenerated catalyst is
introduced at the top of the first reactor and flows by gravity from top to bottom. From here it is
passed on to the next reactor. Partially aged catalyst is removed from the bottom of the lowest
reactor and sent to an external regenerator where the carbon is burned from the catalyst. The
catalyst is reduced and acidified before being returned to the upper reactor. The reaction mixture
from the last reactor is used to preheat the fresh feed and further cooled down prior to being fed
to the low-pressure separator. There the hydrogen-rich gas is separated from the liquid phase.
The gas is compressed and partially recycled to the naphtha feed. The remaining vapour is then
further compressed, recontacted with the liquid from the low-pressure separator, cooled and fed
to the high-pressure separator. The recompression and recontacting is included to maximise
recovery of the C3/C4 fraction from the hydrogen rich gas stream. Any excess of this hydrogen
rich gas is bled to the refinery fuel gas grid. The selection of the reformer operating pressure and
the hydrogen/feed ratio form a compromise between maximum yields and stable operation.
Cyclic process
The cyclic process is characterised by having a swing reactor in addition to those on-stream, in
which the catalyst can be regenerated without shutting the unit down. When the activity of the
catalyst drops below the desired level, this reactor is isolated from the system and replaced by
the swing reactor. The catalyst in the replaced reactor is then regenerated by admitting hot air
into the reactor to burn the carbon from the catalyst. Typically, one reactor is always being
regenerated.
Semi-regenerative process
In this process, regeneration requires the unit to be taken off-stream. Depending upon the
severity of operation, regeneration is required at intervals of 3 to 24 months. High hydrogen
recycle rates and operating pressures are utilised to minimise coke laydown and the associated
loss of catalyst activity.
The products from the coking fractionator are refinery fuel gas, LPG, naphtha and light and
heavy gas oils. Petroleum coke is another product, with the type depending upon the process
used, operating conditions and feedstock used. Coke produced by the coker is called “green”
coke and still contains some heavy hydrocarbons left from incomplete carbonisation reactions.
More information about the properties of the coke can be found in Section 10.3.2.
PROCESS DESCRIPTION
Two types of coking processes exist: the Delayed and fluid coking processes that produce coke
and the Flexicoking process that gasifies the coke produced in a fluid coking process to produce
coke gas.
REFINERY GAS
UNSTABILIZED
NAPHTHA NAPHTHA
FRACTIONATOR
LGO
FEED STRIPPER
400°C
0.5-2.0
barg
COKE FINES
FUEL
Figure 2.7: Simplified process flow scheme for delayed coking unit
The delayed coking feed stream of residual oils is first introduced to a fractionating tower,
where residual lighter materials are drawn off and the heavy ends are condensed (not shown in
Figure 2.7). The heavy ends are removed, heated in a furnace and then fed to an insulated vessel
called the coke drum, where the cracking takes place. In the case of fluid coking, a fluidised bed
is used. Temperature (440 - 450°C), pressure (1.5 - 7.0 barg) and recycle ratio are the main
process variables which contribute to the quality and yields of delayed coking products. When
the coke drum is filled with product, the feed is switched to an empty parallel drum (dotted
drum in Figure 2.7).
When the coke drum is full, steam is then injected to remove hydrocarbon vapours. The coke
bed is then quenched with water and the cokes are cut out with high pressure water. The cutting
water passes to dedicated settlement where coke solids settle out and the clarified water is re-
cycled. The wet green coke is conveyed to designated open stock piles, where water drains out
and is recycled. Green coke can already be sold and used for energy production. The yield of
delayed coking is usually above 80 % conversion of feed to products. The production of
petroleum coke is of the ratio of 0.13 tonnes of petroleum coke produced per each tonne of
feedstock.
Hot vapours from the coke drums, containing cracked lighter hydrocarbon products, hydrogen
sulphide and ammonia, are fed back to the fractionator where they can be treated in the sour gas
treatment system or drawn off as intermediate products. Condensed hydrocarbons are re-
processed and collected water is re-used for coke drum quenching or cutting. Any remaining
vapours commonly pass to the flare system. Normally, products such as naphtha are totally fed
to the naphtha hydrotreater for further processing. The heavier products are suitable feedstock
for catalytic reforming after proper hydrotreatment. Light oil needs further treatment before
being sent to the gasoil blending pool. The heavy gasoil is preferably sent to a hydrocracker unit
(Section 2.13) for further conversion into light components. When no cracking units are
available it is blended in the heavy fuel oil pool.
For certain applications, green coke should be calcinated before being used or sold. The
incinerator kilns are direct fuel gas or coke fine fired at the discharge end and calcinate the coke
at up to 1380 ºC, driving off volatile matter and burning it within the kiln. Exhaust gases
discharge from the feed end and are incinerated to burn off residuals and coke fines. Hot flue
gases pass through a waste heat boiler and gas cleaning by multi-cyclones. Collected fines from
the cyclones are pneumatically conveyed to a silo with exit air filters. The calcined coke
discharges to a rotary direct water injection. Off-gases from the cooler pass to gas cleaning by
multi-cyclones and water scrubber. Collected cyclone fines may be re-cycled to product, which
is oil sprayed as a dust suppressant, or may be incinerated or sold as a fuel.
Flexicoking
The flexicoking process typically converts 84 - 88 %w/w of the vacuum residue to gaseous and
liquid products. Virtually all metals in the feed are concentrated in the 2 % solids purged from
the process. Flexicoking is a very robust process in which coking and gasification are fully
integrated. The process is advanced compared to the classic delayed coker in terms of operation
and labour intensity.
The flexicoking process uses three major vessels: the reactor, the heater and the gasifier. As
auxiliary facilities the system includes a heater overhead cooling system and fines removal
system, a coke gas sulphur recovery unit and reactor overhead scrubber (Figure 2.8). The
preheated vacuum residue feed is sprayed into the reactor, where it is thermally cracked,
typically at 510 - 540 °C. The freshly formed coke is deposited on the surface of the fluidised
recirculated coke particles. In the gasifier, the coke is reacted at elevated temperatures, typically
850 - 1000 °C, with air and steam to form coke gas, a mixture of hydrogen, carbon monoxide
and dioxide and nitrogen. Sulphur in the coke is converted in the gasifier primarily to hydrogen
sulphide, plus traces of carbonyl sulphide (COS). Nitrogen in the coke is converted to ammonia
and nitrogen. Unlike normal gasifiers which are fed with pure oxygen, the flexicoking gasifier is
fed with air, resulting in a coke gas with a relatively low calorific value, as it contains a high
quantity of nitrogen inerts.
COKE
GAS
REACTOR
HOT RESIDUE HEATER GASIFIER LCGO
FEED STRIPPER
480-540°C STEAM
0.7 barg
LCGO
PURGE AIR
COKES
STEAM STEAM STEAM
HCGO
COKES
The cracked hydrocarbon vapour product from the reactor is passed through cyclones to remove
coke particles and is subsequently quenched in a scrubber section located at the top of the
reactor. Material boiling above 510 – 520 °C is condensed in the scrubber and recycled to the
reactor. The lighter material goes overhead to conventional fractionation, gas compression and
light ends recovery sections. The treatments and use of the products are very similar to the ones
already described in the delayed coker. Pressure relief from the fractionator of the coking
processes passes to flare and from the coke drums to the quench tower system.
PROCESS DESCRIPTION
A wide range of techniques is used for cooling purposes in oil refineries. Most refineries use a
combination of the techniques available. The selection of a cooling system depends on required
cooling temperature, cooling capacity, contamination risk (primary or secondary cooling loop)
and local circumstances. Simplified diagram of the cooling techniques are shown in Figure 2.9
and briefly described below:
A B
C D
HOT AIR
E Dry F Wet
PROCESS OTHER
OR COOLING PROCESS COOLING
SYSTEM SYSTEM
COLD AIR
Open Closed
G H water
Wet air
Air
AIR
Air cooling
In an air cooler (forced or induced draught), the process stream in the tubes is cooled against air
delivered by a fan. An induced draught air cooler is illustrated in Figure 2.9E.
Water cooling
(1) Direct cooling (i.e. quenching)
Because the high contamination generated by this type of cooling, nowadays, quenching is only
used in cokers (Section 2.7), in gasifiers and in some sludge incinerators.
2.9 Desalting
PURPOSE AND PRINCIPLE
Crude oil and heavy residues can contain varying quantities of inorganic compounds such as
water soluble salts, sand, silt, rust and other solids, together characterised as bottoms sediment.
The salt in the crude is primarily in the form of dissolved or suspended salt crystals in water
emulsified with the crude.Those impurities, especially salts, could lead to fouling and corrosion
of heat exchangers (crude preheaters) and especially the crude distillation unit overhead system
(Section 2.19). Salts are detrimental to the activity of many of the catalysts used in the
downstream conversion processes and sodium salts stimulate coke formation (e.g. in furnaces).
Some inorganic impurities are chemically bound, such as vanadium and nickel, often referred to
as oil-soluble salts. These cannot be eliminated in a desalter. Moreover, water should preferably
be removed if the water content of the crude is relatively high. Therefore desalting of the
incoming crude oil is generally applied before separation into fractions (in Section 2.19).
The principle of desalting is to wash the crude oil or heavy residues with water at high
temperature and pressure to dissolve, separate and remove the salts and solids.
PROCESS DESCRIPTION
After preheating to 115 - 150 °C, the oily feedstock is mixed with water (fresh and pre-used
water) in order to dissolve and wash out the salts. Intimate mixing takes place by contacting the
wash water in a globe valve mixer, a static mixer or a combination of both. The water must then
be separated from the oil feedstock in a separating vessel by adding demulsifier chemicals to
assist in breaking the emulsion and/or, more commonly, by applying a high-potential electric
field across the settling vessel to coalesce the polar salt water droplets. The separation efficiency
depends on pH, density and viscosity of the crude oil, as well as the volume of wash water used
per volume of crude. Either AC or DC fields may be used and potentials from 15 to 35 kV are
used to promote coalescence. Many refineries have more than one desalter. Multiple stage
desalters also exists. The wash water containing dissolved hydrocarbons, free oil, dissolved salts
and suspended solids is further treated in an effluent treatment plant (Section 4.24). Where
bottoms sediments are critical in downstream process units, desalters are equipped with a
bottom flushing system to remove settled solids. Figure 2.10 shows a simplified process flow
diagram of a modern design crude desalter.
DESALTED CRUDE
TO PREFLASH VESSEL
OPTIONAL
DUAL POLARITY
OIL TO CRUDE
OIL STORAGE
WASH WATER
FROM VARIOUS EFFLUENT
SOURCES
IN LINE
MIXER
CRUDE DISTILLATION UNIT
PRODUCT STREAMS EFFLUENT TO WASTE WATER
SAND/SLUDGE
TREATMENT PLANT
WET CRUDE OIL
FROM STORAGE
FLUE GAS
SECONDARY FUEL
GAS SUPPLY TO USERS
STEAM
LPG VAPORISER
STEAM
LPG MAKE-UP
Liquid refinery fuel (heavy fuel oil, HFO) used in the refinery is normally a mixture of the
residues from atmospheric and/or vacuum distillation and conversion and cracking processes.
Liquid refinery fuels are available in various grades, viscosity being the main parameter. The
lower the viscosity the more expensive the fuel. The heavier (more viscous) grade fuels require
heating to reduce their viscosity before combustion. They contain sulphur (<0.1 - 7%), particle
promoters (eg. V, Ni) and nitrogen (0.1 - 0.8 %) resulting, after direct combustion, in high SO2,
particulates and NOx emissions. They can also be gasified in the IGCC plant where virtually any
refinery residue (visbroken or thermal tars, etc.) can be converted to heat and power.
Provided that the crude oil is properly desalted, ash content of the fuel will be directly related to
the total solids, the amount being proportional to the sum of the nickel and vanadium present
(Ni-V value: 0.03 - 0.15 % w/w depending on the residue source and crude origin). To arrive at
the metal content of the HFO, the metal content of the crude is multiplied by a factor of 4 - 5
(dependent on the yield of residue and the residue content of the crude). The metal content of
the HFO can vary between 40 and 600 ppm for HFO from a North Sea crude and Arabian
Heavy crude respectively, generating particulate concentration in the flue gas of 150 - 500
mg/Nm3. The most prominent metals indigenous to the crude are vanadium and nickel. Other
metals as cadmium, zinc, copper, arsenic and chromium have been detected [43, Dekkers and
Daane, 1999]. Next table shows the metal content of residual fuel oil typically used in
refineries.
FLUE GAS
RECIRCULATION
FROM BURNER
STORAGE TANK WITH OUTFLOW
HEATER AND HEATING COIL TO FUEL OIL USER
DUPLEX
FILTER
STEAM
Solid fuels such as petroleum coke can be gasified as a refinery fuel gas source for refineries
(Flexicoking, Section 2.7). Coke is burnt in the catalytic cracking regenerator (Section 2.5) and
coking process (Section 2.7) and represents a heat production source in the refinery. Coal, as
imported fuel, is not applied in European refineries.
A variety of furnaces and burner types are used in refineries, largely determined by the heat
release characteristics required by a particular process. Many, but not all, furnaces are dual
(oil/gas) fired to allow flexibility in the refinery fuel system. Refinery process heaters are
typically rectangular or cylindrical enclosures with multiple fired burners of specialised design
using mainly low combustion intensity. Boilers (fixed or fluidised bed) are generally fairly
standard steam-producing units of medium or high combustion intensity. Waste heat boilers
may also be present in the furnaces of stacks. Direct-fired heaters and boilers normally achieve
thermal efficiencies of over 85 %. If air preheat is applied and the combustion products (flue
gas) are cooled close to their dew point, the thermal efficiency can be as high as 93 %. Boilers
consume about 10 - 20 % of refinery energy requirements.
The principle is based on the high-temperature and high-pressure reaction of organic carbons or
coke with steam and under-stoichiometric amounts of oxygen (partial oxidation) to produce
syngas (CO+H2). After the combustion chamber the system contains a number of sophisticated
energy recovery systems to produce steam and electricity. Figure 2.13 shows a block flow
diagram of an IGCC plant. In the partial oxidation of hydrocarbons the product gas contains a
certain amount of free carbon (soot). The soot particles are removed from the gas together with
the ash in a two-stage water wash.
AIR
AIR
SEPARATION NITROGEN
OXYGEN
OXYGEN
BFW WATER POWER POWER
HYDROCARBON
FEEDSTOCK HYDROGEN
SEPARATION
GAS ACID GAS AND/OR GAS STEAM
GASIFICATION
COOLING REMOVAL PRODUCTION TURBINE TURBINE
(WATER GAS
SHIFT REACTION)
HOT EXHAUST
STEAM
EXHAUST
HYDROGEN
HEAT
SOOT RECOVERY
SULPHUR RECOVERY
/ GREY WATER
RECOVERY STEAM
TREATMENT BFW WATER GENERATOR
STEAM TO
METAL WASTE SULPHUR REFINERY
CAKE WATER STEAM
STEAM
The gasification plant consists of two integrated complex units. In the first, syngas
manufacturing process plants (SMPP), the gasification of heavy fractions takes place and the
syngas is produced and purified. In the second, the combined cycle power plants, the syngas is
fed to a combined cycle thermoelectric unit. The SMPP includes the following two sections:
• Gasification and carbon extraction: In the gasification section the feedstock is gasified
through a non-stoichiometric reaction with pure oxygen and water; the reaction occurs in
the gasifier, a non-catalytic vessel internally coated with refractory, operating at high
temperature (about 1300 °C) and pressure (about 65 bar).
• Gas cooling and purification: In the gas-cooling section the waste heat from syngas is
recovered by generation of steam at three pressure levels. A small quantity of carbon,
formed in the gasifiers, is removed from the gas by direct contact with water in a scrubber.
Water is then treated in the grey water treatment and then sent to the existing refinery
biotreatment plant. A solid effluent in the form of a filter cake is discharged from this unit
and sent to external plants for the recovery of the metals. In addition, a COS hydrolysis
reactor is provided to convert the small amount of COS produced in the gasifier to H2S.
This section also includes a gas expander for the recovery of the pressure energy of the
syngas (pressure in gasifiers is about 65 bar). This system contains an acid gas removal. In
this unit, a circulating amine stream is used to selectively absorb the H2S formed in the
gasifier and in the COS hydrolysis. It also contains an air separation unit. This unit
produces the oxygen required for the gasification and the Claus plant, and the nitrogen for
syngas conditioning. It is based on conventional cryogenic air fractionation. And finally
contains a sulphur recovery unit. Claus units recover elemental sulphur from the H2S
recovered in the acid gas removal section and a tail gas treatment section, that maximise the
overall sulphur recovery.
After the cooling and purification sections, the purified syngas is sent to the combined cycle
power plant, for power generation. It mainly consists of a conventional cycle with a gas turbine,
a heat recovery steam generator and a steam turbine.
Steam
The different steam qualities generated in the boilers of the refinery have the following general
characteristics (the heat content ranges from about 2700 MJ/t for low pressure steam to
3200 MJ/t for super heated high pressure steam at 50 bar):
• High pressure (HP) steam net (>30 bar, 350 – 500 °C), generated in waste heat boilers
(cooling of hot off-gases and/or hot products at catalytic processes and hydrocrackers) and
in fired boilers. HP-steam is mainly used in turbines to produce electrical power (and MP
steam);
• Medium pressure (MP) steam net (7 - 20 bar, 200 – 350 °C), generated by pressure
reduction of HP steam, is used within the refinery for stripping, atomisation, vacuum
generation and heating (e.g. reboilers, tanks);
• Low pressure (LP) steam net (3.5 - 5 bar, 150 – 200 °C), generated in heat exchangers by
cooling of hot products, and by pressure reduction of MP steam. LP steam is used for
heating, stripping and tracing.
Steam is produced by heating demineralised water, so called boiler feed water (BFW), under
pressure in a steam boiler. Steam raising plants are normally fuelled with refinery fuel gas or
liquid. The refinery is equipped with dedicated steam boilers in virtually all process units,
containing a HP-, MP- and LP-steam distribution network and with HP-, MP- and LP-
condensate collection networks, which are connected with the BFW preparation unit and the
condensate storage tank. (Reference is made to Figure 2.14 and to the LCP BREF [317,
EIPPCB, 2002]).
Steam used in turbines and heaters after cooling is usually recovered as condensate. BFW is
therefore a mixture of fresh demineralised make up water (quality dependent on steam pressure)
and recovered condensate. BFW make up can be bought in but can also be prepared at the
refinery using drinking water, filtered ground water, seawater distillation, surface water or even
treated effluent by using a combination of treatment operations, such as sandfiltration or
microfiltration (to remove suspended solids) and demineralisation which is accomplished by
subsequent cation and anion exchange (see dotted area in Figure 2.14). Reverse osmosis (to
remove ions, colloids and large organic molecules) is usually applied in new plants and is
followed in some cases by mixed bed ion exchange and active carbon filtration for final
polishing. The condensate tank is generally equipped with an oil detection system and an oil
skimming device. To avoid corrosion in the steam and condensate systems, oxygen and carbon
dioxide are removed in de-aerators, and oxygen scavengers and corrosion inhibitors are added.
After conditioning, the BFW is pumped to the boilers. In boilers the hot flue gases and BFW
flow in countercurrent; BFW is preheated in the economiser and further heated in the first and
second superheater. In order to keep the concentration of dissolved compounds and suspended
solids in the steam drum constant, a condensate blown down of 1 - 2 % is normally required.
Drinking
water of
HCl NaOH ELECTRICITY
iron-free
TANK TANK
ground CO2
water
STEAM USERS
Steam
CATION ANION
EXCHANGER DEGASSER EXCHANGER
Stack
Condensate emissions
return
SUPERHEATER 2
DAERATOR
Chemicals
conditioning
tank Boiler feed water (BFW)
Fuel Blowdown
to WWTP
Figure 2.14: Typical layout of a boiler feed water preparation unit and a steam boiler
Electrical power is mainly generated in turbines with high-pressure steam but can be also be
produced in gas turbines. Electricity is necessary to run pumps, compressors, control systems,
valves, etc. Refinery electrical systems are therefore extensive.
ENERGY MANAGEMENT
Good design and management of energy systems are important aspects of minimising the
environmental impact of a refinery, bearing in mind the highly integrated and interdependent
nature of most processes. The normal aim is to match continuously the variable production and
consumption of fuels in processes and utilities at the lowest economic and environmental cost.
This issue is also analysed in this document and in Section 2.15 the integration of all the
techniques that may be used in a refinery is examinated. This section is included here because
the energy efficiency of a refinery can be increased not only by improving the energy efficiency
of the individual processes (which is addressed in each section) or energy efficiency of the
energy production system but also by improving energy management, energy conservation and
heat integration/recovery within the refinery as a whole.
Energy management has long been an important issue for refineries. For example, management
techniques such as the ISO 14000 system series or EMAS can provide an appropiate framework
to develop suitable energy management system and can increase the energy efficiency of the
refinery as a whole. Energy conservation techniques such as reporting and incentivating the
energy savings, carrying out combustion improvements or reviewing the energy integration of
the refinery are some of the techniques that may have a big impact in reducing energy
consumption and consequently in increasing the energy efficiency of a refinery. Other technical
tools to increase that efficiency are the heat integration/recovery techniques, of which examples
are: the installation of waste heat boilers, the installation of expanders to recuperate power and
increasing the insulation of buildings and process units to reduce heat losses. Steam
management is another good tool for increasing energy efficiency.
2.11 Etherification
PURPOSE AND PRINCIPLE
A number of chemicals (mostly alcohols and ethers) are added to motor fuels either to improve
performance or to meet environmental requirements. Since the 1970s, alcohols (methanol and
ethanol) and ethers have been added to gasoline to increase octane levels, reduce carbon
monoxide generation and reduce atmospheric ozone due to the lower reactivity of resulting
VOC emissions. These additives replaced the lead additives which were being phased out as
required by the Auto-Oil I. As a result, a number of different ethers are currently added to the
gasoline and are better able to meet both the new oxygen requirements and the vapour pressure
limits. The most common ethers being used as additives are methyl tertiary butyl ether (MTBE),
ethyl tertiary butyl ether (ETBE), and tertiary amyl methyl ether (TAME). Some refineries
(~30% of EU+ refineries) manufacture their own supplies of those ethers.
PROCESS DESCRIPTION
Multiple variations of commercial processes are available. Most processes can be modified to
react isobutylene or isoamylene with methanol or ethanol to produce the corresponding ether.
All use an acidic ion exchange resin catalyst under controlled temperature and pressure
conditions. Temperature control of the exothermic reaction is important to maximise conversion
and minimise undesirable side reactions and catalyst deactivation. The reaction is usually
carried out in two stages with a small excess of alcohol to achieve isoolefin conversions of over
99 % and the methanol consumption is essentially stoichiometric. The basic difference between
the various processes is in reactor design and the method of temperature control.
RAFFINATE
(UNREACTED HC FEED)
WATER
ISO BUTYLENE METHANOL BLEED
REACTOR
FEED WATER STILL
WASH
COLUMN
STEAM STEAM
60-90°C
14 barg
METHANOL
RECYCLE METHANOL
FEED
MTBE PRODUCT
Figure 2.15: Simplified process flow scheme for a MTBE production process
to RFG
Reflux
H2 drum
Condenser
Fractionator
Hydrogen
reflux
scavenger
Depentaniser
The C5 stream is then scavenged to remove catalyst poisons by passing it through an ion
exchange resin to remove basic nitrogen compounds, eg. ammonia, and any metallic
contamination. A hydrogen feeds is also scavenged to remove any acidic components. The
feedstock, containing injected methanol and hydrogen, is fed to the reactor section. Hydrogen is
used to convert di-enes into monoolefins and prevent gum formation during the reaction. This
takes place over a palladium-impregnated ion-exchange resin and the isoamylenes are converted
to TAME.
The TAME product stream is purified by fractional distillation, washing and phase separation.
Fractionator overheads pass to a reflux drum with the gas phase of low boiling hydrocarbons
(C1, C2, C4, etc.) plus unreacted hydrogen being vented to refinery fuel gas or flare. The bottom
product of TAME gasoline with some methanol is cooled and mixed with recycled water from
the methanol recovery plant, then routed to a settler for phase separation. The TAME gasoline
fraction from this is blended with the depentaniser bottoms C6+ stream and passed to storage.
The methanol/water fraction is recycled to the methanol recovery plant feed drum.
Methanol is recovered by distillation in a common still with the overheads methanol being
condensed and passed to buffer storage for recycling to the TAME plant or other uses. The
bottoms are essentially water with some contaminants and are mainly recycled but with a purge
to effluent treatment to avoid formic acid build up.
PROCESS DESCRIPTION
As a minimum, the gas plant consists of two columns, an absorber/stripper column (de-
ethaniser) to strip all light C2-minus components and to maximise recovery of C3-plus
components from the feed streams. Figure 2.17 shows a simplified process flow diagram of a
gas plant.
(OPTIONAL)
Refinery
Fuel Gas
GAS VIA AMINE
TREATER FROM
VARIOUS UNITS
RECONTACTING SECTION 57°C
55°C 25°C 55°C
29barg 19barg
PROPANE TO
STORAGE
TREATED
DEBUTANISER LPG DEETHANISER DEPROPANISER
FROM VARIOUS UNITS
112°C 110°C
STEAM STEAM
BUTANE TO
STORAGE
Figure 2.17: Simplified process flow scheme for a part of a gas plant
The feed stream to the gas plant is cooled and chilled, the residual vapour phase from this being
routed to a refrigerated absorber, meeting chilled re-cycled light gasoline absorbent. Absorber
overhead is the C1, C2 fraction. The bottoms are combined with the liquid stream from chilling
and pass to the first fractionator or de-ethaniser. The first fractionator overheads (essentially C2-
ethane) are combined with the absorber overheads while the bottoms pass to the second
fractionator to produce a C3/C4 overheads stream and debutanised gasoline bottoms. A
recontacting section, where vapours from various units are compressed and recontacted with the
de-ethaniser overhead can be installed to maximise the C3/C4 recovery. The bottoms of the de-
ethanizer column, mainly C3/C4 compounds, are fed to the depropaniser column. This overheads
stream is routed to a third fractionator to split the C3 and C4 streams as overheads (propane) and
bottoms (butane) respectively. Products pass to pressurised storage via any final sweetening by
Mineral Oil and Gas Refineries 51
Chapter 2 Applied processes and techniques
selective molecular sieve adsorption. Subsequent loading of transport containers is carried out
by means such as closed loop systems or by venting and release to the refinery fuel gas system.
The gasoline bottoms stream passes to a fourth fractionator to produce a depentanised fraction
for use as chilled, recycle gasoline on the absorber. The net output is blended to form the
gasoline product. A drying step before sending to storage can be necessary (not shown). Also, a
vessel containing a bed of caustic pellets (not shown) can be installed in the butane system as an
additional guard, although water and H2S should have been removed in the top of the de-
ethaniser and depropaniser column. If no (or insufficient) upstream treatment has taken place,
this can also be done in the unit itself, e.g. with an amine H2S absorber followed by a mercaptan
oxidation/extraction of the de-ethaniser bottoms with amine H2S absorption of the de-ethaniser
net overhead gas.
If thermal and/or catalytic cracking units are present, recovery of olefinic components may also
be worthwhile. It is also possible to separate isobutane from n-butane. The iso-butane can be
used as feed to an alkylation unit, while the n-butane (or part of it) can be used as a blending
component in the gasoline pool or isomerisation.
Hydroprocessing may also be designed to remove low levels of metals from the feed. The
metals to be removed include nickel and vanadium, which are native to the crude oil, as well as
silicon and lead-containing metals that may be added elsewhere in the refinery.
Hydrocracking
It is one of the most versatile of all refining processes, capable of converting any fraction from
atmospheric gasoils to residual (deasphalted) oil into products with a lower molecular weight
than the feed. The hydrocracking reactions occur under high hydrogen partial pressure in the
presence of a catalyst with a two-fold function, hydrogenation and cracking. Hydrocracking
may also be used for the cracking of superior fuels and the production of lubricants (Dewaxing
in Section 2.3). The type of catalyst maximises the production of naphtha, mid-distillates or lube
production. The presence of hydrogen suppresses the formation of heavy residual material and
increases the yield of gasoline by reacting with the cracked products, giving net products, which
are a mixture of pure paraffins, naphthenes and aromatics. Hydrocracking produce mid-
distillates with outstanding burning and cold flow properties as follow:
Hydrocracking
Hydrogen in substantial quantities is consumed in those processes, making the hydrogen
manufacturing unit mandatory (Section 2.14) in those refineries that contain hydrocracking. In
addition to the treated products, those processes produce a stream of light fuel gases containing
hydrogen sulphide, ammonia and water.
As is reflected in Table 2.4, the main feed stream to a hydrocracker is the heavy vacuum
distillate stream from the High Vacuum unit. Those feedstocks are fractions very difficult to
crack and cannot be cracked effectively in catalytic cracking units. Other process streams such
as heavy cycle oil from the catcracker unit, heavy gasoils from the coker or visbreaker unit,
extracts from lube oil units, mid-distillates, residual fuel oils and reduced crudes may be
blended to the main heavy vacuum distillate stream. The main products are LPG, gasoline, jet
fuel and diesel fuel, all practically sulphur-free. The production of methane and ethane is very
low, normally less than 1 %.
PROCESS DESCRIPTION
are typically 30 to 40 bar and 320 to 380 °C. The reactor effluent is cooled in the feed/effluent
exchanger and reactor cooler and flashed into the high-pressure separator. The flashed vapours,
consisting mainly of unreacted hydrogen, are compressed and recycled to the reactor. The
fractionation part is very similar to the one explained in hydroconversion process.
SOUR GAS
to amine
treatment
FLUE GAS STEAM
Ejector Set
SLOPS OIL
STRIPPER
SOUR WATER
TO SWS
OFF GAS
REACTOR HP LP
SEPARATORS
SLOPS OIL
HYDROGEN SOUR WATER DRIER
to SW Stripper
SOUR WATER
FUEL WASH TO SWS
DISTILLATE
WATER
DISTILLATE TO
STORAGE
Diesel Oil deep desulphurisation (Hydrofining). Because this technique operates at low
pressures, efficient hydrogen utilisation can be achieved within an overall refinery context. Very
low sulphur contents can be achieved (8 ppm) in the raffinate. The unit usually operate at 45
bars and consumes very low amounts of hydrogen. Gasoline deep desulphurisation techniques
with a comparatively low hydrogen consumption are currently under development.
The principle process scheme for residue hydrotreating is the same as normal distillate. The
main difference is the reactor systems which normally consists of two or three reactors in series.
Removal of the metals from the residue feed normally occurs in the first reactor(s) and uses a
low-activity coarser Co/Mo catalyst. Further hydrotreating and hydrogenation occur in the tail
reactor(s), resulting in a higher hydrogen-to-carbon ratio and a lower Concarbon number of the
residue. Since the catalysts operate in a H2S and NH3 rich environment the Nickel/Molybdenum
or Nickel/Tungsten catalysts are usually applied in the tail reactors.
Next table shows the typical hydrotreatment operating conditions for the different feedstocks:
Hydrogenation of light dienes. This process is a highly selective catalytic process that can
hydrogenate acetylenes and dienes to the corresponding monoolefin without affecting the
valuable olefin content of the feedstock. In addition, those processes can be designed to provide
hydroisomerization of some of the olefins. (e.g. conversion of 1-butene to 2-butene).
Hydrogenation takes place in a liquid-phase fixed-bed reactor. Unless the hydrogen purity is
low, no separation step is required for the removal of light ends from the product. Thus, reactor
effluent can be charged directly to downstream units.
Saturation of aromatics. The use of highly active noble-metal catalysts permits the reactions to
take place under mild conditions. Because of the mild conditions and the very selective catalyst,
the yields are high, and hydrogen consumption is largely limited to just the desired reactions.
The process is carried out at moderate temperatures (205 - 370 ºC) and pressures (3500 to 8275
kPa) over a fixed catalyst bed in which aromatics are saturated and in which hydrogenation of
olefins, naphthenic ring opening and removal of sulphur and nitrogen occur.
Hydrocracking
Hydrocracking normally use a fixed-bed catalytic reactor with cracking occurring under
substantial pressure (35 to 200 kg/cm2) in the presence of hydrogen at temperatures between
280 and 475 ºC. This process also breaks the heavy, sulphur-, nitrogen- and oxygen bearing
hydrocarbons and releases these impurities to where they could potentially foul the catalyst. For
this reason, the feedstock is often first hydrotreated and dewatered to remove impurities (H2S,
NH3, H2O) before being sent to the hydrocracker. If the hydrocracking feedstocks are first
hydrotreated to remove impurities, sour water and sour gas streams will contain relatively low
levels of hydrogen sulphide and ammonia in the fractionator.
Depending on the products desired and the size of the unit, hydrocracking is conducted in either
single-stage or multi-stage reactor processes. Hydrocrackers can be classified in three
categories, single-stage once-through, single-stage recycle and two-stage recycle:
light naphtha stream is taken as overhead product, heavy naphtha and kerosene are taken as
side-streams and the bottom stream is fed to a second fractionator. In the second
fractionator, operating at mild vacuum, the diesel product is taken as the overhead product
and the bottom stream, the unconverted oil, is recycled to the reactor section.
LIGHT
NAPHTHA
SW
SWS
H2 Amine
LPG DIESEL
FUEL treating
& C3/C4 HEAVY
DEBUTANISER NAPHTHA
VACUUM splitter 2nd
DISTILLATE FRACTIONATOR
1st KERO
FRACTIONATOR
REACTOR
HP LP
SEPARATORS
Figure 2.19: Simplified process flow scheme for hydrocracker (single stage with recycle)
• In the two-stage recycle configuration, the first hydrocracker reactor operates in a once-
through mode with a typical conversion of around 50 %. The unconverted oil from the first
hydrocracker reactor is fed to a second hydrocracker reactor for further conversion. The
unconverted oil from the second stage hydrocracker is recycled to achieve an overall
conversion of around 97 - 98 %. A small bleed stream of about 2 - 3 % on fresh feed is also
required here. This concept is normally applied when a very heavy high refractory feed
stock such as deasphalted oil is processed.
High-pressure Moderate-pressure
hydrocracking hydrocracking
Conversion, % w/w 70 - 100 20 - 70
Pressure, barg 100 - 200 70 - 100
Liquid hourly space velocity 0.5 - 2.0 0.5 - 2.0
Average reactor temperature ºC 340 - 425 340 - 425
Hydrogen circulation, Nm3/m3 650 - 1700 350 - 1200
Hydroconversion
In principle, three reactions are taking place: hydro-demetallisation, hydrotreating/
hydrogenation and hydrocracking. Removal of the metals from the residue feed predominantly
occurs in the first reactor(s) and uses a low activity Co/Mo catalyst. Hydrotreating,
hydrogenation and hydrocracking occur in the following reactor(s) where the quality is mainly
improved by increasing the hydrogen-to-carbon ratio. Conversion levels to products in a
hydroconversion process are typically 50 - 70 % but depend significantly on the type of
hydroconversion process and the quality of the feedstock.
Figure 2.20 shows a simplified process flow diagram of a moving bed process. The process
comprises 5 reactors in series, catalyst handling facilities and a work-up section. The reactors
operate at high pressure and relatively high temperatures. The first three reactors are bunker
hydrodemetallisation (HDM) reactors. The Ni+V conversion typically exceeds 60 % for the first
60 days of operation and then gradually trends towards the expected equilibrium conversion
level, 50 to 70 percent. This way, high-metal feeds can be treated. The last two reactors are
fixed bed desulphurisation and conversion reactors. The moving bed technology employs a
bunker flow/moving bed technology to replenish the HDM catalyst continuously. Catalysts are
transported through a slurry transport system in which the rate of catalyst replenishment is
controlled in accordance with the rate of metal deposition. The catalyst in the HDM reactors
flows concurrently downward with the process fluids. Screens separate the catalysts from the
process fluids before leaving the reactor. Sluice systems are present at top and bottom of the
reactors to enable catalyst addition and withdrawal. The following conversion section consists
of two fixed-bed reactors in series, containing catalysts, that are highly active for sulphur
removal and conversion. The ebullated bed reactor operates as a fluidised-bed three-phase
system with back mixing of the unconverted liquid and the catalyst.
REJUVENATED CATALYST
Vacuum
residue
feed HYDROCONVERSION REACTORS
Recycle gas
Off-gases
SEPARATOR
SECTION Gases
Naphtha
Kerosene
Gasoil
FRESH
CATALYST
Figure 2.20: Sample of simplified process flow scheme for hydroconversion process (moving bed)
The concept of the fractionation section is also very licensor-dependent as it depends on the
configuration of the separator system and the resulting temperatures. Typically it consists of a
main fractionator, a vacuum distillation column and some form of a gas plant to fractionate and
stabilise the lighter fractions. Products from the fractionation section are normally refinery fuel
gas, LPG, naphtha, kerosene, light gasoil, vacuum distillate and a low-sulphur/ metal vacuum
residue (bottoms) stream. The refinery fuel gas and LPG streams are amine washed to remove
H2S. The naphtha product is normally fed to a naphtha hydrotreater for further processing
identical to the straight run naphtha product. The kerosene and light gasoil products are
normally fed to a hydrotreater for further purification. The vacuum distillate is converted further
in a hydrocracker or FCC unit. The bottoms stream is normally blended into the heavy fuel oil
pool or used as delayed coker feedstock.
In steam reforming, light hydrocarbons are reacted with steam to form hydrogen. In principle,
all products of a refinery could be used for hydrogen production by partial oxidation. However
the most interesting option from the economic point of view is to use products with a low
market value. In some refineries, heavy oil residues are transformed to petroleum coke and
subsequently gasified (Section 2.7) to produce syngas.
PROCESS DESCRIPTION
Steam reforming
This is the most commonly used method for hydrogen production. The best feedstocks for steam
reforming are light, saturated, and low in sulphur; this includes natural gas (the most common),
refinery gas, LPG, and light naphtha. This is usually combined with a hydrogen purification
process to produce very pure hydrogen (>99 % v/v). A simplified scheme of a steam reforming
process is shown in Figure 2.21. The process is carried out in the presence of a catalyst which is
very sensitive to poisoning. Desulphurisation of the feedstock is required in order to protect the
catalyst in the reformer furnace against poisoning and deactivation. The reaction is typically
carried out at 760 - 840°C and a pressure of 20 - 30 barg over a fixed catalyst bed. It is common
practice to operate at excess steam-hydrocarbon ratios to prevent carbon formation. Heat for the
endothermic reforming reaction is provided by the furnace burners. The reformed gas, a mixture
of hydrogen, carbon dioxide, carbon monoxide, methane and steam, is cooled down to about
350°C by raising steam. After reforming, the CO in the gas is reacted with steam to form
additional hydrogen (shift reaction). The oxidation of the CO to CO2 is done in a two-step shift
converter, reducing the CO content to less than 0.4 %. The product gas passes to a CO2 absorber
after being cooled, where the CO2 concentration is reduce to 0.1 % v/v by a suitable regenerable
liquid absorbent (e.g. MEA, hot potassium carbonate or sulfinol). The solvent, enriched with
CO2, is stripped in a solvent regenerator. Residual CO and CO2 in the absorber overhead gas is
methanated, reducing the CO and CO2 content to about 5 - 10 ppm. Unlike CO, a small amount
of CH4 is usually not objectionable in hydrocracking units and other hydrotreating.
REFORMER FLUE
FURNACE GAS
METHA-
H2 NATOR
HT LT
SHIFT CONVERTORS HYDROGEN
DESULFURISATION
REACTORS
SOLVENT
ABSORBER
840 °C ATM.
22 barg
NAPHTHA STEAM
Heat input
PROCESS
from furnace CONDENSATE
BUTANE
PROCESS
GAS STEAM BFW CONDENSATE
FUEL SOLVENT
PRETREATMENT REGENERATOR
REFINERY GAS
BFW
Gasification of coke
The processes used for the gasification of petroleum coke are the same as those used in the
gasification of coal and they are integrated in the Flexicoker (Section 2.7). In an oxygen-blown
operating mode of a gasifier, the gas produced can be processed to recover hydrogen or
synthesis gas, or can be used as a medium-calorific value fuel. The gasifier product gas (syngas,
CO, H2, CO2, CH4 and H2O) after it has passed the cyclones, contains hydrogen sulphide (H2S)
and carbonyl sulphide (COS). With a sulphur adsorbent such as limestone (CaCO3) or dolomite
(Mg, CaCO3) in the gasifier, the sulphur content of the gas can be drastically reduced. If no
sorbent is used, the sulphur content of the gas will be in proportion to the sulphur in the feed.
The particulates in product gas are removed in the barrier filter. Volatile metals and alkali tend
to accumulate on the particulate as the gas is cooled. The particulates contain a high percentage
of carbon and are usually sent with the ash to a combustor, where the remaining carbon is
burned and the calcium sulphide is oxidised to sulphate. In this hot-gas clean-up system, there is
no aqueous condensate produced, although some may be produced in subsequent processing of
the gas.
Table 2.7: Example of composition of petroleum coke used as well as the composition of the
syngas produced in a oxygen-blown fluidised bed gasification process
Source: [166, Meyers, 1997]
Hydrogen processing in this system depends on how much of the gas is to be recovered as
hydrogen, and how much is to be used as fuel. Where hydrogen production is a relatively small
part of the total gas stream, a membrane is normally used to withdraw a hydrogen-rich stream.
That stream is then refined in a purification unit.
Purification of hydrogen
A wide variety of processes are used to purify hydrogen streams. Since the streams are available
at a wide variety of compositions, flows, and pressures, the method of purification will vary.
They are wet scrubbing, membrane systems, cryogenic separation and pressure-swing
adsorption (PSA). This last technique is the most commonly used. In the PSA plant, most
impurities can be removed to any desired level. An adsorbent (molecular sieves) removes
methane and nitrogen from the out-stream. Nitrogen is the most difficult to remove of the
common impurities, and removing it completely requires additional adsorbent. Since nitrogen
acts mainly as a diluent, it is usually left in the product if the hydrogen is not going to be used in
a very high-pressure system such as a hydrocracker. Hydrogen purity is 99.9 - 99.999 % v/v
after the PSA unit. The residual constituents of the product gas are mainly methane and less
than 10 ppm CO. Several adsorber beds are used, and the gas flow is periodically switched from
one vessel to another to allow regeneration of the adsorbent by pressure reduction and purging,
thus releasing the adsorbed components. The desorbed gas is normally accumulated in a vessel
and used as fuel at a convenient location.
Techniques to prevent air emissions, such as fugitive emissions, odour and noise that are
relevant to the whole refinery are included in Sections 23. Soil contamination prevention
techniques are also included in Sections 25.
The development of Environmental Management Systems (EMS) started by building upon the
experience gained with other business parameters. Quite often, management responsibility for
Safety is combined in one person with responsibility for the Environment (and Health, and
sometimes Quality). Environmental Management is also named Environmental Care. In this
section the importance of Good housekeeping and Management is highlighted. It is remarked
that systems exist for improving performance in many fields such as Safety, Maintenance,
Product Quality. Environmental Management Systems have also been developed for the
improvement of refinery performance in environmental matters.
UTILITIES MANAGEMENT
Energy management including steam management and cooling are included in other Sections (8
and 10).
Ballast water is relevant to those refineries that have crude receipt facilities from ships or handle
big product tankers or inland barges. This ballast water can be high in volume and salt content
(seawater), and heavily polluted with oil. However, volumes of ballast water to be treated are
declining with the gradual introduction of segregated ballast tankers.
Water and drainage applies to any industrial site. It entails the complete system of fresh water
supply, rainwater, ballast water, process water, cooling water and ground water as well as
effluent collection, storage and the various (primary, secondary and tertiary) waste water
treatment systems. The design is based on local factors (rainfall, receiving water bodies, etc.),
effluent segregation, source reduction, first flush philosophies, flexible routing and re-use
options.
Recirculated process water streams and cooling water streams are often manually purged to
prevent the continued build up of contaminants in the stream (blowdown system).
2.16 Isomerisation
PURPOSE AND PRINCIPLE
Isomerisation is used to alter the arrangement of a molecule without adding or removing
anything from the original molecule. Typically, low molecular weight paraffins (C4-C6) are
converted to isoparaffins having a much higher octane index. Isomerisation of olefins is also
included in this section.
PROCESS DESCRIPTION
Figure 2.22 shows a simplified process flow diagram of a low temperature isomerisation unit.
The isomerisation reactions occur in the presence of hydrogen and catalyst. The atmosphere of
hydrogen is used to minimise carbon deposits on the catalyst, but hydrogen consumption low.
LPG
STABILISER
REACTORS
C5/C6 FEED
DRIER
AB-/DESORPTION
HIGH OCTANE
AND/OR
ISOMERATE
DEISOHEXANISER
nC5/nC6
The reactions normally take place in two reactors in series. This allows replacement of the
catalyst during operation. A further advantage of the two-reactor scheme is that the first reactor
may operate at higher temperatures to kinetically drive the reactions, while the tail reactor may
be operated at lower temperatures to push the desired products closer to equilibrium conversion.
There are several isomerisation processes in use today. In general they can be classified in two
categories, 'once-through' or recycled.
• In 'once-through' isomerisation, only fresh feed is processed in the isomerisation unit. The
octane number which can be achieved is only around 77 - 80 RON when using a zeolite
based catalyst and 82 - 85 RON using a chloride promoted catalyst. Conversions of 80 %
can be expected.
• In recycle isomerisation, the unconverted, lower octane paraffins are recycled for further
conversion. Depending on the recycle option, the recycle may be normal paraffins or
methylhexanes and n-hexane. The octane number obtained can be up to 92 RON, dependent
on the feedstock composition, configuration and catalyst used. The yield in isomerate is
around 95 to 98 % depending on the targeted octane number of the final stream.
There are three distinctly different types of catalysts currently in use: chloride promoted,
zeolitic, and sulphated zirconia catalysts. The zeolite catalyst operates at significant higher
temperatures (250 – 275 °C and 28 barg) and is more tolerant to contaminants though resulting
octane improvements is lower. The zeolite catalyst is mainly used when higher octane
isomerate product does not justify the additional capital required to reduce feed contaminants
for the chlorided alumina catalyst or is a better match for a retrofit unit. The highly active
chloride promoted catalyst, operates at a relatively low temperature (190 – 210 °C and 20 barg)
and gives the highest octane improvement.This type of catalyst requires the addition of small
amounts of organic chlorides converted to hydrogen chloride in the reactor to maintain the high
activity. In such a reactor, the feed must be free of oxygen sources including water to avoid
deactivation and corrosion problems. Furthermore, this catalyst is very sensitive to sulphur, so
deep desulphurisation of the feed to 0.5 ppm is required. Lower reaction temperatures are
preferred to higher temperatures because the equilibrium conversion to the desired isomers is
enhanced at lower temperatures.
After isomerisation, the light ends are stripped from the product stream leaving the reactor and
are then sent to the sour gas treatment unit. In a once-through isomerisation unit the bottom
stream from the stabiliser is, after product cooling, sent to the gasoline pool. In a recycle stream,
the bottom stream from the stabiliser is fed to a separation unit, which is either a de-
isohexaniser column or an adsorption system.
In the de-isohexaniser column, a split can be made between the higher octane dimethylbutanes
and the lower octane methylpentanes. The dimethylbutanes and lower boiling C5 components
(isomerate product) are taken overhead from the column and are sent to the gasoline pool. The
methylpentanes and normal hexane are taken as a side-stream close to the bottom, and are
recycled to the reactor. The bottom stream from the de-isohexaniser is a small quantity of heavy
by-product which is sent with the isomerate product to the gasoline pool or to a catalytic
reformer if the refinery recovers benzene as chemical feed. The principle of adsorption is that
the unconverted normal paraffins are adsorbed on the molecular sieve whereas the iso-paraffins
pass the adsorbent. Desorption takes place with heated hydrogen-rich gas from the separator or
a butane mixture. The desorbent is separated from the net hydrogen-rich recycle stream in a
separator vessel.
PROCESS DESCRIPTION
As it is shown in Figure 2.23, the purification plant consists on a gas sweetening plant where
acid gases as CO2, H2S, SO2 are separated. Natural gas is considered "sour" when contains
significantly greater amounts of hydrogen sulphide than those quoted for pipeline quality or
when contains such amounts of SO2 or CO2 to make it impractical to use without purification.
The H2S must be removed (called "sweetening" the gas) before the gas can be utilized. If H2S is
present, the gas is usually sweetened by absorption of the H2S in an amine solution. Amine
processes are the most common process used in the United States and Europe. Other methods,
such as carbonate processes, solid bed absorbents, and physical absorption, are employed in the
other sweetening plants.
Natural gasoline, butane and propane are usually present in the gas, and gas processing plants
are required for the recovery of these liquefiable constituents (see Figure 2.23). The type of
processes found are very similar to those described in Section 2.12. As summary, they are
physical separations at very low temperatures (typically distillation).
Reinjection
Flare or
Emergency incineration
flare
Sour
Elemental
Flare (only during gas Gas sweetening Acid gas SRU
CO2+H2S sulphur
well testing and plant
completion) Natural gas
Sweet (C1/C2)
gas Gas processing
LPG (C3/C4)
Emergency plant
flare or vent
Heavier
Emergency (C5+)
flare or vent
Separators
and
dehydrators Pipeline
Reinjection
Gas, Oil
and water if sweet
Hydrocarbon Water
condensates
2.18 Polymerisation
This section covers up the polymerisation, dimerisation and condensation of olefins.
PROCESS DESCRIPTION
The reactions typically take place under high pressure in the presence of a phosphoric acid
catalyst adsorbed onto natural silica and extruded in pellets or small cylinder form. All reactions
are exothermic, and therefore the process requires temperature control. The feed must be free of:
sulphur, which poisons the catalyst; basic materials, which neutralize the catalyst and oxygen,
which affects the reactions. The propene and butene feed is washed first with caustic to remove
mercaptans, then with an amine solution to remove hydrogen sulphide, then with water to
remove caustics and amines, and finally dried by passing through a silica gel or molecular sieve
dryer. A simplified scheme of a polymerisation unit is shown in Figure 2.24.
DEPROPANIZER DEBUTANIZER
CATALYST CHAMBER
BUTANE to
STORAGE
C3 Quench
C3 recycle
FEED
POLYMER
PROPANE
WATER GASOLINE
TO STORAGE
TO STORAGE
When the polymerisation yield drops, the catalysts need to be replaced. After nitrogen purging,
the polymerisation unit is opened and the catalyst removed by means of a high-pressure water
jet. It can also be removed using steam (compression dumping). The phosphoric acid goes in the
water medium, while the natural silica pellets break down to form a slurry, which is usually
pumpable.
Increased demand for light products and a reduced demand for heavy fuel oil have resulted in
refiners upgrading the atmospheric residue in more valuable lower boiling fractions such as
naphtha, kerosene and middle distillates. Vacuum distillation is simply the distillation of
petroleum fractions at a very low pressure to increase volatilization and separation whilst
avoiding thermal cracking. The high vacuum unit (HVU) is normally the first processing step in
upgrading atmospheric residue followed by downstream refining units. HVU produces
feedstocks for cracking units, coking, bitumen and base oil units. The contaminants from the
crude oil stay predominantly in the vacuum residue.
The main feed stream to the HVU is the bottom stream of the crude oil distillation unit referred
to as atmospheric or long residue. In addition the bleed stream from the hydrocracker unit (if
applicable) is normally sent to the HVU for further processing. The products from the HVU are
light vacuum gasoil, heavy vacuum gasoil and vacuum residue. Light vacuum gasoil is normally
routed to the gasoil hydrotreater(s), heavy gasoil is normally routed to a fluid catcracker and/or
hydrocracker unit. The vacuum residue can have many destinations such as visbreaking,
flexicoking or delayed coking, residue hydroprocessing, residue gasification, bitumen blowing
or to the heavy fuel oil pool.
PROCESS DESCRIPTION
Atmospheric distillation
Distillation involves the heating, vaporisation, fractionation, condensation, and cooling of
feedstocks. The desalted crude oil is heated to about 300 - 400 ºC and fed to a vertical
distillation column at atmospheric pressure where most of the feed is vaporised and separated
into its various fractions by condensing on 30 to 50 fractionation trays, each corresponding to a
different condensation temperature. The lighter fractions condense and are collected towards the
top of the column. The overhead hydrocarbon vapours are condensed and accumulated in the
overhead reflux drum of the main fractionator. In this drum sour water, light fractions (about
0.5 % on crude charge) and stripping steam (1.5 % on crude), are separated from the
hydrocarbon liquid. The overhead hydrocarbon liquid, the so-called naphtha minus stream, is
commonly fed directly to the downstream naphtha treater. Within each atmospheric distillation
tower, a number of side-streams of low-boiling point components are removed from different
trays in the tower. These low-boiling point mixtures are in equilibrium with heavier components
which must be removed. The side-streams are each sent to a different small stripping tower
containing four to ten trays with steam injected under the bottom tray. The steam strips the
light-end components from the heavier components and both the steam and light-ends are fed
back to the atmospheric distillation tower above the corresponding side-stream draw tray. Most
of these fractions generated in the atmospheric distillation column can be sold as finished
products after a hydrotreatment, or blended with products from downstream processes. In
Figure 2.25 a simplified process flow diagram of a crude distillation unit is shown. Many
refineries have more than one atmospheric distillation unit.
NON
CONDENSABLES
CRUDE
TPA COLUMN UNSTABILISED
NAPHTHA
PREFLASH
CRUDE
Figure 2.25: Simplified process flow diagram for a crude distillation unit
The operating conditions of the tower are based on the properties of the crude oil and the
desired products yield and quality. To maximise the yield of distillates, the pressure is
minimised, but the temperature is increased to the maximum. Every refinery has a crude
distillation unit designed for a selected crude (mix). There are therefore many different crude
distillation configurations with varying product cuts and heat integration.
Vacuum distillation
Figure 2.26 shows a simplified process flow diagram of the high vacuum unit. Atmospheric
residue is heated up to 400 °C, partially vaporised (30 - 70 % by weight) and flashed into the
base of the vacuum column at a pressure between 40 and 100 mbar (0.04 to 0.1 kg/cm2). The
vacuum inside the fractionator is maintained with steam ejectors, vacuum pumps, barometric
condensers or surface condensers. The injection of superheated steam at the base of the vacuum
fractionator column further reduces the partial pressure of the hydrocarbons in the tower,
facilitating vaporisation and separation. The unvaporised part of the feed forms the bottom
product and its temperature is controlled at about 355°C to minimise coking. The flashed vapour
rising through the column is contacted with wash oil (vacuum distillate) to wash out entrained
liquid, coke and metals. The washed vapour is condensed in two or three main spray sections. In
the lower sections of the column the heavy vacuum distillate and optional medium vacuum
gasoil are condensed. In the upper section of the vacuum column the light vacuum distillate is
condensed. Light (non-condensable) components and steam from the top of the column are
condensed and accumulated in an overhead drum for separating the light non-condensables, the
heavier condensed gasoil and the water phase. Figure 2.26 shows a simplified process flow
diagram of a crude distillation unit. The most important operational aspect of a vacuum unit is
the quality of the heavy vacuum gasoil, especially when this is fed to a hydrocracker unit. The
Concarbon level and/or metal content is very critical for a hydrocracker unit and depends very
much on the operation and performance of, especially, the wash oil section in the vacuum
distillation unit as well as the desalter in the crude distillation unit.
EJECTOR
SET
Steam Waste
gas
Slops
VACUUM
COLUMN Sour water
LIGHT
VACUUM
GASOIL
VACUUM
FLUE GAS ATMOSPHERIC GASOIL
RESIDUE
ATMOSPHERIC
RESIDUE FURNACE WASH OIL
VESSEL
VR VACUUM VR FUEL
DISTILLATE
VACUUM
RESIDUE
(VR)
ATMOSPHERIC ATMOSPHERIC
RESIDUE RESIDUE
Figure 2.26: Simplified process flow scheme for high-vacuum distillation unit
As a result no reduction in the total sulphur content takes place during sweetening and
consequently it is only applied to those streams where sulphur content is not a problem.
PROCESS DESCRIPTION
Mercaptans are removed from light hydrocarbon streams by a concentrated alkaline wash in an
extraction column under elevated pressure (5 barg). If H2S or acids are present a caustic
prewash is required. The treated and odour-free hydrocarbon stream leaves the reactor as an
overhead stream. The aqueous bottom phase is heated to 50 °C, mixed with air and fed to the
oxidation reactor. The dissolved NaSR is now converted into disulphides (which are insoluble
in caustic soda water solution) at 4.5 barg. By using an excess of air and adding catalyst, a high
rate of reaction is maintained. In this way the alkaline solution is regenerated. The liquid from
the reactor is routed to a vessel where spent air, the disulphide fraction insoluble in the caustic
solution and the caustic solution are separated. The spent air is routed to an incinerator or
process furnace, the disulphides are usually recycled to the crude and the regenerated caustic is
circulated to the extraction column. Figure 2.27 shows a simplified process flow diagram of an
extraction process.
Another oxidation process is also a mercaptan oxidation process that uses a solid catalyst bed.
Air and a minimum amount of alkaline caustic ("mini-alky" operation) are injected into the
hydrocarbon stream and the caustic cannot regenerated. As the hydrocarbon passes through the
mercaptan oxidation catalyst bed, sulphur mercaptans are oxidized to disulphide.
PURIFIED LIGHT
HC PRODUCT
EXCESS AIR TO
FURNACE,
INCINERATOR OR FCC
MERCAPTAN CO-BOILER
EXTRACTION MERCAPTANS
COLUMN OXIDATION
AIR INJECTION COLUMN
DISULPHIDE
RECYCLED TO
CRUDE
DISULPHIDE
SEPARATION
VESSEL
CAUSTIC
PRE-WASH
VESSEL
STEAM
Figure 2.27: Simplified process flow diagram of the mercaptan oxidation extraction process
Both processes can be integrated in a so-called caustic cascading system to achieve the desired
product quality improvements at minimum caustic make up and spent caustic disposal costs.
Figure 2.28 shows a simplified process flow diagram of the mercaptan oxidation extraction
process.
NaOH NaOH
DILUTED CONCENTRATED
GAS
RESIDUAL H2S GAS
REMOVAL
C 4 / C 4= MERCAPTAN MAKE-UP
C4 / C 4=
REMOVAL
AMINE TREATED
CAUSTIC
FOR BULK H2S REGENERATION
REMOVAL UNIT
DISULPHIDES
BLEED
SULPHIDIC PHENOLIC
SPENT CAUSTIC SPENT CAUSTIC
Figure 2.28: Simplified process flow diagram of the caustic cascading system (mercaptan oxidation
extraction and mercaptan oxidation sweetening)
Another process that can be found in refineries is the selective hydrocracking (hydrofiner),
where one or two zeolite catalysts are used to selectively crack the wax paraffins (n- and near-
to-n-paraffins). This technique can be used for the dewaxing of middle-distillate components
that may be blended into products for extreme winter conditions. A variation of this is the use of
isodewaxing catalyst that isomerizes the n-paraffin wax to desirable isoparaffin lube molecules,
while co-producing low quality middle-distillate transportation fuels. The design and operation
of this unit is similar to that of a normal hydrotreater unit (Section 2.13).
Feed blending may be applied to prepare the optimum feed stream to a refinery unit, thereby
ensuring the optimum performance of the refinery unit. For example, a refiner processes a
mixture of four different low-sulphur crudes in his atmospheric distillation unit. The feed blend
is selected so that the throughput in the distillation unit and further downstream units are
optimised/ maximised with the objective of maximising overall profit. Alternatively, it also
happens that the different quality crudes are processed separately for a short period (a few days)
in so-called “blocked-out” operations.
Product blending is applied to produce the optimum mix of finished refinery products. The
majority of the product streams as produced in different refinery units, normally identified as an
intermediate product stream, can be blended into more than one finished product stream. For
example (hydrotreated) kerosene product is normally blended into diesel oil, light heating oil
and even heavy fuel oil, with the remainder blended into the jet fuel oil. The amount of kerosene
allocated to the different products is selected so that product demands and specifications are met
with the objective of maximising overall profit. Blending products involves mixing the products
in various proportions to meet specifications such as vapour pressure, specific gravity, sulphur
content, viscosity, octane number, cetane index, initial boiling point and pour point and to add
distinctive smells (LPG).
PROCESS DESCRIPTION
The crude oil storage systems can be located either at a separate oil terminal or within the
refinery complex. More than 50 % of the refinery area is occupied by oil movement facilities.
Storage tanks can be divided into four main types: Pressure vessels, fixed roof tanks, fixed roof
tanks with floating cover and floating roof tanks. Figure 2.29 shows a drawing of the different
types of storage systems found in a refinery.
BALANCE
INERT LINE
INERT PROCESSING
GAS e.g. VRU
GAS VAPOUR
RETURN LINE
EXTERNAL INTERNAL DESTRUCTION
FLOATING FLOATING FIXED ROOF
ROOF ROOF
PUMPS
TO PROCESS
CAVERN
Pressure vessels are normally used to store gases at high pressures (>91 kPa, e.g. LPG). Fixed
roof tanks can be open to atmosphere, or designed as a pressure tank, with several classes of
allowed pressure build-up, from 20 mbarg (low pressure) to 60 mbarg (high pressure). The
pressure tanks are provided with Pressure/Vacuum Relief Valves to prevent explosions and
implosions, the vacuum setting being -6 mbarg. Floating roof tanks are constructed in such way
that the roof floats on the liquid, and moves with the liquid level (>14kPa to < 91 kPa).
Above-ground storage tanks (ASTs) are used at refineries for holding either the raw feedstock
(crude oil) or end-products generated by the refinery processes (gasoline, diesel, fuel oils etc.).
Underground storage tanks are used much less frequently (if at all) at refineries - primarily for
storing fuel for onsite boilers and vehicles, or for capturing liquids at low level drain points.
Storage of crude oil and products in caverns is also applicable in some European countries.
Blending can be carried out in-line or in batch blending tanks. An in-line blending system
consists of a manifold where individual streams are blended on flow control, and the blend ratio
is normally controlled and optimised by a computer. When a certain volume of a given quality
product is specified, the computer uses linear programming models to optimise the blending
operations to select the blending components to produce the required volume of the specified
product at the lowest cost. To ensure that the blended streams meet the desired specifications,
in-line stream analysers for flash point, RVP, boiling point, specific gravity, research and motor
octane number (RON and MON), sulphur, viscosity, cloud point and others are inserted to
provide feedback to the computer which in turn corrects blending ratios where necessary.
Batch blending involves mixing the feed streams in a blend tank, from where the relevant
process units are fed. The same applies to intermediate product streams, which are first sent to
intermediate storage tanks, from where they are batch blended into the final product tanks.
Additives and odorants. The odorant is stored as a liquid, normally in fixed tanks. The odorant
is not added to the gas stream prior to liquefaction but is usually added to the LPG whilst the
LPG is being loaded into the delivery tankers, although in-tank odorising is also carried out.
Allowance may be made for any residual mercaptans already present in the LPG. The pump
addition rate is carefully controlled. In the case of liquid propane, methanol may be added with
the odorising agent in order to prevent hydrate icing in propane evaporators.
Pipes, valves and auxiliary systems, such as vacuum recovery units are found throughout a
refinery. Gases, liquids and even solids are transferred from unit operation to unit operation by
pipes. Process pipes are normally overground but some pipes are underground.
2.22 Visbreaking
PURPOSE AND PRINCIPLE
Visbreaking is a well-established noncatalytic thermal process that converts atmospheric or
vacuum residues to gas, naphtha, distillates and tar. It uses heat and pressure to break large
hydrocarbon molecules into smaller, lighter molecules.
When vacuum residue is directly blended to the heavy fuel oil pool, significant quantities of
cutter stock (normally high-value gasoil) need to be blended to the residue to meet the viscosity
specifications of the heavy fuel oil. By thermally cracking the vacuum residue at relatively mild
conditions, about 10 - 15 % of the feed is cracked to lighter fractions and, more importantly, the
viscosity of the vacuum residue is reduced significantly. For this reason, the thermal cracker
unit is normally called the “visbreaker” gas oil unit.
PROCESS DESCRIPTION
Thermal cracking is one of the oldest conversion processes to upgrade heavy oil fractions. At
present it is mainly used to upgrade vacuum residue. Figure 2.30 shows a simplified process
flow diagram of the Visbreaker unit. The most important factor in controlling the cracking
severity should always be the stability and viscosity of the visbroken residue fed to the fuel oil
pool. In general, an increase in the temperature or residence time results in an increase in
severity. Increased severity produces higher gas-plus-gasoline yield and at the same time a
cracked residue (fuel oil) of lower viscosity. Excessive cracking, however, leads to an unstable
fuel oil, resulting in sludge and sediment formation during storage. Thermal cracking
conversion is maximum 20 % of the feed. Thermal cracking units to upgrade atmospheric
residue have conversion levels significantly higher (35 - 45 %) and the viscosity of the
atmospheric residue is reduced.
RECONTACTING
COMPRESSOR
FRACTIONATOR
SOUR WATER REFINERY GAS
MPA RECONTACTING
VESSEL
STRIPPER WILD NAPHTHA
SOAKING
DRUM
STEAM
FLUE GAS GAS OIL
450°C
9-10 barg
STEAM
STEAM BFW VISBROKEN
RESIDUE
FEED
MPA VISBROKEN
RESIDUE FUEL FEED
The feedstock is heated above 500 ºC and then fed to a reaction chamber which is kept at a
pressure of about 9.65 barg. Following the reactor step, the process stream is mixed with a
cooler recycle stream, which stops the cracking reactions. The product is then fed to a flasher
chamber, where pressure is reduced and lighter products vaporise and are drawn off. The lighter
products are fed to a fractionating tower, where the various fractions are separated. The
"bottoms" consist of heavy residue, part of which is recycled to cool the process stream leaving
the reaction chamber; the remaining bottoms are usually blended into residual fuel.
There are two types of visbreaker operations, “coil or furnace cracking” and “soaker cracking”.
Coil cracking uses higher furnace outlet temperatures (470 – 500 °C) and reaction times from
one to three minutes, while soaker cracking uses lower furnace outlet temperatures
(430 - 440 °C) and longer reaction times. The product yields and properties are similar. Run
times of 3 - 6 months are common for furnace visbreakers and 6 - 18 months for soaker
visbreakers.
Pollutants such as NOx, particulates, H2S, SO2, other sulphur compounds and VOC, among
others, are typically abated by end-of-pipe techniques. One of the largest systems within a
refinery is the abatement of H2S produced across the refinery. These systems typically contain
an amine scrubbing system and a sulphur recovery unit to convert H2S into sulphur, a by-
product produced within refineries. Flares are also another technique used within the refinery
for safety and environmental reasons. Techniques for the abatement of odour and noise are also
relevant for refineries.
Refineries also contain waste water plants with different unit operations. Oil separators,
flotation, floculation and biological treatment are typical within refineries. Biological treatment
is required when biodegradation of some organic chemicals that may be present within the
waste water stream is necessary. Final water cleaning systems can also be present.
Refineries also generate solid wastes. Some of them are recycled within the refinery, others are
recycled by specialist companies (e.g. catalysts) and others are disposed of. Soil contamination
prevention techniques that are also relevant to the whole refinery.
Section 3.1 gives an overview of the main emissions and consumptions of European refineries
as a whole. It is not a simple aggregation of the emissions and consumptions in the other
sections: most of them cannot be aggregated because of the integration of processes in
refineries.
Section 2 to 22 cover the emissions and consumptions from the various processes/activities
covered by this document. Sections 23-25 covers the emissions generated by the techniques
used to abate emissions, including the emissions from sulphur recovery units. The chapter ends
with a section on monitoring, which covers the monitoring systems typically applied in
refineries and provides some discussion on their application.
This chapter addresses those three media in turn, and the pollutants that affect them, and is a
summary of the emissions from all refinery processes. The main environmental problems were
mentioned in Chapter 1 (Section 1.4), but without figures. This section aims to quantify refinery
emissions.
Although the principal raw material input to petroleum refineries is crude oil, they use and
generate an enormous number of chemicals, some of which leave the facilities as discharges of
air emissions, waste water or solid waste. Pollutants generated typically include ammonia
(NH3), carbon dioxide (CO2), carbon monoxide (CO), hydrogen sulphide (H2S), metals,
nitrogen oxides (NOx), particulates, spent acids (e.g. HF, H2SO4), sulphur oxides (SOx), volatile
organic compounds (VOC) and numerous organic compounds (some very toxic). Figure 3.1
shows a simplified example of what it is consumed and emitted by a refinery.
Crude = 1 Mtonne
PRODUCTS
Energy = 470 - 1500 GWh
Water = 0.1 - 4.5 Mtonnes
Chemicals = 300 tonnes
Figure 3.1: Example of specific emissions and consumptions found in European refineries
Note: European refineries range from 20 to less than 0.5 Mt of crude oil per year. Low values in the
ranges normally correspond to refineries with abatement techniques and good environmental
performance, whereas high values typically correspond to refineries without abatement techniques.
Figure 3.1 shows only the main pollutants generated by refineries, but more than 90 specific
compounds have been identified [108, USAEPA, 1995]. The great majority are pollutants to air.
The next table shows the environmental account of refinery processes. Next table shows a
summary of the impact on the different media from the different activities found within
refineries.
The capacity of the combustion plants in a refinery varies widely from less than 10 to up to 200
megawatts thermal input (MWth). The total installed capacity ranges from several hundred to
more than 1500 MWhth in the largest refineries (equals 1.7 to 5.4 GJ per tonne crude processed).
This capacity equals to an overall power installation from 20 to 1000 MW. These ranges depend
strongly on the degree of energy integration of the refinery, its complexity and the integration of
units such as cogeneration and petrochemical and lubricant oil plant.
Water and steam are used in the various refinery processes, e.g. to assist the distillation process
or the cracking of hydrocarbons and in scrubbing, quenching or (steam) stripping. Petroleum
refineries use relatively large volumes of water: in the processes; in steam generation and,
especially, in cooling systems. Wash water is used in a number of processes, of which the
desalter is the biggest user and therefore also the main producer of waste water in oil refineries
(exceptions are base oil refineries). Wash water is also used in the overhead systems of many
processes. Tank and flare seal drains and cleaning water all contribute to the amount and quality
of process water streams. The amount of water used depends on the type of refinery and the
cooling system used. It ranges from 0.01 up to 5 tonnes per tonne of crude oil processed.
The amount of chemicals used in a refinery is around 300 tonnes per each million of tonnes of
crude oil processed.
The main air emissions from a refinery are CO2, SOx, NOx, VOC and particulates (dust, soot
and associated heavy metals (mainly V and Ni)). However, noise, odour, H2S, NH3, CO, CS2,
benzene, toluene, dioxins, HF and HCl also contribute to the air emissions. They emerge
typically from sources such as stacks of process furnaces and boilers, regenerators (FCC),
individual items such as valves and pump seals and to a lesser extent from flares and incinerator
stacks. Some documents have already been published that establish emissions factors for the
calculation of air emissions from refineries [208, USAEPA, 1996] [136, MRI, 1997].
600
Specific CO2 emissions
500
(kt CO2/Mt crude)
400
300
200
100
0
0 2 4 6 8 10 12
Nelson complexity Index
Figure 3.2: Specific CO2 emissions from some European refineries as function of the Nelson
complexity index
Source: [268, TWG, 2001]
Combustion processes are the main source of NOx. The main emission sources for NOx are
process furnaces/ boilers, gas turbines, FCC regenerators, incinerators, gas turbine and the flare
system. Roughly 60 - 70 % of the NOx of a refinery derives from the use of fuels in furnaces,
boilers and gas turbines. 16 % relates to the FCC, 11 % to engines and the rest to gas turbines
and combined cycle systems. NOx emissions from refineries depend on the fuel type, fuel
nitrogen or hydrogen content, combustor equipment design, and operating conditions.
Accordingly, large differences in the NOx emission level can be expected between refineries and
even between different combustion installations at the same refinery at different times. The
range of NOx emissions in refineries is from 50 to 5000 tonnes per year. The specific emission
range of NOx varies from 60 to 500 tonnes of NOx per million tonnes of crude oil processed
[101, World Bank, 1998]. Next figure gives the plot of the specific NOx emissions from some
European refineries as a function of the Nelson complexity index. As it can be seen, the figure
does not show that specific emissions of NOx depends on the refinery complexity.
500
Specific NOx emissions
Refineries
with FCC
(t NOx/Mt crude)
400
300
200
100
0
0 2 4 6 8 10 12
Figure 3.3: Specific NOx emissions from some European refineries as function of the Nelson
complexity index
Source: [268, TWG, 2001]
As it has been mentioned above, FCC generates NOx emissions and they represent around 16%
of the NOx emissions in refineries. As consequence, it is possible that FCC refineries may emit
more NOx than the ones that they do not have. In the graph, refineries with FCC have been
plotted with red points. As it can be seen, the figure does not show that specific emissions of
NOx depends on having or not having a FCC.
Refinery capacity Total flue gas volume Specific volume of flue gas
(Mt) (109 Nm3/yr) (109 Nm3/yr per Mt)
15 22.173 1.48
12.4 10.171 0.82
10 11.9 0.84 gas fired refinery
9.6* 6.2 1.55 More liquid fuel
9.6* 6.6 1.46
9.6* 7.4 1.30
9.6* 9.9 0.97 Less liquid fuel
9 8.28 0.92
9.0 9.2 0.98
5 5.098 1.02
4.5 5.535 1.23
4.2 1.996 0.48
2.5 0.866 0.35
1.2 0.288 0.24
Range in a EU+ 1 – 1.5
country
* Values that correspond to the same refinery but for different years. The proportion of RFG increases as you go
down in the table.
As it can be shown in this table, the specific volume of flue gas is very variable. However, it has
been found a correlation that speciality refineries (last two raws in the table) tend to give lower
values. The average for other nine refineries is around 1 and it varies between 0.48 and 1.55.
This factor depends on fuel distribution (gas/liquid ratio as shown in the asterisk example within
the table), fuel composition, oxygen content of the flue gas, etc. and consequently the range is
wide. However, it may be calculated, considering those values, that a refinery emitting 200
tonnes of NOx/Mt of crude processes may emit flue gases with a NOx concentration of 130-420
mg/Nm3 (average: 200 mg/Nm3).
Particulate emissions
The concern with particulate emissions (including of heavy metals) stems from health effects.
The main emission sources are process furnaces/ boilers (mainly those fired with (liquid) heavy
fuel oil), catalytic cracker regenerators, coke plants, incinerators, decoking & sootblowing of
furnaces and the flare. The range of emissions found in European refineries is from 100 to
20000 tonnes of particulates emitted per year. The specific emission range found is from 10 to
3000 tonnes of particulates per million tonnes of crude oil processed. The lower emission value
can be achieved in simple refineries burning substantial amounts of gas or when effective
dedusting devices (ESP, bag filters) are installed.
Important heavy metals in crude oils are arsenic, mercury, nickel and vanadium (See
Section 10.3). Nickel and vanadiun are enriched in the residues during distillation and are
removed with the particulate matter by ESP or fabric filters after combustion in the furnaces or
after regeneration of catalysts by burning off. Next table shows data on air emissions of Nickel
and Vanadium from two EU+ refineries:
Nickel Vanadium
Refinery capacity (Mt) Specific load Range of Specific load Range
(t/Mt) concentration (t/Mt) concentration
(mg/Nm3) (mg/Nm3)
15.3 0.196 0.2 – 1.2 0.327 0.3 – 2.1
12.3 0.772 0.02 – 2.35 1.666 0.05 – 5.1
There are both continuous and non-continuous sources of SOx emissions. The number of sources
may vary from refinery to refinery. The main emission sources of SO2 are the process furnaces/
boilers, gas turbines, sulphur recovery units, FCC regenerators, flare system, incinerators,
decoking operations and the treating units. CONCAWE [49, CONCAWE, 1998] gives an
average distribution of the SOx emissions of some 70 refineries studied. Mainly as a result of the
use of liquid refinery fuels in boilers and furnaces, 59.4 % of refinery SOx emissions stem from
combustion processes. FCC units account for 13.5 %, sulphur recovery units for 10.7 % and
other miscellaneous sources for 11.4 %. Next table shows the ranges found for those refineries.
Table 3.3: SO2 discharged to atmosphere as an average from 70 European refineries and a split
of total refinery emissions as an example for SO2 emissions.
Sources: Corinair W-Europe and CONCAWE report 3/98
The same report mentioned above shows that, in 1995 the average SO2 concentration in
European refineries amounted to 1350 mg/Nm3; concentrations ranged from 226 – 2064
mg/Nm3 flue gas. Some data on average bubble calculations have been included in Section 3.15.
The range of annual emissions is from 49 to 10000 tonnes per year and specific emissions range
from 30 - 6000 tonnes of SO2 per million of tonne of crude processed. That report also shows
that, despite similar sulphur levels in the crude oil, major differences exist between average
refinery SOx emissions. In order to complement this information the next three figures relating
SOx emissions from EU+ refineries are plot. Next figure gives the plot of the SO2 emissions
from some European refineries as a function of the crude capacity. Figure shows that bigger
refineries tend to emit more SO2.
30000
25000
20000
SO2 (t/y)
15000
10000
5000
0
0.0 5.0 10.0 15.0 20.0
Crude capacity (Mt)
Figure 3.4: SO2 emissions from some European refineries as function of the crude capacity
Source: [268, TWG, 2001]
However, if the emissions of SO2 are divided by the capacity of the refinery, the figure does not
show that specific emissions of SO2 depends on the crude capacity. The following figure shows
the results for the same refineries of the picture above.
2500
Specific SO 2 emissions (t SO2/ Mt)
2000
1500
1000
500
0
0.0 5.0 10.0 15.0 20.0
Crude capacity (Mt)
Figure 3.5: Specific SO2 emissions from some European refineries as function of the crude
capacity
Source: [268, TWG, 2001]
No correlation is also found when the specific SO2 emissions are plotted against Nelson
complexity index. As it can be seen, the figure does not show that specific emissions of SO2
depend on the refinery complexity.
2500
Refineries
0.8 with FCC
2000
Specific SO2 emissions
(t SO2/Mt crude) 1500
1000
0.6 1.35
500 2.5 1 0.51
0.24 1.02
0
0.7 2.5 0.8 0.26
0 2 4 6 8 10 12
Figure 3.6: Specific SO2 emissions from some European refineries against the Nelson complexity
index
Source: [268, TWG, 2001]
As it was mentioned in the beginning of this section FCC contribute between 7 to 14 % of the
SO2 emissions within a refinery. Then possibly, refineries with FCC may tend to emit more SO2
than refineries without FCC. However, red points in the above graph do not show that refineries
with FCC tend to emit more SO2. Another variable that typically is refer to to explain behaviour
of SO2 emissions in refineries is the content of sulphur in the crude oil processed. Figures in the
above figure, show that refineries with similar sulphur content in the crude oil that they process
(e.g. 0.8 %) can have very high differences in specific emissions. Moreover, it is shown in the
figure that no correlation is demonstrated with this figure. This also reinforces the conclusion of
the Concawe report that shows that, despite similar sulphur levels in the crude oil, major
differences exist between average refinery SOx emissions.
Table 3.4 shows the historic development in the distribution of sulphur output in European
refineries. It also presents the average sulphur recovery percentage in the western European
refineries that has grown from about 10 % in the late 70’s to more than 36 % now. This table
also shows that the sulphur reductions achieved in the oil products (so called S in fuels) sold to
customers are not mirrored in direct sulphur emissions from the refineries, which have remained
static.
Table 3.5 shows that since 1985 the sulphur recovered in refineries has increased, which reflects
the steady increase in sulphur recovery capacity of European refineries. This is also
demonstrated in the decrease in the sulphur content of the oil products leaving the refineries.
However, Table 3.5 also shows that since 1985 direct SOx emissions from European refineries
have been stable. It demonstrates that on average the reductions in SO2 emissions have clearly
lagged behind the sulphur reductions achieved in the oil products sold to the customers of the
refineries.
Table 3.5: Historical trend sulphur output distribution European refining industry
Source: [115, CONCAWE, 1999]
Table 3.5 depicts the development of the sulphur balance in European refineries from 1979 to
1995. It shows that in 1995 36.3 % of the 6.56 million tonnes of sulphur intake is recovered as
elementary sulphur and 8.6 % is emitted as SO2 by the refinery. About 50 % w/w is emitted as
SO2 via the combustion of the oil products, while 14.6 % w/w is accumulated in non-fuel
products, such as bitumen, lube oils, coke and chemical feed stocks. Those last compounds may
be burnt to some extent or can finish their life as solid or liquid waste and consequently might
contribute to emissions to the air, water and soil. The historic trend is clearly one of a steady
decrease in the sulphur levels in oil products and an equally steady increase in the sulphur
recovered in the refinery. The decrease of sulphur levels in oil products is partly the result of
tighter European and national legislation, especially with respect to middle distillates. A major
part of the decrease is the result of changes in the demand for fuel oil.
Some accidental spillages can be of a continuous nature, such as leaking pump seals, leaking
valve glands and leaks from pipework or others which are of a one-off nature, can occur from
sources such as equipment failures, overfilling of tanks and overfilling of road and rail cars.
Fugitive emissions from process equipment are the largest single source of VOCs emitted to the
atmosphere in a refinery and can frequently account for 50 % of the total emissions. Fugitive
emissions include the emissions that occur from items such as valves, pump and compressor
seals, flanges, vents and open ends. Valves are considered to account for approximately
50 - 60 % of fugitive emissions. Furthermore, the major portion of fugitive emissions comes
from only a small fraction of the sources (e.g. less than 1 % of valves in gas/vapour service can
account for over 70 % of the fugitive emissions of a refinery). Some valves are more likely to
leak than others, e.g.:
• valves that are operated frequently, such as control valves, may wear more quickly and will
allow emission paths to develop. However, newer, low-leak control valves provide good
fugitive emissions control performance.
• valves with rising stems (gate valves, globe valves) are likely to leak more frequently than
quarter-turn type valves such as ball and plug valves.
Factors driving these releases of hydrocarbons are equipment design, quality of the sealing
system, maintenance programme and properties of the line contents. Poorer designs (with wider
tolerances), poor sealing systems (e.g. leak-prone valve packings) and limited maintenance will
lead to higher emissions. Those emissions can be estimated using the factors in Table 3.6.
For example it can be estimated, using the USAEPA method, that a pump for light liquid that
has an emission measured in the range of 1001 - 10000 ppm, has an emission factor of 33.5 g/h.
Other methods have been applied to calculate the fugitive emissions from refineries. They are
the adsorption method and DIAL method (see Section 3.26). Emission factors measured with
the latest method are presented in Table 3.7.
Table 3.6: Emission factors according to stratified USAEPA-method for the assessment of
fugitive emissions
The mercury and components of arsenic are volatile and are emitted to some extent with the
clean gas. A part of these compounds react with plant material or deposit on the catalysts in the
reactors of the conversion plants. Therefore protective layers are used for the interception of the
catalyst poisons.
Flares, compressors, pumps, turbines and air coolers require particular attention as regards
sources of noise.
Odours in a refinery are mainly created by sulphur compounds as H2S, mercaptanes but also by
some hydrocarbons (e.g. aromatics). The main focus of odour in refineries are storage (e.g. sour
crudes), bitumen production, desalter water, sewers, dissolved air flotation and biotreatment and
flaring.
Waste waters consist of cooling water, process water, sanitary sewerage water and storm water.
The quantity of waste waters generated and their characteristics depend on the process
configuration. Waste waters are treated in on-site waste water treatment facilities and then
discharged. The water and blowdown streams from refinery operations can become
contaminated with dissolved gases, dissolved and suspended solids and hydrocarbons. Water
can also become contaminated with dissolved gases in the process, compounds that could be
toxic and/or could give to water an unpleasant smell. Nearly all refinery processes have steam
injection to enhance the distillation or separation processes. This leads to the production of sour
water (containing ammonia, hydrogen sulphide and hydrocarbons). Sour water needs stripping
prior to further treatment or re-use as wash water.
Depending on the prevailing climate, storm or rainwater can also generate a significant effluent
stream in refineries. Rainwater may come into contact with potentially oil-polluted surfaces
(‘surface water run-off’). The effect of rainwater on both the amount and the quality of the
effluent water and the problems that the water discharges to either fresh continental waters or to
the sea are issues to consider. Occasionally, ballast water from crude carriers and from other
seagoing vessels or barges transporting oil products is also treated in refineries. Furthermore,
sanitary waste water and fire-fighting water are waste water sources which all need attention
and quality control prior to deciding on treatment, direct discharge and/or potential re-use.
Some water emission parameters are the pH, dissolved solids, suspended solids, total organic
carbon, total nitrogen, total phosphorous, COD, BOD, H2S, NH3, oil, aromatics (BTEX),
phenols, salt, AOX, mercaptans, cyanides, MTBE, fluoride, PAH, heavy metals, temperature,
92 Mineral Oil and Gas Refineries
Emissions and consumption levels Chapter 3
conductivity, bacteria and fish toxicity. As a general guide, approximately 0.1 - 5 m3 of waste
water (process waste water, cooling water and sanitary waste water) per tonne of crude is
generated when cooling water is recycled.
Table 3.9 shows the composition and annual averages of the water effluents from European
refineries.
Table 3.9: Typical refinery influent/effluent annual average composition and load
Sources: [118, VROM, 1999], [261, Canales, 2000], [101, World Bank, 1998], [293, France, 2000]
The amount of oil discharged is usually expressed in grams of oil per tonne of refinery
throughput. Most of the European refineries already meet the standard of the Oslo/Paris
Commission of 3 grams oil/tonne of refinery throughput.
Refineries also generate solid wastes and sludges with a range of 0.01 - 2 kg per tonne of crude
processed (before waste treatment). According to the World Bank, 80 % of those solid wastes
may be considered hazardous because of the presence of toxic organics and heavy metals [101,
World Bank, 1998]
The refinery waste generation was covered in a report [82, CONCAWE, 1995] representing the
1993 European refinery waste situation (see Table 3.10). A summary of the waste generated in a
refinery is 45 % sludge, 35 % non-refining wastes and 20 % other refining wastes. From the one
million tonnes of wastes identified in Europe in 1993, 39.9 % was placed in landfills, 21.4 %
was recycled or re-used, 14.9 % was incinerated with energy recovery, 8.4 % was incinerated
without energy recovery, 4.9 % was landfarmed, 1.7 % was used as alternative fuel and the rest
(0.6 %) was sent to unidentified disposal routes [82, CONCAWE, 1995].
Sludges originate from various sources such as crude and product tanks bottoms, desalters,
alkylation units, boiler feed water preparation, biotreaters, the cleaning of heat exchanger
bundles and equipment, oil spills and soil remediation. In terms of volume oily sludges
represent the largest waste category from refineries. Partly this is due to the presence of basic
sediment and water in the crude, which can vary from crude to crude. Bio-sludge production
takes place only if a refinery operates a biotreater.
Other wastes are generated from many of the refining processes, petroleum handling operations
and waste water treatment. Both hazardous and non-hazardous wastes are generated. Spent
catalysts originate from reformers, catalytic crackers, hydrocrackers, hydrodemetallisation,
hydrodesulphurisation and hydrotreating units. The regeneration of catalysts is a well-
established technique.
3.2 Alkylation
This section contains data on consumption and emissions generated by alkylation processes.
Consumptions
Table 3.11 gives a summary of the utility and chemical consumptions in the two techniques
currently used in the alkylation processes.
Alkylation technique
Values per tonne of alkylate Sulphuric acid Hydrofluoric
produced
Utilities
Electricity (kWh) 4 20 - 65
Fuel (MJ) n.a. 1000 - 3000
Steam (kg) 830 100 - 1000
Cooling water (m3) 72 62
(∆T = 11 ºC)
Industrial water (m3) 0.08
Chemicals
Fresh acid (kg) 78 - 120 1.15
Caustic (as 100 % NaOH) (kg) 0.41 0.57
Lime n/a n.a.
Note: AlCl3 and water-free CaCl2 consumption depend on operating
conditions (water content of feed- and side-reactions)
Table 3.11: Estimated utilities and chemical consumption for the various alkylation techniques
Sources: [166, Meyers, 1997], [118, VROM, 1999], [261, Canales, 2000]
Emissions
Emissions generated by the alkylation processes are summarised in next tables. [108, USAEPA,
1995]
Deasphalting
Consumption
A conventional solvent-based base oil lube complex is energy intensive. The main reason is the
high amount of heat required to evaporate the solvents and to separate them from the raffinate
and extract streams. Solvent losses are usually in the order of 1 % despite extensive solvent
recovery. In the tower method, for example, four to eight volumes of propane are fed to the
bottom of the tower for every volume of feed flowing down from the top of the tower. As an
example, the figures below give the utility requirements of a solvent deasphalting absorption
(SDA) unit.
Emissions
Air Waste water Solid waste
Air emissions may arise from The solvent recovery stage results Little or no solid waste
fugitive solvent emissions and in solvent-contaminated water generated.
process vents. which is typically sent to the
waste water treatment plant.
Heater stack gas (in Section 3.10). Oil components
Aromatic extraction
Consumptions
Emissions
Air Waste water Solid waste
Fugitive solvent VOC The total amount of process water is around Little or no solid
emissions can come from the 2 - 4 m3 per tonne of product, containing wastes generated.
furfural and NMP from its around 15 - 25 ppm furfural, 10 - 15 ppm
storage tanks. NMP.
Flue gas from fired heater (in The water stream leaving the fractionator is
Section 3.10). likely to contain some oil and solvents.
Source: [118, VROM, 1999]
Solvent dewaxing
Consumptions
The utility requirements in this process together with the aromatic extractions are:
Table 3.18: Typical utility consumption of solvent dewaxing and aromatic extraction units
together
Sources: [118, VROM, 1999], [261, Canales, 2000]
Emissions
Air Waste water Solid wastes
VOC may arise from The solvent recovery stage results in solvent-contaminated Little or no solid
solvent vapour extraction water which is typically sent to the waste water treatment wastes
from the wax filters as plant. Potential releases into water are hydrocarbons, generated.
well as fugitive emissions. sulphur compounds and organic compounds from spillages
Fugitive solvent VOC and leaks, and of organic compounds from process waters
emissions can come from from solvent recovery operations. Waste water contains
the MEK/ Toluene from 1 - 3 ppm MEK/toluene.
its storage tanks.
Flue gas from fired heater
(in Section 3.10).
Hydrofinishing
Consumptions
The utility needs for a hydrofinishing unit are:
Emissions
Air Waste water Solid wastes
Pressure relief valves of the hydrofiner, Leakages from flanges, glands and n.a.
solvent recovery systems and refrigerant seals on pumps, compressors and
systems; valves.
Leakages from flanges, glands and seals
on pumps, compressors and valves.
Emissions
FCC RCC
Fuel (MJ/t) 120 - 2000 120 - 2000
Electricity (kWh/t) 8 - 50 2 - 60
Steam consumed (kg/t) 30 - 90 50 - 300
Steam produced (kg/t) 40 - 60 100 - 170
Cooling water (m3/t, ∆T=17 °C) 5 - 20 10 - 20
Catalyst make-up (kg/t) 0.4 - 2.5 2-4
Virtually all the heat required in a FCC unit is generated in the regenerator. The catalyst used
depends greatly on the type of product required and can be silica-alumina substrate carrying rare
earth and/or precious metals or can be based on zeolites.
Emissions
Air emissions
Within a refining complex, one of the sources with major potential for atmospheric emissions is
the catalytic cracking unit. Air emissions come principally from the regenerator and are mainly
CO, CO2, NOx, particulates (mainly catalyst fines) and SO2. The emissions from catalytic
crackers are very variable. The wide variation reflects both the wide variations of feedstocks
used in the catcracker units (nitrogen, sulphur, metals contents), and the regenerator and waste
heat boiler operating conditions. Emissions from an FCC can be 20 - 30 % of total refinery SO2
emissions, 15 - 30 % for NOx and 30 - 40 % of particulates. However, these figures can be
subject to wider variation [112, Foster Wheeler Energy, 1999]. Air emissions generated by the
combustion of fuels in the FCC furnaces are treated in Section 3.10.
The next table shows a summary of the emission factors and emissions from FCCs
Residue catalytic crackers (RCC) use mainly as feedstock atmospheric residue. Hence the
Conradson carbon, sulphur content and nitrogen content is higher. Therefore, RCC are
potentially a stronger source of SOx and NOx, and contaminated catalyst than normal FCC. As
consequence of the type of feedstock used in the RCC, RCC produces more coke and hence
more heat should be removed from the regenerators.
Because the catcracker is a significant emitter of pollution in a refinery, there follows a more
detailed analysis by type of pollutant.
Carbon dioxide
Carbon dioxide is generated in the catalyst regeneration train and its load depends on the size.
Carbon monoxide
Relatively high concentrations of carbon monoxide can be produced during regeneration of the
catalyst and it is typically converted to carbon dioxide either in the regenerator (total
combustion) or further downstream in a carbon monoxide boiler (partial combustion).
Incomplete combustion may lead to releases of carbon monoxide in the total combustion. CO ex
full burn units ranges typically between 50 and 1200 mg/Nm3 (depending on temperature, CO
promoter level, size of unit). In partial combustion mode, depending on the conditions under
which the regenerator is operating, the levels of CO in the regenerator exhaust gases feeding the
CO boiler may be in the region of 5 to 10 %. Levels of CO in the exhaust from the CO boiler
can be maintained to less than 100 mg/Nm3 depending on the type of CO boiler used [80, March
Consulting Group, 1991].
Nitrogen oxides
The wide range in NOx emissions (levels) from catcracker regenerators (gases) reflects the main
effects of different combustion conditions from FCC regenerators or FCC CO-boilers, the
spread in unit capacities and the coke (feed) nitrogen contents; these latter depend on crude type
and upstream process configuration. Lower emission values can be found when deNOx
treatments are applied. Some examples of NOx emissions from catcrackers as well as the ranges
found are shown in the next table.
Table 3.24: Range and examples of nitrogen oxide emissions from catcrackers
Sources: [250, Winter, 2000], [136, MRI, 1997], [268, TWG, 2001]
Particulates
Generally a FCCU is the biggest single emitter of particulates although the calciner in a coking
unit is also a significant emitter. Particulate matter arises from catalyst fines from the catalyst
regeneration exhaust gases and catalyst handling and disposal. The emitted catalyst is fine and is
produced in catcracker as a result of the constant movement of the catalyst grains. In the
absence of moisture or sulphuric acid condensation, the normal cause of stack plume opacity is
the presence of the fine catalyst particles. In practice, operating conditions may have a large
effect on particulate emissions.
Table 3.25: Examples of particulate emissions from catcrackers (lower values correspond to FCC
with abatement techniques)
Sources: [250, Winter, 2000] [136, MRI, 1997], [268, TWG, 2001]
Sulphur oxides
The sulphur in the feed to the catcracker is split between liquid product streams, H2S in the
gaseous products and SO2 emission from the regenerator in the approximate ratio of 50/45/5.
The degree of sulphur dioxide leaving the exhaust gases from the catalytic cracker depends on
the sulphur content of the feedstock used as well as the technology used to control these
emissions. SO3 aerosols will also contribute to the catcracker stack plume opacity.
The next table shows some examples of yearly averages of SO2 emissions from catcrackers.
Table 3.27: Sulphur oxide emissions from catcrackers (with and without abatement techniques)
Sources: [261, Canales, 2000], [250, Winter, 2000], [107, Janson, 1999], [112, Foster Wheeler
Energy, 1999], [268, TWG, 2001]
Other compounds
Hydrogen sulphide, mercaptans and ammonia may arise from sour waters from reflux
condensers. Hydrocarbons (typically 80 % alkanes and typically 15 % of olefins) may be
released from pressure reliefs, storage and handling operations, spillages and water discharges.
Measured dioxin emission values in two European FCC reported levels of < 0.016 ng TEQ/Nm³
after CO - combustion. [268, TWG, 2001]
Solid wastes
Potential releases into land are of catalyst in the form of fines from particulate matter arrestment
equipment and intermittent spent catalyst discharge. Those solid residues are enriched in the
heavy cycle oil and clarified oil (distillate of FCCU) fractions.
Catalyst addition rates for resid operation depend on feed metal levels. Addition rates shown are
typical values covering a feed metal range of 10 - 20 ppm Ni+V+Na.
Emissions
(well below 10 mg/Nm3 [117, VDI, 2000]), SO2 and dioxins and furans (113 ng/Nm3 TCDD-eq
EADON, 131 ng/Nm3 TCDD-eq I-TEF [215, Jansson, 1999]). A study from USAEPA [315,
USAEPA, 2000] shows that the CDD/CDF emission factors from petroleum catalytic reforming
units are 0.196 ng/t for semi-regenerative units and 1172 ng/t for continuous regeneration units.
Dioxin emissions measured in four Belgian refineries are <0.1-0.13; 3.3-6.7; <0.01 and <0.01
ng TEQ/Nm3 (higher values are derived from an installation with a flow of 2000Nm3/h (ca 1800
h/yr) [268, TWG, 2001]. The regenerator gas is normally treated in a water scrubber to remove
dust, chlorine compounds and SO2 prior to release to atmosphere. The storage and handling of
organic chlorides used during the regeneration may also lead to releases.
Waste water
The amount of waste water generated in the catalytic reforming is around 1 - 3 litres per tonne
of feedstock. The waste water contains high-levels oils, suspended solids, COD and relatively
low levels of H2S(sulphides), chloride, ammonia and mercaptans. Those compounds may also
be found in the stripping tower used to remove light ends from the reactor effluent.
Hydrocarbons from spillages and leaks may occur. A study shows that the untreated waste
waters and spent caustics from the reformers contain a wide range of CDD/CDF concentrations
of 0.1 pg I-TEQDF/l to 57.2 ng I-TEQDF/l [315, USAEPA, 2000].
Table 3.32: Example of composition of the sludges generated in the unit separator sludge of an
ultraformer
Source: [80, March Consulting Group, 1991]
Delayed coking
Table 3.33 shows the energy and process materials usage in the delayed coking.
The amount of required make-up water for the coking process depends on the evaporation losses
and the bleed to the desalter. Treated effluent can be used for this. The water used for coke
cutting during decoking of the drums is highly contaminated with (un)saturated hydrocarbons
and coke particles.
Flexicoking
Table 3.34 shows the energy and process materials usage in flexicoking.
Calcinator
The calciner specific energy consumption figures are given per tonne of calcined coke
produced.
Emissions
The most important health and safety aspect of these processes is the handling of the coke fines.
Consulting Group, 1991] [251, Krause, 2000]. Storage, crushing and handling of green coke
takes place in the wet state with no releases to air. Typical size distribution profiles for
particulate matter in calciner exhaust gases after passing through a cyclone system is:
Waste water
Waste water is generated from the coke removal, water bleed from coke handling, sour water
from fractionator overhead, cooling operations and from the steam injection and should be
treated. The amount of waste water generated in the coking processes is around 25 litres per
tonne of feedstock. It contains H2S, NH3, suspended solids (coke fines with high metal contents),
COD, high pH, particulate matter, hydrocarbons, sulphur compounds, cyanides and phenols.
Detailed analysis of the contaminants of such waste waters is not available because they are
transferred directly to the refinery main waste water system for treatment.
Solid wastes
Solid wastes generated in the coking processes are coke dust (carbon particles and
hydrocarbons) and hot oil blowdown sludges containing hydrocarbons. Table 3.36 shows a
typical analysis of those sludges.
Table 3.36: Typical chemical characterisation of the sludge generated in the coker
Source: [80, March Consulting Group, 1991]
The need for cooling in a refinery depends on the processes used and on the degree to which
they are integrated but, where water cooling systems are applied, it certainly accounts for the
majority of water usage. In a refinery, it is important to maximise heat integration at a plant
level and at process/activity level to minimise cooling duty. Therefore, energy requirements for
cooling will depend on the cooling system used together with the cooling strategy applied. The
next table shows a typical split of the cooling needs in a refinery (Hydrocracking refinery with a
throughput of 7 Mt/yr)
Cooling capacity
Application MW %
Process streams 400 94
Pumps, compressors 10 2
Vacuum systems 15 4
TOTAL 425 100
The next table gives a distribution of the cooling capacity according to temperature range within
a typical refinery. (Hydrocracking refinery with a throughput of 7 Mt/yr)
Cooling duty
End temperature (T) process fluid (ºC) MW %
T> 43 380 95
43>T>38 15 4
38>T>30 0 0
30>T 5 1
Total 400 100
Consumptions
Power is consumed by pumps in a water cooling system and by fans in an air cooling system.
The water cooling system uses water and requires chemicals as corrosion and bacterial growth
inhibitors. For more detailed information and annual consumption figures, reference is made to
the IPPC BREF on cooling which can be regarded as the state-of-the-art document on cooling
systems. Both recirculating and once-through cooling water systems require additives, to
prevent fouling and/or corrosion. Because once-through systems generally use surface water
(either fresh or salted), the chances of fouling are higher, compared to recirculating systems.
Therefore, in these systems, more anti-fouling additives (i.e. chlorinated biocides) should be
used. On the other hand anti-corrosion additives are mainly applied in recirculating systems and
not used in once-through systems.
Emissions
The main “contamination” generated by the cooling systems is the heat, increasing the
temperature of the cooling fluid used. In refinery water cooling, the increase of the temperature
(∆T) is around 10 – 15 ºC.
Water consumption (mentioned above), energy consumption (pumps, air-cooler fans) and water
pollution are the main environmental issues of cooling systems. Other environment-related
effects include noise production (cooling tower, pumps, air-cooler fans (97 - 105 dB(A) at
source)) and plume formation (cooling tower).
The main pollutants to be considered in water cooling systems are the chlorinated and/or
brominated antifouling additives, and anticorrosion additives containing zinc, chromium,
molybdenum etc. Special attention has to be paid to the use of dispersing additives in closed
cooling water systems, particularly when the bleed is routed to an oil separating water treatment
unit, where they can interfere with the oil-water separation process. A once-through system
coupled with a low rate of leaks and a high volume of water mean that cooling water discharges
contain between 0.1 and 1 mg/l oil. Hydrocarbon emissions to the air from cooling towers (as a
result of leakages and stripping) may occur. The emissions to the air were reported to vary
between 0.5 and 85 g hydrocarbons per m3 cooling water recirculating over the cooling tower
[119, Bloemkolk and van der Schaaf, 1996]. Cooling by quenching (only used in delayed cokers
within refineries) results in high vapour emissions, significant energy losses, large water use and
severe water pollution.
Table 3.39: Environmental impact of different cooling systems for a typical refinery
Source: [119, Bloemkolk and van der Schaaf, 1996]
3.9 Desalting
The quantity of inorganic impurities in the crude oil depends very much on both the crude
origin and the crude handling during transport from the crude well to the refinery.
Consumptions
The water used in crude desalting is often untreated or partially treated water from other
refining process water sources. Table 3.40 shows the typical operating conditions and water
consumptions in the desalters, depending on the type of crude oil used.
For desalting long CDU residue, 10 % w/w water on feed is common. Once the water and the
oily phase have been well mixed, the water should be separated from the oil feedstock in a
separating vessel by adding demulsifier chemicals (5 - 10 ppm) to assist in breaking the
emulsion and/or, more commonly, by applying a high-potential electric field across the settling
vessel to coalesce the polar salt water droplets. The electricity consumption used in the desalting
processes varies typically from 0.075 to 0.15 kWh per tonne of crude oil.
Emissions
Air emissions
No major emissions to the air are produced during the desalting processes. Air emissions from
the heating processes are expected (in Section 3.10), and fugitive emissions (hydrocarbons) may
be expected.
Waste water
The desalter is a big contributor of process waste water (30 - 100 litres / tonne feedstock
desalted). The desalting process creates an oily desalter sludge and a high-temperature salt water
waste stream (possible the most polluted in the refinery) which is typically added to refinery
waste water treatment facilities. The waste water generated is highly contaminated. Table 3.41
shows the ranges to be expected from waste waters from desalters.
• specific energy consumption (SEC). This is the most simple index. It is calculated as the
ratio between the energy consumed by the refinery and the tonnes of feedstock processed.
The specific energy consumption of European refineries ranges from 1 to more than 4 GJ
per tonne of throughput. Because it is a simple index, this ratio does not take account of the
complexity of the refinery (more complex refineries tend to consume more energy).
• products method. [318, Phylipsen, Blok et al., 1998] This method considers the products
and internal energy products produced in the refinery, giving a specific energy consumption
benchmark per tonne of energy product produced. Multiplying those specific energy
consumptions by the amount produced within the refinery and adding all those totals gives
an energy consumption benchmark for that refinery. Some calculations indicate that the best
specific energy consumption figures are between 2.4 and 2.9 GJ/tonnes whereas actual
values are from 1 to 4.8. That means that some European refineries are doing better than
figures that are considered a good benchmark.
• energy intensity index (EII) is a measure used to compare energy consumption in refineries.
The standard energy use is based on the energy consumption at about 300 refineries
worldwide. The average EII obtained in a worldwide market survey (Solomon study, 1994)
was 92, with a range from 62 to 165. [107, Janson, 1999]. More energy efficient refineries
correspond to the lower values of EII. So, some refineries are almost three times as energy
efficient as others. This index does reflect the types of processes and throughput of each
process in the refinery. This data is not available for all refineries and typically is
considered confidential by refineries. 10 refineries have reported its figure and the values
have been plotted against the complexity index. Lowest EII values are typically achievable
under local circumstances where low grade heat can be externally exchanged.
150
100
World average
75
50
0 2 4 6 8 10 12
Nelson Complexity Index
Figure 3.7: Energy intensity index vs. Nelson complexity index for some EU+ refineries
Source: TWG
Gasification plant
An example of IGCC units produces 130 t/h of syngas mainly composed of CO and H2 in the
ratio 1:1 (calorific values of 9600-10000 kcal/kg) starting from 58 t/h of feedstock (calorific
value of the heavy residue varies between 8800 and 9200 kcal/kg, sulphur content: 3.5 - 7 %
and metal content: 300-800 ppm ). A solid effluent in the form of a filter cake (about 160 - 400
kg/h dry) is discharged and sent to external plants for the recovery of the metals. Two Claus
units recover 4 t/h of elemental sulphur from the H2S recovered in the acid gas removal section.
A tail gas treatment section, that permits an overall sulphur recovery of 99.9 %, follows the
units [297, Italy, 2000].
Refinery fuels
The baseline ratio of gas to liquid refinery fuel used in a refinery is a function of a number of
factors, important amongst which are the size, degree of complexity, LPG recovery efficiency
and extent to which refinery fuel gas is processed into other products (e.g. olefins) or exported
to adjacent chemicals plants (either directly or as energy from common facilities). It varies from
80/20 or 70/30 (gas/liquid) on a stand-alone, moderately complex refinery to 40/60 on a highly
complex site which also services a chemicals complex. However these ratios can be increased
when energy conservation measures are applied and the gas availability becomes sufficient for
the energy supply of the refinery.
Next table illustrates the calorific value and the sulphur content of the different fuels as well as
its calorific value.
There are great differences in the characteristics of the liquid refinery fuel used in European
refineries. The data from a CONCAWE report show that in 1995 some 20 % of the total liquid
refinery fuel used in the refineries had a sulphur level between 3 and 4 %, another 20 % a
sulphur level of 2 to 3 %, some 40 % a sulphur level between 1 and 2 %, and 20 % a sulphur
level below 1 %. However, it is worth to mention that some European refineries use heavier
liquid refinery fuels (up to 7%). Next table shows the chemical characteristics of three different
liquid fuels.
The nitrogen and sulphur contents are two of the most important chemical parameters and
responsible amongs other to the emissions of NOx and SOx. The following graph shows the
distribution of nitrogen and sulphur contents in various types of vacuum residues according to
their geographical origin.
Figure 3.8: Sulphur and nitrogen contents in HFO (vacuum residues) according to their
geographical origin
Source: [345, Molero de Blas, 2000]
Steam generation
Steam generation requires a energy input of 2700 - 3200 MJ per tonne of steam produced.
Conditioning chemicals are dosed to the BFW in low concentrations and comprise the following
groups of chemicals: anti scaling agents, corrosion inhibitors and anti foaming agents. A 100 t/h
steam generation system requires approximately 1.5 - 3 t/yr corrosion inhibitors and 2 - 4 t/yr
anti-scaling agents. Those conditioning chemicals are:
• corrosion inhibitors (mainly oxygen scavengers and alkaline compounds). Sulphite (< 60
bar), oximes, hydroxyl amines, and hydrazine (declining use due to safety issues) etc. are
commonly applied as oxygen scavengers for deaerated boiler feed water prior to pumping
into the boiler. Commonly applied alkaline compounds are sodium phosphates (which are
also hardness binders), caustic, ammonia and neutralising amines.
• anti-scaling agents such as polyacrylates and phosphonates that are rest hardness binders
and dispersing agents
• anti-foaming agent, in general intermittently dosed, to combat foaming in case the
condensate contains oil or organics.
3.10.3 Emissions
The main releases to air from combustion processes are stack gases containing oxides of
sulphur, oxides of nitrogen, oxides of carbon (carbon monoxide and carbon dioxide) and –
particularly important when liquid refinery fuel or coke is combusted – particulate matter
(including PM10 and metals (e.g. V, Ni)). When operating properly and when burning cleaner
fuels such as refinery fuel gas, low-sulphur fuel oil or natural gas, these emissions are relatively
low. If, however, combustion is not complete, or heaters are fired with refinery fuel pitch or
residuals, emissions can be significantly higher. Incomplete combustion may lead to releases of
carbon monoxide, smoke and, if heavy fuel oils are used, particulate matter. As a result of that,
combustion plants are major contributors to the release of air emissions from refineries. The
level of pollutants emitted will depend on the quality of the fuels burnt, which can vary
considerably. There are conflicting factors in a combustion process which contribute to
emission levels. For example, combustion conditions which favour low particulate emissions for
liquid refinery fuel, i.e. excess air, high temperature, good air/fuel mixing and good fuel-
atomisation, are not favourable for low- NOx emissions.
Refinery fuel gas, if properly treated, is a low polluting fuel. Liquid refinery fuels generate more
emissions to the air than refinery fuel gas. Gas-fired heaters and boilers generate little dust and
lower SO2 emissions, as the refinery fuel gases are usually cleaned in amine scrubbers. NOx
emissions are also much lower than those of liquid-fired boilers and heaters. By way of
example, tables 3.44 to 3.46 show, respectively, the air emissions generated by gas and liquid
refinery fuels in two power plants used in European refineries.
Table 3.45: Air emissions from a power plant fired with heavy fuel oil (primarily visbreaker
residue, up to 7 % S)
Source: [247, UBA Austria, 1998]
Table 3.46: Air emissions from the power plant of the Mider refinery buring refinery liquid
residue
Source: [247, UBA Austria, 1998].
These follows an analysis of the air emissions produced by the energy system, pollutant per
pollutant.
CO2
Combustion processes of fossil fuels produce CO2 as a result of the combustion of
hydrocarbons. The amount of CO2 emitted to the atmosphere by European refineries varies from
20000 to 20 Mtonnes per year (range depending on the type of refinery and the energy
integration). The specific CO2 emissions range from 0.02 to 0.82 tonnes of CO2 per tonne of
crude oil processed. CO2 emissions from power plants in the refineries correspond to around
42 % of the CO2 emitted by the refinery. The use of liquid fuels results in lower thermal
efficiencies and higher CO2 emissions than the use of gaseous fuels.
Table 3.47 shows the sources (type of fuels) of the CO2-emissions of the refining industry. The
table also includes the calculated emission factors for CO2 for various refinery fuels
CO
One of the products of partial combustion processes is carbon monoxide. The CO emissions,
which range from 20 - 42 mg/Nm3 @ 3 % O2, are very dependent on the type of fuel used and
the completeness of the combustion process. Specific emission values for CO are 0.58 kg CO
per tonne of refinery fuel gas and 0.24 kg CO per tonne of liquid fuel. [247, UBA Austria,
1998]
NOx
NOx emissions from a refinery energy system depend on the fuel type, fuel nitrogen or hydrogen
content, combustor equipment design and operating conditions. The formation and release of
NOx from combustion processes arises from the oxidation of nitrogen present in the fuel source
and/or the air used. Accordingly, large differences in the NOx emission level can be expected
between refineries and even between different combustion installations at the same refinery at
different times. Differences in temperature, residence time and oxygen concentration result in
varying levels of thermally formed NOx. The influence of temperature is most important, with
NOx emissions increasing exponentially with temperature.
Gaseous fuels typically release less NOx per unit of energy compared to liquid fuels, especially
liquid refinery fuels. Oil burning normally leads to higher levels of NOx releases for several
reasons, especially the problem of fuel NOx arising from the nitrogen content (0.03 - 1 %), the
way in which plant is often operatied so as to balance NOx and particulate releases and the
frequent design requirement for firing in combination with gas. However, the latter statement is
only true for non-abated emissions, since the inclusion of secondary measures can reduce both
NOx and particulate emissions. One refinery in 1996 found that fuel oil can generate about three
times more NOx than gas fuels. NOx factors (NOx produced per tonne of fuel combusted), used
by some refiners to report NOx emissions, indicate NOx from oil firing to be 2 to 3 times that
from gas firing. However data from other refiners, particularly those based on stack monitoring,
indicate that NOx from oil can be as much as 5 or 6 times that from gas (e.g. see Table 3.44 and
Table 3.45)
The fired heaters, boilers and gas turbines running on refinery blend gas produce lower NOx
emissions than FCCs. Specific NOx emissions related to energy consumption range from
15 - 200 mg/MJ for furnaces with 100 % gas fuel. Specific NOx emissions related to the
throughput range from 84 to 700 tonnes of NOx per million tonnes of crude oil processed. The
NOx load ranges from 20 to more than 2000 tonnes per year.
Particulates
Under normal conditions, the solids in the flue gases of a furnace or boiler consist mainly of
coke particles with some cenospheres, depending on firing conditions. The particulate emission
of oil-fired equipment may vary considerably since they depend on a number of more or less
independent parameters such as: fuel type, burner design, oxygen concentration at the outlet of
the radiant section, the flue gas outlet temperature of the radiant box, and the residence time of
the fuel droplets. The particle size of the particulate matter (PM) from furnaces and boilers on
HFO is in the order of 1 µm. The PM in flue gases can take any of the following four forms:
• soot; the particle size is below 1 µm-visible smoke from a stack is caused by all particles
but mainly those with a particle size between 0.5 and 5 µm.
• cenospheres: they originate from the liquid phase wastes of combustion of heavy oil
droplets at relatively low temperature (< 700 °C). The size is equal to or larger than that of
the original oil droplets
• coke particles, formed through liquid phase cracking in combustion at high temperatures
(>700 °C). The particle size is generally from 1 to 10 µm
• fine particles (< 0.01 µm): their contribution to the total mass emission is negligible.
The PM in the flue gas of furnaces and boilers fired with heavy fuel oil is a mix of metal
compounds and soot/coke. Metals (mainly V and Ni) are a natural constituent (indigenous) of
the crude. Soot and coke result from imperfections of the combustion process. Contaminants
such as sand, rust and other metals, and coke particles of the fuel itself, may also contribute to
the formation of particulates. The indigenous metal content of the heavy fuel oils (see Section
2.10, and which can be calculated from the metal content of the crude oil from which the HFO
is made) is the basis for calculating the minimum ash content of the HFO and therefore also the
basis for the PM in the flue gas. Ash contents are higher than the metal content itself (as the
metal compounds are in the ash), and generally lie in the range 500 – 1500 ppm (0.05 - 0.15 %
w/w). In practice the PM content is normally higher by a factor of 2 - 4, because unburnt fuel
(soot) adheres to the metallic PM, and because aerosols of SO3 are also monitored as PM.
The range of the unabated PM (150 - 500 mg/Nm3) is typical for the burners applied nowadays
(steam atomization and low- NOx), assuming all measures are taken to achieve good combustion
(optimum oxygen content and lowest practicable NOx in the flue gas) and directly related to the
ash content of the fuel. For older furnaces, liquid refinery fuel-fired flue gas levels may range
from 500 to 1000 mg/Nm3. For boilers, all these figures are on average lower. For a new
optimal burner design with steam atomisation this can be well below 200 mg/Nm3. The range of
150 - 500 mg/Nm3 represents the current typical range with liquid refinery fuels employed (ash
contents) and burners installed (low-NOx with steam atomisation). Table 3.45 shows the
particulate emission range found in current European refineries.
Sootblowing is an operation that is carried out at regular intervals to remove soot from the
furnace that is accumulating on the furnace equipment and hampers proper functioning. During
this operation the PM content of the exhaust gas can reach values of 2000 mg/Nm3.Abatement
techniques installed for normal operations will be effective to reduce the PM emissions from
this operation.
SOx
The release of sulphur dioxide is directly linked to the sulphur content of the refinery fuel gas
and fuel oils used. Heavy fuel oil residues normally contain significant proportions of sulphur
and nitrogen, depending mainly on their source and the crude origin. Data from CONCAWE
show that the average SO2 content in the flue gas of refineries (from oil/gas firing) is 1350
mg/Nm3. 55 % of the flue gas from oil/gas that is burnt, is at less than 1000 mg/Nm3.
SOx (mg/Nm3) @ 3 % O2 Refinery fuel gas (*) Liquid fuel oil (#)
Process furnaces 3 - 1700 50 - 7000
Boilers 3 - 1700 50 - 7000
Turbines (@15 % O2) 3 - 1700 n.a.
* The lower range relates to natural gas firing. Higher value corresponds to
untreated refinery fuel gas without sulphur abatement.
# The lower range relates to combustion process of very low sulphur liquid
refinery fuel with abatement technique. Higher value corresponds to liquid
sulphur fuel (4.1 %) without abatement.
Table 3.50: Sulphur dioxide emissions ranges for existing European refineries
Sources: [45, Sema and Sofres, 1991], [247, UBA Austria, 1998], [198, (Hellas), 1999; 197, Hellenic
Petroleum, 1999; 199, Petrola Hellas, 1999]
In air-blown gasification of coke, the sulphur content will be about 4000 mg/Nm3 per percent
sulphur in the feed.
VOC
Fugitive VOC emissions may occur from storage of fuels as well as uncombusted fuels. No data
have been made available on that.
Waste water
Steam used for stripping, vacuum generation, atomisation and tracing is usually lost to waste
water and to the atmosphere. The waste water generated in the energy processes comes mainly
from the boiler feed water (BFW) system. The main streams are the boiler blowdown (1 to 2 %
of the BFW intake) and the BFW preparation regeneration wash (2 - 6 % of the BFW
produced). The main pollutants and composition of the first stream are COD: 100 mg/l N-Kj;
0 - 30 mg/l PO4; 0 - 10 mg/l. The BFW preparation regeneration stream is combined with
NaOH/HCl for pH neutralisation. Normally no biotreatment is required.
3.11 Etherification
Consumptions
The etherification reaction is exothermic and cooling to the proper reaction temperature is
critical in obtaining optimal conversion efficiency. Methanol is needed for the production of
those ethers. The utility consumptions in Table 3.51 are specified per tonne of MTBE processed
in an oxygenate unit.
Emissions
Air emissions
Potential releases into air are of hydrocarbons from pressure reliefs on vessels, depentaniser
column overheads drum and distillation column reflux drum, methanol plant, steaming vents on
scavengers and reactor catalyst.
Waste water
Potential releases into water are of hydrocarbons, methanol and ethers from spillages and water
bleed from methanol recovery. The water bleed is generated at a flow of 1 - 2 m3/t and with a
composition of COD: 50 - 200 mg/l and a N-Kj: 5 - 20 mg/l. Some components that can be
found in this waste water are methanol (ethanol), ethers and formic acid (acetic acid).
Emissions
Air emissions
Potential releases into air are of hydrocarbons from pressure reliefs on vessels, regeneration vent
gas from molecular sieves, C1/C2 refinery fuel gas, refrigeration system leaks, storage and
handling operations.
Mercury if present in the feedstock of the refinery, will concentrate in overhead sections,
particularly coolers. LPG, tops and naphtha coolers are most likely involved. Steaming that
contaminated equipment may give mercury emissions to atmosphere. Upon opening equipment
for inspection and maintenance mercury is sometimes found.
The final process in LPG production is the addition of some odorants. The odorants used are
typically volatile organic sulphur compounds, i.e. mercaptans and sulphides. Potential releases
into air include leaks or spillages of odorant and vapours displaced during tank filling or from
thermal expansion of blanket gas. Also the products of incinerating or flaring displaced vapours
if these techniques use used, including a small release of sulphur dioxide from combustion of
the odorant.
Waste water
Potential releases into water are spillages of hydrocarbons, H2S, NH3 and amines.
HYDROTREATMENTS
Consumptions of hydrotreatment units
The H2 consumption, and consequently the energy requirement, significantly increase in the
order naphtha (0.05 % H2), distillate (0.3 % H2) and residue hydrotreating (1.8 % H2). Next
table shows the utility requirements for different hydrotreatments.
Emissions to air
Air emissions from hydrotreating may arise from process heater flue gas (dealt with in
Section 3.10), vents, fugitive emissions and catalyst regeneration (CO2, CO, NOx, SOx). The
off-gas stream may be very rich in hydrogen sulphide and light fuel gas. The fuel gas and
hydrogen sulphide are typically sent to the sour gas treatment unit and sulphur recovery unit.
Hydrocarbons and sulphur compounds from pressure relief valves; leakages from flanges,
glands and seals on pumps, compressors and valves, particularly on sour gas and sour water
lines; venting during catalyst regeneration and replacement procedures or during cleaning
operations. The following tables show two examples of emissions from hydrotreating processes.
These air emissions include the emissions generated by the combustion of fuel required in those
processes.
HYDROCRACKING
Consumptions of hydrocracking units
Hydrocracking is an exothermic process. The heat generated in the reactors is partially
recovered in a feed/product heat exchanger. A significant amount of heat is required in the
fractionation section. Catalysts used in this process are typically cobalt, molybdenum, nickel or
tungsten oxides impregnated on alumina. Fluorided alumina, activated clay, silica-alumina or
zeolites are also present. Average consumption of catalysts is 57.4 t/Mt of feedstock. A small
amount of coke-like material is formed which accumulates over time on the catalyst. This
deactivates the catalyst, requiring regeneration off-site every one to four years by burning off
the coke. Water is removed by passing the feed stream through a silica gel or molecular sieve
dryer. The utility requirements for hydrocracking are the following:
Emissions to air
Heater stack gas contains CO, SOx, NOx, hydrocarbons and particulates that generate smoke,
grit and dust in flue gas (covered in Section 2.10), fugitive emissions (hydrocarbons) and
catalyst regeneration (CO2, CO, NOx, SOx, and catalyst dust). Fuel gas and bleed stream will
contain H2S and should be further treated. VOCs are generated by the non-condensable from
vacuum ejectors set condenser.
Waste water
Hydrocracking generates a flow of waste water of 50 - 110 l per tonne processed. It contains
high COD, suspended solids, H2S, NH3 and relatively low levels of BOD. The sour water from
the first stage HP separator, LP separator, and overhead accumulator should be sent to the sour
water stripper/treatment. Effluents from hydroconversion processes may contain occasionally
metals (Ni/V).
Solid wastes
Hydrocracking also generates spent catalysts fines (metals from crude oil, and hydrocarbons).
Catalyst should be replaced once per <1 - 3 years generating an average of 50 - 200 t/yr for a
refinery of 5 Mt/yr. Hydroconversion normally generates between 100 and 300 t/yr of spent
catalysts which contain more heavy metals than hydrocracking catalysts.
and on average those processes produce 2600 Nm3 of hydrogen (210 kg) per tonne of feedstock.
Catalyst for reforming is 25 - 40 % nickel oxide deposited on a low-silica refractory base. The
catalyst used in the reformer furnace is not regenerable and is replaced every four or five years.
A refinery with a capacity of five million tonnes of crude per year will consume 50 tonnes per
year.
The composition of the product depends on the purification technique, as shown in Table 3.58.
Coke gasification
Coke gasification produces between 2600 and 3500 Nm3 of hydrogen (210 - 300 kg) per tonne
of coke gasified. Figure 3.9 shows a scheme of the hydrogen production from a coke gasifier
and the amount of hydrogen produced per tonne of petroleum coke.
Petroleum
Shift and Acid Hydrogen
coke
Gasifier
2000 tonnes/day 100 tonnes/day
Gas removal
Sulphur removal
Methanol
Compression
260 tonnes/day
Synthesis
Steam and
Sulphator 100 MWe power
Power
or
Combustor
generation 300 tonnes steam/h
Bitumen and heavy fuel can be gasified. Bitumen characteristics that have been used in the
gasification are shown in next table
Table 3.59: Bitumen characteristics that can be used as partial oxidation feedstock
Source: [166, Meyers, 1997]
Emissions
Air emissions
Steam reforming
NOx emissions are the most important to consider. Other emissions such as SOx or water
emissions are minimal, because low-sulphur fuel is typically used and there are few emissions
other than flue gas. The choice of heat recovery system can have a major effect on NOx
production, since both the amount of fuel fired and the flame temperature will be affected. NOx
emissions from a steam reforming unit using gas or light gasoline as fuels and with low-NOx
burners are 25 - 40 mg/MJ (100 - 140 mg/Nm3, 3 % O2) [107, Janson, 1999]. Other emissions,
such as CO2, originate from carbon in the feed.
Coke gasification
Sulphur sorbents, such as limestone (CaCO3) or dolomite (Mg, Ca carbonate), are normally used
in the gasifier, reducing drastically the sulphur content. Sulphur composition in the exhaust gas
ranges from 600 to 1200 mg/Nm3 of H2S and COS. If no sorbent is used, the sulphur content of
the gas will be in proportion to the sulphur in the feed. In oxygen-blown gasification, the
sulphur content will be about 10000 mg/Nm3 per percent sulphur in the feed. Ammonia is
formed in the gasifier from the fuel-bound nitrogen. Ammonia in the product gas typically
contained less than 5 % of the fuel-bound nitrogen when limestone was present in the gasifier.
Solid waste
Coke gasification
The solid waste from the process consists mainly of spent limestone and metals from the
petcoke. The particulates in product gas are removed in the barrier filter to a level of less than 5
ppm. Volatile metals and alkalis tend to accumulate on the particulate as the gas is cooled. The
particulates contain a high percentage of carbon and are usually sent with the ash to a
combustor, where the remaining carbon is burned and the calcium sulphide is oxidised to
sulphate. In this hot-gas clean-up system, there is no aqueous condensate produced, although
some may be produced in subsequent processing of the gas. The solid waste from the process
consists mainly of spent limestone and metals from the petcoke. After processing in the
combustor/sulphator, this material is send to disposal. Vanadium and nickel are found in most
petroleum cokes in concentrations far greater than in coals or lignites. They remain in the ash
product.
Hydrogen purification
Several adsorber beds are used, and the gas flow is periodically switched from one vessel to
another to allow regeneration of the adsorbent by pressure reduction and purging, thus releasing
the adsorbed components. The desorbed gas is normally accumulated in a vessel and used as
fuel at a convenient location.
The breakdown of these annual emissions globally for the Western European sites is about:
§ 65 % from the combustion, representing a consumption of 28.5 % liquid fuel of 1.7 %S
content and 65 % of fuel gas
§ 13 % from the FCC units
§ 11 % from the sulphur recovery units
§ and the 11 complementary % from flare and miscellaneous sources.
Bubble values are based on calculated (*) yearly averages for all refineries participating: total
emissions divided by total (calculated stoichiometric required) flue gas emitted. Thus å(S unit1 +
S unit 2 + S FCC + S SRU + S FLARE) divided by å(Volume flue gas resulting from burning different
types of fuels). The results are shown in next table:
Bubble values (mg/Nm3) European ATL area MED area NWE area OTH area
Global average 1600 > 2000 > 2200 < 700 < 400
Combustion weighted 1000 1290 1540 570 310
average
Geographical contribution 100 % 26 % 25 % 39 % 10 %
3
of the areas (% FOE)
Expressed in term of “Global Average Bubble”, the mass of sulphur emitted per volume of flue gas
calculated for the reporting sites represents an annual Western European average concentration of
1600 mg/Nm3, all units of the refineries included for all types of refineries.
Expressed in term of “Combustion Bubble” - only due to fuels burned within the refinery site, thus
excluding FCC and SRU - the distribution of calculated annual value of each individual site is
reported in a graph according to their respective and cumulative contribution to the total fuel use in
Europe.
The geographical splitting in the three major zones confirms variation between areas,
representative of these different operational conditions through Europe.
The results of last three surveys from years 1992, 1995 and 1998 are given in Figure 3.10 a
weighted average for the annual bubble of 1000 mg/Nm3 is calculated representing a significant
decreases from previous surveys for which the average combustion bubble was 1350 mg/Nm3.
3
The four areas are :
- NWE - Northwest Europe (mainland France, United Kingdom, Belgium, the Netherlands, Germany and
Denmark)
- ATL - Atlantic (Ireland, Portugal and the Atlantic coasts of France and Spain)
- MED - Mediterranean (Spain, the med coast of France, Italy and Greece)
- OTH - “Others” (Norway, Sweden, Finland, Austria, Hungary and Switzerland).
7000
4000
3000
2000
1000
0
0 20 40 60 80 100
The lower part of the curves represents sites which are combining:
§ burning more gas than liquid fuels
§ lower sulphur content in liquid fuel used
§ high sulphur content marketed products
§ and/or control techniques of SO2 emissions.
Then it is a combination of what kind of operating window they have been running in
(crude/feedstock, supply and demand situation, plant design limitations, etc.) and the local
availability of refinery fuels and natural gas for burning in the refinery. It is only in this
complex context that any reference to “good” performance should be seen.
according to the type of unit and the purpose of the shutdown. If a particular unit is only to be
depressurised and heavy components flushed out for a temporary stop, the emissions will be far
less than if all equipment must be purged, steamed out and filled with air to allow workers to
enter. Stringent safety and health care procedures are normal practice in present-day refineries.
Not only own personnel but also contractors have to adhere to these procedures. Nevertheless,
occasional accidents take place and safety precautions have to be practised regularly. Shut-down
or blowdown can also have an impact on neighbouring dwellings (noise and light flaring).
UTILITIES MANAGEMENT
The daily sanitary waste water production of an individual is about 120 litres, so the total
sanitary effluent of the refinery can be easily calculated. Normally this domestic effluent is
collected in a septic tank and then treated in the waste water treatment plant.
Ballast water is relevant to those refineries that have crude receipt facilities or handle big
product tankers or inland barges. This ballast water can be high in volume and salt content
(seawater), and heavily polluted with oil. It can easily lead to an upset of existing effluent
treatment systems. The use of ballast water tanks is therefore an important equalisation tool for
feeding the water in a controlled way to either the process water system or the continuously oil-
contaminated system, if the COD is lower than 100 ppm. As more and more tankers are
equipped with segregated ballast tanks, the ballast water problem is slowly phasing out.
Water used in processing operations also accounts for a significant portion of the total waste
water. Process waste water arises from desalting crude oil, steam stripping operations, pump
gland cooling, product fractionator reflux drum drains and boiler blowdown. Because process
water often comes into direct contact with oil, it is usually highly contaminated.
The water usage depends both on purpose and complexity of the refinery. While there are
usually opportunities to economise on waster use, the scope for major reductions in existing
refineries is sometimes limited. Data from 63 European refineries gave the following data on
water usage (data in annual averages).
Seal drum blowdown generates an effluent of 1 - 2 m3/h and 10 times as much in case of
emergency, with a composition of: COD, 500 - 2000 mg/l, H2S, 10 - 100mg/l, phenol, 5 - 30
mg/l, NH3, oil and suspended solids, 50 - 100 mg/l
3.16 Isomerisation
The main environmental problems, for this process, apart from the energy consumption are the
catalyst promotion and the reduction of side-effects.
Consumptions
The total energy requirement depends mainly on the configuration (recycle operation 2 - 2.5
times higher than once-through), and the facilities to separate the n- from the iso-paraffins;
absorption/desorption process or a de-isohexaniser column. The isomerisation processes require
an atmosphere of hydrogen to minimize coke deposits but the consumption of hydrogen is
negligible. The utilities requirements for an isomerisation process are:
Catalyst life can range from 2 to 10+ years depending on unit operation. The chloride promoted
catalyst, chlorinated alumina containing platinum, requires the addition of very small amounts
of organic chlorides to maintain high catalyst activities, to avoid catalyst deactivation and
potential corrosion problems. See Section 2.16 for further information about the type of
catalysts used.
The next table shows the utility and chemical requirements for the adsorption process for a feed
rate of 600 tonnes per day.
Emissions
Air emissions
Air emissions may arise from the process heater, vents and fugitive emissions. Emissions from
process heaters are analysed in Section 3.10. Other emissions are HCl (potentially in light ends
from organic chloride added to increase catalyst activity), vents and fugitive emissions, pressure
reliefs on surge drum, separators and column reflux drum, regeneration vent gas from dryers,
storage and handling operations, spillage (hydrocarbons). Releases of hydrogen may occur from
pressure relief on the hydrogen system. Product stabilisation, however, does result in small
amounts of LPG (C3+C4, rich in i-C4) and in stabilizer vent (H2+C1+C2) products. The stabilizer
vent products are usually used as fuel. Fuel gas is normally caustic-treated to remove HCl.
(desiccant, water free CaCl2) contains dissolved CaCl2 and hydrocarbons. Its flow depends on
moisture content and is usually drained to the effluent treatment plant.
Emissions
Emissions to the air
The major emission sources in the natural gas processing industry are compressors, boilers and
furnaces, acid gas wastes, fugitive emissions from leaking process equipment and, if present,
glycol dehydrator vent streams. Regeneration of the glycol solutions used for dehydrating
natural gas can release significant quantities of benzene, toluene, ethylbenzene and xylene, as
well as a wide range of less toxic organics.
Potential release routes for prescribed substances and other substances which may cause harm
are shown in Table 3.64.
Table 3.64: Releases to the environment commonly associated with the natural gas processes
Source: [144, HMIP UK, 1997]
Emissions of SO2 will result from gas sweetening plants only if the acid waste gas from the
amine process is flared or incinerated. Most often, the acid waste gas is used as a feedstock in
nearby sulphur recovery (see Section 4.23.5.2) or sulphuric acid plants. When flaring or
incineration is practised, the major pollutant of concern is SO2. Most plants employ elevated
smokeless flares or tail gas incinerators for complete combustion of all waste gas constituents,
including virtually 100 percent conversion of H2S to SO2. Little results from these devices in the
way of particulates smoke or hydrocarbons and, because gas temperatures do not usually exceed
650 ºC, no significant quantity of nitrogen oxides is formed. Emission factors for gas
sweetening plants with smokeless flares or incinerators are presented in Table 3.65. Factors are
expressed in units of kilograms per 1000 cubic meters. Emission data for sweetening processes
other than amine types are very meagre, but a sulphur mass balance will give accurate estimates
for sulphur dioxide (SO2)
Emissions to water
Routine process-derived liquid effluent produced by natural gas plants is normally derived
mainly from formation water arising from the drying of the gas and associated condensate. The
effluent principally occurs as condensed steam from the glycol or methanol regeneration plants
and usually contains a number of organic contaminants which may include glycol, methanol,
aliphatic and aromatic hydrocarbons, amines and mercaptans. The presence of these
contaminants typically causes the effluent to have a very high biochemical and chemical oxygen
demand. The process of natural gas containing mercury can generate waste water containing
that metal.
Solid waste
Natural gas refining gives rise to relatively small quantities of process wastes. These typically
include occasional disposal of spent catalysts, activators, corrosion inhibitors, absorbents,
adsorbents, filter cartidges, separator dusts, etc.; they may be contaminated with trace amounts
of hydrocarbons. Some natural gas sources contain traces of Hg which should be collected
(generating spent adsorbent) and safely disposed off. For every 1000 kg of sludge treated, 12 kg
of metallic mercury is produced. Total sludge production within some Dutch natural gas plants
is on average 250 tonnes/year with a maximum of 400 tonnes/year. This corresponds to 3-5
tonnes of mercury annualy.
3.18 Polymerisation
Consumptions
The reactions typically take place under high pressure in the presence of a phosphoric acid
catalyst. Typically 0.2 gr. of phosphoric acid are consumed to produce 1 tonne of polymer or
1.18 kg of catalyst (phosphoric acid + support) per tonne of product. Catalyst life is normally
limited to 3 to 6 months, depending on unit size and operating conditions. A typical unit,
capable of producing 25 tonnes of polymerised gasoline per day, is customarily filled with 12
tonnes of catalyst.
Caustic solution is used to remove mercaptans from the propene/butene feed, then amine
solution to remove hydrogen sulphide, then water to remove caustics and amines. It is finally
dried by passing through a silica gel or molecular sieve dryer.
Emissions
Emissions to air
Potential releases into air are of hydrocarbons from pressure reliefs, storage and handling
operations, spillages and water and waste discharges, and of particulate matter from catalyst fine
from handling and disposal of spent catalysts. Sulphur dioxide and H2S may arise during the
caustic washing operation.
Table 3.67: Utility requirements for atmospheric and vacuum distillation units
Sources: [118, VROM, 1999], [261, Canales, 2000], [268, TWG, 2001]
Emissions
Air Emissions
Potential releases into air are from:
§ Flue gases arising from the combustion of fuels in the furnaces to heat the crude oil. Those
emissions are treated in Section 3.10.
§ Pressure relief valves on column overheads; relief from overhead accumulator are piped to
flare as well as the vent points.
§ Poor containment in overhead systems, including barometric sumps and vents.
§ Glands and seals on pumps, compressors and valves.
§ De-coking vents from process heaters. During furnace decoking (once or twice a year) some
emission of soot can occur if operation is not properly controlled in terms of temperature or
steam/air injection.
§ Venting during clean-out procedures.
§ Some light gases leaving the top of the condensers on the vacuum distillation column. A
certain amount of non-condensable light hydrocarbons and hydrogen sulphide passes
through the condenser to a hot well, and is then discharged to the refinery sour fuel system
or vented to a process heater, flare or other control device to destroy hydrogen sulphide. The
quantity of these emissions depends on the size of the unit, the type of feedstock and the
cooling water temperature. If barometric condensers are used in vacuum distillation,
significant amounts of oily waste water can be generated. Oily sour water is also generated
in the fractionators. The emissions of non-condensables from vacuum ejector set condensor
are between 50 and 200 kg/h, depending on furnace design and type of crude and
throughput. They are hydrocarbons and H2S.
§ Fugitive emissions from atmospheric and vacuum distillation units account for 5 - 190 t/yr
for a refinery with a crude capacity of 8.7 Mt/yr [79, API, 1993].
The following table gives some examples of the air emissions generated by the atmospheric and
vacuum distillation units by two European refineries. Those tables include the emissions from
combustion of fuels in the furnaces.
Table 3.68: Examples of air emissions generated by crude oil and vacuum distillation units
Source: [250, Winter, 2000]
Waste water (sour water) is generated in the vacuum distillation units from process steam
injection in furnace and vacuum tower. It contains H2S, NH3 and dissolved hydrocarbons. If
steam ejectors and barometric condensers are used in vacuum distillation, significant amounts of
oily waste water can be generated (± 10 m3/h) containing also H2S, NH3.
Catalytic dewaxing
Catalytic dewaxing catalysts typically last 6 to 8 years. During that time, they are regenerated as
needed. Typical cycles last 2 to 4 years between regenerations. Utilities needed in the catalytic
dewaxing process are summarised in the following table.
Emissions
Air emissions
Potential releases into air from oxidation-extraction processes are of hydrocarbons, sulphur
compounds (e.g. disulphides) and nitrogen compounds from sour water drains, pressure relief
valves, vents from feed vessels, spillages and leakages from flanges, glands and seals on pumps
and valves – particularly from overhead systems and fugitive emissions. Emissions may occur
from the spent air from the disulphide separator that contains less than 400 ppm of disulphides
and is normally incinerated.
Emissions
Air Emissions
Air emissions, and more specifically VOC emissions, are the main emissions that occur during
the storage and handling activities of refinery materials. Emissions from hydrocarbon liquids in
storage occur because of evaporative loss of the liquid during storage and as a result of changes
in the liquid level. Even if equipped with floating tops, storage tanks account for considerable
VOC emissions at petroleum refineries. Moreover, fugitive emissions in storage systems are
mainly from imperfect seals or from tank fittings. A study of petroleum refinery emissions
found that the majority of tank losses occurred through tank seals on gasoline storage tanks
[108, USAEPA, 1995]. However, it is generally the case that the amount of emissions depends
much more on the vapour pressure of the product (higher vapour pressure tend to produce
higher VOC emissions) than on the type of tank.
VOC emissions from storage represent more than 40 % of the total VOC emissions and are
normally the biggest emitter in a refinery. The emissions from tanks have been estimated by
using API estimations [245, API, 1983/1989/1990]. These estimations suggests an annual
emission of 320 tonnes per year from the storage systems, a much lower figure than the one
calculated from the DIAL measurements: 1900 tonnes per year for a refinery processing 11
million tonnes per year [107, Janson, 1999]. However, API methods have been updated and
storage losses can be predicted with acceptable accuracy [259, Dekkers, 2000]. A CONCAWE
study [229, Smithers, 1995] which was the only trial of DIAL where measurements were
conducted over a reasonably long period showed that there was good agreement between DIAL
and the latest API estimation methods when the test times were long enough.
When transferring liquids to vessels at atmospheric pressure, the vapour phase (often air, but
also inerts) in the receiving vessel is often emitted to atmosphere. Such loading operations are
recognised as having an impact on the environment due to the presence of VOC. VOC
emissions from blending come from blending tanks, valves, pumps and mixing operations. The
amount depends on design and maintenance of the systems. Although barge loading is not a
factor for all refineries, it is an important emissions source for many facilities. One of the largest
sources of VOC emissions identified during the Amoco/USAEPA study was fugitive emissions
from loading of tanker barges. It has been estimated that during loading of products with high
partial pressure, 0.05 % of the throughput can be emitted to the air.
VOC emissions to the air from cavern storage of hydrocarbon materials can be prevented
connecting the gas spaces of several caverns. So the breathing gases when filling one cavern are
leaded to the other caverns. In case of only one cavern the breathing gases must be leaded to the
air. However the VOC-emissions in this case are low because the temperature of storage is low
(5 - 10 ºC).
Waste water
Leaking flanges and valves may contaminate rainwater. The amount depends on the
maintenance service. Liquid tank bottoms (primarily water and oil emulsions) are periodically
drawn off to prevent their continued build up. That waste water is contaminated with tank
product. Oil levels up to 5g/l in tank water bottoms can be found [101, World Bank, 1998].
However, if the draw off rate of the water is too high, oil can be entrained with the water,
thereby giving a false high reading of oil in the water.
The groundwater leaking into underground storage systems (caverns) should be pumped out and
led to the waste water treatment system of the refinery. The amount of the water depends on the
solidity of the rock and how carefully the fractures in the rock are sealed by injection of
concrete [256, Lameranta, 2000]. The quality of waste water depends on the product (or crude)
wich is stored in the cavern. Typically the waste water contains hydrocarbons as emulsion and
water soluble components of stored liquid. Some examples of waste water generated by cavern
storage is given in the next table:
Table 3.71: Composition of the sludge taken from four different sewer systems close to the storage
tanks
Source: [80, March Consulting Group, 1991]
3.22 Visbreaking
Consumptions
The following represents typical utility consumptions for a visbreaker:
For visbreaking, fuel consumption accounts for about 80 % of the operating costs with a net fuel
consumption of 1 - 1.5 %w/w on feed. Fuel requirements for soaker visbreaking are about
30 - 35 % lower.
Emissions
Air Emissions
Air emissions from visbreaking include emissions from the combustion of fuels in the process
heaters (discussed in Section 3.10), vents and fugitive emissions. A sour water stream is
generated in the fractionator. The gas produced contains H2S and should be further treated.
Hydrogen sulphide and mercaptans are released from the sour water stream from reflux
condensers. Hydrocarbons are released from pressure reliefs on reflux drums and vessels,
storage and handling operations, spillages and discharges of waste/water. Particulate emissions
occur during furnace decoking and cleaning operations about twice a year.
The next table shows the air emissions from visbreaking processes in two European refineries.
The flow of waste water generated in the visbreaking process from the sour water drain is
around 56 litres per tonne of feedstock. It represents between 1 and 3 percent volume on feed
intake. Table 3.74 shows the range of composition of the visbreaker waste water.
Next sections cover the emissions arising from refinery processes that are used for the treatment
of waste gas, waste water and solid waste. Those techniques are described in Chapter 4
(Sections 4.23 to 4.25). The main objective of those processes is to reduce the concentration of
pollutants in the effluents. Obviously the loads and concentrations of pollutants in those streams
will be reduced by those processes, but some pollutants will remain in the streams and other
may be generated. That is the purpose of the following next three sections. Emissions from the
refinery as a whole are included in Section 3.1.
Flares
Consumptions
The main utilities that flare systems will use are refinery fuel gas or nitrogen for purging the
flare headers, fuel gas for the pilot burners and steam for heating purposes. However, the
consumption of these utilities is very low under normal operating conditions.
Emissions
They emit CO, CO2, SO2 and NOx. The composition depends on the system being relieved and
combustion efficiency at the flare tip. Flares are a very important point source of refinery
emissions. Some plants still use older, less-efficient waste gas flares. Because these flares
usually burn at temperatures lower than necessary for complete combustion, larger emissions of
hydrocarbons and particulate, as well as H2S, can occur. No data are available to estimate the
magnitude of these emissions. Emissions of VOC and soot are also present and they depend on
the system being relieved and combustion efficiency at the flare tip.
Process effluent volume can be compared to the volumes of process effluent discharged
(excluding once through cooling water). Although related, the values are not identical given that
there are other sources of water to the effluent system including rainwater on plant surfaces,
water separated from crude, ballast water etc. On the other hand, some water will evaporate, be
used in chemical reactions, etc. Figures for effluent for 63 European refineries are (all in annual
averages):
Emissions
Air emissions
The air emissions from waste water treatment plants stem from evaporation (hydrocarbons,
benzene, H2S, NH3, mercaptans) from the numerous tanks, ponds and sewerage system drains
that exposes its surfaces into air. The sewerage system and the waste water treatment in a
refinery may be sources of odour nuisance, particularly from open drains and from oil
separators. VOCs are also emitted during air stripping in flotation units and in the biotreater.
The HC emissions from the waste water systems can be determined by calculation from the
exposed surface area of the oil-contaminated untreated water tank (API separator) and an
empirical oil evaporation factor of [117, VDI, 2000]:
Emissions to water
Potential releases into water consist of the residual substances contained in the effluent after
treatment. Depending on the nature of the refinery and the treatment efficiency, these include
hydrocarbons (dissolved and suspended), organic compounds (notably phenols), sulphur
compounds (notably sulphides), ammonia/ammoniacal compounds and their derivatives after
treatment. Traces of heavy metals are also likely to be present, arising from the desalting of
crude oil feedstock and from caustic soda if this contains trace contaminants, notably mercury.
More information in this subject is included in Section 3.1.3.
year with a unit operating at 600 m3/h. Table 3.76 shows the composition of the different
sludges generated in the waste water treatment plants.
The amount of sludge generated depends on the types of processes and the availability of
incineration. As a common figure, the generation rate of solid waste and sludges is normally
less than 0.5 % of crude processed, but in some refineries is less than 0.3 %. Annual sludge
generation was 1250 kt per year (in 1993), some 0.2 % of refinery intake.
In 1993, CONCAWE carried out a survey of waste in Western European refineries [82,
CONCAWE, 1995]. In this survey 89 refineries reported their sludge production from all
sources as one million tonnes. That means an average of 11000 t per year per refinery. This
figure represents a 0.20 % of the crude oil processed by those refineries. However, a variation
range from 2.5 % to less than 0.02 % was found in European refineries, with the value being
independent of the type of refinery. From that million tonnes of waste, 45 % was sludge, 20 %
other refining wastes and the rest non-refining wastes. Sludges come from stabilisation
processes (4.7 % of the total sludge generated after treatments), waste water sludges (39.8 %)
and sludges with no treatment (55.5 %). The total amount of identified other refining wastes
which are specific to the refining process (e.g. spent catalysts, tank scales, contaminated soils,
etc.) produced in 1993 by the 89 European refineries reporting was 201983 tonnes (i.e. 0.04 %
w/w of the total refinery throughput). Data provided by 16 EU+ refineries show that the specific
production of waste ranges from 133 to 4200 t/Mt of crude. Next table shows an example of the
waste production of a refinery.
Spent catalyst
Hydroprocessing requires the use of catalysts to remove impurities and to convert the oil into
more useful products. The catalysts used consist mainly of oxides of nickel, cobalt and
molybdenum on an alumina carrier. During hydroprocessing operations the catalysts are loaded
with sulphur, vanadium and coke. The porosity of the catalysts decreases and their activity is
reduced, which eventually requires replacement of the catalyst in the reactors. Consequently, the
catalysts become spent and must be replaced.
Ni/Mo catalysts are typically used in hydrotreaters and hydrocracking units. A 5 Mt/yr refinery
generates typically 20 - 100 t/yr spent Ni/Mo catalyst.
Ni/W catalysts are used in lube oil hydrofinishing. This category experiences limitations to
dispose off in view of the high Tungsten content (24 % w/w). A 50000 t/yr Lube oil plant can
generate 50 t/yr of this catalyst waste.
FCC spent catalysts, also including heavy oil and residue cracking spent catalysts (RCC), are
the largest catalyst waste category in refineries (world production about 0.5 Mt/yr). A 1 Mt/yr
FCC unit generates 400 - 500 t/yr predominantly as FCC fines if an ESP is installed. For a RCC
this amount can be a factor 5 - 10 higher, depending on the feed as spent catalyst is withdrawn
from the regenerator.
Reformer and Isomerisation catalysts are exclusively reprocessed by the suppliers of the fresh
catalysts. Replacement contracts have been concluded since the introduction of these processes
due to the very expensive noble Pt metal involved. Typical average amounts involve 20 - 25 t/yr
for a 5 Mt/yr refinery. If necessary guard beds of HDS catalysts are used for protection and
extension of the life-time of the expensive catalysts.
Hydrodemetallisation catalysts typically have high vanadium contents (10 - 20 %) and are
currently alumina based (used to be silica). The annual amount for regeneration is in the order
of 500 - 1000 t/yr for Hycon (Shell Pernis NL) dependent on the feed quality.
Zn containing beds from H2 plants are typically recycled to the Zn-industry where ZnS ores are
processed. Amounts are some 50 t/yr.
Process S C Mo V Ni Co Al Other
FCC, RCC <1 <1 4 - 8000 2 - 3000 30
* *
Cat.Reforming 30 0.5 Pt,
& Pa, Rh
Isomerisation
Hydro 6 - 16 10 - 30 4-8 2 - 12 1-2 1-2 20 - 30
processing
Claus plant 5 5 4-8 2-3 20 - 30
Hydrofinish 5 1-2 2-4 30 24 W
Hydrogen 5 - 15 0 30 Zn
Hydrodemetal. 5 - 15 10 - 30 10 - 20 2-5 30
(*: ppm)
3.26 Monitoring
There is an IPPC BREF and an expanding series of literature (e.g. HMIP Technical Guidance
Notes) on monitoring, containing information on theory and methodology, and providing
technical information on monitoring techniques. It should be considered that the requirements
and frequency of sampling, analysis and type of monitoring requirements are site and/or process
specific, influenced by consideration of the anticipated volume and composition of the waste
streams and normally are included in the permits. Monitoring BREF includes guidances to
permit writters as compliance assessment and environmental reporting. That document also
encorage the comparability and reliability of monitoring data across Europe.
This section aims to give recommendations on the scope and frequency (continuously or
discontinuously) of monitoring streams in a refinery. Monitoring should be undertaken during
commissioning, start-up, normal operation, and shutdown unless it is agreed that it would be
inappropriate to do so.
The monitoring system should allow adequate processing and emission control. Some elements
of a monitoring system could include:
§ continuous monitoring of pollutants for high volume flows with a high variability in
pollutant concentrations
§ periodic monitoring or the use of emission relevant parameters for flows with a low
variability
§ regular calibrating of measurement equipment
§ periodic verification of measurement by simultaneous comparative measurements.
In order to have a good idea of the emissions generated by an installation (e.g. a refinery), the
emissions need to be quantified. This will give to the refinery and permit writter the possibility
to know where the actions to reduce a certain pollutant may be more economical. Typically
higher emissions at high concentrations are more economic to be abate that lower emissions at
low concentrations. As consequence, a quantification of the emissions for each single refinery is
the first starting point in any environmental assessment. This quantification may contain a full
mass balance also considering other outpus (e.g. products).
SO2, NOx, particulate matter, and CO are typically monitored in continuous (on-line or
predictive) in the refineries. Records of the volumes are also required for the calculation of the
load (tonnes of pollutant per year) or for the application of the bubble concept. Following table
shows the locations where the main air pollutants are typically continuous monitored in a
refinery.
Next information gives some good monitoring practices that have been found in European
refineries.
Sulphur monitoring
It can be expected from the operators to compute site sulphur balances over appropriate time
periods as a part of monitoring their operations; the time period may vary with the
circumstances (e.g. how often feedstocks change) but could typically be quarterly. In the case of
SOx, the amount of SOx emitted by combustion processes may be calculated from the analysis
of the fuel being burnt.
Additionally, flare systems can be equipped with adequate monitoring and control systems
necessary to operate smokelessly and should be observed at all times under non emergency
conditions. Such systems can include flow measurement, for which proven non-intrusive
systems at the flare base are available, with automatic steam control, luminosity measurement
with automatic steam control, remote visual observation using colour TV monitors in relevant
plant control rooms with access to steam control, pilot flame detection, etc.
Gaseous releases from gas turbines used for electrical generation or gas compression are
typically also monitored.
VOC monitoring
VOC emissions come mainly from fugitive emissions. The site may be equipped to monitor all
process vents and possible sources of fugitive losses. A Leak Detection and Repair (LDAR)
programme can be implemented (see Section 4.23.6.1) in order to prevent those. An assessment
of the total hydrocarbon emissions from the site on a regular basis can be made. Records can be
kept of the individual contributions to each total.
VOC emissions can be estimated by USAEPA Method 21 (emission factors for different type of
equipment) or by mass balance (feedstock – products). Perhaps one should add that another way
of monitoring HC emissions from all sources is through the mass balance between feed stock
and products.Other methods have been developed and proven in recent years for the monitoring
of VOC emissions as well as locating sources of VOC emissions, typically using laser
absorption techniques (DIAL) of considerable sophistication. Mobile versions are available
which can be used to plot pollutant concentration plumes and calculate pollutant flux, including
three-dimensional resolution. Whilst having recognised limitations, they have been used
successfully in the oil and gas industries. In most of this monitoring campaigns the
measurements have been divided into four areas: Process area, product tank farm, crude oil tank
farm and water treatment plant. DIAL method has been used in a number of refineries and in
many cases, the results appear to be higher than those derived from the API estimations.
However, in these trials, the sampling period was very short. In the only trial where a long-term
survey was conducted, the results gave reasonable agreement with the API estimations [229,
Smithers, 1995]. Next paragraph gives an example of monitoring of VOC within a refinery as
well as the discrepancies that they have found using estimation method and DIAL measurement.
Before the measurements, it was estimated a VOC emissions of about 200 tonnes per year at a
certain European refinery. The DIAL measurements and annualisation of the results suggested
a yearly emission exceeding 1000 tonnes. Especially emissions from tanks were higher than
expected. In all cases, emissions estimates using API methods gave lower emissions than
estimates based on the DIAL monitoring. In some cases, the discrepancies are very large. By
using the method for estimating fugitive emissions proposed by USAEPA [244, USAEPA, 1992]
the emissions from the process area 1Mtonnes per year refinery have been estimated to 125
tonnes per year. Extrapolations of the DIAL measurements to a yearly emission give emissions
of 500 - 600 tonnes per year [107, Janson, 1999]. For the total emission from the refinery, the
use of estimation methods suggests that the emissions are 600 - 1100 tonnes per year.
Extrapolations of DIAL measurements indicate a yearly emission of 1600 - 2600 tonnes for
medium refineries. During the same exercise of measurements it was also found that the
dominating fraction of fugitive emissions was alkanes within interval C8 - C10. Aromatics
contributed to about 9 - 15 % of the total emission [107, Janson, 1999]. However, it is important
to notice that the composition of the vapour will depend heavily on refinery configuration and
what happens to be leaking at the time of the measurement of the VOC.
Monitoring of process effluents released to controlled waters and sewers will commonly be
made for the following: flow rate, pH, temperature, TOC (surrogate for COD/BOD). Samples
are also being monitored for appropriate other parameters such as COD, BOD, hydrocarbon oil,
ammoniacal and total nitrogen, suspended solids, phenols, sulphides, dissolved oxygen,
phosphates, nitrates, nitrites, metals (typically Cd, Hg, Cr, Ni, Zn, Cu, As). The periodicity can
be typically daily, weekly or monthly depending on risk assessment and on local circumstances.
One problem about water discharge from refineries is analysis and especially analysis of oil.
There are a number of different methods (e.g. IR one wave lengths, IR two wave lengths,
gravimetric) which give completely different results. Another problem is the toxic substances
used in the COD analysis.
Losses of oil through leaks can also be minimised through continuous monitoring of the cooling
water system. At its simplest, this comprises monitoring of the cooling water separators for oil
build-up. If this is observed, then it will be necessary to trace back through the system to
identify the source of the leak so that corrective action can be taken. Detailed system drawings
are essential for this activity. Fingerprinting of the oil can also speed up identification of the
leak. A further refinement is to install oil in water monitors at various points in the cooling
water system. This allows leaks to be detected rapidly and corrective action taken. For this
procedure to be effective, sparing of critical heat exchangers will be required.
In case of a once-through system the cooling fluid outlet system may be divided into a suspect
system, where the process pressure is higher than the cooling system pressure and a clean
system, where the cooling system pressure is higher than the process pressure. The suspect
system needs to be monitored for contamination.
Soil Monitoring
Quoted costs for installation and operation of monitoring wells was highly variable ranging
from 25000 EUR for 100 wells (250 EUR per well) to 140000 EUR for 50 boreholes
(2800 EUR per well) but with a minimal operating cost. Costs of groundwater monitoring
systems were quoted as 1400 EUR per year for 18 wells (78 EUR per well) in one case to an
average of 45000 EUR (including 7000 EUR for analysis) for three surveys at intervals of some
years.
As in previous Chapters 2 and 3, each section of this chapter deals with a refinery process or
activity and contains the process and abatement techniques worth considered in the
determination of BAT. If different techniques may be applicable for one process/activity, they
are discussed there. Sections from 4.2 to 4.22 in this chapter have been structured internally in
the same way. Each of those sections contain first the pollution prevention techniques
applicable in the process/activity section referred to and secondly the end-of-pipe techniques
that may be applicable to reduce the emissions coming from the process/activity. Those end-of-
pipe techniques are grouped in media/pollutant bases to clarify the sequence of techniques
applicable as, in some cases, the number of EOP techniques that may be use is quite extensive.
At the end of this chapter, three sections appear containing the end-of-pipe (EOP) techniques
applied to waste gas, waste water and waste management. Those sections contain the
descriptions of the EOP techniques that may be applicable to more than a refinery
process/activity as well as some other EOP processes. Consequently, the descriptions of those
EOP techniques that may appear in this chapter are included in Sections 4.23, 4.24 and 4.25 and
not in the activities/process sections.
Each technique (both preventive (including process techniques) and control) is well documented
and includes information on consumption and emission levels considered achievable by using
the technique, some idea of the costs and the cross-media issues associated with the technique
and the extent to which the technique is applicable to the range of installations requiring IPPC
permits, for example new, existing, large or small installations. Each technique in this chapter is
analysed in the same way in order to give a good background for the determination of BAT for
the refinery sector. That determination is done in the next chapter. Table 4.1 shows the structure
of the information for each technique included in this chapter.
When possible, this chapter provides information from real activities that can be, or are being,
implemented by this sector, including associated costs. When possible, information provided
gives the context in which the technique can be used effectively.
Table 4.1: Information contained in the description of each technique included in Chapter 4
Techniques applied to
Chapter Activity/Process production TOTAL
section and gases and waste solid
prevention waste gas water waste
2 Alkylation 3 0 0 0 3
3 Base oil production 14 4 2 1 21
4 Bitumen production 2 5 1 2 10
5 Catalytic cracking 17 13 2 5 37
6 Catalytic reforming 3 3 0 0 6
7 Coking processes 9 19 8 3 39
8 Cooling 3 - - - 3
9 Desalting 13 0 4 1 18
10 Energy system 56 22 2 0 80
11 Etherification 1 0 1 1 3
12 Gas separation processes 3 2 0 0 5
13 Hydrogen-consuming 8 0 0 2 10
processes
14 Hydrogen production 6 0 0 0 6
15 Integrated refinery 33 0 24 6 63
management
16 Isomerisation 3 0 0 0 3
17 Natural gas plants 0 12 5 3 20
18 Polymerisation 1 0 0 2 3
19 Primary distillation units 3 2 3 3 11
20 Product treatments 5 2 4 0 11
21 Storage and handling of 21 19 2 12 54
refinery materials
22 Visbreaking 3 1 1 1 6
23 Waste gas treatments - 76 - 1 77
24 Waste water treatments - - 41 - 41
25 Waste management - - - 58 58
TOTAL 207 180 100 101 588
4.2 Alkylation
The techniques considered in the determination of BAT for the alkylation process are the
following:
Description
A brief description of the process can be found in Section 2.2.
Cross-media effects
Effluent gases: Scrubber using alkaline solution (NaOH or KOH) is necessary to remove HF
from the incondensable gas stream. The acid relief neutraliser is operated so as to minimise the
hydrogen fluoride content of the incondensable gas stream. Emissions levels of <1mg HF/Nm3
can be achieved. The vent gas should pass to flare not to the refinery fuel gas system; a
dedicated flare/stack is normally retained for this. Fugitive emissions are also generated by this
process.
HF is a very dangerous compound because of its severe corrosive nature and burning effects of
both liquid and fumes to skin, eyes and mucous membranes. Consequently, storage and
handling it should comply with all safety rules.
KF (or NaF) is formed during the neutralisation process. The spent solution is stored and then
requires regeneration with lime (or alumina). The regenerated KOH (or NaOH) is recycled.
CaF2 is regularly cleaned out and disposed of, usually to a landfill. If alumina is used instead of
lime as the neutralising agent, then, recovery of the aluminium is possible by the aluminium
industry. Fumes can originate in the unit’s neutralising basin. To prevent the discharge of these
odorous gases to the surroundings, the neutralising basins are tightly covered and equipped with
a gas scrubber to remove any offensive agent.
Water: HF alkylation effluents are a potential cause of acid excursions in refinery effluents and
a high standard of control should be exercised on the neutralisation treatment system, e.g. on-
line pH monitoring. The effluent containing HF acid can be treated with lime (CaO-Ca(OH)2),
AlCl3 or CaCl2 or it can be neutralised indirectly in a KOH system to produce the desired CaF2
or AlF3 (insolubles) which is separated in a settlement basin. The supernatant liquor flows to the
water treatment process. KOH may be regenerated for re-use. After this treatment, the
supernadant contains still 10 – 40 ppm F and some hydrocarbons and is directed to the WWTP.
KOH may be regenerated from the aqueous solution for re-use.
Wastes: The HF process also yields tars (polymeric material) but these are essentially free from
HF. HF-containing tars are neutralised (with lime or alumina) and disposed of by incineration or
blended as a fuel-oil component in small amounts because its pronounced odours. However,
technology and special operating techniques such as internal acid regeneration have virtually
eliminated this liquid-waste stream.
Bad operation of the system (excessively cold acid, improper acid settler level, low reactor
pressure, excessive acid recycle rate, low i-butane or acid concentration, high reactor
temperature and improper mixing) increases the amount of waste generated.
Operational data
The potentially corrosive and toxic properties of HF require special engineering design and
special construction materials, advanced process controls and stringent safety, personal
protection and maintenance requirements. These precautions include:
a) special seals on acid-containing equipment such as pumps and valve stems
b) special safety clothing which must be worn by personnel entering an acid area
c) extensive operator training
d) special facilities (e.g. sprinkler systems, acid-inventory reduction, HF-detection systems,
isolation valves, rapid acid transfer systems) should be included in the design of this kind of
system to mitigate the effects of an accidental emission of HF vapour clouds. For leak
detection, HF sensitive colours on pipes and flanges are suitable as alternative to continuous
air analysers and video control. The potential reduction of airborne HF acid has also been
demonstrated by the addition of additives. Special care should be taken during maintenance
and cleaning of the plant, e.g. for major shut-downs not to release odours from the drainage
system and/or ponds.
Same developments (such as additives) have been already tested in order to lower the vapour
pressure and to minimise the amount of HF released to the atmosphere in case of accident. On
the downside of this is that handling and recyling the additive add to the complexity of the
process.
Applicability
Fully applicable.
Economics
The next table shows the investment and production cost summary for HF alkylation. Disposal
costs are not included.
Example plants
Common technique.
Reference literature
[166, Meyers, 1997], [113, Noyes, 1993], [261, Canales, 2000], [330, Hommeltoft, 2000], [268,
TWG, 2001]
Description
A description of the process can be found in Section 2.2.
Cross-media effects
Technologies using suphuric acid as catalyst produce very large quantities of spent acid
(sulphuric and sulphonic acids) that has to be regenerated. The transport of spent and fresh acid
to and from the sulphuric acid regeneration has give rise to some concern and increased the
pressure on refiners to establish sulphuric acid regeneration plants near the alkylation unit. In
some cases this transport to/from the regeneration facility is by pipeline. However, no major
new improvements have been introduced in sulphuric acid alylation technology dealing with the
spent acid issue. Fugitive emissions from this process is similar to the HF alkylation.
Acid-soluble oil should be recycled to process. Tars from the H2SO4 process contain sulphuric
and miscellaneous sulphonic acids, which may cause disposal problems. This process generates
waste water that should be neutralised before be routed to WWTP, gum, spent acid, caustic
solutions from regeneration and fugitive emissions.
Operational data
The product quality from butyplenes (primary feedstock) is higher than in HF alkylation. To
produce an equivalent product quality, HF alkylation would require similar energy consumption
to sulphuric acid alkylation. This type of alkylation is considered relatively safer when
compared to HF catalysed alkylation.
The sulphuric acid removed need to be regenerated in a sulphuric acid plant which is generally
not a part of the alkylation unit and is normally located off-site.
Applicability
Fully applicable. This type of process technique is driven by the economics of the proximity of
regeneration facility of sulphuric acid.
Economics
Example plants
Commonly used.
Reference literature
[166, Meyers, 1997], [212, Hydrocarbon processing, 1998], [330, Hommeltoft, 2000], [268,
TWG, 2001]
Description
More information about the processes can be found in Section 2.13, 4.13.5, 2.16 and 4.16.
Cross-media effects
Drawbacks for the implementation of such technique is the increased fuel/energy consumption,
the increased fugitive emissions and need for spent catalyst handling used in such a processes.
More information about those issues can be found in Sections 4.13 and 4.16.
Operational data
It requires hydrogen and consumes energy.
Applicability
No restrictions for application.
Economics
See Section 4.13 and 4.16.
Reference literature
[113, Noyes, 1993], [268, TWG, 2001]
Description
Solvent based processes used in the manufacture of base oils are energy intensive because large
volumes of solvent must be recovered by flash distillation for recycle in the process. Number of
stages used for evaporation of the solvents has a significant effect on the energy cost for these
processes and as many as five evaporation stages were used in some early liquid SO2 extraction
units in Europe.
Operational data
Utilities needed in the process are listed in the economic section. The energy saving of the triple
effect systems is on the low level heat side only (in most cases LP-steam) where refineries often
already have a surplus. This means that the actual saving depends on the local circumstances
and a retrofit/change should be evaluated for each individual location.
Applicability
Multiple effect evaporation requires both a pressure and a temperature driving force. In some
instances, the increase in pressure and temperature required to go from double to triple effect is
not feasible and could result in higher emissions and energy consumption because of larger
residual amounts of solvent left in the oil leaving the evaporation section and entering the
stripping section.
Triple effect systems are typically used for non-fouling feedstocks only (e.g. wax) as they are
known to be more vulnerable in this respect, so their application is restricted.
Economics
Investment (basis: 318 - 6360 m3/d, 4th Q, 1998, US Gulf): 5000 - 18900 $ per m3/d.
Example plants
Common technology used in refineries having lubricant oil production. Double effect units are
the most predominant type of extraction processes. Units built in 1950-1975 used double effect
and a few used single. Because the cost of energy increased rapidly during the 1970, most new
units built since about 1980 have been designed (and older units converted to) triple effect
evaporation to reduce the cost of energy consumed.
Reference literature
[166, Meyers, 1997], [212, Hydrocarbon processing, 1998], [319, Sequeira, 1998]
Description
The selection of the type of solvent used in the aromatic extraction has an impact on the energy
consumption of the system and the use of a less toxic solvent (furfural and n-methyl pyrrolidone
(NMP) can be used in preference to the more toxic selective solvent as phenol and sulphur
dioxide).
The advantage of NMP versus furfural is a higher selectivity leading to a higher raffinate yield
and a lower solvent ratio which both result in a lower energy consumption of some 30 - 40 %.
Cross-media effects
Care should be taken to avoid solvent contamination of aqueous effluents by applying efficient
solvent recovery techniques. Some contamination of aqueous effluents occurs, so care should be
taken not to upset the activated sludge treatment unit of the waste water treatment. NMP is
easier to be degraded in the WWTP than furfural however is more corrosive than clean furfural.
Operational data
The performance of NMP and furfural techniques are similar. Well maintained and operated
furfural extraction units perform just as well as NMP plants. Similar quantities of furfural,
compared to NMP, are required when the plants are run in the correct manner.
Applicability
Some technical dificulties for applicability or change of solvent are the following:
• phenol is easier to recover than furfural. Furfural forms a different ratio azeotrope with
water, so a straight switch of solvent of a phenol unit may not possible, much larger
quantities of furfural are used per unit throughput
• NMP switching of phenol is also difficult. NMP has a higher boiling point (by 22 ºC), a
lower melting point (by 64 ºC) and no azeotrope. Because those reasons, furfural is
typically used when producing lower boiling point base oils (e.g. transformer oil distillates)
Major modifications would be required for existing units to change from furfural to NMP as the
process conditions are different (change from hot oil heat exchangers to a fired furnace to
achieve the required temperatures in the solvent recovery section).
Economics
The change of solvent used does not entail significant extra costs in some cases. In other cases,
for instance, the change from a SO2 aromatic extraction to another solvent needs a completely
new unit. Cost comparison of major base oil refining processes is given below
The switch from furfural to NMP process need economical calculations because major
modification may be required. As consequence, the energy saving against the cost of the revamp
has to be carried out individually. Changing from furfural to NMP requires significant
investment, including, but not limited to the extraction tower, furnaces, metallurgy etc. Cost for
conversion of a phenol unit (277 m3 per operating day) to a NMP unit (563 m3 per operating
day) is estimated in EUR 0.9 million (cost including solp and solvent tankage). Other estimates
says that the investment cost of a switch of a phenol units to NMP unit is around 70 % of a
completely new unit.
Example plants
Common technologies.
Reference literature
[110, HMIP UK, 1993], [319, Sequeira, 1998], [268, TWG, 2001]
Description
Information about the process can be found in Section 2.3. Different prevention techniques may
be applied to the solvent dewaxing process:
• vapours from the wax filters are contained in the inert gas system. In this process the inert
gas is circulating and solvent vapours are recovered by refrigeration as part of the process.
Solvent incineration in furnaces should be minimised. Alternative to refrigeration is
pressure absorption of solvents in oil (e.g. fresh feed or finished)
• solvents used for the process include propane and mixtures of methyl ethyl ketone (MEK)
with methyl isobutyl ketone (MIBK). Toluene or chlorinated hydrocarbons should be
avoided
• the solvent recovery stage results in solvent-contaminated water, which should be sent to
the waste water treatment plant
• use of inert gas in place of steam for stripping the last traces of solvent from the dewaxed
oil and waxes. When chlorinated solvents are used this procedure is typically applied.
Cross-media effects
Large refrigeration systems are used and refrigerant losses from these should be minimised.
Energy consumption for refrigeration / pressure and heat. Care should be taken to avoid solvent
contamination of aqueous effluents because most of the solvent used in solvent dewaxing are
problematic for water micro-organisms and waste water plants. VOC fugitive emissions may be
generated during the processes.
Operational data
Applicability
Fully applicable.
Economics
A new solvent recovery unit cost 66000 US$ per m3/day (basis 44000 m3/day feed rate capacity,
1998 US Gulf Coast). Data on implementation of inert gas stripping in refineries shown
paybacks of 9-14 months.
Example plants
Many refineries having lubricant production contain this type of processes.
Reference literature
[212, Hydrocarbon processing, 1998], [19, Irish EPA, 1993], [319, Sequeira, 1998], [268, TWG,
2001]
Description
Hydrofinishing or clay treating can be used in the wax processing.
Cross-media effects
When hydroprocessing is used, hydrogen is necessary. Hydrotreatments consume energy,
hydrogen and the nickel, tungsten or even platinum catalyst needed for this process. Spent clay
results in a disposal problem.
Operational data
Hydrogen finishing yields approach 100 % whereas the clay treating product yield ranges from
75 to 90 % for the microcrystalline waxes and as high as 97 % for light parafin waxes.
Hydrogen consumption is around 15 Nm3 per m3 of wax produced. Clay consumption runs from
2 to 45 kg/t of product.
Applicability
Hydrofinishing is fully applicable especially when hydrogen is available. However, in some
cases, hydrofinishing processes cannot reach certain very special product quality specifications.
Economics
Investment cost for a wax hidrofinishing process is EUR 1 million for a feed capacity of 20000
t/yr.
Example plants
Many examples exists.
Reference literature
[319, Sequeira, 1998], [268, TWG, 2001]
Description
A short description of the process can be found in Section 2.3. The use of hydrotreating or
sulphuric acid and clay treatments are treatments used in the lube oil clean-up.
Cross-media effects
Hydrotreatments consume energy, hydrogen and the nickel, tungsten or even platinum catalyst
needed for this process.
Operational data
In stand-alone luboil refineries, the small amounts of H2S generated due to hydrotreatment are
normally incinerated because the erection of a Claus unit is not justified economically.
Applicability
Fully applicable. However, in some cases, hydrofinishing processes cannot reach certain very
special product quality specifications.
Example plants
Sulphuric acid and clay treatments are typically used in old plants.
Reference literature
[19, Irish EPA, 1993], [268, TWG, 2001]
Description
Storage of intermediates and products from base oil production is typically found in base oil
production units.
Description
Solvent used in the solvent extraction of aromatics should be stored in blanketed tanks to
minimise fugitive emissions of furfural, NMP and MEK/toluene. The practice of internal and/or
external benchmarking may reduce the solvent consumption.
Applicability
Those tehniques can be seen as part of the VOC abatement programme (4.23.6) or soil
contamination prevention programme (4.25.1).
Example plants
Benchmarking of solvents and leakage prevention is widely applied in refineries.
Reference literature
[297, Italy, 2000], [268, TWG, 2001]
Description
Hydrogenation processes generate H2S. As consequence off-gases containing H2S are recovered
in a sulphur recovery plant (See Section 4.23.5) or incinerated.
Cross-media effects
Amine recovery of H2S consumes energy and chemicals. Sulphur recovery units consumes
energy (see Section 4.23.5). Incineration requires fuel.
Operational data
See Section 4.23.5
Applicability
In stand-alone luboil refineries, the small amounts of H2S generated in this process are typically
incinerated. Sulphur recovery is less commonly applied in those specialised refineries.
Economics
See Section 4.23.5. Data from EU refineries have found that the implementation SRU in stand-
alone lubricant refineries producing more than 2 tonnes of sulphur per day is cost-effective.
Example plants
At least one specialised European refinery has a SRU which recover 99.1 % of the sulphur in
the acid gas.
Reference literature
[147, HMIP UK, 1995], [268, TWG, 2001]
Description
Stripping of waste water from aromatic extraction prior to sending to waste water treatment
plant. Other technique to consider is dehydration/settling.
Cross-media effects
Stripping consumes energy.
Operational data
Stripping consumes energy.
Applicability
Fully applicable
Example plants
Applied in some refineries.
Reference literature
[147, HMIP UK, 1995], [319, Sequeira, 1998], [268, TWG, 2001]
Description
When the base oil production units have a single energy system, the introduction of clean-up
facilities in the flue gases is more cost-effective.
Cross-media effects
See different cross-media for the abatement of SOx, NOx and particulates in Section 4.23
Operational data
See Section 4.23
Applicability
This technique may be very difficult to be applied to existing refineries.
Economics
See Section 4.23
Example plants
Some stand-alone lubricant refineries have already applied this concept to reduce emissions
coming from the burning of liquid fuels.
Reference literature
[268, TWG, 2001]
Description
Bitumen should be stored in proper storage tanks normally under heating conditions and
insulated. Bitumen is not handled as a solid because is too difficult and is very intensive in man
power. Loading and unloading of the tank is typically done as follow. When the tank is filled,
nitrogen does not flow to the tank, and the pressure is lowered by letting part of the gas go to the
atmosphere. When the tank is unloaded on a low speed, a little bit nitrogen is led to the tank.
However when the speed of unload is higher, higher amounts of nitrogen must be used. If the
tank would be equipped with some cleaning system, it must be mechanically very simple and
easy to clean.
Cross-media effects
Hydrocarbons and sulphur compounds may emanate from leakages (particularly in overhead
systems) and pressure relief valves and in the form of aerosol-containing liquid droplets from
the venting of tanker top-loading operations.
Operational data
Electricity hot oil and low pressure steam can be used for heating. The heat exchanger can be
inside the tank or outside the tank in which case bitumen is circulated through it. Difference in
temperature should not be too high due to the surface temperature and coking.
Applicability
Storage of bitumen products is done in refineries that have bitumen production. Prevention
techniques as mentioned are typically applied.
Example plants
Some refineries in Europe use those techniques mentioned here.
Reference literature
[268, TWG, 2001]
Description
Oxidiser overhead can be routed to a scrubber rather than direct water quenching for
contaminant removal prior to incineration.
Cross-media effects
Additional contaminated water. Scrubber water is dirty and requires oil and solids separation
prior to re-use as desalter wash water and/or biotreatment. The sour water from the scrubber is
routed to and stripped in a sour water stripper prior to re-use and/or purification.
Applicability
Typically applied to overheads from the bitumen blowing systems
Reference literature
[147, HMIP UK, 1995]
Description
• Both incondensable products and condensates from the separator, hydrocarbon and aqueous,
can be burnt in a purpose-designed incinerator, using support fuel as necessary or in process
heaters.
• Oxidiser overhead slop oil can also be treated in the sludge processing or recycled in the
refinery slop oil system.
Cross-media effects
On scrubbing, aerosols can give rise to plugging. Additional contaminated water streams. In a
well operated unit, SO2 or odour from incineration of bitumen blowing incondensables do not
pose any problem.
Operational data
The incinerator should operate at a temperature of at least 800 ºC and the combustion chambers
should have a residence time of at least 0.5 second. The oxygen concentration at the combustion
chamber outlet should be greater than 3 % v/v. Low-NOx burners can be fitted in those
incinerators.
Applicability
Widely used to get rid of bitumen fumes. Incondensable and/or condensates can be burnt in
process heaters. However they should be treated or scrubbed to remove compounds of sulphur
or combustion products that may cause odours or other environmental problems.
Example plants
Many bitumen oxidisers have associated facilities to handle gas and liquid wastes.
Reference literature
[147, HMIP UK, 1995]
Description
Techniques that may be applied to prevent VOC emissions and odours are:
• odorous gases vented during the storage of bitumens and the tank blending/filling operation
can be vented to an incinerator
• the use of compact wet-type electrostatic precipitators has been proven capable of
successfully removing the liquid element of the aerosol generated during the top-loading of
tankers.
Cross-media effects
Energy consumption and in case of wet ESP waste generation.
Operational data
In the waste gas of a cleaning unit, a mass concentration of VOC of a total of 150 mg/Nm³ can
be complied with. In the waste gas of a incineration plant, a mass concentration of VOC, given
as total C, of a total of 20 mg/Nm³ can be complied with (Half-hourly mean values attainable in
continuos operation).
Applicability
Fully applicable.
Reference literature
[147, HMIP UK, 1995], [117, VDI, 2000]
Description
SOx abatement techniques and SRU are techniques to reduce sulphur emissions. Those are
extensivelly documented in Section 4.23.5.
Cross-media effects
See Section 4.23.5 for the different techniques applicable.
Operational data
See Section 4.23.5 for the different techniques applicable.
Applicability
The implementation of those techniques are extremely linked to the rest of the refinery. In
stand-alone bitumen refineries its application is more restricted.
Economics
See Section 4.23.5 for the different techniques applicable.
Example plants
Those techniques are applied in many refineries. However less applications are found in stand-
alone bitumen production refineries.
Reference literature
[268, TWG, 2001]
Description
Oxidiser overhead waste water accumulated in the overhead condensate collection drum can be
sent to a sour water stripper prior to sending it to the effluent water treatment facilities.
Cross-media effects
Increase of oil and particulates load to sour water stripper.
Applicability
Typically applied to waste water from bitumen blowing.
Reference literature
[147, HMIP UK, 1995], [268, TWG, 2001]
Description
In the bitumen production facilities, including storage, leakages can be produced. Those
leakages typically generate waste when mixed with other components as sand.
Operational data
This can be seen as a part of a waste management programme described in Section 4.25.1.
Applicability
Fully applicable
Example plants
Many refineries apply those type of systems.
Reference literature
[268, TWG, 2001]
Description
When the bitumen production units have a single energy system, the introduction of clean-up
facilities in the flue gases is more cost-effective.
Cross-media effects
See different cross-media for the abatement of SOx, NOx and particulates in Section 4.23
Operational data
See Section 4.23
Applicability
This technique may be very difficult to be applied to existing refineries.
Economics
See Section 4.23
Example plants
Some stand-alone bitumen refineries have already applied this concept to reduce emissions
coming from the burning of liquid fuels.
Reference literature
[268, TWG, 2001]
Description
This section gives emission information from the FCC when it is run under favourable
conditions and the regenerator in total combustion mode. Operational techniques that affect the
emissions from the FCC are also discussed. A description of the process can be found in Section
2.5.
Cross-media effects
The catcracker is the source of SO2 and NOx, CO2, CO, dust (particulates), N2O, SO3, metals,
hydrocarbons (ex. aldehydes), ammonia emissions, as discussed in Section 3.5. For example,
the basic design of a FCC includes two-stage cyclones in the regenerator vessel, which prevent
the bulk of the fine catalyst used from escaping from the system. However, smaller catalyst
particles, some of which are introduced with fresh catalyst and some created by attrition in the
circulating system, are not easily retained by the two-stage cyclone system. Consequently, in
many cases, other abatement techniques can be included to complement the process abatement
techniques discussed here. Next table gives a summary of the lowest emissions of pollutants to
the atmosphere due to an uncontrolled catcracker (No end-of-pipe technology is used to abate
air emissions).
A reduction of the regeneration temperatures under the usual conditions (700-750 °C) would not
have an significantly impact in the NOx emissions but could need a CO boiler and increase coke
formation. A change in the design or operation of the regeneration may increase the CO
concentration.
Operational data
Operational data of the FCC process can be found in Sections 2.5 and 3.5. More operational
data about FCCs can be found in [325, Gary and Handwerk, ; 326, Nelson, ]. A high-severity
FCC process enables higher yields of low olefins (C3, C4) that may be used for alkylation,
oxygenates production or direct sale to the market. This is carried out increasing the temperature
at which the cracking process is done and decreasing the contact time.
Applicability
Total combustion mode is typically applied to vacuum distillate feedstocks. Only low
Conradson carbon feeds are typically processed in FCCs.
Economics
Investment cost for FCCs in Basis of calculation
EUR/(t·yr)
48 – 64 1.5 Mt/yr fresh feed including reaction/regeneration system
and product recovery. Excluding offsites, power recovery and
flue gas scrubbing. (1998)
45 – 50 2.4 Mt/yr fresh feed including converter, fractionator, vapour
recovery and amine treating but not power recovery battery;
battery limit, direct material and labor, (1994)
Maintenance 2-3 % of investment per year
The operating cost for the addition of CO promoter to a 1.5 Mt/yr FCC (flue gas generation of
around 109 Nm3/yr) providing an efficiency around 30 % (700 mg/Nm3) is EUR 0.5 million /yr.
A major revamp of a FCC to change the design and/or operation is very expensive and typically
is not justified only by environmental reasons.
Example plant(s)
Common technique. Around 50 FCCs are installed in Europe. Catalytic cracking has largely
replaced thermal cracking because it is able to produce more gasoline with a higher octane and
less heavy fuel oils and light gases. FCC is a very typical operation in medium-complexity
refineries but RCC units are becoming more prevalent.
Reference literature
[212, Hydrocarbon processing, 1998], [113, Noyes, 1993], [117, VDI, 2000], [297, Italy, 2000]
[136, MRI, 1997], [247, UBA Austria, 1998], [45, Sema and Sofres, 1991], [268, TWG, 2001]
Description
This section gives emission information from the RCC when it is run under favourable
conditions and the regenerator runs in partial combustion mode. Operational techniques that
affect the emissions from the RCC are also discussed. A description of the process can be found
in Section 2.5.
Cross-media effects
Some feedstocks may need hydrotreatment.
Applicability
Fully applicable. FCC can be retrofitted to RCC. Medium Conradson carbon feeds would need a
CO boiler and higher ones a catalyst cooler. Metal content (Ni, V) should be limited; residue
demetallisation/hydrotreating may extend the range of residual feedstocks. It can run with
carbon Conradson index lower than 6 - 8 % and the total metal content less than 20 ppm.
Example plants
Several examples in Europe.
Reference literature
[212, Hydrocarbon processing, 1998]
Description
FCC regenerators are operated either in a complete (considered in this document as the basis
operational mode) or partial combustion mode. In the partial combustion mode considerable CO
is present in the flue gas and it is consumed downstream of the regenerator in a CO boiler, both
to recover the energy produced in a CO boiler and to meet environmental requirements. This
system can be seen as two stage regeneration in comparison with the total combustion mode,
where a single regeneration step is present. Another technique to consider is the modification of
the design and operation of regenerator especially to avoid high temperature spots that tend to
increase the NOx formation.
The emission factors for a cracking unit with an electrostatic precipitator and a CO boiler are
shown below.
Table 4.3: Emission factors for cracking (units in kg/1000 litres of fresh feed)
Cross-media effects
Fuel gas is needed for the incineration of CO. If fuel contains sulphur, emissions of SO2 may
occur. Other benefits are greater throughputs and heat/power recovery.
Operational data
Where a CO boiler is installed it should be carefully designed and operated to minimise CO and
NOx releases.
Applicability
This mode of combustion is fully applicable. In general it can be said that feed having a
Conradson carbon figure up to 2 - 3 % can be processed in a ‘standard’ FCC without CO boiler
as well as partial-burn FCC. A Conradson figure from 2 - 5 % would require a CO boiler and
from 5 - 10 % a cat cooler would be needed. It should be noted that the above figures are only
indicative and depend also on the severity of cracking.
Economics
Investment cost necessary to convert a FCC from total to partial combustion mode is around
EUR 2.5 - 4 million. A major revamp of a FCC to change the design and/or operation is very
expensive and typically is not justified only by environmental reasons.
Example plants
Carbon monoxide boiler is normally a standard practice in fluidised catalytic cracking units in
the USA. Many examples also in Europe and the rest of the world.
Reference literature
[297, Italy, 2000], [316, TWG, 2000], [268, TWG, 2001]
Description
Hydrotreatment of feed to the catcracker operates at similar conditions to those used for gasoil
and atmospheric residue hydrotreatment (see Sections 2.13 and 4.13).
Cross-media effects
Cross-media effects are the increase of energy consumption and the consequent increase in CO2
emissions, due mainly to the hydrogen, as discussed in Section 3.14. As discussed in Section
4.13, hydrotreatment processes generate catalyst for disposal and increase the H2S production
with direct consequences for the sour water stripper and sulphur recovery units (which may
need to be enlarged or replaced).
Operational data
Installation of catcracker feed hydrotreatment (or a feed hydrocracker) will reduce the sulphur
content in catcracker products and improve their quality, so that they require less final
processing. Hydrogen and energy are required for this process. The water used for the different
catalytic sections yields 20 - 40 m3/h of sour water. The sulphur removal efficiency achieved by
hydrotreatment depends on the boiling range of the catcracker feed. The heavier the feed, the
more energy is required for the same sulphur removal efficiency.
Applicability
Fully applicable. This process is more cost-effective when hydrogen, SWS and SRU capacities
are available in the refinery.
Economics
Size/capacity Typical capital/installed cost (EUR million) Operating cost
ktonnes/yr Costs include necessary interconnections to existing plant for EUR million/yr
integration purposes.
1250 65 n.a.
2500 106 18
3750 150 n.a.
Note: The cost assumes that there is adequate space and existing SRU (sulphur recovery units) and sour
water stripping capacity. If further hydrogen production is needed, a new hydrogen plant for a 2500 kt/yr
FCCU feed hydrotreater or hydrocracker would typically cost in the range of EUR 60 - 75 millions
Table 4.4: Catalytic feed hydrotreatment (typical feeds being atmospheric residue and vacuum
gas oil)
Source: [112, Foster Wheeler Energy, 1999]
Example plants
Many examples.
Reference literature
[296, IFP, 2000], [115, CONCAWE, 1999], [112, Foster Wheeler Energy, 1999], [45, Sema and
Sofres, 1991], [247, UBA Austria, 1998], [297, Italy, 2000], [268, TWG, 2001]
4.5.5 Waste heat boiler and expander applied to flue gas from the FCC
regenerator
Description
Heat recovery from the regenerator flue gas is conducted in a waste heat boiler or in a CO-
boiler. Heat recovery from the reactor vapour is conducted in the main fractionator by heat
integration with the unsaturated gas plant as well as generation of steam with the residual heat
from product rundown streams and pumparound streams. The steam produced in the CO boiler
normally balances the steam consumed. Installing an expander in the flue gas stream from the
regenerator can further increase the energy efficiency. Figure 4.1 gives a simplified scheme of
application of a waste heat boiler.
Flue gas
ESP
CO BOILER
WASTE EXPANDER
HEAT
BOILER Regenerator
Flue gas
COMPRESSION
Air to REGENERATOR
Cross-media effects
Significant quantities of catalyst fines are collected in the waste heat boiler (WHB). The newer
WHBs have facilities which permanently remove the collected fines, but older WHBs are
normally soot-blown once a shift. During this operation the emission of catalyst fines is
especially high if the unit is not equipped with a particulate abatement technique.
Operational data
Power recovery on regeneration gas reduces CO boiler duty.
Applicability
Retrofitting of of this equipment can be very difficult because of space limitations in the
refinery. On small or low pressure units, expanders are not justified economically.
Economics
Applying expander on the regenerator gas could be costly because the additional particulate
systems under high temperature that they need. Turbo expanders are big cost items, and so are
waste heat recovery units.
Example plant(s)
Energy recovery by utilising an expander in the regenerator flue gas is only applied in the
larger, more recently built units.
Reference literature
[136, MRI, 1997], [268, TWG, 2001]
Description
Techniques to consider are:
a) Use a higher quality FCC catalyst, process efficiencies can be increased while catalyst waste
can be reduced and rate of replacement reduced.
b) Use of an attrition resistant catalyst to reduce the frequency of replacing catalyst and reduce
emission of particulates from the regenerator.
Applicability
A change in the catalyst may have an adversely effect to the performance of the FCC.
Economics
Investment expenses: none.
Operating expenses: Negligible.
Reference literature
[80, March Consulting Group, 1991], [115, CONCAWE, 1999]
Description
Some techniques to consider are:
a) Some catcracker designs contain a cascading overhead washing section designed to minimise
water usage.
b) re-use waste water generated in FCCU within the refinery (e.g. desalters) or ultimately route
to WWTP.
Applicability
Both techniques are fully applicable to catcrackers
Example plants
Used in some catcrackers
Reference literature
[80, March Consulting Group, 1991]
This section includes NOx abatement techniques that may be applied to FCCs.
Description
See Section 4.23.3.
Cross-media effects
Use of NH3 (storage/handling), risk of NH3 emissions when operating outside stoichiometric
proportion and depending on the age of the catalyst (<2 - 10 mg/Nm3) and SCR catalyst
regeneration and disposal. Ammonia needed for this technique may be supplied by two-stage
sour water strippers (see Section 4.24.2). Another disadvantage of SCR is that a few percent of
SO2 is oxidised to SO3 over the deNOx cat and as consequence it can potentially create plume
opacity problems.
Operational data
Catalyst performance is expected to deteriorate with time, due mainly to the dust and SOx.
However, current examples have shown lifetimes well over the expected ones (6 years).
Information from operators claim that not all SCR applied to FCC have operated to design loads
leading that fact to increase catalyst lifetimes. Some other concerns from operators is that SCR
catalyst could potentially be fouled by the particulates in the flue gas stream. More general
operational data for SCRs can be found in Section 4.23.3.
Applicability
Because the temperature window is wide (300 - 400 ºC), it is very flexible for retrofit
applications. However, considerable space is needed for installation. SCR applications often
requires a new waste heat boiler (full burn) or CO-boiler (partial burn) if they are not present.
Preferably the deNOx unit is integrated with the waste heat boiler. Because SCR operates at
oxidizing conditions it cannot be installed upstream of a CO boiler (partial burn).
Economics
Table 4.5 shows some examples of the economics of the application of SCR to FCCs
A more detail economic assessment of a SCR in a FCC unit is shown in Table 4.6.
EUR/unit EUR/yr
Operating hours (h/yr) 8000
Investment costs (EUR million) 1.45
Input factors for annual expenditure:
Number of years 15
Rate of interest (%) 6
Annual repayment incl. interest (EUR/yr) 150000
Proportional investment costs incl. interest 150000
Volume of catalyst (m3) 20
Endurance (years) 8
Renewing of catalyst (m3/yr) 2.5 15000 €/m3
Average renewing of catalyst (EUR/yr) 36300
Catalysts 36300
Maintenance + wear and tear 2
(% of investment costs)
Maintenance + wear and tear (EUR/yr) 29000
Maintenance + wear and tear 29000
Pressure drop (mbar) 8
Energy for reheating (MJ/h) 0 3.6 €/GJ 0
Electrical energy (kWh/h) 88 0.065 €/kWh 46000
NH3 liquid (kg/h) 36.96 0.25 €/kg 75200
Total costs 336269
Notes: In a refinery with a volume of exhaust gas of 100000 Nm3/h achieving a NOx emission reduction of
1000 mg/Nm3 in relation to actual oxygen content and for a clean gas concentration of <200 mg NOx/Nm3.
Inlet concentration may vary between 200 - 2000 mg/Nm3 @ 3 %O2. The outlet concentration of NOx is
reduced to 80 - 120 mg/Nm3 by SCR.
Table 4.6: Main cost factors for a Selective Catalytic Reduction (SCR) installation (raw gas)
after the FCC-plant
Examples
At least six SCR are in operation in FCC worlwide and two in Europe (Scanraff-Sweden and
Pernis-The Netherlands).
Reference literature
[107, Janson, 1999], [45, Sema and Sofres, 1991], [254, UKPIA, 2000] [247, UBA Austria,
1998], [115, CONCAWE, 1999], [316, TWG, 2000], [136, MRI, 1997], [348, Ashworth
Leininger Group, 2001]
Description
See Section 4.23.3.
Cross-media effects
Use of NH3 (storage/handling), risk of NH3 emissions when operating outside stoichiometric
proportion. Ammonia needed for this technique may be supplied by two-stage sour water
strippers (see Section 4.24.2). The use of urea generates more ammonia (from urea) slip and
some N2O formation.
Operational data
High temperatures (800 - 900 ºC) of the flue gases are needed.
Applicability
It is applicable in partial combustion FCCs with CO boiler; retrofitting in existing CO boilers is
relatively simple. It is also applicable to full combustion units. Space requirements are mainly
related to NH3 storage.
Economics
Size of the FCC Efficiency NOx in/out concentration Investment Operating cost Specific
(Mt/yr) (%) (mg/Nm3 @ 3 % O2 (EUR (EUR removal cost (4)
million) million/yr) (EUR/t of NOx
removed)
1.5 60 800 / >320 5.4 0.1 – 0.4 1300
0.6 (1)
1.5 60 - 80 200 / 40-80 0.35 - 1.5(1) 0.05 - 0.4 1700
1. Cost data not including the cost of the CO boiler
Example(s)
It has been applied in Japan on an FCCU where a CO boiler provides a sufficient temperature
window.
Reference literature
[115, CONCAWE, 1999], [45, Sema and Sofres, 1991], [316, TWG, 2000]
The selection of the catalyst can be seen as a particulate abatement technique. That technique
has been included in Section 4.5.6 together with other possible effects of catalyst replacement.
Description
Highly specialised cyclones are used (third-stage and multicyclones), which are designed to suit
the arrangement, dimensions, contours, velocities, pressures and densities of the particles to be
removed. This is the natural first choice of clean-up device for particulates: these are
conventional cyclones, fitted externally to the regenerator but operating on the same principle as
the internal first and second cyclones. They are high-velocity devices and recovered catalyst is
returned to a dust hopper.
Cross-media effects
Catalyst fines recovered are considered as waste (300 - 400 tonnes/yr per unit). Consequently
pollution is transferred from air to the soil.
Operational data
It generates a pressure drop in the flue gas. Good performance and reliability in many FCCs.
Applicability
Applicable to any FCC.
Economics
Size of the FCC Efficiency Downstream Investement Operating cost
(Mt/yr) (%) particulate (EUR million) (EUR million/yr)
concentration
(mg/Nm3)
1.5 30 - 40 40 - 250 1 - 2.5 0.7
1.5 30 - 90 60 - 150 * 0.5 - 1.5 0.1
1.2 75 50 - 100 # 1.5 - 2.5
Note: Operating costs include only direct cash operating expenses i.e. do not include depreciation of
investment, nor financial charges. Investment costs refer to new plant installation. Economics do not
include the cost of disposal of the waste generated.
* Initial concentration: 450 mg/Nm3 (300 - 600 mg/Nm3 range)
# Initial concentration: 200 - 1000 mg/Nm3
Economics of disposal of the fine catalyst is around EUR 120 - 300 per tonne, including
transportation.
Example plants
Many FFC units are operating with these systems.
Reference literature
[45, Sema and Sofres, 1991], [80, March Consulting Group, 1991], [297, Italy, 2000], [115,
CONCAWE, 1999]
Description
A short description of an electrostatic precipitator can be found in Section 4.23.4.
Cross-media effects
Because these systems recover the fine particulates (mainly catalyst) coming from the FCC, the
refinery may need extra facilities to manage the fine particulates recovered. ESPs also consume
electricity.
Operational data
Electrostatic precipitators cause a very slight pressure drop. The extra energy consumption is
relatively low, but they consume electricity. They also require high maintenance to keep the
capture efficiency high. Continuous monitoring of particulates is not amongst the most reliable
continuous monitoring systems. Information from a EU+ refinery shows that the use of deep-
desulphurisation of the feedstock have a great impact of the electrostatic properties of the
particulates (because content of sulphur and metals are smaller) reducing the availability of the
ESP to catch those particulates. In those cases, they have reported that the best values
achievable are 30 – 35 mg/Nm3.
Applicability
Because the FCC outflows to be treated are large (1.5Mt/yr FCC generates 2.8 million Nm3/day)
and the flow gas speeds should be low to increase the capture of particulates, these systems
require a lot of space. ESP may be difficult to apply if high electrical resistivity particulates are
to be captured.
Economics
Size of the Efficiency Downstream PM Investement Operating cost
FCC (Mt/yr) (%) concentration (EUR million) (EUR million/yr) *
(mg/Nm3)
1.5 90 30 4-6 0.25 - 0.5
1.5 85 - 95 <50 3-5 0.25 - 0.5
1 95 10 - 20 * 5.5 small @
1.5 90 - 95 <50 # 4-6 0.25 - 0.5
Note: Operating costs include only direct cash operating expenses i.e. do not include depreciation of investment, nor
financial charges. Investment costs refer to new plant installation.
* Initial concentration: 250 - 300 mg/Nm3 range
# Initial concentration 4000 mg/Nm3
@
Disposal of fines not included
Example plants
More than 10 European refineries use ESP in their FCCs. They are also used in many refineries
in the USA and Japan.
Reference literature
[45, Sema and Sofres, 1991], [297, Italy, 2000], [117, VDI, 2000], [247, UBA Austria, 1998],
[222, Shell Pernis, 1999], [115, CONCAWE, 1999], [268, TWG, 2001]
4.5.9.3 Filters
Description
A further option for regenerator exhaust gases are filters (bag, ceramic or stainless steel filters).
Operational data
Bag filters are temperature-sensitive (<200 ºC) and hence their use is limited. They cause a
high-pressure drop.
Applicability
Fabric or bag filters can be retrofitted to the underflow of third-stage cyclones. However bag
filters are not favoured for use in the FCC because of the pressure drop, the potential for
"blinding" of the bags, the large space required and their inability to cope with upset conditions.
Example plants
At least one successful application in European FCC.
Reference literature
[80, March Consulting Group, 1991]
Description
Significant quantities of catalyst fines are often present around the FCCU catalyst hoppers,
reactor and regeneration vessels as well as those collected by the abatement techniques
mentioned above. The fines can be collected and recycled before being washed to the sewers or
migrating off-site via the wind. Those techniques may be considered as a housekeeping
measures, included in Section 4.15. Some techniques to consider are:
§ dry sweeping the catalyst fines and sending the solids to be recycled or disposed of as non-
hazardous waste.
§ use of vacuum ducts in dusty areas (and vacuum hoses for manual collection) which run to a
small baghouse for collection.
§ recycling opportunities for the fines (cement production).
Applicability
Fully applicable especially during catalyst loading/unloading processes.
Example plants
Techniques already in use in USA refineries.
Reference literature
[80, March Consulting Group, 1991]
Description
Two types of techniques have been applied to improve separation of catalyst from slurry decant
oil from the slurry settler used in catcracker: a) One system incorporates high voltage electric
fields to polarize and capture catalyst particles from decant oil. b) The amount of catalyst fines
reaching the decant oil can be also be minimized by installing high-efficiency cyclones in the
reactor to shift catalyst fines losses of the decant oil to the regenerator, where they can be
collected by any particulate abatement technique.
Operational data
Catalyst fines suspended in the separator underflow are recycled back to the reactor.
Applicability
Hydrocyclones are applied in many FCC units.
Example plants
Hydrocyclones, for removal of catalyst fines, are successfully applied in many FCCs.
Reference literature
[316, TWG, 2000]
This section contains sulphur oxides abatement techniques that may be applicable to FCCs.
Description
Sulphur dioxide in the regenerator off-gas of an FCC unit can be reduced by using a catalyst
(metal (e.g. Al/Mg, Ce) oxide) which transfers a significant portion of the sulphur associated
with the coke on the catalyst to the reactor, where it is liberated as hydrogen sulphide. This
leaves the reactor with the cracked vapour product for capture in the refinery’s amine scrubbing
system and hence conversion of sulphur in the SRU. DeSOx is a 3-step process: a) oxidation of
SO2 to SO3 catalysed by cerium, b) adsorption of SO3 produced in the regenerator to give a
sulphate which, on return to the reactor, c) reverts to the oxide and releases hydrogen sulphide
into the product gas stream for recovery.
Cross-media effects
Disadvantages of the process are:
• deSOx additive is more efficient in full burn mode. However, full burn implies more SOx
and NOx formation than partial combustion mode
• possible poorer yield of products from FCC unit
• reduced FCC unit operating flexibility
• other impacts are the increased energy consumption and the possibility of bottlenecks in the
H2S handling facilities.
Applicability
This technique is sensitive to unit design, especially regenerator conditions. It operates most
effectively under full oxidising conditions, when as much as 50 % of the sulphur in the flue
gases can be removed. Unit performance may, however, be affected and transfer catalyst
replacement rate may be significant.
Economics
There are no major investment costs required for this option, save for the dosing equipment for
introducing the additive to the catalyst system. Operating expenses are around 0.34 - 0.7 EUR/t
(EUR 1 million/yr for a 1.5 Mt/yr FCC). Another reference gives a typical cost of EUR 3
million/yr (1997) for the catalyst for a gas flow of 200000 Nm3/h (2.3 Mt/yr FCC). Cost is
highly dependent on the unit, the initial SO2 emissions and the SO2 target.
Example plant(s)
Many refineries have applied deSOx additives to the catalyst. The method is well established
commercially.
Reference literature
[115, CONCAWE, 1999], [45, Sema and Sofres, 1991], [112, Foster Wheeler Energy, 1999],
[297, Italy, 2000], [316, TWG, 2000]
Description
There are several wet scrubbing processes. A brief description is included in Section 4.23.5.4.
Cross-media effects
The wet scrubbing system creates secondary problems of aqueous slurry waste disposal and
increases the refinery energy consumption. The effluent water purge contains Na2SO4. Other
drawbacks are the expensive raw material, e.g. caustic soda, if sulphur and other loadings are
high, and they may need flue gas reheat to prevent plume mist. These systems are generally
sensitive to other contaminants such as particulates, salts and sulphur trioxide. Other impacts are
the possible debottlenecking of H2S handling facilities (e.g. SRU, amine scrubbers), the
production of a by-product as well as the raw material supply and handling.
Wellman-Lord system: Increased energy consumption, By-product, raw material supply and
handling, possible bottlenecking of H2S handling facilities.
Operational data
They are flexible and reliable. Day-to-day operating changes can be readily handled. They
generate low pressure drop, operate at low temperature and do not create solid deposition
problems. Some CO2 will be removed as well by wet scrubbers, reducing the capability of the
media to dissolve SO2. More information can be found in Section 4.23.5.4
Applicability
They are compact systems and fully applicable. The system ranges from 93 to 465 m2 for FCCs
ranging from 55.2 to 276 dm3/s feed rate (1.5 to 7.5 Mt/yr). Recent trends include forced
oxidation to process the SO2 in the SRU, reaction with lime to produce gypsum, a commercial
product where there is a market. Spin-off application of power plant, equipped with Wellmann-
Lord FGD. More information can be found in Section 4.23.5.4.
Economics
Objective of the process FCC capacity Investment expenses Operating expenses
(Mt/yr) (EUR million) (EUR million/yr)
SO2 reduction 1.5 10 2-5
Particulates reduction 1.5 4-6 2-5
2-5
SO2 reduction 3 15 - 20 2-3
SO2 reduction (regenerative) 3 24 - 28 1.5
SO2 reduction 5 13 * no data
SO2 and particulate reduction 0.5 - 4 3 - 15 no data
* Investment expenses include the water purge treatment unit.
Example plants
It exists a considerable operating experience. This technique is widely used on FCC unit plants
in the USA. Wellman Lord system has been successfully applied to power plants. More
information can be found in Section 4.23.5.4
Reference literature
[181, HP, 1998], [269, Confuorto, 2000], [45, Sema and Sofres, 1991], [247, UBA Austria,
1998]
Description
Two variants exist, the jet venturi and the high-energy venturi.
Cross-media effects
They produce liquid and solid waste and a wet gas plume.
Operational data
These systems are generally sensitive to other contaminants such as particulates, salts, sulphur
trioxide etc.
Applicability
The high-energy venturi is the design usually favoured by recent installations. Widely applied
for the treatment of small emissions flows.
Example plants
Venturi scrubbing, using an alkaline liquid, has been used successfully on a less widespread
scale.
Reference literature
[112, Foster Wheeler Energy, 1999].
Description
Two types of techniques exist: dry and semi-dry. The key component of the semi-dry process is
a spray dryer in which the hot flue gas is contacted with a spray of fine droplets of lime slurry.
The SO2 is absorbed into the droplets, forming reaction products which are dried to a fine
powder by the hot flue gas. Both dry and semi-dry processes require downstream dust
arrestment systems, such as an electrostatic precipitator or bag filtration. More information can
be found in Section 4.23.5.4.
Cross-media effects
Deposition of solid waste results in case that the reaction products cannot fullfill the quality
needed by consumers. Other drawbacks are:
• high pressure drops across the bag filters if are used
• increased dust load in gas stream; need for dust capture
• operational difficulties in water/heat balance (spray dryers only)
• possibly significant pressure drop across bag filter dust arrestment plant, e.g. bag filters
• generation of solid waste: 1 tonne of SO2 reduction entails about 2.5 tonnes of solid wastes
being generated
• with the dry- and semidry scrubbers a mixture of CaSO3, CaSO4, fly-ash and lime is
produced.
Operational data
These systems are generally sensitive to other contaminants such as particulates, salts, sulphur
trioxide etc.
Applicability
Operate at low-temperature end of system. Waste generated can be difficult to re-use (no market
for gypsum), and difficult to landfill (landfill use to be severely restricted in near future).
Economics
Dry process is a relatively low cost solution. Raw materials for those processes are cheap.
Capital cost and operating costs typically less than wet scrubbing. Investment cost around EUR
15 - 20 million, and operating cost around EUR 2 - 3 million/yr (lime cost + landfill waste
disposal).
Reference literature
[112, Foster Wheeler Energy, 1999], [257, Gilbert, 2000], [45, Sema and Sofres, 1991], [297,
Italy, 2000]
Description
This scrubbing uses the natural alkalinity of the seawater to remove SO2. More information can
be found in Section 4.23.5.4.
Cross-media effects
Particulates containing metals (e.g. V, Ni, Sb) and other pollutants from the flue gas are
transferred to the seawater. The sulphur content of the effluent seawater from the process itself
is increased by only approximately 3 %.
Operational data
The system needs electrical power (seawater pumps, aeration fans and flue gas fan) and
seawater. It is feasible to use spent cooling water, for example, when seawater is used as cooling
water. The need for seawater is roughly 15000 - 20000 m3/h per tonne of sulphur removed.
Special attention should be paid to minimise corrosion and the maintenance derived from it.
Applicability
The feasibility of applying seawater scrubbing is site-dependent. The process needs access to
seawater in large amounts at a competitive cost. Fully applied for fuels up to 1.5 % of sulphur.
This system normally is applied after a particulate abatement technique to avoid contamination
of sea environment by particulates (containing metals).
Economics
Seawater scrubber annual costs depend on size and SOx in/out. The cost is comparable to wet
scrubbers. Typical investement cost is in the order of EUR 60 /net kWe output of the power
plant. The main operating cost is the electrical power for operation of the technique.
Example plants
System applied to some power plants aroung the world and one proven commercial application
in FCC, Norway.
Reference literature
[278, Alstom Power, 2000]
Description
See Section 4.23.5.4.
Reference literature
[247, UBA Austria, 1998]
Description
There are processes on the market that include an ESP, a SCR for NOx, catalytic oxidation of
SO2 and production of sulphuric acid. More information about this technique can be found in
Section 4.23.3.
Cross-media effects
See each of the particular techniques included in this process.
Operational data
See each of the particular techniques included in this process.
Applicability
Process qualified for treating very high SO2 concentrations in the flue gas.
Economics
Design capacity 1000000 Nm3/h
Investment cost 100 EUR million
Example plants
Applied to FCC and coking flue gas.
Reference literature
[297, Italy, 2000]
Description
Waste is generated during the FCC processes. The selection of catalyst used in the FCC, control
of the emissions of particulates and the proper management of product and slurry tank bottoms
can reduce the generation of waste. Also the implementation of benchmark performance on the
waste generation may be an incentive on how and how much the total amount of waste can be
reduced.
Operational data
Those techniques can be seen as part of the waste management programme of a refinery.
Applicability
See Section 4.5.6 for applicability restriction of selection of catalyst.
Example plants
Applied in many refineries.
Reference literature
[268, TWG, 2001]
Description
In the continuous process, the catalyst is regenerated continuously and maintained at high
activity. Catalyst is regenerated in one external reactor. More information can be found in
Section 3.6.
Cross-media effects
It generates more dioxins than semi-regenerative during the regeneration of the catalyst. For
more information, see Section 3.6.
Operational data
It operates at lower pressures (6 - 14 bar) than and at the same temperature as semi-regenerative
processes. It uses bi-metallic catalysts (Pt-Re, Pt-Sn). Hydrogen is also produced in those
processes. Continuous regeneration reformer produces around 350 Nm3 of hydrogen per tonne
of feedstock. As increased coke laydown and thermodynamic equilibrium yields of reformate
are both favoured by low-pressure operation, the ability to maintain high catalyst activities and
selectivities by continuous catalyst regeneration is the major advantage of the continuous type
of unit.
Applicability
Fully applicable.
Economics
Investment cost based in 3975 m3/d continuous octanizing unit, battery limits, erected cost, 1998
Gulf Coast: 10000 USD per m3/d. The erected cost based in 3180 m3/d has been of USD 48.3
million ($15200 per m3/d). In addition, catalyst costs add another USD 2.6 million. Costs are
4Q95 US Gulf Coast.
Example plant(s)
Nowadays, only continuous catalytic reformer units are built.
Reference literature
[212, Hydrocarbon processing, 1998], [166, Meyers, 1997], [268, TWG, 2001]
Description
A brief description can be found in Section 2.6.
Cross-media effects
In semi-regenerative reforming, the unit catalyst is regenerated at intervals of 3 to 24 months,
depending upon the intensity of the process. The emission factor of dioxins per tonne of
feedstock is much lower than in the case of continuous regeneration. For more information, see
Section 3.6.
Operational data
The reactors operate at temperature in the range of 400 - 560º C and pressures in the range of
20 - 50 bar. This type of process produces between 130 and 200 Nm3 of hydrogen per tonne of
feedstock.
Applicability
Fully applicable.
Economics
The erected cost of a semiregenerative unit of 3180 m3/d is 33 million USD (10400 $ per m3/d).
In addition, catalyst costs add another 3.4 million USD. Costs are 4Q95 US Gulf Coast.
Example plants
Many applications exist in Europe as well as in the rest of the world.
Reference literature
[261, Canales, 2000], [166, Meyers, 1997]
Description
Ozone depleting substances (e.g. carbon tetrachloride) are sometimes used during the
regeneration of the catalyst of the reformer. Emissions of such substances should be avoided or
at least minimized by using less harmful substitutes or by using them in confined compartments.
Cross-media effects
Due to the use of chlorinated compounds, the possibility of emissions of dioxins and furanes
during generation has been detected. Refer to Section 3.6.
Operational data
Promoters should be handled in enclosed systems.
Example plants
Most refiners have switched to lower ozone depleting substances such as perchloroethylene
during the regeneration of the reforming catalyst.
Reference literature
[80, March Consulting Group, 1991]
Description
Regenerator flue gas containing HCl, H2S, small quantities of catalyst fines, traces of Cl2, SO2
and dioxins can be sent to a scrubber prior to to release to atmosphere. The storage and handling
of organic chlorides used during the regeneration may also lead to releases. In some reformers a
further filter trap of Cl2 (ZnO/NaCO3 or NaOH over alumina) is installed.
Cross-media effects
The recirculated and bleed streams from the regeneration flue gas wash should be sent to the
waste water treatment plant. Due to low pH of this waste water stream, neutralisation prior to
biological treatment may be necessary. The use of scrubber can move some dioxins from the air
to a water emission.
Applicability
Fully applicable. Cl2 trap has been used in continuous reformers.
Reference literature
[112, Foster Wheeler Energy, 1999], [268, TWG, 2001]
Description
Regenerator flue gas containing HCl, H2S, small quantities of catalyst fines, traces of Cl2, SO2
and dioxins can be sent to an electrostatic precipitator prior to release to atmosphere. Other
activities as venting during catalyst regeneration or replacement and plant cleaning may be sent
to the ESP.
Cross-media effects
See Section 4.23.4.2
Applicability
Emissions from continuous regeneration sections require particular attention. The use of this
type of systems in semiregenerative units is less extended because almost no exists mechanism
for the generation of catalyst fines.
Reference literature
[112, Foster Wheeler Energy, 1999], [268, TWG, 2001]
Description
Referring to Sections 3.6 and sections in 4.6 on catalytic reforming, dioxins are typically formed
in the three types (continuous, cyclic and semiregenerative) of catalytic reforming during the
regeneration of the catalyst.
If the regenerator flue gas is treated in a water scrubber (e.g. Section 4.6.4), the dioxins have
been appeared detectable in waste water from the scrubber, but they have not been detected after
the WWTP, possibly because dilution effects.
In some other cases, because the use of other techniques as fixed bed filters have result in a
combine reduction of chlorine and dioxins. In some cases, activated carbon has been used to
remove dioxins. Other techniques that have been used is the recirculation of regeneration vent
gases. However this last one is not clear how it can reduce the emission of dioxins.
The study of dioxins emissions from reformers and how the regeneration conditions may affect
those emission can be a good technique to start to understand and solve the problem.
Cross-media effects
Some dioxins from the regenerator gases may be transferred to water via scrubbing.
Example plants
Some EU+ refineries have already applied and monitored dioxins emission from catalytic
reformers. Techniques as chlorine trap and recirculation of vents have not been reported that
have been applied to semiregenerative reformers.
Reference literature
[268, TWG, 2001]
Description
A description of this process can be found in Section 2.7. The techniques that may be applied to
the delayed coker to prevent emissions are the following:
1. uncondensable vapours generated in the coking processes should not pass to the flare
system
2. pressure reliefs from the coke drums should pass to the quench tower
3. techniques for coke drum quench tower arrestment train to include consideration of venting
the final emission to a flare header with the possible elimination of the second tower and
blowdown pond
4. water use in the delayed coker is treated in Section 4.7.7
5. steam generated in this process can be used to heat up other refinery processes
6. the delayed coking process itself has a low level of heat integration. The heat to maintain
the coke drums at coking temperature is supplied by heating the feed and the recycle stream
in a furnace. However, the atmospheric residue and/or vacuum residue can be fed straight
into the delayed coking unit without intermediate cooling, which results in a high heat-
integration level between the different units and saves a considerable amount of capital on
heat exchangers
7. use of coking gas. The energy efficiency of the coker can be further increased if the coke
gas is combusted in a gas turbine of a combined cycle unit. Extra information on the
application of refinery fuel gas in combined cycle units appears in Section 4.10.
Operational data
Some operational data of delayed cokers can be found in Section 3.7.
Applicability
Delayed coking is fully applicable. It is normally applied when the production of petroleum
coke has a market or when high quality coke is required to produce anodes for the non-ferrous
metal industry.
Economics
Typical investment of a full delayed coker (basis 1Mt/yr straight run vacuum residue feed, U.S.
Gulf Coast 1998, fuel grade coke, includes vapour recovery): 136250 - 218000 USD per t/yr.
Example plants
Many delayed processes exist in Europe as well as in the rest of the world. Delayed coking is
the most commonly used technology in the refinery sector.
Reference literature
[112, Foster Wheeler Energy, 1999], [19, Irish EPA, 1993], [118, VROM, 1999]
Description
A description of this process can be found in Section 2.7. Another technique that can be used to
prevent emissions or increase energy integration in the fluid cokig is to use the coking gas in a
gas turbine of a combined cycle unit. Extra information on the application of refinery fuel gas in
combined cycle units appears in Section 4.10.
Operational data
This system operates under fluidised bed conditions.
Applicability
Fully applicable. Although the coke from fluid coking is of a higher grade than the coke from
delayed coking, it is not suitable for some specialist applications (e.g. carbon graphite
electrodes).
Economics
Investment (U.S. Gulf Coast 1996): 10000 - 13200 USD per m3/d.
Example plants
Some examples exist around the world.
Reference literature
[136, MRI, 1997]
Description
A brief description of this process can be found in Section 2.7. Certain techniques can be
applied to the calcination process to prevent emissions. Some of those considered to be
beneficial are the following:
1. The kilns can be directly fuelled by coke gas or coke fines, driving off volatile matter and
burning it within the kiln.
2. The hot stream of waste gas from the calcination of petroleum coke in rotary kilns contains
considerable quantities of particulate, which - possibly after heat recovery from the waste
gas - is separated with suitable filter devices, e.g. high efficiency multi-cyclones, bag filters
and electrostatic precipitators (see also Chapter 4.7.8.2). During calcination in multiple-
hearth furnaces, particulate abatement techniques are not usually used because of the
comparatively low emissions in the waste gases.
3. The unburnt gases from the calcinator are burned in an incinerator, then passed through a
waste heat boiler before being released to the atmosphere via a dust collection system.
4. The calcined coke discharges to rotary cooler where it is cooled by direct water injection.
Off-gases from the cooler pass to gas cleaning by multi-cyclones and water scrubber.
5. Collected fines from the dust abatement techniques should be conveyed to a silo with exit
air filters. Collected cyclone fines may be recycled to product or may be used within the
refinery or may be sold as product.
Table 4.11: Emission values for petroleum coke production plants (calcination of green coke)
Operational data
Some operational data of petroleum coke calciners can be found in Sections 2.7 and 3.7.
Applicability
Applied to coke produced by delayed cokers and fluid cokers.
Example plants
Many examples exist in Europe and around the world. Multiple-hearth furnaces have been
extensivelly applied to waste incineration. A majority of facilities that calcine coke utilise a
rotary kiln. The last two calcining facilities built in the world are rotary kilns.
Reference literature
[117, VDI, 2000], [147, HMIP UK, 1995], [268, TWG, 2001], [347, Services, 2001]
4.7.4 Flexicoking
Description
A brief description of this production process can be found in Section 2.7. Some techniques that
may be applied to prevent emission from the flexicoker are:
1. The flexicoking process has a high level of heat integration. The only source of heat in the
Flexicoker process is the gasifier, where the coke is partially oxidised. The remainder of the
heat in the coker gas is recovered by generating steam. The energy efficiency can be further
increased if the coke gas is combusted in a gas turbine of a combined cycle unit. See
Section 4.10.
2. As coke cutting from drums is not needed, emissions and dirty effluent generation are
avoided compared to delayed coking. Moreover, the sulphur components from the coke gas
are easy to eliminate. Some 84 - 88 % w/w of the hydrocarbon feed is recovered as
hydrocarbon product, the remainder is converted into CO, CO2 and H2O.
3. The coke gas from the heater is passed through a cyclone set to remove the coarser coke
particles and is subsequently cooled by generation of high-pressure steam and by preheating
boiler feed water. About 75 % of coke fines which pass through the two stages of heater
cyclones are recovered in tertiary cyclones. Almost all the coke fines which escape the
tertiary cyclones are scrubbed out in the venturi scrubber. The water slurry from the venturi
scrubber is steam stripped to remove absorbed hydrogen sulphide and ammonia.
Operational data
Some operational data for flexicokers can be found in Sections 2.7. and 3.7.
Applicability
Fully applicable. However, because the products from flexicoking are different than other
coking processes (e.g no coke is produced) the implementation of this option typically should
consider the refinery product requirements.
Economics
Investment (basis U.S. Gulf Coast 1996): 15100 - 19500 US$ per m3/d (typical).
Example plants
Many examples exist in Europe, USA and the rest of the world.
Reference literature
[212, Hydrocarbon processing, 1998], [271, Martinez del Pozo, 2000]
Description
In refineries with a coker, oily sludges, sludges from the waste water treatment and waste can be
destroyed in the coker (delayed, fluid or flexicoker). In the case that coke is produced, the coke
quality produced should remains acceptable (acceptable with respect to further use as fuel
within/outside refinery or as material for other purposes). Many oily sludges can be sent to a
coking unit where they become part of the refinery products. In the case of delayed coker, the
sludge can be injected into the coke drum with the quench water, or injected into the coker
blowdown contactor used in separating the quenching products.
Cross-media effects
Typically reduction of the quality of the coke produced.
Applicability
The quantity of sludge or waste that can be sent to the coker is restricted by coke quality
specifications which may limit the amount of solids in the coke. Coking operations can be
upgraded, however, to increase the amount of sludge that they can handle.
The quantity of solids injected depends on the slurry’s solids content, being typically 10 to
15 %. Loading rates exceeding 82 kg of oil-free dry solids per tonne of coke have been
established. Higher rates are achievable. At this loading rate, the coke is below anode-grade
specifications, but within fuel-grade specifications. Cokers in general are an attractive process-
integrated reprocessing outlet for oily sludges, provided the sludge-on-feed ratio is kept below
1 - 2 %, depending on the required coke quality. If that ratio is not achieved, neither the
operability nor the coke quality is affected.
If sludges from the waste water treatment are included as feedstock to the coker, part of the
water need to be eliminated (e.g. by vacuum evaporation or in the blowdown system) if the
amount of residues would like to be maximised.
Economics
Typical applications recycle 2 to 16 t/d of oil-free dry solids at a total programme cost of EUR
0.6 - 2.5 million/yr.
Example plants
Use of sludge as a feedstock has increased significantly in recent years and is currently carried
out by most refineries.
Reference literature
[147, HMIP UK, 1995]
Description
After the coking gas has been cleaned to remove particulates and to recuperate part of its heat, it
is heated and passed over a catalyst bed in the COS converter, where COS is converted to H2S.
The gas is then cooled and most of the water is condensed. The H2S is recovered from the coke
gas in an amine treater for ultimate recovery of sulphur (see amine treating of refinery fuel gas
in Section 4.23.5.1). Clean low-sulphur coking gas can either be burned in boilers and furnaces
or can be sold as low calorific-value gas. The next figure shows a scheme of that process.
RICH SOLVENT TO
REGENERATOR
COS
CONVERTOR SOUR WATER TO
SWS
Cross-media effects
Increased use of the amine scrubbing system and increased energy consumption of the COS
convertor.
Operational data
For amine scrubbing see Section 4.23.5.1.
Applicability
Amine scrubbing is applied to all types of cokers and COS convertor to flexicokers.
Economics
Economics of the amine absorption of H2S can be found in Section 4.23.5.1
Example plants
These systems can be found in many cokers.
Reference literature
[118, VROM, 1999]
Description
The water used in the cutting/cooling operations is continuously recirculated with a bleed-off to
the refinery waste water treatment. Settling and filtering over a vacuum filter enables the re-use
of this water, resulting in a 'closed water loop'. For water make-up to the quenching and cutting
water loop, various effluent streams can be used. For this duty treated effluent could be used.
Collected water from any condenser, and for draining of wet green coke, should be re-used for
coke drum quenching or drilling. This type of cooling is not used in any other process in the
refinery because of environmental concern. More information about direct cooling (quenching)
is found in Section 2.8.
Cross-media effects
Cutting and cooling process result in vapour emissions (partially recycled in the main
fractionator), significant energy losses, large water use and severe water pollution. Bleeding of
cutting water to the desalter has the potential to add solids to the effluent, so it is best to keep
coker waters separate.
Operational data
The amount of required make-up water depends on the evaporation losses and the bleed to other
processes or to the waste water treatment plant and amounts to about 10 - 20 m3/h for a 106 t/yr
delayed coker unit.
Applicability
Applicable when delayed coking is used. Re-use of water is subjected to restrictions of coke
quality.
Example plants
Many examples of application exist in the world.
Reference literature
[268, TWG, 2001]
Description
Several techniques are suitable to reduce particulate emissions that may occur during the
handling of the coke (green and calcined):
1. providing the green coke is kept wet, the handling between the discharge and the calciner
feed hopper is not problematic. There is the possibility of parts of the receiving pit drying
out although this is limited as the coke is very wet and is covered with residual
hydrocarbons which tend to make it stick together. The construction of wind-breaks around
the pit might be considered. Careful design is essential, as bad design could make the
situation worse. Complete enclosure of the pit and associated handling facilities is another
possibility. The pits are very large and such an enclosure would be very expensive. A
monitoring programme should be carried out to evaluate if this option is in fact necessary or
desirable
2. cut the coke into a double-roll crusher, for example, and convey it to an intermediate storage
silo. The railway waggons are then loaded from the storage silo. This approach would
collect, filter, and dewater the coke prior to storage
3. spray the coke with a very fine layer of oil, which sticks the dust fines to the coke. The use
of oil is limited by its acceptability in further processing. The oil option has the added
advantage of reducing unloading problems
4. covered and de-pressurised conveyor belts
5. aspiration systems to extract or collect dust
6. use of an enclosed hot blowdown system
7. loading areas could be enclosed and positive/negative pressures maintained, exhausting
through bag filters. As alternative dust extraction systems can be incorporated with loading
equipment.
8. collected fines from the cyclones are pneumatically conveyed to a silo with exit air filters.
Dust collection systems are provided for handling, storage and loading, using bag filters.
Collected fines are re-cycled to storage by enclosed means.
Operational data
Additionally, the storage silo provides a buffering capacity for process fluctuations and enables
a controlled loading of railway waggons.
Applicability
3. Oling coke is sometimes practised in fluid and calcined coke, but is rarely applied with
delayed coke.
8. That technique is mostly applicable to calciners, fluid cokers and flexicoking units.
Economics
The investment cost of about EUR 30 million for a typical coking unit with a capacity of 1.5
Mt/yr may be expected for the storage system.
Example plants
Many examples available in European refineries.
Reference literature
[261, Canales, 2000], [80, March Consulting Group, 1991], [117, VDI, 2000], [268, TWG,
2001]
This chapter mentiones where the abatement techniques may be used and describes the
techniques used for abatement of air pollutants.
Description
Flue gases and coke gas from coking processes normally contain coke particulates (including
metals). Dust collection systems are then used in some parts of the coking processes.
1. cleaning of coke gas
2. cleaning of the gas produced during the coke cooling
3. incinerated exhaust gases from the calcinator also produce coke fines. Hot flue gases pass
through a waste heat boiler and gas cleaning by a dust collection system.
In addition to the particulate abatement technique used in the FCC (Cyclones or ESP in
Section 4.5.9) that may be also used here, bag filters can be used in this processes.
Cross-media effects
See Section 4.5.9.
Operational data
See Section 4.5.9. The application of ESP for the control of particulate emission from petroleum
coke calciner have shown some difficulties to achieve the lower end of the range. The main
reason is that coke is a very good conductor of electricity and consequently its surface is very
difficult to be electrically charged and consequently difficult to be catch by the ESP.
Applicability
See Section 4.5.9. The application of high efficiency cyclones are easier to be applied than ESP.
Economics
Bag filters can be used at a cost of about EUR 5 million EUR. Calciner boiler cyclones cost
EUR 225000 (1999). Economics for other type of systems are not available. An economic study
done in a EU+ refinery in 1992 shown that the cost for an ESP for a calciner was high.
Example plants
Many examples are available in European refineries that have coking processes.
Reference literature
[147, HMIP UK, 1995] [60, Balik and Koraido, 1991], [297, Italy, 2000], [272, Shawcross,
2000], [117, VDI, 2000], [268, TWG, 2001]
Description
Sulphur oxides are emitted during the coking processes, especially during the calcination
processes. The principal option to reduce sulphur dioxide releases from the process is the use of
the lowest possible sulphur-content feedstocks. In practice, low-sulphur feeds are typically used
for product quality reasons, since a substantial part of the sulphur remains fixed in the product.
However, this option is not always possible and emissions of SO2 may be significant, especially
during the calcining process. To control those sulphur oxide emissions, the same abatement
techniques that may be used in the FCC may be also used in this process (Section 4.5.10 and
4.23.5.4) with the exception of the DeSOx catalyst additive.
Cross-media effects
Section 4.5.10
Operational data
Section 4.5.10
Applicability
Normally applied to calciner flue gases.
Economics
Section 4.5.10
Example plants
Some examples of these techniques can be seen in some European and worldwide refineries.
Reference literature
[297, Italy, 2000] and references in Section 4.5.10
The same types of techniques used in the FCC can be used in the coker. More information can
be found in Section 4.5.11.
This chapter analyses the ways to prevent emissions to effluents. Techniques used in the final
cleaning of the waters generated in the coking processes are in Section 4.24.
In the coking processes, sour water is generated (steam condensate). Consequently, all water
from the coking process is send to the sour water stripper (Section 4.24.2) before being sent to
the waste water treatment plant.
Description
The proposed pollution prevention alternative was to retrofit the sump where the oil/coke fines
are collected with an inclined plate separator to increase the separation efficiency.
Economics
The annual cost saving associated with the increased recovery of product (coke) and the
reduction in oil/water separation solids was approximately EUR 300000 (7.5 Mt/yr refinery).
Reference literature
[60, Balik and Koraido, 1991]
Description
Coke fines are often present around the coker unit and coke storage areas. The coke fines can be
collected and recycled before being washed to the sewers or migrating off-site via the wind.
Collection techniques include dry sweeping the coke fines and sending the solids to be recycled
or disposed of as non-hazardous waste. Another collection technique involves the use of
vacuum ducts in dusty areas (and vacuum hoses for manual collection) which run to a small
baghouse for collection.
Cross-media effects
Electricity consumption by vacuum ducts/hoses.
Applicability
Fully applicable.
Example plant(s)
Techniques already in use on USA refineries.
Reference literature
[60, Balik and Koraido, 1991]
Description
Because process waters are normally more contaminated than cooling waters, it is important to
maintain the segregation of the two. Only in cases where cooling waters need treatment
(recirculation systems) should they be mixed, and then only at the right point (after the primary
treatment of the process waters).
Cross-media effects
No cross-media effect has been found.
Applicability
Waste water treatment plant need to be sized to handle to the process water flow and the cooling
water flow. Segregation can be very costly in some parts of existing installations.
Reference literature
[107, Janson, 1999], OSPAR, HELCOM recommendations.
Description
See Section 2.8 and BREF on cooling.
Cross-media effects
Tend to produce more noise than water cooling. Noise production due to air cooler fan is
97 - 105 dB(A) at source.
Operational data
A major disadvantage is, that normally a large plot is required when compared to water coolers
(5 – 30 m2/MWth duty). It requires electricity but minimal maintenance cost
Applicability
Air cooling can be sufficient for the cooling needs in some parts of the refinery processes.
Ambient conditions restrict the temperature levels that can be attained. Climatic conditions (hot
climates or temperatures below 0 ºC) typically restrict its use. However the plant location is
independent of water availability. They cannot be located near buildings because air short-
circuiting may occur.
Economics
Air coolers can be expensive. Minimal maintenance cost.
Example plants
Many examples exists in EU refineries.
Reference literature
[316, TWG, 2000], [119, Bloemkolk and van der Schaaf, 1996]
Description
Looses of oil throught leaks can be minimised through continuous monitoring and a kind of
Leack Detection and Repare system (similar as that reference for VOC in Section 4.23.6.1). At
its simplest, this comprises monitoring of the cooling water separators for oil build-up. If this is
observed, then it will be necessary to trace back through the system to identify the source of the
leak so that corrective action can be taken. Detailed system drawings are essential for this
activity. Fingerprinting of the oil can also speed up identification of the leak. A further
refinement is to install oil in water monitors at various point in the cooling water sytem. This
allows leaks to be detercted rapidly and corrective action taken. For this procedure to be
effective, sparing of critical heat exchangers will be required. More information in the cooling
BREF.
Reference literature
[316, TWG, 2000], [119, Bloemkolk and van der Schaaf, 1996]
4.9 Desalting
4.9.1 Good desalting practices
Description
A short description of desalters can be found in Section 2.9. Several good desalting designs
have been included in this section:
1. multistage desalters and the combined use of AC and DC fields provide high desalting
efficiencies as well as energy savings
2. recycling, in multistage desalters, part of the brine effluent water of second desalters to the
previous one minimises the wash water quantity
3. use a low shear mixing device to mix desalter wash water and crude
4. avoid turbulence in desalter vessels by using lower water pressure.
Operational data
Two-stage processes achieve 99 % or better efficiency (more than 99 % of the salts/solids are
removed from the crude oil). This high efficiency provides process benefits because it allows
less corrosion in the topping units.
Applicability
Two-stage or even three-stage desalting is used either if the crude oil salt content is higher than
0.02 %w/w, or if heavy residue is further catalytically processed.
Reference literature
[113, Noyes, 1993], [297, Italy, 2000], [268, TWG, 2001]
Description
Techniques that may be applied are:
1. transfer the water effluent from desalting units to a settling drum where a further separation
between oil and water can be achieved. The oil can be directly recovered in the slop system
2. choice of optimum interface level controllers. As a function of specific gravity and range of
crudes processed, it is possible to consider the most accurate level sensors among
displacers, capacitance probes or radio wave detectors. The accuracy of interface level
control is fundamental for the good operation of a desalter.
3. a good improvement in water-oil separation can be achieved using “wetting”agents, whose
aim is to de-oil the solid contaminants that are usually responsible for a significant
entrainment of oil in the water
4. use of non-toxic, biodegradable, safe to use, non-flamable specific demulsifying chemicals
to promote coalescense of the water droplets.
10 - 20 % less oil is sent to the API separators. The second one can separate off some 5 - 10 %
oil from the water phase.
Cross-media effects
Some of the techniques proposed need the application of chemicals.
Applicability
Fully applicable.
Reference literature
[297, Italy, 2000]
Description
Solids entering the crude distillation unit are likely eventually to attract more oil and produce
additional emulsions and sludges. The amount of solids removed from the desalting unit should,
therefore, be maximized. Then the objective is to minimize solids leaving the desalter with the
crude oil. A number of techniques can be used such as:
1. use low-shear mixing devices to mix desalter wash water and crude oil
2. use low-pressure water in the desalter to avoid turbulence
3. replace the water jets with mud rakes. They cause less turbulence when removing settled
solids
4. the water phase (suspension) can be separated in a pressurised plate separator. Alternatively
a combination of a hydrocyclone desalter and a hydrocyclone de-oiler can be used
5. evaluate the effectiveness of a sludge wash system. The sludge washing is a batch operation
meant to stir the water phase in the desalter to suspend and remove the solids accumulated
on the vessel bottom. This cleaning operation increases the efficiency of desalters during
normal run, particularly for long cycle runs.
Cross-media effects
The application of these techniques increases the sludge generation within the refinery. At a
bottom sediment content of the crude of 0.015 % w/w, theoretically 1500 t/yr sludge can be
collected in a 10 Mt/yr refinery.
Applicability
Where bottoms, sediments and water are critical in downstream process units, desalters are
equipped with a bottom flushing system to remove settled solids.
Example plant(s)
Desalters with desludging facilities are scarce. Atmospheric residue desalters are also few, but
the number is increasing with the increasing complexity of refineries with respect to residue
conversion. In a few refineries, desalters have been equipped with a bottom flushing system
(See design in Figure 2.10).
Reference literature
[297, Italy, 2000]
Description
The desalting process plays an important role in the waste water management in a refinery (see
Section 4.15.7). The water used in other processes can be re-used in the desalter. For instance, if
stripped sour water is used as desalter wash water, the ammonia, sulphides and phenols that it
contains can be reabsorbed by the crude to some extent.
The following process water streams can be suitable for use as desalter wash water:
1. the accumulated water in the crude distillation unit overhead drum, usually 1 - 2 % w/w on
crude feed from steam injection
2. the (unstripped) steam condensates from the light and heavy gasoil dryers and the vacuum
distiller overhead (about 3.5 % w/w on feed)
3. stripped sour water and also other solid-free process water streams. Scrubber or quench
water is dirty and requires oil and solids separation prior to biotreatment and/or re-use as
desalter wash water. The sour water is routed to and stripped in a sour water stripper prior to
re-use and/or final purification in effluent water treatment facilities
4. slowdowns from cooling water and boilers.
Cross-media effects
Recycling of water streams that may form emulsions need to be avoided, because it results in
deterioration of the oil/water phase separation in the desalter, which in turn leads to excessive
oil carry-over to water.
Operational data
The scrubber water would be hard-piped to the desalter water storage tank without being
stripped first, in most cases.
Applicability
Examples of effluents that may form emulsions in the desalter are: bitumen blowing units,
hydrocrackers, cokers (fines stabilise emulsions), other deep conversion facilities (insoluble
metal sulphides that can stabilise emulsions) and HF-Alkylation units (corrosive fluoride
deposits). This option is fully applicable to new refineries and more difficult to apply in existing
refineries. This option is also limited in use when the salinity of the waste water may reach
values that would negatively impinge on the Biox unit.
Economics
The costs of collecting, treating, pumping and piping these waters need to be considered.
Reference literature
[79, API, 1993], [268, TWG, 2001]
Description
Strip desalter brine for hydrocarbons, sour components and ammonia removal before sending
brine to waste water treatment. Recovered hydrocarbons can be blended with several refinery
streams. Acid injection can also be employed to enhance the stripping of oil from emulsions.
Applicability
Pretreatment of desalter brine is normally applied when processing a very heavy crude. In one
refinery: 8.7 Mt/yr refinery produces a waste water flow in the crude unit of 1.3 m3 per minute
containing 90 kg/day of phenols. The water discharge from the desalters contains a maximum of
20 ppm benzene, and the quantity of desalter wash water is equivalent to 4 - 8 %v/v of the crude
charge. The discharge from the stripper bottoms, with benzene levels in the 20 ppb range, is
then routed to the waste water treatment system. A brine de-oling vessel is installed ahead of the
stripping column to decant any free oil undercarry from the desalter. For the above-mentioned
refinery, with this system the phenols can be reduced to an annual emission of 3.29 kg/yr.
Reference literature
[113, Noyes, 1993], [79, API, 1993]
The energy system is an important activity from the environmental perspective. Although heat-
producing plants are an essential and integral part of most refining processes, the systems tend
to be very similar and they are therefore dealt with in this horizontal section. This section
therefore includes a detailed analysis of the energy systems that are specific to refineries, i.e.
only energy technologies that run (partially or totally) on refinery fuels. When a refinery uses a
commercial fuel that can be found on the open market (e.g. natural gas, fuel oil) the emission
levels that can be achieved by a certain energy technique (e.g. gas turbines, boilers) are found in
the BREF on Large Combustion Plants. Techniques to reduce the energy consumption of a
process or activity are addressed in each of the other sections of this document because the
reduction of energy consumption is considered an environmental benefit. However, issues such
as how the energy system can be integrated within a refinery and how the different
processes/activities may share energy are addressed in this section.
Table 4.12 summarises and shows how the main pollutants may be affected by the techniques
that have been considered in this section.
Table 4.12: Possible effect on the main pollutants of the techniques to consider for the
determination of BAT in the energy system
Good design and management of energy systems are important aspects of minimising the
environmental impact of a refinery, bearing in mind the highly integrated and interdependent
nature of most processes. The normal aim is to match continuously the variable production and
consumption of fuels in processes and utilities at the lowest economic and environmental cost.
This section should be seen as an integration of all techniques proposed below, since an increase
in the energy efficiency of a refinery can be achieved through energy conservation technique
and heat integration/recovery techniques.
Description
Some techniques to improve and calculate the energy efficiency within the refinery are
mentioned below.
b) Another way to match production and consumption is to try to improve the energy efficiency
within the refinery. More efficient refineries make better use of the energy produced within the
refinery. To evaluate that, several methodologies exist, including the Solomon energy efficiency
index (the most detailed one), specific energy consumption and (less accurate and more simple)
the index relating the energy consumption to the amount of feedstock processed (Methods
described briefly in Section 3.10.1). Solomon data is also available in that section. Solomon
data are available in a form, that only makes sense in its context, and one needs to understand
certain underlaying (benchmarking) concepts that are proprietary to Solomon.
Techniques to increase the energy efficiency are available in next sections of this chapter.
Cross-media effects
Some difficulties have been identified in the collection of Solomon energy efficiency data from
refineries. The main reason is that the figure is considered as confidential by both refineries and
the provider of the market study (Solomon Associates). Futhermore, Solomon Associates has
not provided to the TWG data on the methodology used to calculate the figures. Moreover, it
was identified that not all refineries participate in this benchmarking exercise because different
reasons (e.g. cost, confidence in their figures was not high). Also not all refineries participate
every year.
Applicability
Fully applicable. The wide range of energy consumption in refineries means that in certain
refineries a considerable reduction in energy consumption can be realised. A concerted and well
managed effort based on an integrated approach in reducing energy use, improved operations,
good housekeeping, adequate management and selective investment are good practices that are
worth highlighting in this context.
Economics
Energy consumption may account for some 50 % of the total operating cost of refineries. As
consequence a reduction in the energy consumption, or an increase in the refinery efficiency,
reduce the total operating cost.
Example plants
Many refineries have installed an energy management system, they issue annual reports on
energy consumption performance and participate in benchmarking of energy consumption.
Worldwide market studies of the wide variety of refineries (configurations/capacities) on the
basis of Equivalent Distillation Capacities has proved useful in helping refineries to compare
their performance.
Reference literature
[118, VROM, 1999], [316, TWG, 2000]
Description
A concerted and well managed effort based on an integrated approach to reduce energy use,
improved operations, good housekeeping, good management and selective investments are the
best practices worth highlighting here. There follows a list of techniques that are considered for
the determination of BAT in the refinery sector. As can be seen, some of them are related to
each other in some way.
Applicability
Especially applicable to those refineries with very high specific energy consumption. The wide
range of specific energy consumptions in refineries (by a factor of 4, see Section 3.10) means
that in certain refineries a considerable reduction in energy consumption can be realised.
Reference literature
[297, Italy, 2000], [118, VROM, 1999]
Description
Actions that can be taken to improve integration and recovery of heat within a refinery to
increase efficiency can be (list do not try to be comprehensive):
• general measures to reduce refinery CO2 emissions, such as optimised heat integration and
improved furnace efficiency combined with computer controlled combustion. This will lead
to lower fuel consumption per tonne of crude processed
• installation of waste heat boiler in heaters
• installation of expanders/power recovery
• expanded heat exchanger areas in heat exchangers where cold streams are preheated by
warm product streams directly from processes
• direct feed of “semiproducts” to processes without cooling and storage. It is always useful,
from an energy conservation point of view, to recover the waste heat of hot products of the
crude distillation unit, for instance, by later feeding them directly to the downstream units,
rather than cooling them for storage and later feeding the downstream units from tankage
• balancing of vapour and refinery fuel gas systems
• use of high-efficiency pumps and compressors
• use of heat pumps
• decreased film temperature and increased turbulence on heat transfer surfaces
• waste heat delivery to adjacent buildings. The identification and use or opportunities for
synergy outside the refinery fence (e.g. district/industrial heating, power generation) may
reduce the cooling needs within the refinery and will reduce the consumption of fuel
somewhere else.
• application of advanced process control to optimise energy utilisation
• insulation of buildings and process units (Minimisation of heat losses by thermal radiation)
• optimisation of energy production (within 4.10.3).
• optimise recycle gas rates, operating temperatures and pressures and steam pressure levels
• minimising or even avoiding slops production and their necessary reprocessing
• keeping heat exchanging surfaces clean or cleaning them regularly (good housekeeping)
• repairing leaks and leaking steam traps
• increase of heat exchanger surface areas of new sections of existing plants (so-called “add-
ons“). In the case more add-ons have been added to an existing plants: re-routing of
intermediate process streams and re-definition of the number of steam pressure levels and
steam duties
being burnt. Heat integration/recovery techniques directly result on lower emissions of CO2,
NOx, particulates and SO2 (see Section 3.10.3 for air emissions from energy system).
Cross-media effects
The interchange of heat between processes implies the transference of disturbances from one
process to another. This can affect safety, so stability control systems may be required.
Applicability
Waste heat is abundant at refineries, and so is low/medium-pressure, low-temperature steam.
Any effort to recover waste heat as low-pressure/low-temperature steam is pointless if there is
no additional use for the extra steam produced. Options for using that heat need to be carefully
quantified and qualified. Heat exchangers require space. The identification and use or
opportunities for synergy outside the refinery fence is some times difficult and need the
compromise not only by the refinery but also for the other part.
Economics
It makes economic sense to maximise heat integration in a plant, and as a result minimise
heating requirements and the cooling system burden. Heat integration/recovery offers an
opportunity for energy cost savings (50 % of total operating cost of refineries), but the cost of
heat exchangers and piping need to be considered when heat integration is analysed.
Example plants
Techniques widely applied in refineries.
Reference literature
[107, Janson, 1999], [118, VROM, 1999], [268, TWG, 2001]
Description
Several techniques have been considered:
• steam used for stripping, vacuum generation, atomisation and tracing is usually lost to waste
water and to the atmosphere. Steam used for mechanical and/or electrical energy generation
and heating is usually recovered as condensate in the HP-, MP-, and LP-condensate systems
and collected in the condensate storage tank. Reducing the amount of steam stripping is an
option to reduce the generation of waste water. Stripping steam is normally used to meet
flash point specification and to improve front end fractionation and yields distribution. In
order to reduce the sour water loads to SWS strippers and reduce chemical treatments in the
overhead systems, an option would be to strip sidedraw products, particularly the lighter
cuts, using reboiled sidestrippers instead of steamed strippers. However, the majority of
steam is used to strip the column bottom, which could not be reboiled by any other means,
so the reduction of condensed steam would be limited anyway; besides, the stripping is far
better under vapour stream than under reboiling conditions, because the more volatile
fraction distillates off
• where inert gas as N2 is available at an economic price, it can be an alternative to steam for
stripping operations, especially for the lighter ends
• optimising steam production by waste heat recovery in waste heat boilers (WHB or CHP)
from hot flue gases (e.g. stacks) and hot product streams (related to the techniques shown in
Section 4.10.1.3).
Example plants
Applications are found in some refineries
Reference literature
[297, Italy, 2000]
As mentioned above, this document offers a detailed analysis only of refinery fuels produced in
the refinery. Whist the use of marketable fuels such as LPG, commercial fuel oil and gasoil in
refineries is a technique to consider that may be included in this chapter of the BREF, such use
is extensively analysed in the LCP BREF [317, EIPPCB, 2002], which also gives achievable
emissions levels using those fuels.
Description
An alternative to reduce SO2, NOx, CO2 and metals emissions from a refinery would be to
replace or to decrease the use of liquid refinery fuel with LPG – often produce on site -, refinery
fuel gas (produced by some conversion techniques) or natural gas. This increase of the use of
gas is typically accompanied with a balance and control of the RFG system between suitable
pressure limits to give the system flexibility, with make-up available from clean fuels such as
LPG or imported gas. In those cases, an state-of-the-art controls to optimise the performance of
the RFG is necessary.
As a matter of summary the benefits for a switch to 100% gas-fired refinery are:
§ SO2 emissions will be reduced to nil for the single unit and for the refinery bubble will
be given by the emissions from other sources (very small from refinery gas, from the
SRU, from the FCCU, the flare etc)
§ particulate emissions including heavy metals will be reduced
§ NOx will be reduced to the levels typical for natural gas firing for energy production
techniques, and consequently other sources as catcrackers will become predominant
emitters in the refinery
§ CO2 emission reduction is achieved mainly because of the lower carbon content of gas,
its higher calorific value and in addition because of a higher attainable efficiency (Stack
gases can be cooled further).
Cross-media effects
It is acknowledged that the residual fuel replaced by gas results in a further surplus of residue,
which has to be considered in any integrated solution for the fuel system of a refinery. It can be
the case that those residual fuels can be burned unproperly outside, so emissions produced by
that can be seen as an emission shift to outside the refinery. Furthermore, the conversion of
heavy fractions into light products and the targets for lower sulphur specifications of fuels
require considerable extra energy. This will lead to an inevitable increase in CO2 emissions. As
a first approximation, NOx emissions can be magnified by the use of hydrogen and residual
fuels containing fuel bound nitrogen. High-hydrogen fuels result in higher flame temperatures,
which typically lead to higher NOx levels. Although not all the fuel nitrogen ends up as NOx
emissions, the fuel NOx contributions can range from non-existent, as in the case of natural gas
fuelled equipment, to several times the thermal NOx contribution of the equipment for refinery
fuels. Refinery fuel gas may contain amines (nitrogen compounds) and other compounds.
Further information on conversion techniques of heavy fractions to refinery fuel gas can be
found in Section 4.10.3.5, coking sections and hycroconversion processes.
Operational data
The move from liquid to gas fuel would require process upgrades and gas grid connections.
Applicability
Some gases are used locally, i.e. in the process of origin or an adjacent process, but most
refineries operate a common RFG main into which most RFG is fed and transferred to the gas
users. In a modern refinery the RFG main is carefully "balanced" with respect to supply and
demand; necessary flexibility is obtained by control of production (i.e. reformer throughput,
LPG evaporation). The relationship to the refinery flare system is important and RFG will
normally include gas received from flare gas recovery. It may also release excess gas to flare if
the upper limit pressure is exceeded. The application of energy conservation concepts (see
Section 4.10.1) can help refineries to meet all their needs with refinery gas produced in-house.
Economics
The cost of switching to gas can be up to EUR 30 million per year for a 10Mt/yr refinery.
Use of LPG in lieu of fuel oil: Approximate capital cost is small (some reburnering) and the
approximate operating cost per year is 120EUR per tonne of fuel (cost differential between LPG
and fuel oil). However, operating costs can vary significantly, depending on the season of the
year and on the price of LPG in the market.
Use of natural gas in lieu of fuel oil: Approximate capital cost of installation is around GBP
4 million. The approximate operating cost per year may vary from below 50 EUR per tonne to
above 100 EUR per tonne (cost differential between natural gas and fuel oil). Again, operating
cost can vary significantly, depending on the season of the year and on the market.
Example plants
The amount of refinery gas plus natural gas combusted by European refineries typically account
from 60 to 100 % (in terms of megawatts fired). However data from some single European
refineries shows that the amount of heavy fuel oil fired can approach to 60 %.
Reference literature
[118, VROM, 1999], [292, HMIP UK, 2000], [317, EIPPCB, 2002], [249, BMUJF, 1999], [268,
TWG, 2001]
Description
Some refinery fuel gases may be sulphur-free at source (i.e. from catalytic reforming and
isomerisation processes) or sulphur-containing at source (most other processes, i.e. from crude
distillation, cracking and all hydrodesulphurising processes). In the latter case the gas streams
are normally treated by amine scrubbing to remove H2S before being released to the refinery
fuel gas system. More information about amine scrubbers can be found in Section 4.23.5.1.
Cross-media effects
Possibility of bottlenecking the amine scrubbing system. More information in Section 4.23.5.1.
Operational data
See Section 4.23.5.1.
Applicability
Fully applicable.
Economics
See Section 4.23.5.1.
Example plants
The treatment of refinery fuel gas in amine scrubbers are commonly used in all refineries.
Reference literature
[297, Italy, 2000]
Description
The nitrogen, sulphur, particulates and metals content of the fuel used in refineries are
determined by the crude that is used at the refinery and by the process units it has passed
through. Liquid refinery fuel streams originate from various processes such as crude distillation
units, vacuum distillation, thermal cracking, catcracking and hydrocracking of residues. Except
for the latter one, the sulphur content of these residues can only be controlled by feedstock
choice. In general, the liquid refinery fuel may comprise one or more of the above-mentioned
fractions and the sulphur content can vary greatly. Table 4.13 shows the sulphur content of
different fractions suitable to be used as liquid refinery fuel.
Table 4.13: Sulphur, nitrogen and metal content of fractions suitable to be used as liquid refinery
fuels
Hydrotreatment of fuels can reduce the sulphur, nitrogen and metal content of the refinery
fractions. Hydrotreatment of liquid fuels can reduce the sulphur content to 0.03 – 1 %. For more
information on hydrotreatments see Section 2.13. This technique is a pre-combustion
operational change, treating the feed before it is used.
Cross-media effects
The hydrotreatment of fuels is a very energy intensive process. Moreover, effluent water and
waste (used catalyst) are generated (see Section 3.13).
Operational data
Information on performance and consumption of hydrotreatments can be found in Section 2.13
and 3.13.
Applicability
Fully applicable
Economics
A severe hydrotreatment is very expensive, increasing the cost of the liquid refinery fuel. Table
4.14 shows an example of cost of liquid refinery fuel desulphurisation process.
Example plants
The amount of refinery gas combusted by refineries compared to liquid fuel, in terms of
megawatts fired, is typically 60 % refinery gas to 40 % liquid refnery fuel. However data from
the public register for 1996 in the UK, shows that, in some refineries, the amount of fuel oil
fired can approach 60 %.
Reference literature
[45, Sema and Sofres, 1991], [292, HMIP UK, 2000], [118, VROM, 1999], [248, Ademe, 2001]
All types of energy production techniques found in refineries are included here. However, the
emission levels associated with each technique are different from those included in Chapter 3 of
this document, because in only good performance levels are included in this section. Techniques
to prevent emissions from those energy production techniques are also included.
Description
Primary measures considered in this section for furnaces and boilers are (see also LCP BREF):
• installation of combustion air preheaters, which would increase the furnace efficiency
significantly (more than 5 %)
• optimise furnace operation, and hence combustion efficiency, by an advanced control of the
operations variables (Air/fuel ratio for the fuel mix, avoid sensible heat losses by optimising
excess air)
• high thermal efficiency heater/boiler designs with good control systems (e.g. oxygen trim)
• minimisation of heat losses via radiation or exhaust gas (e.g. minimisation of heat losses via
unburnt gases (H2, CO) or unburnt residues (ignition loss))
• continuous monitoring of temperature and O2 concentration of flue gas for combustion
optimisation. Monitoring of CO may be also considered
• high boiler pressure
• preheating of fuel charged to the boilers
• preheating of boiler feed water with steam (See also 4.10.3.2)
• prevention of the condensation of exhaust gas on surfaces
• minimisation of own requirements by high efficient pumps, vents, and other equipment
• optimisation of combustion conditions (from 4.15.2)
• Techniques to control CO emissions are:
§ Good operation and control
§ Constant delivery of liquid fuel in the secondary heating
§ Good mixing of the exhaust gases
§ Catalytic afterburning.
Table 4.15 to Table 4.20 provide the achievable emission levels when primary measures are
implemented in furnaces and boilers for each air pollutant. Other measures, such as low- NOx,
flue gas desulphurisation and others are reviewed later in this chapter. Values in the tables are in
mg/Nm3 attainable in continuous operation (half-hourly mean values) and based on a 3 %
oxygen volume in the waste gas, except where specified. Gas: The lower values in the ranges
given below relate to natural gas firing. Liquid refinery fuel: Thermal cracked residue, vacuum
residue, etc.
Table 4.16: Expected CO2 emissions from furnaces and boilers with optimal burner and design
Metals (As, Pb, Cd, Cr, Co, Ni, V and their Gas Liquid refinery
components) (given as the sum of the elements) fuel
Process furnaces 0 5 - 10
Boilers 0 5 - 10
Table 4.19: Expected metal emissions from furnaces and boilers with optimal burner and design
Cross-media effects
Air preheaters typically increase the production of NOx.
Operational data
Direct gas-fired heaters and boilers normally achieve thermal efficiencies of over 85 %. If air
preheat is applied and the combustion products (flue gas) are cooled close to their dew point, the
thermal efficiency can be as high as 90 - 93 %.
Applicability
Fully applicable
Example plants
Each refinery contains many furnaces and boilers of different sizes.
Reference literature
[117, VDI, 2000]], [195, The world refining association, 1999], [45, Sema and Sofres, 1991],
[317, EIPPCB, 2002], [249, BMUJF, 1999], [297, Italy, 2000], [118, VROM, 1999], [268,
TWG, 2001]
Description
Techniques to consider for the BFW production and re-use are:
1) the condensate tank is generally equipped with an oil detection system and an oil skimming
device
2) to avoid corrosion in the steam and condensate systems, oxygen and carbon dioxide are
removed in de-aerators, and oxygen scavengers and corrosion inhibitors are added. For
further conditioning of BFW, dispersing, alkaline and sometimes anti-foaming agents are
dosed
3) minimisation of heat losses via flue gas (unburnt gases, such as H2, CO), via residues
(ignition loss), via ashes and slags, via thermal radiation
4) preheating of boiler feed water (to de-areator) by waste heat
5) reduction of own energy requirement in the BFW production
6) production of boiler feed water: common techniques are ion-exchange, microfiltration and
reversed osmosis. Unlike ion-exchange techniques, membrane processes do not generate
waste water containing high loads of salts. Technologies should be preferred which generate
recyclable residues (e.g. iron-containing sludges). Preferred use of non-toxic chemicals for
water preparation, which are easily degraded by micro-organisms. Use of mineral oil-free
flocculation agents. Use of chemicals with no or only low content of chloroorganic
compounds. Rejection of the following compounds: EDTA (and homologous compounds)
and its salts; aminopolycarbonic acids and their salts; metal-organic compounds; chromates,
nitrites, organic polyeletrolytes with a monomer content >0.1 % w/w. Treatment of waste
water from boiler feed water preparation in a well designed WWTP, especially in case of
regenerates that contain high amounts of NH3 derived from condensate recovery applying
ion exchange. [317, EIPPCB, 2002]
7) conditioning of boiler feed water: Good operation is the combined operation with oxygen
dosage: By addition of ammonia the pH is adjusted to an alkaline milieu and a small amount
of oxygen is added. With this measure, addition of hydracine (which is considered
cancinogenic substance) can be avoided and the demand for ammonia is decreased.
Furthermore a protective layer of magnetite and haematite is formed on the inner surface of
the pipes, which is less rough thereby decreasing the pressure drop within the pipes and
decreasing the energy demand of the pumps. Waste water from water conditioning need to
be neutralised and treated in a well designed WWTP [317, EIPPCB, 2002].
8) repeated reheating of steam
Cross-media effects
Biotreatment may be required if anti-foaming inhibitors are used. Corrosion inhibitors are not
biodegradable in the waste water treatment plant.
Operational data
BFW should not be recycled or routed to the desalter.
Applicability
When the re-use of condensate water as deareator feed water or preheating the BFW is applied,
in some instances, where the condensate and the deareator are located far apart, their integration
is not always possible because economical reasons.
Economics
Site-specific availability and economics determine the choice of the BFW source.
Example plants
BFW production occurs in all refineries, in some measure.
Reference literature
[327, Broughson, ], [317, EIPPCB, 2002], [316, TWG, 2000], [268, TWG, 2001]
Description
A description of gas turbines can be found in the LCP BREF [317, EIPPCB, 2002]. A brief
description can also be found in Section 2.10. Some techniques that can be applied to gas
turbines in order to reduce air emissions are listed below:
• steam injection
• gas turbines with exhaust gas as combustion air
• optimised transformation of steam into electrical energy (highest possible pressure
difference in the steam turbine, generation of steam with high temperature and pressure,
multiple reheating of the steam)
• other primary techniques as dry low NOx burners included in the following sections (4.10.4
to 4.10.6).
• use high-efficiency turbines by, for example, optimisation of the design of the turbines,
reduce as low technically feasible the outlet steam pressure in the back-pressure turbine
Cross-media effects
Steam injection typically produces higher emissions of CO and hydrocarbons. Steam should be
produced if it is not available in the refinery.
Applicability
Fully applicable.
Economics
Steam injection applied to a 85MWe output turbine. Uncontrolled NOx emissions of 500
mg/Nm3 @ 15 % O2. Down to 50 - 80 mg/Nm3 @ 15 % O2. Investment cost (1998) EUR
million 3.4 (including cost of steam production) Operating Cost: 0.8 EUR million (excludes
capital charge).
Example plants
Many examples exist in refineries. A number of refineries have, or are currently installing,
combined cycle gas turbine (CCGT), designed to produce steam and power for the refinery.
This is usually done to replace in whole or part old oil-fired boiler plant, to reduce operating
costs and to decrease reliance on other power generators.
Reference literature
[45, Sema and Sofres, 1991], [118, VROM, 1999], [115, CONCAWE, 1999].
Description
A brief description can be found in Section 2.10.
Cross-media effects
No cross-media effects have been detected.
Operational data
Most turbines require a particularly stable mix of fuel in order to be sure of flame stability and
are basically designed to burn natural gas. Refinery fuel gas components can vary considerably,
especially when surplus hydrogen is produced, such as when a hydrotreatment unit is
temporarily shut down, resulting in excess hydrogen being sent to the fuel gas system. However,
these problems can usually be overcome, up to a limit of about 70 % hydrogen in the fuel.
Applicability
Generally applicable. The steam and power cogeneration concept can be also be applied to
boilers firing, for instance, liquid refinery fuel. They can be designed to generate high-pressure
steam and to let the pressure down over an expander/turbo-generator. Economisers and the
optimisation of air-to-fuel control are also techniques applicable in cogeneration plants.
Example plants
A number of refineries have, or are currently installing, combined cycle gas turbine (CCGT) or
Combined Heat and Power (CHP) plant, designed to produce steam and power for the refinery.
This is usually done to replace in whole or part old oil-fired boiler plant, to reduce operating
costs and to decrease reliance on other power generators.
Reference literature
[118, VROM, 1999]
Description
Integrated gasification combined cycle (IGCC) is another technique and its purpose is to
produce steam, hydrogen (optional) (see Section 2.14) and electric power from a variety of low
grade fuel types with the highest conversion efficiency possible. More information can be found
in Section 2.10.
The IGCC is a highly integrated and efficient process which can supply power, hydrogen and
steam. Furthermore, it offers in principle an acceptable outlet for heavy residues and feedstocks
or even refinery sludges, provided the latter are less than 1 % of feed. Hot gas clean-up systems
have the potential to increase system efficiency and lower system costs. The achieved emissions
to the atmosphere from that system are: SO2 50 mg/Nm3, NOx 65 mg/Nm3 @ 3 % O2,
particulate 5 mg/Nm3 and carbon monoxide 10 - 30 mg/Nm3.
The emissions from the IGCC show a significant decrease compared with conventional
power/steam plants. The SO2 concentration in the refinery exhaust is reduced by 80 % but CO2
emissions increase.
The use of by-product and residual streams to meet the fuel requirement of refineries cannot
only be cost-effective, but also is environmentally beneficial in that it makes use of what would
otherwise be a waste refinery stream that would be flared without recovering the energy content.
Cross-media effects
In some cases, some difficulties may occur in burning low calorific value gas produced. Water
effluent is normally sent to the existing waste water treatment plant of the refinery. It may
contain significant amounts of metals as V, Cr or Ni and PAHs.
Operational data
Utility requirements for the gasification processes are 1800 - 4900 kWh/t of power and 1140
kg/t of steam consumption. The soot product is about 50 - 75 % w/w V2O5 residue, which can
be sold to metal reclaimers. The IGCC complex is also equipped with all the necessary auxiliary
systems, including cooling water (mixed system with an open seawater circuit for large users
and a closed clean water circuit for the other users), demi-water, air, nitrogen, water and fuel gas
networks, fire fighting, flare, storage, electrical distribution, buildings, etc.
IGCC is a technology with high flexibility in start-up, shutdown and part-load operation,
depending on the level of integration between the different sections. Typically, systems that use
heat exchange equipment are more efficient than those using quench cooling; however, the
capital cost of the system with heat exchange is higher and there is a risk of fouling. Handling of
the soot and filter cake should be done with care to avoid dust (even at 80 % moisture) due to
the toxic properties of the residue.
Applicability
This technique can be seen as an alternative approach to removing sulphur using feed
hydrotreatment (see Section 4.10.2.3). During normal refinery operation the gasifiers of the
IGCC plant are able to convert almost any refinery residue (atmospheric residues, vacuum
residues, visbroken or thermal tars, etc.) to heat and power. These feedstocks can have a high
sulphur content.
Economics
Table 4.22 gives the economics of two examples of IGCC applied in European refineries.
Example plant(s)
The IGCC concept as such is quite a recent technological application for power stations. The
main sections of an IGCC plant, i.e. Gasification, Air Separation, Gas Cleaning and Combined
Cycle are well known technologies that have previously been used separately for different
applications and with different feedstocks. Integration (the “I” in IGCC), though, is a more
recent idea. Oil gasification is a process that has been applied for many years. Gasification of
heavy residues, according to the principles of IGCC, is rather new. At least 4 IGCC plants are
already in operation within European refineries and some others are in the design/construction
phase. Coal gasification, also applying the IGCC concept, is new as well and a few plants are in
operation.
Reference literature
[45, Sema and Sofres, 1991], [297, Italy, 2000], [320, Italy, 1996]
Description
An alternative method for disposal of heavy residual oils or petroleum coke is combustion in a
fluidised boiler with limestone injection for sulphur capture.
Cross-media effects
The resulting calcium sulphate and unconverted calcium oxide together with the nickel and
vanadium in the fuel are discharged from the boiler as a solid residue which can be used as road
aggregate or disposed of to landfill.
However, such schemes have a lower sulphur capture performance than gasification and they do
not provide the option of producing hydrogen. There could also be environmental objections to
the mining and transport of limestone and disposal of the residue. For these reasons, gasification
may be generally more attractive in the long-term.
Applicability
A combination of fluidised bed boilers with upstream solvent deasphalting or delayed coking
can be a cost effective solution for refineries with existing FCCU capacity and steam/power
deficiencies.
Economics
Typically cheaper than gasification.
Reference literature
[118, VROM, 1999]
Techniques to reduce NOx emissions fall into two broad categories. Primary techniques include
NOx control techniques, such as pre-combustion operational changes and combustion
modifications. Secondary techniques include the post-combustion flue gas treatments or NOx
abatement techniques. More information about NOx abatement techniques can be found in
Section 4.23.3. Primary and secondary techniques are considered in this section. The types of
NOx control and abatement techniques to be considered in the energy systems are summarised
in Table 4.23.
Table 4.23: NOx control and abatement techniques considered to be applied into energy systems
Description
Low- NOx burners, either air staged or fuel staged, have the aim of reducing peak temperature,
reducing oxygen concentration in the primary combustion zone and reducing the residence time
at high temperature, thereby decreasing thermally formed NOx Staging of fuel addition is also
thought to provide a reburning effect, further reducing the NOx Ultra-low-NOx burners add
internal recirculation of flue gases to the features of the low-NOx burner, enabling further NOx
reductions.
Table 4.24: NOx emissions achieved with low and ultra-low-NOx burners for different type of
equipment
Cross-media effects
For oil firing there is a direct link between NOx and particulates i.e. reduction in NOx as the
flame temperature falls will lead to an increase in particulates. For low-NOx fuel oil burners, as
with conventional fuel oil burners, a further reduction of thermal NOx results in an increase in
carbon particulates. CO emissions are also increased.
Operational data
Refinery gas firing with ultra-low-NOx burners in both forced or natural draft conditions may
show signs of instability at some points, particularly at low turndown and low excess air. Care
needs to be taken during the installation of these burners. Burner testing to explore the limits of
combustion prior to site installation is highly recommended for reliable operation and to
evaluate that this technique is applicable in specific case.
Applicability
Application is straightforward for new installations of both fired heaters and boilers. Some
liquid fuels are not suitable for the latest generation of low-NOx burners and some older fired
heaters are fitted with large high intensity burners which cannot be retrofitted with new
low-NOx burners. Retrofitting of low-NOx burners depends on the furnace design and may be
simple, difficult or, because of the increased flame volume, impossible without changing the
furnace. For instance the increased length of low-NOx burners may restrict applicability in
furnaces built low above-ground. NOx abatement on older furnaces and boilers may also be less
effective due mainly to the need to avoid flame impingement on the furnace tubes.
Economics
The following tables show the economics of several applications of low- and ultra-low-NOx
burners in different refineries.
Example plants
There are many examples of applications in refineries.
Reference literature
[115, CONCAWE, 1999], [45, Sema and Sofres, 1991], [112, Foster Wheeler Energy, 1999],
[107, Janson, 1999], [348, Ashworth Leininger Group, 2001]
Description
More information can be found in the LCP BREF.
Table 4.25: NOx emissions achieved with dry low-NOx combustors for different types of
equipment.
Applicability
Applicable to gas turbines. Dry low-NOx combustors are not available for gas turbines fired
with refinery blend gas that contains more than 5 to 10 % v/v of hydrogen.
Economics
The investment cost is EUR 2.2 million (1998) and the operating cost is nil for an 85MWe
output turbine.
Reference literature
[115, CONCAWE, 1999], [316, TWG, 2000]
Description
External flue gas recirculation (FGR) is applied to boilers and heaters to increase the diluent
effect, hence to reduce combustion temperature. Typically 20 % of the available flue gas from
the boiler stack is ducted to mix with fresh combustion air.
Table 4.26: NOx emissions achieved with flue gas recirculation for different types of equipment
Operational data
The process is difficult to control especially during turndown.
Applicability
It is applied to boilers and fired heaters. In a boiler retrofit, FGR increases hydraulic loads, and
shifts the heat load towards the convective section(s) and may not be practical.
Economics
Higher cost than other primary measures.
Reference literature
[115, CONCAWE, 1999], [316, TWG, 2000]
Description
Inert diluents, such as flue gas, steam, water and nitrogen added to combustion equipment
reduce the flame temperature and consequently the concentration of NOx in the flue gases.
Table 4.27: NOx emissions achieved with diluent injection for different types of equipment
Cross-media effects
Energy to produce steam, higher emissions of CO and hydrocarbons.
Operational data
The use of steam generates more corrosion in the system than the use of nitrogen.
Applicability
Steam and water injection is widely applied to gas turbines both in new installations and
retrofits and is also applicable to fired heaters and boilers (see also Section 4.10.5.2). There are
technical difficulties in applying water injection in boilers and furnaces. Nitrogen dilution is
only applied when nitrogen is already available in the refinery.
Economics
Capital cost of the steam and water injection is less than that of SCR, making the technology a
good first choice for substantial levels of NOx reductions, with SCR often added on if higher
NOx reduction is needed. Substantial recurring operating costs are, however, encountered for
producing high purity steam, and maintenance costs for re-blading may be high.
Other sources give that operating costs in a gas turbine using 80 t/h steam injection would cost
EUR 0.9 million per year.
Example plants
By-product nitrogen from the air separation plant in refinery residue gasification projects has
recently been commercially demonstrated as a diluent for gas turbine NOx reduction. Within the
refining industry, steam injection predominates.
Reference literature
[112, Foster Wheeler Energy, 1999], [115, CONCAWE, 1999], [268, TWG, 2001]
4.10.4.5 Reburning
Description
Fuel staging, also called reburning, is based on the creation of different zones in the furnace by
staged injection of fuel and air. The aim is to reduce NOx emissions, which have already been
formed back to nitrogen. This technique adds to the flame cooling a reaction by which organic
radicals assist in the breakdown of NOx.
Applicability
It is applied at the burner level.
Reference literature
[112, Foster Wheeler Energy, 1999]
Description
This technique also called THERMAL deNOx reduces the NOx that has been formed in the
combustion process. This process is a non-catalytic process which utilises ammonia or urea to
reduce NOx to nitrogen and water. More information can be found in Section 4.23.3.
Table 4.28: NOx emissions achieved with SNCR for different types of equipment
Cross-media effects
Risk of NH3 or urea emissions (storage and non-reacted). Turndown is a problem. The
emissions of N2O may be increased by this technique. A side reaction of particular concern is
the formation of ammonium sulphates when firing sulphur-containing fuels such as liquid
refinery fuel. Sulphates give rise to the fouling and corrosion of downstream equipment. Storage
of gaseous ammonia has a great hazardous potential. Therefore a liquid solution of ammonia
(25 %) is preferred. The use of urea causes higher emissions of CO and N2O and can cause
high-temperature corrossion.
Operational data
More information can be found in 4.23.3.2.
Applicability
SNCR requires temperatures above 650 ºC. Application is complicated when flue gases to treat
come from the combustion of heavy fuel oil. Space requirements are mainly restricted to
ammonia storage.
Economics
Table 4.29 shows the economics of application of SNCR for the treatment of flue gases coming
from the combustion processes from different sources.
Values corresponding to a 100 GJ/h Fired Heaters and Boilers Firing Residual
installation – Retrofit of existing Boilers Firing Fuel Oil
installation. Refinery Blend Gas
Down to ppm @ 3 % O2 50 100
Investment cost (1998) (EUR million) 0.4 - 0.5 0.4 - 0.9
Operating costs per year (excludes 0.025 0.05 - 0.07
capital charge) (EUR million)
EUR/unit EUR/yr
Note: installation on a volume of exhaust gas of 250000 Nm3/h after a power plant in a
refinery achieving a NOx emission reduction of 500 mg/Nm3 in relation to actual oxygen
content and for a clean gas concentration of <200 mg NOx/Nm3
Table 4.29: Examples and main cost factors for a Selective Non Catalytic Reduction (SNCR)
Investment cost for a SNCR applied to an existing boiler was EUR 0.56 million including
equipment for storage of urea. Storage of gaseous ammonia is even more expensive than urea
storage.
Example plants
This process has been applied in heaters and boilers of refinery plants. The experience of
application of SNCR on oil-fired heaters is more limited. An USA study reports that SNCR is
less used than SCR for control of NOx and that only 12 of the 150 boilers/heater installations of
8 refineries use this type of technique.
Reference literature
[107, Janson, 1999], [115, CONCAWE, 1999], [45, Sema and Sofres, 1991], [250, Winter,
2000], [348, Ashworth Leininger Group, 2001]
Description
Another secondary technique is known as catalytic deNOx. As in THERMAL deNOx, ammonia
is used to reduce the NOx to nitrogen and water vapour. The ammonia vapour is mixed with the
flue gas and the mixture passed through a catalyst to complete the reaction. More information
can be found in Section 4.23.3.
Table 4.30: Power plant of the Mider refinery in Germany, raw gas and clean gas concentration
Cross-media effects
Risk of NH3 emissions when operating outside stoichiometric conditions and catalyst disposal.
The risk of ammonia emissions during storage of gaseous ammonia can be minimised if
ammonia is stored and used as an aqueous solution (25 %). It may increase the emissions of
N2O. Operational slip of NH3 is around 5 ppm, a value that normally increases during the
catalyst’s lifetime. A side reaction of particular concern is the formation of ammonium
sulphates when firing sulphur-containing fuels such as liquid refinery fuel. Sulphates give rise
to deactivation of the catalyst and to fouling and corrosion of downstream equipment.
Operational data
More information can be found in 4.23.3.2.
Applicability
The application of this technology to existing process furnaces can be limited by problems of
space, pressure and temperature. However SCR is considered in California technologically
feasible in boilers and heaters. Oil fly ash contains metal oxides, soot and coke. Unabated fly-
ash concentrations for oil firing are in the range of 100 - 600 mg/Nm3 (with highest values for
vacuum residues). SCR applied under these conditions may suffer from clogging by fly ash and
sulphates (with coal firing, the sandblasting effect of the fly ash keeps the catalyst clean). The
potential for sulphate precipitation is generally higher with vacuum residue, due to its high
sulphur content (2.5 - 4 %).
Economics
Costs of SCR vary over a range, depending on the fuels, exhaust volume and required reduction.
The range may be EUR 25 - 110/kW for new power generation installations. Maintenance costs
are minimal, the bulk of which are from NH3 consumption. Some examples of the application of
SCR in refinery power plants are shown in Table 4.31 and Table 4.32. The use of a SCR-unit
after the power plant has been estimated in raw-gas and clean-gas operation mode. The two
cost-tables are given below.
EUR/unit EUR/yr
Endurance (years) 15
Catalysts 19379
Note: Installation on a volume of exhaust gas of 250000 Nm3/h in the dedusted off-gas
after a power plant in a refinery achieving a NOx mission reduction of 500 mg/Nm3 in
relation to actual oxygen content and for a clean gas concentration of <100 mg NOx/Nm3
Table 4.31: Main cost factors for a Selective Catalytic Reduction (SCR)
EUR/unit EUR/yr
Endurance (years) 8
Catalysts 63589
Note: Installation (raw gas) on a volume of exhaust gas of 250000 Nm3/h after a power
plant in a refinery achieving a NOx emission reduction of 500 mg/Nm3 in relation to actual
oxygen content and for a clean gas concentration of <100 mg NOx/Nm3
Table 4.32: Main cost factors for a Selective Catalytic Reduction (SCR)
Example plant(s)
They have been applied to FCC off-gases, gas turbines, process boilers, process heaters such as
naphtha reformers, steam reformers, crude and vacuum distillation units, thermal cracking and
hydroprocessing units. SCR has been also successfully applied to gas streams with high
particulate content in high sulphur coal-fired utility industry and FCC units combined with
upstream particulate removal. A report on USA refineries show that the 20.7 % (or 31
boilers/heaters) of 150 boiler/heater installations at 8 refineries employs SCR as a control
technology. The same study shows that 3 SCR installations have been applied to the 4 turbines
found in the 8 refineries analysed.
Reference literature
[181, HP, 1998], [107, Janson, 1999], [45, Sema and Sofres, 1991], [250, Winter, 2000], [115,
CONCAWE, 1999], [348, Ashworth Leininger Group, 2001]
Description
A separate spray tower is usually required prior to the main SO2 absorption spray tower. An
additive applied in this extra tower oxidises the NO to NO2, which is then absorbed with the
SO2. The reaction for the absorption of NO2 is 3NO2 + H2O à 2HNO3 + NO.
Cross-media effects
If using wet gas scrubbing, slurry can be expected as a solid waste from this process. Nitrates
generated should be treated in the WWTP.Because the gas effluent from this process generates
NO, it may need further reduction of NOx emissions.
Economics
Wet deNOx scrubbing has advantages including lower capital cost, if installed in conjunction
with SO2 wet gas scrubbing.
Examples
It is normally used in nitric acid production.
Reference literature
[108, USAEPA, 1995]
Description
The technology uses a single catalyst that operates in two cycles: oxidation/absorption and
regeneration. The catalyst works by simultaneously oxidising CO to CO2, NO to NO2, and then
absorbing NO2 onto its surface through the use of a potassium carbonate absorber coating. The
regeneration of the catalyst is accomplished by passing a controlled mixture of regeneration
gases across the surface of the catalyst in the absence of oxygen. The regeneration gases are
steam, hydrogen and carbon dioxide. Consequently the nitrates are reduced to nitrogen. An
additional catalyst layer is required to reduce SO2 emissions and preserve the CO- and NOx -
reducing catalyst when sulphur is present in the fuel.
Because the regeneration cycle should take place in an oxygen-free environment, a section of
catalyst undergoing regeneration must be isolated from exhaust gases. This is accomplished
using a set of louvres, one upstream of the section being regenerated and one downstream.
During the regeneration cycle, these louvres close and a valve opens, allowing regeneration gas
into the section.
Cross-media effects
The system emits CO2, H2O, N2 and trace levels of SO2 to the stack. Due to masking and
poisoning of the catalyst, annual catalyst cleaning is required with de-ionised water and a
potassium carbonate solution (K2CO3). The spent cleaning fluids can be neutralised and
disposed of through the sewerage system and are harmless to the water and soil. Precious metals
on the spent catalyst have residual value and can be reclaimed. There is no waste disposal issue
caused by spent catalyst.
Operational data
This system requires electricity for control systems, valves and actuators, natural gas for
regeneration gas production and steam both for regeneration gas production and as a dilution
carrier. The typical pressure drop is between 8.5 mbar and 15 mbar. The system can operate
effectively at temperatures ranging from 150 °C to 370 °C. The utility requirements for a 25
MW gas turbine are steam (333 to 389 ºC): 1590 kg/h and natural gas: 14 kg/h.
Applicability
Applicable to both new and retrofit applications, the system can operate effectively at
temperatures ranging from 150 °C to 370 °C and does not limit gas turbine performance. Such
unit can be installed at the back of the boiler or in the Heat Recovery Steam Generator within
the same envelope reserved for a conventional SCR system.
Economics
The estimated cost numbers presented refer to a typical 400 MW size combined-cycle gas-fired
power plant. The cost estimate numbers below are based on 8000 h/yr of operation and a NOx
reduction from 25 ppm to 5 ppm (50 to 10 mg/Nm3 as NO2 at standard conditions 0 ºC; 1013
mbar), which equals approx. 666 tonnes annually of NOx removed. Included in the numbers are
investment cost, operation & maintenance cost and indirect annual costs.
A reduction of NOx from 25 ppm to 2 ppm (50 to 4 mg/Nm3 as NO2 at standard conditions 0 ºC;
1013 mbar) will contribute to an increase in the investment cost due to the need for additional
catalyst. It will also somewhat increase the operation & maintenance cost-due to increased
consumption of natural gas and steam and increased pressure drop.
A second example gives an installation cost of EUR 6.2 million for the application of this
technique to a gas turbine of 25 MW. The operating costs are estimated to be approximately
EUR 0.42 million/yr, including maintenance, steam and natural gas, pressure drop through the
system, and the cost of catalyst replacement.
Example plants
There are example applications in USA in small gas power plants.
Reference literature
[276, Alstom Power, 2000], [268, TWG, 2001]
The next table provides some examples of combination of NOx control and abatement
techniques that may be applicable to refineries.
Reference literature
[115, CONCAWE, 1999], [268, TWG, 2001]
Particulate loads (containing metals) from refinery combustion plant are normally fairly low
unless heavy residues are burnt. Those particulates may be abated using a number of methods
including electrostatic precipitators (ESP), bag filters and wet scrubbing, as already described in
Section 4.23.4.
Description
Refinery fuel gas and hydrotreated liquid refinery fuels contain less particulate loads than heavy
liquid refinery fuels. Switching to those fuels has a positive impact in the reduction of
particulates emissions. The achieved emission particulate levels as well as cross-media effects,
operational data, applicability and economics have already been described in Section 4.10.2
Description
Steam injection together with the liquid refinery fuel to lower the particulate concentration.
Applicability
Steam injection in fired furnaces/boilers is widely practised with liquid fuel.
Reference literature
[118, VROM, 1999]
4.10.5.3 Filters
Description
See Section 4.23.4.
Cross-media effects
Pressure drop. Filter material life is limited and may add to disposal problem.
Operational data
Limited utilities (e.g. compressed air) are needed. For more information see Section 4.23.4.
Applicability
Bag blinding can arise from "sticky" soot particles from boilers when burning fuel oil. Filters
are normally used for cleaning flue gases with flows below 50000 Nm3/h.
Economics
Investment cost details depend on installation. Operating costs are low, but the filters normally
need to be changed every 1 - 2 years.
Example plants
There are many power plants in Europe operating with bag filters.
Reference literature
[250, Winter, 2000], [118, VROM, 1999]
Description
For more information see Section 4.23.4.
Cross-media effects
Electricity consumption. More information in Section 4.23.4
Operational data
ESPs consume electricity. More information in Section 4.23.4
Applicability
ESPs are widely applied in FCCs, in thermal power plants and incinerators. They can be
installed both in new and existing plants. They require space for their application.
Economics
Typical installed capital costs for ESPs are from 1 - 3.8 EUR million.
Examples
Wet ESPs are the most common type used for the abatement of particulates in heavy fuel oil
combustion processes.
Reference literature
[247, UBA Austria, 1998], [45, Sema and Sofres, 1991]
The release of sulphur dioxide is directly linked to the sulphur content of the refinery fuel gas
and fuel oils used. Techniques to reduce sulphur oxide emissions are fuel optimisation
(described in the 4.10.1), fuel desulphurisation (hydrotreatments described in Section 4.10.2.3)
or flue gas desulphurisation. The lower the sulphur content of the fuel, the lower the emission of
sulphur dioxide. In this section, techniques for the capture of SO2 from the flue gas after
combustion or flue gas desulphurisation techniques and the use of additives are included. More
generic information on those techniques is given in Section 4.23.5.4.
Description
Use of lime or limestone as additive to fuels to capture the sulphur oxides.
Cross-media effects
Limestone/ gypsum mixture has to be disposed of.
Applicability
For conventional refinery boiler and furnaces, the injection of lime or limestone into the furnace
for sulphur capture cannot be applied. If the heaters and boilers are fluidised bed or circulating
fluid boilers to burn petroleum coke or heavy bitumens resulting from solvent deasphalting,
then lime/limestone injected into the furnace as abatement would be practical.
Reference literature
[45, Sema and Sofres, 1991]
Caustic soda scrubbers can reduce emissions of NOx, SO2 and particulates significantly, as well
as removing CO2 as carbonate. A report suggests that SO2 could be cut by up to 67 % and NOx
by up to 47 %.
SNOx process
Particulate removal rate less than 10 mg/Nm3 at ESP outlet
NOx removal rate 90 - 94.7 %
SO2 removal rate 94 - >96 % with 5 %v/v O2 and a temperature of 410°C inlet the
SO2 converter.
Cross-media effects
Name of Flue gas desulphurisation Regenerative flue gas desulphurisation
technology
Other impacts Increased energy consumption, by- Increased energy consumption, possible
product, raw material handling bottlenecking of H2S handling facilities
Operational data
More information can be found in Section 4.23.5.4.
Applicability
More information can be found in Section 4.23.5.4. It can be applied to both new and existing
installations.
Economics
Annual costs for a caustic soda scrubber would range from GBP 1.1 million to nearly GBP 7
million per refinery.
The SNOx plant is designed for a flue gas load of 1.0 million Nm3/h at the exit of the existing
induced draught fans. Investment cost EUR 100 million
Example plants
Wellman Lord system after three oil fired heaters of the power plant is used in Austria. The
Gela Refinery SNOx plant was designed for cleaning the flue gas produced in the boilers at a
power plant burning a mixture of high-sulphur Fuel Oil and high-sulphur Petroleum Coke. The
Mider refinery in Germany uses a wet desulphurisation process with lime after the three oil fired
heaters of the power plant. The fuels used in this plant are vacuum residues, visbreaker residues
and FCC slurry.
Reference literature
[112, Foster Wheeler Energy, 1999], [45, Sema and Sofres, 1991], [292, HMIP UK, 2000],
[258, Manduzio, 2000], [297, Italy, 2000]
This scrubbing uses the natural alkalinity of the seawater to remove SO2. Information about
these techniques is already included in Section 4.23.5.4.
4.11 Etherification
4.11.1 Catalytic distillation
Description
More information in Section 2.11.
Operational data
The advantages of the process that include the reaction catalyst within the distillation column
include essentially complete isoolefin conversions. It can be applied for MTBE, ETBE or
TAME production. Conversions of 98 % of isobutylene are typical for refinery feedstocks.
Conversion is slightly less for ETBE than for MTBE. For TAME, isoamylene conversions of
95 %+ are achievable. It requires an additional reactor.
Catalyst handling in elevated places during installation and catalysts change may cause safety
risks.
Economics
Investment, EUR 25000 per m3/d of MTBE product.
Reference literature
[212, Hydrocarbon processing, 1998]
Description
Water effluents from MTBE and TAME production contain releases of methanol, formic acid
and ethers. Those compounds or degradation products of them can be toxic for the fauna of the
biotreater. So high concentrations of those components need to be prevented from entering. This
can be easily controlled by the use of a storage tank or by planning of production in order to
dosify the right flow to the waste water treatment.
Applicability
Fully applicable.
Economics
The implementation of this technique is not expensive.
Reference literature
[272, Shawcross, 2000]
Description
Ethers produced in the etherification processes are very soluble in water. If they leak may
contaminate the soil, raining water and ground water. It can be seen as a part of a programme
for prevention, detection and control of the soil and groundwater (See Section 4.25.1)
Reference literature
[349, Finnish Environmental Institute, 2001]
Description
The gas plant is a relatively simple process and the temperature levels are not high enough to
create the opportunity for steam generation. No great efficiency improvements are to be
expected within a gas plant, but heat integration with upstream plants will normally make use of
e.g. reboiling heat available from upstream plant streams, saving direct heating capacity in the
gas separation plant.
Cross-media effects
No cross-media effects.
Example plants
Technique widely applied.
Reference literature
[282, Conoco, 2000]
Description
Re-sizing/improving condensate knock-out system on gas streams & on-site recycling.
Applicability
Relatively minor modifications.
Reference literature
[115, CONCAWE, 1999]
Description
Gas plants process light compounds and are consequently more likely to produce fugitive
emissions. The use of double mechanical seals on pumps, compressors and agitators can reduce
VOC emissions. Use of low-leaking valves also can contribute to reduce VOC emissions. This
topic is described as an integrated measure within Section 4.23.6.1 because fugitives occur in all
refinery processes, but it is mentioned here explicitly because the light compounds used lead to
a relatively high level of fugitive emissions.
Cross-media effects
Section 4.23.6.1.
Operational data
Section 4.23.6.1.
Applicability
Section 4.23.6.1.
Economics
Section 4.23.6.1.
Example plants
Programmes of minimisation of VOC emissions can be found in many European and non-
European refineries.
Reference literature
Section 4.23.6.1.
Description
Molecular sieve dryers used in LPG production are regenerated using hot fuel gas.
Applicability
Easily applicable.
Example plants
Many examples exist.
Reference literature
[18, Irish EPA, 1992]
Storage and handling of odorants used in the LPG production is going to be covered in the
horizontal BREF on Storage. However the dosing of odorant to the LPG product need to be
handled in a proper way to prevent emissions to the environment. This technique is included in
the Section 4.21.21.
Description
See Section 2.13. Another technique that can be used is the use of heat recovery from high-
temperature process streams in WHB and power recovery in the high-pressure units (letting
down liquid) to increase the heat integration.
Cross-media effects
See next technique in this chapter.
Operational data
See Section 2.13.
Applicability
Fully applicable.
Economics
The investment associated with the installation of a hydrotreating unit depends on the feed
characteristics and the product specifications. Generally speaking, as the feed gets heavier or the
individual product specifications are reduced, the processing requirements are increased. These
more severe processing conditions can result in more pieces of equipment, larger equipment,
and higher operating pressure, all of which increase the cost of the unit. The required capital
investment for a hydrotreating unit can vary from EUR12 to EUR48 per tonne per year of
capacity.
Example plants
Many examples exist in refineries.
Reference literature
[166, Meyers, 1997], [112, Foster Wheeler Energy, 1999], [212, Hydrocarbon processing, 1998]
Description
See Section 2.13. Another technique that can be used is the use of heat recovery from high-
temperature process streams in WHB and power recovery in the high-pressure units (letting
down liquid) to increase the heat integration.
Cross-media effects
Energy consumption, waste generation, effluent water and air emissions.
Operational data
Gasoil hydrotreating units need the additional measure of a high-pressure amine washer in the
hydrogen recycle stream to avoid recombination of H2S from the recycle hydrogen stream with
hydrocarbons from fresh feed. The unit will operate at 45 bar, utilizing 40 Nm³ Hydrogen/t of
gasoil feed and anticipating a run length of 30 months.
Applicability
Applicable to distillates from naphtha to heavy residues.
Economics
The estimated erected cost for a reduced-crude desulphurisation process (basis 132 m3/h unit)
unit is EUR 47 millions. Another estimated investment summary for residue desulphurisation
system hydrotreater to prepare residue fluid catalytic cracking feed is shown below.
refiners do not operate a hydrocracker unit. Two-stage units are also mandatory for refiners who
need to lower the aromatics content and/or the CFPP in their gasoil pool.
Example plants
Many desulphurisation processes exist in refineries.
Reference literature
[166, Meyers, 1997].
Description
Catalytic distillation for desulphurisation of gasoline in a two stage process.
Cross-media effects
Energy consumption, waste generation, effluent water and air emissions.
Applicability
Fully applicable. This technique may also be used for the reduction of benzene in the reformate.
Economics
A two-stage process is designed to treat 7950 m3/d of FCC-gasoline containing 1800 ppm of
sulphur. Its efficiency has been of 95 % and its investment costs are approximately EUR 20
millions.
Example plants
At least one catalytic distillation plant for desulphurisation of 7950 m3/d of FCC-gasoline is
available in Europe.
Reference literature
[247, UBA Austria, 1998]
Description
While the main objective of this process is to extend catalyst life in a down-stream fixed-bed
reactor by maintaining high hydrodemetallisation performance, the catalyst also achieves high
hydrodesulphurisation/hydrodemetallisation and hydrodesulphurisation Conradson carbon
removal/hydrodemetallisation activity ratios.
Operational data
Reactor typically operates at 390 ºC and 2000 psia.
Applicability
Fully applicable.
Example plants
Till 1997, there were 15 on-stream worldwide.
Reference literature
[166, Meyers, 1997]
Description
The hydrogenation of light dienes serves the purpose of producing a more stable refinery stream
reducing upstream, gum formation. The process converts dienes to useful olefins. More
information in Section 2.13.
Cross-media effects
Low energy consumption.
Operational data
Low temperature, liquid-phase operation means that no utilities are required in most cases if the
feed and hydrogen are both available at suitable conditions. Heating or cooling duties are
limited to cases with relatively high diene concentrations in the feed. Commercial operation for
a period of more than two years without the need to regenerate or replace the catalyst is typical.
Applicability
The design of the units is simple. Most of the units are designed for diene reduction to less than
5 ppm in the product.
Economics
The design of the units is simple and requires low capital and nominal operating costs. The cost
of the catalyst is low compared with the installation cost. The capital investment, including
catalyst, for the unit is small, usually within the range of EUR 0.6 - 1.2 million. The investment
base on recent projects is of some EUR 3 millions for a plant of 100000 t/yr capacity.
Reference literature
[166, Meyers, 1997], [261, Canales, 2000]
4.13.6 Hydrocracker
Description
Hydrocracking is an exothermic process and heat integration techniques can be applied. For
more information on hydrocrackers see Section 2.13. Some techniques to consider are:
a) The heat generated in the reactors can be partially recovered in a feed/product heat
exchanger. A furnace heats the feed to its required temperature. The reactor temperature is
controlled via the injection of cold hydrogen between the catalyst beds.
b) A significant amount of heat is required in the fractionation section. Heat integration is
applied to minimise heat consumption.
c) The energy efficiency can further be increased by applying a four-stage separator system.
The feed to the fractionation section is in that case rendered significantly hotter, and
consequently less heat is required in the fractionation section.
d) The use of heat recovery from high-temperature process streams in WHB and power
recovery in the high-pressure units (letting down liquid)
Cross-media effects
No cross-media effects known.
Operational data
Control system need to be enhanced.
Applicability
In some cases, the heat integration can be difficult to apply. Replacement of existing
hydrocrackers by new and highly heat integrated hydrocrakers is in many cases not justified
environmentally and economically.
Economics
The cost of a hydrocracker with heat integration varies from 36 - 84 EUR/(t/yr). Annual
operating cost varies between 0.6 and 1 % of the investment cost. In many cases the economics
of the process do not justify the replacement of an existing hydrocracker, because payback
period can be very long.
Reference literature
[166, Meyers, 1997], [212, Hydrocarbon processing, 1998], [112, Foster Wheeler Energy,
1999], [268, TWG, 2001]
Description
See Section 2.13. Another technique that can be used is the use of heat recovery from high-
temperature process streams in WHB and power recovery in the high-pressure units (letting
down liquid) to increase the heat integration.
Cross-media effects
Cross-media effects are increased energy consumption and a consequent increase in CO2
emissions, due mainly to the hydrogen. As discussed in Section 3.14, the production of
hydrogen is very energy-intensive, generating large quantities of CO2. As discussed in Section
4.13, hydrotreatment processes generate catalyst for disposal and increase H2S production with
direct consequences (enlargement or new) for the sour water stripper and sulphur recovery units.
Other cross-media effects are the increase of waste generation (catalysts) and effluent waste
water (containing H2S).
Operational data
Hydroconversion is an exothermic reaction, the heat generated in the reactor system being
partially recovered in the feed product exchanger. Normally a high level of heat integration and
heat recovery is applied in the fractionation section. The amount of hydrogen required depends
on feed (heavier feeds tend to increase hydrogen consumption) and process objectives (lower
sulphur, nitrogen, aromatics content in the products, higher hydrogen consumption). In most
cases extra energy is required.
Applicability
Commercial experiences range from low-sulphur atmospheric residues to high-metal, high-
sulphur vacuum residues with over 300 ppm metals.
Economics
Investment costs for the various options depend strongly on the feed properties and process
objectives of the residue hydroprocessing unit. Investment costs for a typical new single string
5000 t/day unit will range from 200 - 300 million US$, the higher figure including an integrated
hydrocracker.
Example plant(s)
Many different concepts of residue hydroconversion processes are licensed. The selection of the
type of process depends on feedstock quality (metal content and Concarbon number) and the
required conversion and product quality.
Reference literature
[115, CONCAWE, 1999], [212, Hydrocarbon processing, 1998]
Description
• Gas Heated Steam reforming replaces the reformer furnace: the reforming heatload is
provided by cooling the exit gas from a secondary reformer which uses oxygen. The feed
consists of natural gas or light naphtha.
• The steam reformer has to supply a large amount of heat at high temperature for the steam
reforming reaction by burning fuel, with the result that a large amount of heat is lost in the
flue gas. In order to minimise the amount of heat lost in the stack, a large heat recovery
system is used. A large proportion of the recovered heat is by raising and superheating
steam. Some heat integration is possible around the solvent absorber and the methanator.
Cross-media effects
The environmental regulations with the most impact on plant design are typically NOx
limitations. Other impacts, such as SOx or water emissions, are minimal, because low-sulphur
fuel is typically used and there are few emissions other than flue gas. The choice of heat
recovery system can have a major effect on NOx production, since both the amount of fuel fired
and the flame temperature will be affected. Preheating combustion air will reduce firing, but
since NOx formation is strongly influenced by flame temperature, there will be an overall
increase in NOx formation. Other methods of reducing firing, such as pre-reforming or heat
exchange reforming, do not affect the flame temperature and will therefore reduce NOx
production. The production of CO2 is also important. Each tonne of H2 produced also produces
a minimum of 12 tonnes of CO2 (7.5 tonnes if related to the steam generation).
Operational data
The combination of high temperature and pressure imposes severe demands on the steam
reformer tubes, which must have thick walls to withstand the stresses and must be fabricated
from expensive alloys. The furnace box must be large to obtain the high rates of heat transfer
from the fluegas to the tubes. These combined factors make the steam reformer the largest and
most expensive single item of equipment in a hydrogen plant. The large size of the unit means
that it takes a long time to heat up and cool down, and this makes it the most time-consuming
part of the start-up and shutdown procedure. The large thermal inertia makes it liable to damage
during plant upsets.
Applicability
Fully applicable.
Economics
Relative hydrogen production costs of the two processes are primarily a function of feedstock
cost. A new hydrogen plant for a 7950 m3/day FCCU feed hydrotreater or hydrocraker would
typically cost in range of EUR 60 - 75 millions (1997).
Example plants
A new alternative, already proved commercially for ammonia and methanol production, is the
Gas Heated Steam Reformer Technology.
Reference literature
[112, Foster Wheeler Energy, 1999]
Description
The IGCC plant can also function as a hydrogen supplier, in which case hydrogen is removed
from the syngas (after the sulphur removal). For more information see Section 2.14, 2.10 and
4.10.3.5.
Cross-media effects
Partial oxidation requires an oxygen plant (95 - 99 % purity), which increases costs.
Operational data
Normally, an IGCC plant includes a water purification section with the following units:
• soot recovery either by washing with naphtha and recycle of naphtha/soot mixture to
gasification section and/or by filtration.
• water pretreatment to remove solids (carbon, metals, salts) by filtration before final
treatment in biological plant (GTW)
The grey water treatment (GWT) is designed to treat water blowdown from a carbon extraction
unit with syngas condensate and possibly the amine regeneration overhead system; water from
grey water treatment is finally sent to biological treatment. The GWT may employ physical and
chemical treatment for cyanides destruction and removal of heavy metals followed by a sour
water stripper, which removes sulphides, ammonia, and carbon dioxide from the water.
The gaseous streams are totally washed to remove any contaminant and the process normally
includes the following ancillary units for gas purification:
• washing with water and sometimes oil to remove soot and particulate
• reactor for COS or cyanides by hydrolysis
• sour gas absorbing with amine or equivalent system and production of elemental sulphur
from H2S in Claus unit.
Applicability
Relative hydrogen production costs are primarily a function of feedstock cost. The main
problem with investment in gasification plant is its capital and operating cost. To be attractive
commercially it usually needs to be undertaken on a large scale. Typical requirement for
chemicals production are the generation of more than 200 MWe of electrical power through
IGCC or the use of hydrogen, carbon monoxide and steam on a large scale.
Economics
Generally accepted investment costs for an oil gasification based combined cycle plant in the
200+ MWe range are 1300 - 1700EUR per kW installed, with a thermal efficiency of around
40 % and close to 99 % sulphur emission abatement. As average electrical demand in a refinery
is usually less than 80 MWe, the investment in IGCC often depends on the opportunity to export
the surplus of electrical power generated. However high pressure steam can also be a valuable
gasification product and there are examples of much smaller gasification units that fit in well
with a refinery.
Relative hydrogen production costs of the process are primarily a function of feedstock cost.
Steam reforming of methane usually produces hydrogen at lower costs than partial oxidation of
fuel oil, if the cost of methane is less than about 65 % of fuel oil cost on an energy basis.
Example plants
At least five gasification processes are currently running in European refineries.
Reference literature
[166, Meyers, 1997], [297, Italy, 2000]
Description
Information about the processes can be found in Section 2.14. Some techniques that may be
applied to the hydrogen purification units to achieve better environmental performance are:
• several adsorber beds are used, and the gas flow is periodically switched from one vessel to
another to allow regeneration of the adsorbent by pressure reduction and purging, thus
releasing the adsorbed components. The desorbed gas is normally accumulated in a vessel
and used as fuel at a convenient location
• one technique to reduce air emissions is to use PSA systems only for purification of
hydrogen
• use the PSA purge gas as refinery fuel gas in the reforming furnace instead of higher C/H
ratio fuels.
Cross-media effects
Because of loss of hydrogen to the PSA tail gas, the reformer and front end of a PSA plant are
larger than in a wet scrubbing plant. However, a PSA plant uses less process steam and does not
require heat for the reboiler. PSA is a high energy consumer due to the high vacuum/pressure
systems that it uses.
Operational data
The PSA system is completely automatic and offers a higher purity product. The PSA unit is
simpler to operate than a wet scrubbing system, since it has no rotating equipment or circulating
solutions. The conventional process produces a hydrogen product of maximum 97 - 98 % v/v
purity, while the pressure swing adsorption process produces a very pure hydrogen product
between 99.9 % v/v and 99.999 % v/v. The residual constituents of the product gas are mainly
methane and less than 10 ppm CO. Some data on the utility consumption of the PSA system are:
3700 Nm3/h with a H2 content of at least 85 % and a pressure between 37 and 47 bar. Product-
stream: 2400 Nm3/h purified H2 with a concentration of at least 99.5 %, Temperature: 45°C,
Pressure: 35 bar; waste gas: 1300 Nm3/h, P: 3 bar, T: 30°C, contains 60 % H2, 1.4 %H2S and
40 % C1 - C6 compounds with higher boiling points.
Driving force for implementation
Production process to purify the hydrogen. The choice between the PSA system and the
scrubbing system depends on product purity required, reliability and process economics. The
extra investment for the PSA system for large plant capacities can normally be offset by the
lower operating cost. Also a PSA system will yield a higher H2 purity.
Example plants
Many examples exist.
Reference literature
[166, Meyers, 1997], [297, Italy, 2000], [211, Ecker, 1999]
Description
An environmental management system (EMS) can contain the following elements:
• the implementation and adherence to international accepted system such as those from the
ISO 14000 series or to EMAS. Those systems ensure the continuous improvements,
complaints handling training, performance reporting, process control and improvement
planning etc.
• proof can be enhance by e.g. internal and/or external auditing, certification.
• EMS does not stand in isolation from the other systems (safety, maintenance, finance etc.)
but is to be integrated with the other systems where relevant and beneficial. More
specifically the management systems for Energy and Environment can be combined, which
is in line with the fact that a reduction in energy consumption goes hand in hand with
improvement of environmental performance.
• the preparation and publication of an annual environmental performance report, to be
externally verified. Such a report will also enable the dissemination of improvements to
others, and will be a vehicle for information exchange (to neighbouring environments, to
authorities etc).
• The delivery to stakeholders on an annual basis of an environmental performance
improvement plan. Continuous improvement is assured by such a plan.
• It is common practice in the quest for continuous improvement to compare own
performance with that of others in the field, and to identify the current “best practices”.
Taking these “best practices” from others to heart and applying them will lead to
environmental performance improvements. Systems are available for comparing own
performance with others for energy, for efficiency, for maintenance, in which performances
of refineries with different capacities and complexity are normalized. (e.g quivalent
Distilling Capacity). This technique of performance measurement, comparisons with others
and identification of “best practices” and the adoption in one’s own workplace is also called
“benchmarking”. This technique has scope for application in a number of areas of
environmental performance improvements as well.
• practise of such benchmarking on a continuous basis, including energy efficiency and
energy conservation activities, emissions to air (SO2, NOx, VOC, and particulates),
discharges to water and generation of waste. Benchmarking for energy efficiency typically
involve an internal system of energy efficiency improvements, or intra- and inter-company
energy efficiency benchmarking exercises, aiming for continuous improvements and
learning lessons.
• perform an Environmental Impact Assessment (EIA) for major new activities.
§ mass balance data on sulphur input and output via emissions and products (including low-
grade and off-spec products and further use and fate), to be reported on an annual basis.
Applicability
Those techniques can be typically applied to all refineries
Example plants
Many of those techniques are applied in many refineries
Reference literature
[118, VROM, 1999], [285, Demuynck, 1999]
Description
The bubble concept usually refers to air emissions of SO2, but can also be applied to NOx, dust,
CO and metals (Ni, V). The bubble concept is a regulatory tool applied in several EU countries.
As represented in the picture below, the bubble approach for emissions to air reflects a “virtual
single stack” for the whole refinery. The treatment of process water in the refinery is typically
done in a single waste water treatment plant (for technical and economical reasons) and
conceptually can be regarded as a similar approach to the “bubble concept”4 for air.
The technical justification of the bubble concept in refineries is the complexity of the
installation. Aspects of this complexity include the number of emission points and the technical
and economic interrelationships in a modern refinery between feedstock, types of processes and
the variations in different operating conditions in relation to product slate and product quality
requirements.
The economic justification is that the refinery should be allowed to use, at different plants and at
different times, the various fuels it produces in-house. At the same time, it enables the operator
to:
4
The requirements on a refinery to reduce its discharges to water can be also conceived as a bubble approach.
Water from various process units is collected into one water treatment system. The load in g/tonne crude oil
processed in the refinery can be defined in terms of a “bubble”, aimed at the minimisation of water consumption,
re-use of water in different processes and BAT at the waste water treatment unit in an integrated way and for all
processes together with the objective to reach as low a load as is technically and economically feasible for the
whole refinery.
The environmental justification is that the authority should be allowed to define emission
conditions in line with its environmental targets while allowing the refinery operator flexibility
in the operation of its processing units and in the selection of the various technical and
economic options to fulfil the energy requirements of the refinery in competitive market
conditions. Moreover, it allows easy comparison of the emission performance of a refinery and
between refineries.
2. Different approaches (point 2 above) can be used to define the bubble, i.e:
§ in a concentration units of the flue gas emitted by the refinery, whereby the complexity
of the refinery is reflected in the energy consumption of the refinery, or
§ in a annual load per tonne of crude oil processed which can be combined with the
complexity factor (Nelson index) of the refinery
The load bubble includes energy efficiency of a refinery and the concentration approach
does not. This is the major advantage of the load bubble. There are simplified factors (See
Table 3.2) for switching between the two approaches. It should be notised that such a factor
calculated on an average European energy consumption value could, however, be
misleading because energy efficiencies in European refineries vary greatly (See
Section 3.10.1). The conversion between the two approaches in each specific refinery is
more accurate.
Other methods may be used. It is more a question of convenience and custom than of
principle how and in which units of measurement the bubble is defined, as long as it is clear
which approach is used, which methodology is used to define the unit of measurement and
which units are included.
3. There are different interpretations in various countries of the units to be considered (e.g.
combustion only, including or excluding FCC and/or SRU and/or flares). As an example, a
Member State has reported that for an existing EU+ refinery the bubble concentration
increases by 20% when flares are included in the bubble. Another example was given by
Concawe were the calculated average of the SO2 fuel bubble was 1000 whereas the average
was 1600 if all installations are included within the bubble.
4. Self-explanatory
5. When the short averaging time is defined (e.g. daily, monthly), additional flexibility for
outages, changes of feedstock and other technical problems may be granted by allowing the
high values in the current period to be corrected by lower values in the next period, when
the technical problems have been sorted out. The mass bubble limit (g/tonne crude oil
processed) is typically defined over a long period (e.g. one year). In one EU+ refinery, a
period of three years is applied. Also, it has been recognised by the TWG that the averaging
time may depend on the pollutant. (e.g. environmental and health effects of NOx and SOx
are different). The averaging period is also related to the stringency of the bubble emission
level (the more stringent the levels, the longer the averaging period, to allow flexibility).
The above paragraphs imply that a large amount of data is needed to continuously calculate
the bubble value. For example, for SO2:
§ RFG sulphur contents need to be continually measured (e.g. by mass spectrometer) and
hourly values produced; SO2 is then calculated by recording hourly fuel flows to
furnaces and boilers.
§ Fuel oil S content is measured daily.
§ FCCU figures are estimated based on a percentage of the S feed to the unit (the daily
flow of feed to the unit; the S content is measured after any change of crude slate). This
needs months of data collection to establish a relationship.
§ SRUs, calciners etc all have different algorithms to estimate releases continously based
on data collection.
§ Data are also needed to continuously estimate moisture and O2 content in the flue gases
to correct to the reference conditions.
The last paragraph indicates the scope for considerable variability in the way refiners
calculate the bubble and thus in the accuracy and reliability of the data when comparing
bubble values. However when a good sample schedule is establish together with an
approved calculation method, the follow up of the emissions become very easy.
Environmental concerns
As in some situations refineries are physically or logistically connected to a nearby chemical
plant, there may be a temptation to extend the bubble to the “non-refinery” processes which then
may create misuse of the bubble concept because of dilution effects. These dilution effects can
occur due to the high volumes of flue gases generated by some non-refinery processes that
typically contain low levels of SO2 and NOx. An example of such dilution was submitted by a
Member State where a refinery bubble profits from a large chemical plant and the bubble
concentration is reduced by 50% (included in the LVOC BREF). Dilution effects do not affect
the load approach.
A second concern may be that the concentration bubble is defined as a relative standard, directly
related to the fuel use in the refinery, and therefore not enhancing efficient use of energy in the
refinery. This can be addressed by defining the bubble according to a load per tonne of crude oil
processed or by defining an absolute ceiling (in tonnes/year), in addition to emission values in
mg/Nm3 flue gas in the bubble. However, there is a possible trade-off between the stringency of
the environmental requirements and the need for flexibility in the operation of a refinery.
Special care should be taken not to introduce additional complexity into the regulatory system,
thereby losing part of the attractiveness of the bubble concept. For example, in a MS a 3-year
average is chosen for the SO2 bubble (load) for the refinery. In the case of NOx, however, the
differences in long-term environmental and health effects of SO2 and NOx need to be borne in
mind.
Another possible environmental concern with this concept, is the possibility of high emissions
at the lower stacks (e.g. SRU). This can easily be redressed by requiring that the flue gases of
the cleaner fuels be discharged at the lower stacks and the high stacks are used to avoid high
concentrations of pollutants at ground level, affecting workers and the neighbourhood. Of
course, the regulator should be aware that traditional measures used to ensure that ground level
concentrations of hazardous pollutants are within Air Quality Standards are still required when
using the bubble concept. This comment can be extended to all legislation, because the use of
the bubble concept as a regulatory tool does not prevent other existing/new limit values for
specific units/facilities within the refinery for being applied.
This may imply different emission values associated with the use of BAT, depending on the
type of fuels used and their contribution to the total refinery fuel use, together with emissions of
process units such as the FCC and the SRU and their contribution to the total refinery emissions.
Associated emission values in the case of a gas-fired refinery will be inherently lower than the
emissions in a situation where a major part of the fuel use of the refinery is provided by liquid
fuels. In this latter case BAT may be defined by assessing the emissions associated with the
gaseous fuels used in combination with the emissions associated with using liquid residual fuels
and flue gas treatment, or using clean liquid fuels, i.e. gas oil with limited S-content. If deemed
necessary, these emission values for a bubble based on concentrations can easily be translated
into a yearly or monthly load bubble, i.e. tonne of SO2 and/or NOx per million tonnes of crude
oil processed.
In Annex V, several examples show how, emission values for the bubble concept can be defined
which reflect the SO2 and NOx performance that is technically feasible for the refinery as a
whole. The assessment of BAT-associated emission values involves however also an economic
and an environmental assessment of these values. The investment decision to convert an
existing refinery into a full conversion, gas-fired refinery involves much more technical,
economic and environmental considerations than only the SO2 and NOx emissions from the
refinery itself. Similarly, the decision to invest in flue gas desulphurisation and/or NOx
abatement via SCR may also involve the concentration of liquid fuel firing of the refinery in one
or two larger combustion plants, which will have major implications on the way the refinery
will be able to respond to commercial pressures. The same holds true for the other technical
options that a refinery will be faced with. However, maximum flexibility will be high on the
agenda in assessing the various technical considerations open to the refinery. By defining
emission values in the bubble that respect this need for operational flexibility, the permitting
authority can support the proper environmental outcome of the investment decision process by
the refinery
The bubble concept is an environmental regulatory instrument that may be used to enforce the
application of BAT in the refinery without becoming entangled in the complex technical-
economic interrelationships in the refinery. As mentioned before, there is a link to the BAT
concept provided that the bubble is set at a level that reflects BAT performance of the refinery
as a whole.
Examples
TWG has provided some reference numbers when bubble concept is applied to refineries. Data
has been divided in two groups. One group represents actual performance of refineries and the
second one represents proposals provided by different TWG member(s).
Data on legislation applying bubble concept in some countries are provided in Annex I.
Reference literature
Austria, Belgium, Concawe, France, Greece, Italy, Netherlands, Norway, Spain, Sweden, UK,
[248, Ademe, 2001]
Description
This section includes good practices applied to maintenance, cleaning and other horizontal
issues within refineries. What was considered “good housekeeping” in the past may nowadays
seem self-evident, as it is embedded in day-to-day refinery practice. New measures, based on
experience gained, are however included over time. “Good housekeeping” is often used in the
broader context of good management and conduct of the business. Quite a number of business
performance parameters, such as safety, maintenance, efficiency (of energy and manpower),
product quality, information, unit/equipment reliability, human resources and finance require
approaches that start with a concept of “good housekeeping”. For refineries, safety provides
numerous examples of this approach, often grouped under the heading Safety Management.
Similarly, Product Quality Management stands for a well-developed approach to assure
customer satisfaction and Maintenance Management for improved unit reliability/on-stream
time and maintenance effectiveness. There are also systems for the other business parameters.
All these systems are in place to assure that the correct practices are adopted, and that the
“lessons learned” are reflected in these systems (e.g. ISO 9000 series).
The more traditional aspects of good housekeeping are included in this section. These systems
can also be regarded as the tools to ensure that the right procedures are adhered to in order to
safeguard the proper functioning of the refinery hardware. The following techniques are
considered to be good environmental practices within the industrial sectors.
• Regular cleaning of equipment and refinery premises. Cleaning, by its nature, generates
waste. By choosing the right procedure and technique this waste may be minimised or its
nature altered so as to make it more easily disposable:
§ drain equipment to the maximum
§ recycle "spent" rinse water
§ use high pressure water cleaning rather than detergents or chemicals
§ use on-site pretreatment whenever possible, e.g. wash/steam filter material (e.g. filter
clays) prior to dumping
§ minimise tank sludge prior to cleaning (solvent & mixers)
§ during cleaning, vents and vessel steaming to be directed to flare. This technique may
raise serious safety issues if oxygen is present. In addition when Hg is present –
particularly when steaming overhead coolers- mercury emissions may occur.
§ in many refineries, using high-pressure water to clean heat exchanger bundles
generates and releases water and entrained solids to the refinery waste water treatment
system. Exchanger solids may then attract oil as they move through the sewer system
and may produce finer solids and stabilized emulsions that are more difficult to
remove. Solids can be removed at the heat exchanger cleaning pad by installing
concrete overflow weirs around the surface drains or by covering drains with a screen.
Other ways to reduce solids generation are by using anti-foulants on the heat
exchanger bundles to prevent scaling and by cleaning with reusable cleaning
chemicals that also allow for the easy removal of oil
§ cleaning at designated places in order to have control over run-off material and
contaminted water
§ prompt clean-up of minor spillages with adsorbents.
§ install weirs at cleaning places, in order to hold back solids, which otherwise might
attract oil in CPI separators
§ minimise and re-use cleaning chemicals.
• Use techniques in common industrial equipment (pumps, compressors, etc.) that can reduce
the environmental impact as for example:
§ implementation of operational improvements in new designs
§ equipment reliability improvement
§ use lubrication technique based on the lubrication by an oil mist. The technique
consists of a vortex generator, which generates, with the help of dry pressurised air, a
mixture of air and lubricant oil, This mixture is called oil mist, and lubricates the
dynamic equipment. Once lubricated the air is recirculated to the vortex.
Some techniques that may be considered as good housekeeping techniques are already included
in other sections. Those include for example:
• application of advanced process control (see Section 4.15.5) that can include the on-line
continuous (equipment and process) performance measurement, the measurement of
performance and comparison with targets
• prevention of soil contamination (see Section 4.25.1)
• leak detection and repair (LDAR) (see Section 4.23.6)
• minimisation and prevention of spillage of hydrocarbons (see Section 4.23.6)
• minimisation of waste generation (included in each of the process sections).
Operational data
These systems are dynamic; they are kept alive by the incorporation of continuous improvement
opportunities.
Applicability
The lubricant technique can be applied to lubrication of dynamic equipment with lubricated
bearings, e.g. centrifuge pumps, electrical engines, gears assemblies, intermediate supports.
Economics
Less than two years pay-off is claimed by the technology provider in the lubricant technique.
Example plants
The lubricant technique has been already used by the main petrochemical companies
Reference literature
[118, VROM, 1999], [285, Demuynck, 1999], [19, Irish EPA, 1993], [316, TWG, 2000], [324,
Sicelub, 2001]
4.15.4 Training
Staff training plays an important role in the prevention of contamination within the refinery.
Some important aspects are:
• include environmental concern and environmental issues in the training of the refinery staff
• train staff to reduce solids in sewers. A facility training programme which emphasizes the
importance of keeping solids out of the sewerage systems will help reduce that portion of
waste water treatment plant sludge arising from the everyday activities of refinery staff
• train staff to prevent soil contamination. Contaminated soil can be reduced by educating
staff on how to avoid leaks and spills.
Reference literature
[118, VROM, 1999], [256, Lameranta, 2000]
Description
Some techniques to consider in this respect are:
• pollution control equipment should, in general, be kept running during start-up and
shutdown, for as long as is necessary to ensure compliance of authorisations, provided no
other safety or operating considerations prevent it
• reduce the start-ups and shutdowns to a minimum. Proper production planning can reduce
the frequency and duration of process shutdown
• implemention of advanced process control for the planning and control of production
• advanced process control to optimise energy use within the refinery (related to 4.10.1.2).
Reference literature
[118, VROM, 1999]
Safety is an important cross effect of environmental measures. Any direct relation of influence
exists. Measures may affect safety in a:
§ positive way. Closed loop sampling minimizes accidental releases of material
§ no way. The process water recycle to desalters has no relevant impact on safety
§ negative way. Gas recovery systems often have to deal with explosive atmospheres, where
the untreated offgases are outside the explosibility limits. Negative impacts on plant safety
can reach such an extend, that the additional dangers exceed what a plant is willing to
handle. Even by others reasons promising techniques may lose applicability by that criterion
– at least, until significant improvements of the safety properties of such techniques are
developed.
Reference literature
[118, VROM, 1999], [285, Demuynck, 1999], [302, UBA Germany, 2000], [260, Sandgrind,
2000]
Studies for the optimisation of water, effluent and drainage are often relevant. Similar to the
identification of energy saving opportunities, water pinch studies can be carried out to identify
process water integration options and water reduction and re-use opportunities. In most
refineries some internal water streams are already used as desalter wash water, such as
condensate water and steam-stripped sour water. There is scope for increased water reduction
and re-use in refineries, which will lead to reduced size and costs of both water make up and
end-of-pipe treatment facilities.
The principle of WSI is maximising prevention, reduction, recycling and re-use of process
water, rainwater, cooling water and sometimes contaminated ground water with the aim of
minimising the amount of process water for end-of-pipe treatment. WSI need to be carried out
as part of a Water Masterplan for the refinery. This plan aims at optimisation of the use and re-
use of all water streams in the refinery. For greenfield refineries and in existing refineries, when
new treatment installations are installed, a win-win situation is created when the amount of
imported water, the size of effluent treatment facilities and the discharge are all minimised.
Ideally it may become feasible to use treated effluent as a source for make-up of process water,
cooling water and/or boiler feed water. Every refinery has its specific optimum WSI scheme.
Reduction of fresh water use is a goal for most refineries for two main reasons. Firstly, fresh
water, particularly high quality water, is a valuable resource which in many parts of Europe is
becoming increasingly scarce. Where water of lower quality is used, treatment to acceptable
standards also requires the use of energy and chemicals. Secondly, water used has to be
discharged. Minimising the volume of water discharged reduces the size of effluent treatment
plant required, the amount of energy and chemicals used, and the quantity of pollutants
discharged to the environment. Data on actual fresh water usage is included in Section 3.15. As
a general guide, approximately 0.1 - 5 m3 of waste water (process waste water, cooling water
and sanitary waste water) per tonne of crude is generated when cooling water is recycled. The
process water generated in a refinery is of the magnitude of 0.1 to 0.2 m3 per tonne of crude oil
input. Some techniques to consider for the minimisation of fresh water consumption, can be the
following:
Concept Description
The concept or methodology to design an effluent/water integration system for a refinery is to
inventorise water demand and effluent production of each refinery unit operation and to assess
the water losses. This inventory provides the water balance across the refinery. It is
recommended to visualise the water balance by a block scheme as presented as an example in
Figure 4.5. The next step is to identify all possibilities to reduce the amount of water required
and to re-use as much as possible by matching (quantity and quality) effluents by means of a
water optimisation scheme. This “water-pinch” methodology includes giving consideration to
the re-use of treated effluent. The WSI concept aims at “closing the water loop”.
Spent caustic
FLARE/TANK ~5
DRAINS DESALTER CDU
~ 20 STRIPPERS ~ 20
SPECIAL ~ 55
UNITS 5-20 SOUR WATER VDU / VBU
(MTBE, HF- ~ 10 ~ 10
STRIPPERS
Alkylation, or more
BBU, lubes
groundwater)
FCCU
~ 20 ~ 20
SLOPS 5
SYSTEM
AMINE REG.
~5 SRU & TAIL ~5
GAS TRTMT.
Total
CPI PPS HDS/HYDRO
process drains
~ 120 ~ 20 TREATERS
~15 ~ 90
Figure 4.5: Example of a process water integration scheme for a catcracker refinery
configuration (10 Mt/yr)
Notes: dotted lines only to indicate the hypothetical “closed loop”. Figures within the graph are
water flows in tonnes per hour.
WSI is dependent on the refinery configuration, the crude quality and the level of desalting
required, the cost of potable water, the availability of rainwater and the quality of cooling water.
Within a refinery a number of standard process-integrated effluent/water treatment provisions
are available as well as a number of standard possibilities for water reduction and re-use. In
most refineries a number of these options have already been implemented to some extent, either
in the original design or by retrofit:
(1) in a refinery a number of clean condensate streams are generated from condensed steam
that has not been in contact with product. These condensate streams are suitable for direct
re-use as boiler feed water (BFW), leading to cost savings
(2) the standard provisions for process water treatment are the sour water stripper (SWS, to
treat the sour water produced in various process units) and the desalter, which is a major
water consuming unit whereas the water quality requirement in the desalter does not
constrain the use of SWS effluents. Sour water is produced if wash water and/or steam
comes into contact with hydrocarbon products which contain H2S and NH3. In the SWS
the H2S and NH3 are largely eliminated. This renders a treated effluent into a suitable
quality wash water for the desalter
(3) the re-use of stripped sour water as wash water for crude desalting is a first step in any
WSI. A crude desalter uses 5 - 10 % water on crude, dependent on the crude quality and
the level of desalting required
(4) process units that produce significant quantities of waste water are the Crude Distillation
Unit (CDU) and the Fluid Catalytic Cracker (FCC). In the CDU overhead, 2 - 3 % waste
water on crude intake is produced as steam condensate. This water does not require SWS
treatment but can be directly routed as wash water in the desalter. An FCC-unit, if present
in the refinery, also produces significant quantities of waste water in the range of 5 - 10 %
on feed. This amount can be reduced to 2 - 4 % if cascading of wash water is
accomplished. Stripped sour water can in principle be used as wash water and, on the
other hand, the FCC wash water can either be recycled to the SWS or used as desalter
wash water. Depending on the quantities of effluent generated, optimisation of water
integration is possible
(5) boiler blowdowns, cooling water blow downs and treated rainwater run-off are water
streams that may be suitable for desalter wash water as well, dependent on the salt
concentration
(6) a site-specific Water Masterplan can be designed, applying the concept of WSI. It will be
dependent on and will take account of the location and site-specific elements such as
climate/rainfall, river/sea, existing segregation, availability of two desalters/SWS,
specific refinery units, fresh water costs, groundwater situation, etc
(7) use of phenolic spent caustic for stripped water neutralisation and subsequent desalter
wash to allow phenols resorption into the crude
(8) if the stream generated in the boiler feed water production is not recycled to the desalter,
biotreatment is required because of anti-foaming agents and corrosion inhibitors
(9) refinery waste water need to be properly segregated to ensure the most effective treatment
or re-use. A typical segregation scheme for a petroleum refinery would provide a “clean”
water sewer, an oily water sewer, and a highly-contaminated water sewer
(10) water pinch or water optimisation studies
(11) water coolers/condensors may be replaced by air coolers (closed cooling water systems)
(reference to cooling section)
(12) air cooling, reduced cooling water blowdown, using hydrocracking and hydrotreating
processes that produce less waste water than older processes; and using improved drying,
sweetening and finishing procedures to minimise the generation of spent caustics and
acids, water washes and filter solids; recycling and re-using waste water elsewhere in the
refinery
(13) treated waste water that can be re-used in make-up to process and pump cooling systems,
washdown water and water from fire control systems
(14) use of vacuum liquid ring compression instead of steam ejectors to minimize water
consumption; use waste water free vacuum technique (recirculation of water) if otherwise
harmful substances are transfered to the environment
Environmental aspects
The environmental implications of the conceptual approach of a process water integration
scheme are favourable. Enclosed systems and closed water loops limit the hydrocarbon
emissions to air and discharges to surface water, and also lead to reduced generation of waste in
the form of sludges.
Quantification of the reductions of the water used (and products lost) vary from location to
location but can be substantial (>50 %). A newly built refinery in Thailand generates only 40 t/h
process water at a crude throughput of 8 Mt/yr due to such an approach. Values of 0.5 m3 per
tonne of crude oil have been also achieved within Europe.
Reference literature
[118, VROM, 1999], [113, Noyes, 1993], [268, TWG, 2001], [117, VDI, 2000]
System description
A modern and optimised water and drainage system of an industrial site is a result of a
conceptual design of many unit operations lined up in such a way that water management can be
optimised. The basis of the concept is segregation of the different water streams into oil-free
water streams, accidentally oil-contaminated (AOC) streams and continuously oil-contaminated
(COC) streams. The latter need to be subdivided into low and high BOD streams with the
possibility for biotreatment, effluent polishing and re-use.
The application of a segregation concept takes a certain level of housekeeping and operational
discipline into account. The decision to choose a different route for a certain type of effluent
would be based on monitoring of water qualities, both internal streams and the collective
discharge. The environmental performance and the effluent management of the refinery requires
a constant level of alertness and dedication, sufficient training and motivating instructions and
an extensive monitoring (sampling and analysis) scheme. The first flush principle entails the
collection, separation and treatment of the first charge of rainwater from the potentially
contaminated refinery surface. After the first flush the rest of the rainwater is collected, analysed
and discharged without treatment, if acceptable.
Application
A number of elements of the current design philosophy for a new water and drainage system can
be considered for implementation in existing refineries, and more particularly in cases where the
old effluent and drainage system has become obsolete.
Environmental aspects
A new problem in most refineries is the management of contaminated groundwater. Possibilities
exist to manage and re-use this stream either inside the refinery (after iron removal) or to
accommodate and route this stream directly to the WWT facilities. Source reduction and spill
prevention by good housekeeping are also essential elements for VOC and odour emission
reduction at CPI, API, DAF units and biotreaters. Some refineries employ an H2S oxidation tank
to prevent toxic and odorous emissions from process water prior to DAF and biotreatment.
Status
Not many refineries have segregated process water and rainwater run-off. If they have, these
streams are routed to separate and dedicated treatment systems. The extent of segregation varies
from refinery to refinery (by design or retrofit). Most refineries use sour water and some
condensate streams as desalter wash water (see Section 4.9). Other effluent streams are also
candidates for desalter wash water such as cooling water and boiler blowdowns. Re-use of
treated effluent (biotreatment, sand filtration followed by reversed osmosis) as a source for
boiler feed water is technically feasible. Re-use of treated effluent as cooling water make-up is
done at several locations. First flush principles are followed in several refineries. Figure 4.6
shows a block diagram of an example of an optimised Water/Effluent Routing for a refinery.
Domestic Reverse
osmosis
Ballast Buffer tank CPI FFU-3
Recycle
Figure 4.6: Block diagram of an example of an optimised water/effluent routing for a refinery
Dashed lines indicate hypothetical “ closed loop”.
Reference literature
[308, Bakker and Bloemkolk, 1994], [118, VROM, 1999], [19, Irish EPA, 1993]
4.15.7.3 Rainwater
• in specific situations rainwater may be a useful source of rawwater for the preparation of
process water make-up, boiler feed water and cooling water
• some refineries have segregated effluent and drainage systems. In modern refineries the
water and drainage system includes a segregation of process water, condensate blowdowns,
rainwater and cooling water in order to minimise the environmental impact of effluent
discharges at minimum cost. Moreover, adequate process water integration and water
management measures will include a review of all options to recycle water streams with the
aim of achieving a significant reduction of end-of-pipe treatment. Segregation of the process
water discharges, surface water run-off, cooling water and boiler feed water blowdowns and
other effluents may be considered. Separation of storm water and other effluents of different
origin in order to permit appropriate treatment options. In some cases, retroffiting cost can
be very high.
Reference literature
[118, VROM, 1999], [113, Noyes, 1993]
Sanitary waste water effluent is collected in a septic tank. As this stream is small compared to
the total waste water of a refinery, and contains sufficient nutrients (micronutrients and
phosphorus) required for an industrial biotreater, this effluent can be usefully combined with the
process water to a dissolved air flotation-unit preceding a biotreater.
Ballast water is typically discharged at a very high rate provoking generation of high volume
waste water peaks. That waste water contains high salt concentrations (seawater), and are
heavily polluted with oil. It can easily lead to an upset of existing effluent treatment systems.
Therefore the use of ballast water tanks is an important equalisation tool for feeding the water in
a controlled way to either the process water system or the COC system, if the COD is lower
than 100 ppm. Because of this, (and because jetties are often remote from the rest of the
refinery) ballast water is often treated in a separate dedicated plant. As more and more crude
tankers are equipped with double hulls, the ballast water problem is slowly phasing out.
The blowdown liquid system is typically composed of mixtures of water and hydrocarbons
containing sulphides, ammonia, and other contaminants, which are sent to the waste water
treatment plant.
Some of the air emissions issues are by its nature refer to many processes or activities then they
should be considered as a refinery as a whole issue. Under those issues, we can include sulphur
emissions and VOC emissions. The application of a good environmental performance should
take the refinery as a whole as well as in the case of the sulphur the amount of sulphur contained
in the products. A good environmental planning for reduction of those emissions should
consider those aspects. In order to clarify those aspects, the next table shows were in this
document, topics that may have integrated related issues are tackle.
Reference literature
[316, TWG, 2000]
4.16 Isomerisation
4.16.1 Active chloride promoted catalyst isomerisation process
Description
More information in Section 2.16.
Cross-media effects
The chloride promoted, chlorinated alumina catalyst (containing platinum) requires the addition
of very small amounts of organic chlorides to maintain high catalyst activities. This is converted
to hydrogen chloride in the reactor. The chloride promoted catalyst cannot be regenerated.
Operational data
The highly active chloride promoted catalyst, chlorinated alumina catalyst (containing
platinum), operates at a relatively low temperature (190 - 210 °C and 20 barg) and gives the
highest octane improvement. In such a reactor, the feed must be free of oxygen sources,
including water to avoid deactivation and corrosion problems. The desorption process consumes
utilities and energy.
Applicability
This catalyst is very sensitive to sulphur, so deep desulphurisation of the feed to 0.5 ppm is
required.
Economics
Investment estimated (basis ISBL, USA Gulf Coast 1998) 4150 - 10400 EUR per m3/d.
Investment erected cost (basis: 2Q98 USA Gulf Coast) were evaluated for 1590 m3/d of fresh
feed capacity EUR 8.8 millions (±50 %).
Reference literature
[212, Hydrocarbon processing, 1998], [316, TWG, 2000]
Description
More information in Section 2.16. Some refiners need more octane from the light straight run
naphtha fraction than is possible from the O-T zeolitic isomerisation process. Adsorption
technology can then be used to remove the unconverted normal paraffins. The utilities needs for
that process are quite low.
Cross-media effects
Higher temperature of the process, requiring more heating.
Operational data
The zeolite catalyst operates at significantly higher temperatures (250 - 275°C and 28 barg) and
is much more tolerant to contaminants, though the resulting octane improvement is lower.
Applicability
The zeolite catalyst is mainly used for non hydrotreated feedstreams. Lower reaction
temperatures are preferred to higher temperatures because the equilibrium conversion to isomers
is enhanced at the lower temperatures.
Economics
Estimated cost of the reaction process is EUR 4654 per m3/d. For the absoprtion process, the
investment is EUR 18900 - 25160 per m3/d. The catalyst and adsorbent cost is around EUR
1700 per m3/d.
Reference literature
[166, Meyers, 1997], [212, Hydrocarbon processing, 1998], [316, TWG, 2000]
Description
As the benzene content in the gasoline pool is being reduced by new specifications, the practice
will be to fully dehexanise the reformer feed, leaving in the cyclic hexanes to pass to the
isomerisation unit (preventing their conversion to benzene in the reformer).
Reference literature
[247, UBA Austria, 1998]
Description
The reaction occouring in the process is given below and the process is illustrated in Figure 4.10
of Section 4.23.5.1
If the recovered H2S gas stream is not to be used as a feedstock for commercial applications, the
gas is usually passed to a tail gas incinerator in which the H2S is oxidized to SO2 and is then
passed to the atmosphere via a stack. More information can be found in Section 4.23.5.1.
Cross-media effects
Section 4.23.5.1.
Operational data
The recovered hydrogen sulphide gas stream may be: (1) vented, (2) flared in waste gas flares or
modern smokeless flares, (3) incinerated, or (4) used for the production of elemental sulphur or
sulphuric acid.
Applicability
Fully applicable.
Economics
Section 4.23.5.1.
Example plants
At present, the amine process (also known as the Girdler process) is the most widely used
method for H2S removal.
Reference literature
[136, MRI, 1997], [144, HMIP UK, 1997]
Description
Acid gas streams containing carbon dioxide free from sulphur and hydrocarbon compounds may
be vented directly to air, although this may require heating of the stream to air dispersion.
However, carbon dioxide can be used for other purposes. If the CO2 stream has a significant
hydrocarbon content, it can be used within combustion processes, e.g. mixed into site fuel gas or
used in a utility designed to burn gas of low flame stability.
Reference literature
[144, HMIP UK, 1997]
During routine operation of natural gas terminals and other processes, the aim need to be to
prevent releases to air of natural gas. Techniques to consider to reduce those emissions include:
• minimise frequency of pig/sphere use by operating sea-lines at high velocity, ie use “mist
flow”conditions
• minimise recovery of spheres by use of receivers holding several devices
• vent receivers of high-pressure gas to a low-pressure part of the process for gas recovery by
recompression, before opening for access to pig/spheres
• the occasional shutdown and venting of process plant is necessary, e.g. for maintenance,
upset and changeover purposes. This need to be minimised by appropriate plant selection
and design
• avoid the use of refrigerants for gas dew point control that are of significant environmental
concern (CFCs)
• overheads and any gas vented from storage and glycol and methanol regeneration units
should be condensed and incinerated
• leak detection and repair programme. More information can be found in Section 4.23.6.1.
• VOC emissions, including fugitives, can be maintained within a range of 200 to 250 kg/h
(300 - 350 kg/MNm3)
Reference literature
[144, HMIP UK, 1997], [268, TWG, 2001]
NOx emissions come from the combustion processes applied in the natural gas plant. The
techniques that can be applied for control and abatement of NOx are the same as those applied to
refinery fuel gas. See Section 4.10.4 and 4.23.3 for a detailed explanation.
Description
Prevention techniques that may be applied to reduce the releases to water can be:
• if possible, the quantity and contamination level of waste water to be treated on-shore need
to be minimised and controlled “at source”, i.e. wste water coming from off-shore activities
• a three-phase separator can be used on the liquids from the slug cutcher to control and
minimise hydrocarbon content of the aqueous phase
• sour water can be treated in a sour water stripper unit. See Section 4.24.2
• process effluent from glycol or methanol regeneration plants and any other high BOD/COD
effluents should be kept separate from other streams, e.g. surface water, and treated before
discharge to the site effluent system
• water management techniques. See Section 4.15.7.
The techniques that may be used in the waste water treatment are the ones described in
Section 4.24.
Reference literature
[144, HMIP UK, 1997], [268, TWG, 2001]
Description
Techniques that may be appplied to reduce waste generation are:
• catalysts, absorbents, adsorbents, etc can be returned to the manufacturers for recycling
• where desalination of glycol purge stream is practised this will give rise to solids for
disposal and any residual glycol in these should first be reduced to a low level.
• Some of the gas fields contain mercury vapour in very low concentrations. This mercury is
removed from the gas in a 'cold trap' (e.g. by gas expansion) and recovered as a mercury
containing sludge. A specialised company processes this sludge by treatment in a vacuum
distillation unit.
Reference literature
[144, HMIP UK, 1997], [268, TWG, 2001]
4.18 Polymerisation
4.18.1 Process
Description
More information in Section 2.18.
Cross-media effects
No cross-media effects.
Operational data
Utilities
Electric power (kW/tonne of C5+ product) 20 - 28
Steam (t/ tonne of C5+ product) 0.7 - 1.1
Cooling (m3/tonne of C5+ product) 4.4 - 6.0
Applicability
Production process
Economics
The catalytic condensation process is relatively simple to operate and requires a minimum of
labour. Its simplicity is reflected in the operating requirements summarised in the following
table.
Only one operator is required. Overall, the cost of operating a catalytic condensation unit ranges
from EUR16 – EUR22.6 per m3 of C5+ polymer gasoline. This cost includes utilities, labour,
catalyst, chemicals, and an allowance for process royalty, but does not include any direct or
indirect capital-related costs.
Example plants
Some polymerisation processes are running in European refineries. Nowadays alkylation units
are more predominant than polymerisation units, even though polymerisation units are cheaper.
Reference literature
[166, Meyers, 1997], [212, Hydrocarbon processing, 1998], [268, TWG, 2001]
Description
Two techniques can be applied to minimise the impact of catalyst disposal:
• the catalyst to be disposed of from the process is potentially pyrophoric and requires special
treatment. However, removal by steam/water avoids any risk of fire. The catalyst solidifies
in situ, and removal is achieved manually under nitrogen purge, or more recently by
‘explosive’ steam decompression with steam suppression to a catchment system. The
catalyst may then be sent for either off-site disposal as special waste or on-site treatment.
On-site treatment involves neutralisation and fixing with cement. The material may then not
be classified as special waste
• the spent catalyst can be re-used as fertiliser or as phosphorous feed to the biotreater.
Cross-media effects
No cross-media effects.
Operational data
Disposal frequencies of up to 12 times per year may be expected.
Applicability
Fully applicable.
Reference literature
[34, Italy, 1999]
Description
A progressive distillation unit with integrated CDU/HVU, which saves up to 30 % on total
energy consumption for these units. The technology includes atmospheric distillation (topping),
vacuum distillation, gasoline fractionation, naphtha stabilizer if required and gas plant.
Hydrotreater
feed
120 - 160 °C
Petro naphtha
Kerosene
1 bar 154 °C
Light gasoil
90°C 90 - 120°C
Heavy gasoil
416 °C
Vacuum unit
1 bar 131°C
Medium vac.
Flue gas
gasoil
Crude
Oil 357°C
Heavy vac.
HEATER
gasoil
HEATER Fuel
Vacuum
Fuel residue
Progressive distillation is the extreme of heat integration between atmospheric and vacuum
distillation. It also avoids superheating of light cuts to temperatures higher than strictly
necessary for their separation and it avoids degrading the thermal levels associated with the
drawing-off of heavy cuts.
Cross-media effects
No cross-media effects identified.
Operational data
Table 4.33: Energy consumption of the progressive crude distillation when the same plant is used
to process two types of crude oil
Applicability
The process is applicable to all or part of these units when being constructed and can also be
used in revamps for debottlenecking.
Economics
Reduce fuel consumption and thus reduce operating cost of the distillation units.
Investment (basis 11.45 Mt/yr including atmospheric and vacuum distillation, gas plant and
rectification tower): EUR 41000-55000 per t/yr (US Gulf Coast 1998).
Example plant(s)
Some European refineries employ a progressive distillation unit with integrated CDU/HVU.
Reference literature
[195, The world refining association, 1999], [212, Hydrocarbon processing, 1998], [247, UBA
Austria, 1998]
Description
As stated in Section 4.10.1, there is a tendency towards improved heat integration with crude
and other units. The high energy consumption of the CDU makes heat integration very relevant.
To optimise heat recovery from the atmospheric distillation column, two or three reflux streams
are normally kept in continuous circulation at several points per top and middle pumparounds.
In modern designs, integration with the high vacuum unit and sometimes the thermal cracker is
achieved. Some techniques applied are:
• optimise the heat recovery, studying and implementing the optimal energy integration. In
this context, major progress has been made in recent years in understanding the design of
efficient heat recovery networks. "Pinch Analysis" has emerged as a tool for the evaluation
of total systems designs, helping to balance investments against energy savings. See also
Section 4.10.1
• apply pinch analysis to the crude preheat train heat integration. Increase crude preheat
temperature and minimise heat losses to air and cooling water
• increase crude distillation column pumparounds from two to four. Reboil sidestrippers with
a heat transfer oil rather than by steam stripping
• heat transfer in the crude preheating can be improved using specific antifoulant treatments
in crude heat exchanger train. Antifoulants are available from many chemical companies
and, in many applications, are efficient in increasing the run cycle length of the exchangers ;
antifoulant can help prevent plugging of tube exchangers, improving heat recovery and
preventing hydraulic losses, depending on the nature of fouling. Service factors of the
various units/trains, as well as heat recovery (energy efficiency) increase at the same time.
• application of advanced process control to optimise the energy utilisation within the crude
unit.
Cross-media effects
Control of plants is impaired because disturbances are transferred between them, influencing
plant safety.
Applicability
Integration in retrofitting application may will normally depend on plot space available. Except
in a very few cases this technique is commonly applicable.
Example plants
Heat integration procedures are extensively applied in crude units. Progressive distillation is the
stream of heat integration between atmospheric and vacuum distillation.
Reference literature
[147, HMIP UK, 1995], [79, API, 1993], [297, Italy, 2000]
Description
The number of side-streams in a high vacuum unit is chosen to maximise heat integration of
producing streams at different temperatures, rather than to match the number of products
required, with the exception of the luboil vacuum distillation units. Heat integration with the
crude units can be achieved. The atmospheric residue is then taken directly from the crude
distillation unit to the vacuum furnace, and the product streams and circulating refluxes of the
high-vacuum unit are cooled against crude oil.
The main feed stream to the high-vacuum unit is the bottom stream of the crude oil distillation
unit, referred to as atmospheric or long residue, which is either supplied directly hot or
(relatively) cold from storage tanks. The latest option requires higher energy consumption.
The application of advanced process control to optimise the energy utilisation within the crude
unit.
Cross-media effects
Control of plants is impaired because disturbances are transferred between them, influencing
plant safety.
Applicability
Integration in retrofitting application may will normally depend on plot space available. Except
in a very few cases this technique is commonly applicable.
Example plants
Heat integration procedures are extensively applied in vacuum units. Progressive distillation is
the stream of heat integration between atmospheric and vacuum distillation.
Reference literature
[147, HMIP UK, 1995], [79, API, 1993], [297, Italy, 2000]
Description
Vacuum liquid ring compressors instead of steam ejectors.
Cross-media effects
Replacement of the steam ejectors by vacuum pumps will increase the electricity consumption
for vacuum generation, but will reduce the heat consumption, the cooling water consumption,
the electricity consumed for cooling pumps and the consumption of agents used for conditioning
of cooling water. Within the refinery there are many processes where surplus steam can
recovered and be used for the production of vacuum. However, an energy management analysis
will help to decide, whether use of surplus steam for steam ejection instead of applying vacuum
pumps is more efficient than using surplus steam for other purposes.
Operational data
The use of vacuum pumps consumes electricity.
Applicability
Fully applicable.
Example plants
At present, vacuum pumps are used more often than ejector sets.
Reference literature
[79, API, 1993], [268, TWG, 2001]
Description
Lowering the vacuum pressure, e.g. down to 20 - 25 mm Hg, will allow a reduction in the
furnace outlet temperature, while maintaining the same target cut point of the vacuum residue.
Cross-media effects
Energy (electricity or steam) is necessary to generate the vacuum.
Applicability
It is limited usually by tower capacity, condensing fluid temperature or material restrictions.
Reference literature
[297, Italy, 2000], [79, API, 1993]
Description
Techniques for controlling releases from vacuum units include scrubbing, compression into
refinery fuel gas and burning in adjacent process furnaces, or combinations thereof. The gases
from some units may contain significant amounts of air and such gases are normally best
combusted locally. Amine scrubbing techniques may need to be applied carefully as
hydrocarbon contamination may cause foaming problems in amine regeneration units.
Incondensables from overhead condensers can be passed to light ends treatment or recovery
systems or refinery fuel gas systems; sour incondensable gases vented from sealed barometric
pumps of vacuum distillation units should be extracted and dealt with in a manner appropriate to
the nature of the sour gas.
A control technique applicable to the non-condensable emissions vented from the vacuum
ejectors or pumps consisting of venting into blowdown systems or refinery fuel gas systems,
and incineration in furnaces or waste heat boilers.
Cross-media effects
In the incineration technique, the combustion products have to be accounted for.
Applicability
Fully applicable.
Example plants
Applied in some European refineries.
Reference literature
[136, MRI, 1997], [127, UN/ECE, 1998]
Description
This topic is also treated in Section 4.15.7. Two techniques are considered in this section:
• the overhead reflux drum generates some waste water. That water can be re-used as a
desalter wash water
• sour water from atmospheric and vacuum unit condensates should pass to a sour water
stripper in enclosed systems
• optimise water re-use by application of side-stream softening to blowdown streams.
Applicability
Fully applicable.
Reference literature
[79, API, 1993]
Description
Some other techniques to consider are the following:
(1) where applied, ammonia injection should be handled in enclosed systems. Alternative
neutralising techniques are available and can reduce the ammonia/ammonium load on sour
water and sulphur recovery systems
(2) de-coking vents need to be provided with suitable knock-out and dust suppression
facilities;suitable methods of preventing emissions during clean-out procedures need to be
used
(3) many oily sludges can be sent the crude distillation (or in alternative to the coking unit. See
4.7.5) where they become part of the refinery products. This technique typically imply to
send the light slops to floating roof tanks (double sealed), and the heavy slops typically to
fix roof tank. Typically the heavy slop after setting is blend with liquid fuel. The blended
light slops after enough setting may be also blent to liquid fuel or sent to the crude
distillation at a mixing rate to avoid fouling.
(4) Pressure relief valves on column overheads; relief from overhead accumulator should be
piped to flare as well as the vent points.
(5) Use of spent caustic instead of fresh caustic for corrosion control on distillation unit. (More
information in Section 4.20.2
Operational data
The process of oily sludges in the crude distillation may generate problems in the desalter or
may foul up the heat exchangers of the distillation column.
Reference literature
[79, API, 1993], [268, TWG, 2001]
Description
An overall reduction in the caustic consumption of wet treating units can be obtained if semi-
spent caustic from one treating unit can be re-used in another one. A typical example of this
procedure is the use of the bleed of regenerated caustic (e.g. in mercaptan treaters for catcracked
gasoline or for removal of H2S or thiophenols) in a pre-wash step of the non-catalysed
catcracked gasoline sweetening processes. An example of a caustic integration scheme is given
in Figure 2.28.
Operational data
Systems handling spent caustic require particular care regarding sulphur compounds.
Economics
Sweetening treatment is much cheaper than hydrotreating, both in terms of capital investment
and in operation costs. The caustic is normally almost fully regenerated and only a small bleed
results.
Reference literature
[115, CONCAWE, 1999], [166, Meyers, 1997], [83, CONCAWE, 1990]
Description
Caustics are used to absorb and remove hydrogen sulphide and phenol contaminants from
intermediate and final product streams can often be recycled. Spent caustic solutions from some
sweetening units are odorous and need to be handled in enclosed systems and treated as
necessary before release at a controlled rate to the effluent system. Several techniques exist to
maximise the use of caustics within a refinery. They include recycling within the refinery or
outside the refinery or destruction within incinerators. Techniques considered are:
• Treatment of caustic may include neutralisation and stripping
• Because of the very high concentration of organic compounds in spent caustic solutions
(COD>>> 50 g/l), incineration can be an appropriate alternative to waste water treatment.
• Spent catalyst/caustic need to be handled and disposed of in a manner that avoids dust
generation. They should not be disposed of to the land.
• Some techniques are available to re-use spent caustic within the refinery:
(1) corrosion control on crude distillation units using spent caustic rather than fresh caustic.
Ustable chloridic (magnesium) salts that are not extracted from the crude oil in the
desalter will decompose upon heating to the crude distiller and cause chloridic
corrosion. To prevent corrosion of the exposed equipment, small quantities of caustic
(sodium) are injected in the crude oil feed by which the chloridic components are
neutralised due to the formation of stable sodium chloride. For this purpose of
neutralisation of chloridic decomposition products often spent caustic can be used,
which is recommended as well to minimize waste generation.
(2) recycle to crude desalter or sour water strippers
(3) addition to biotreaters for pH control.
(4) Caustics containing phenols can also be recycled on-site by reducing the pH of the
caustic until the phenols become insoluble thereby allowing physical separation. The
caustic can then be treated in the refinery waste water system.
• Re-use spent caustic outside refinery:
(1) in paper mills (sulphidic caustic only)
(2) as raw material for Na2SO3, cresylics and Na2CO3 (may require segregation of
sulphidic, cresylic and naphthenic caustics).
(3) Spent caustics may be saleable to chemical recovery companies if concentrations of
phenol or hydrogen sulphide are high enough. Process changes in the refinery may be
needed to raise the concentration of phenols in the caustic to make recovery of the
contaminants economical.
• Measures to regenerate or oxidise spent caustic are:
(1) Hydrogen peroxide
(2) Fixed bed catalyst
(3) Pressurised air: 120 - 320°C; 1.4 - 20.4 MPa
(4) Biological system
Cross-media effects
The cross-media effects found in the different techniques mentioned above are:
(2) the addition of caustic within the desalter or within the crude unit may enhance the coke
formation within later units.
(3) it adds phenols and BTX to waste water treatment. As a consequence, the degradation
efficiency of the biotreater can be adversely affected or the emissions from the waste water
treatment plant for these components can increase.
Operational data
Systems handling spent caustic require particular care regarding sulphur compounds.
Example plants
Most refineries are able to regenerate their spent caustic, but sometimes they have to dispose of
some excess quantities, mainly from caustic pre-wash activities. Usually these quantities are
small and can be managed within their effluent treatment system or, if this is not the case, they
are disposed of via contractors for use as a bleaching agent in the paper and pulp industry. Some
refineries sell their concentrated phenolic caustic for the recovery of cresylic acids. Some
refineries work up the phenolic caustics themselves. From the extraction process the disulphide
recovered can be sold as product, or be recycles to a hydrometer or incinerator.
Reference literature
[115, CONCAWE, 1999], [259, Dekkers, 2000], [268, TWG, 2001]
Description
Four air vented from sweetening process contain sulphur compounds that typically have strong
odour. The range of quantity of sulphur in the foul air vented from sweetening processes is
around 0.7 – 7 kg/day (disulphide concentration can be as high as 400 ppm) for a 10000 t/d
crude unit and the percentage contribution it makes to the stack gases where it is incinerated has
been estimated to 0.16 – 2.48 %. For this reason, the abatement previous an incineration is not
justified and the foul air vented from sweetening processes is incinerated in local furnaces.
Reference literature
[268, TWG, 2001]
Description
Where a need to remove colour bodies and olefins exists, hydrotreating is replacing clay
filtration as the method of choice. Hydrotreating does a better job of improving colour and
increasing stability against oxidation, and has no yield loss (the oil measured in the spent filter
clay represents a loss of some of the most valuable products produced in a refinery). It also
eliminates the problem of spent clay disposal.
Cross-media effects
Need for hydrogen and energy consumption. See Section 4.13.1.
Operational data
See Section 4.13.1.
Economics
See Section 4.13.1.
Reference literature
[113, Noyes, 1993]
4.20.5 Treating
Description
Treating gas, LPGs, butanes, gasolines, kerosenes and diesels with caustic, amines, water and
acid to remove amines, caustic contaminants, H2S, COS and mercaptans. Dry adsorbents such
as molecular sieves, activated carbon, iron sponge and zinc oxide may also be used to reach
specification (LPG) and prevent odour problems. (See also.4.23.9)
Reference literature
[212, Hydrocarbon processing, 1998]
Description
Catalytic dewaxing process typically produces a lower-pour product than solvent dewaxing.
This system produce fuel components instead of waxes. A short description of the technique can
be found in Section 2.20.
Cross-media effects
Hydrogen consumption.
Applicability
Fully applicable for new units. It is unlikely to be retrofit on another type of dewaxing process
because it is a completely different process. Catalytic dewaxing has a pour point advantage, but
has a viscosity index disadvantage compared to solvent dewaxing.
Economics
The following table compares the relative cost of debottlenecking a 300 kt/yr solvent extraction
complex to 500 kt/yr with the cost involved in the construction of a a new 200 kt/yr solvent
extraction complex and hybrid debottlenecking with a catalytic dewaxing.
Investment (basis: 795 m3/day of lube base oils excluding fuels hydrocracker, 1998 USA Gulf Coast)
$80 million.
Solvent extraction Hybrid desbottlenecking
(200 kt/yr grass-roots) (from 300 to 500 kt/yr)
(% of the total cost) (% of the solvent extraction cost)
Capital change 36 24 - 36
Fixed costs 20 7-9
Variable costs 8 8
Hydrocarbon cost 35 11
Total 100 50 - 64
Example plants
It has been identifed one EU+ refinery that have this process to produce distillates with low wax
content.
Reference literature
[212, Hydrocarbon processing, 1998], [247, UBA Austria, 1998], [268, TWG, 2001]
Description
More information in Section 2.21.
Cross-media effects
Ground water leaking into the cavern must be removed and treated with other oily waters.
Applicability
The geology of the site should be adequate to install a cavern: non porous rock.
Economics
The construction costs of storage caverns are markedly smaller than for above-ground tank
farms. In Finland, profitability begins at 50000 m3 under favourable conditions (depends highly
on the rock type). Maintenance costs have been calculated to be one-sixth those of above-
ground tanks.
Example plants
Used in some European countries.
Reference literature
[256, Lameranta, 2000]
Description
Internal floating roof tank (IFRT)s. Emissions primarily occur during standing storage, with an
additional contribution from withdrawal emissions. In addition to the rim seal area and roof
fitting penetrations, sources of standing loss from internal floating roof tanks include bolted
seams in the floating roof. More information in Section 2.21. Some techniques that might be
applied to IFRT are:
• the replacement of primary/secondary seals with tighter seals, which can also reduce the
VOC emissions
• drain design on floating roof tanks to avoid hydrocarbon contamination of rainwater.
ranges from 60 to 99 % depending on the type of roof and seals installed and the true vapour
pressure of the liquid stored.
Cross-media effects
The net storage capacity of a fixed roof tank is reduced by aprox. 10 %.
Operational data
Performance data on emissions and other useful information about IFRT are available in [323,
API, 1997].
Applicability
An acceptable alternative to an EFRT is to retrofit a fixed roof tank with an internal floating
cover (IFRT).
Economics
The costs of retrofits are given in Table 4.34 below. Costs are dependent on the tank diameter.
Emission
Refinery Storage
Source
Control Internal Floating Roofs in Secondary/Double Seals on Other Roof Fitting
Technology fixed roof tanks Floating Roof Tanks Emission Controls
(roof legs, still wells)
and options (tank paint)
Efficiency 90 – 95 % 95 % Over 95 % when
together with
secondary seals
Investment 0.20 - > 0.40 for 20 - 60 0.05 - 0.10 for 20 - 50 metres 0.006 for 50 metre
Costs EUR metres diameter tanks(1) diameter tanks(2) diameter tanks(1)
million
Operating Costs Minor Replacement every 10 years Minor
Other Impacts Requires tank to be taken Can reduce maximum storage Not suitable for high
Comments out of commission capacity of tank sulphur crude oils due
Reduces the net storage by to pyrophoric scale
5 – 10 % possibility
References: (Installed and Retrofitted)
(1)
UN-ECE/IFARE, and Industry Propriety Information
(2)
UN-ECE/IFARE and Industry Propriety Information (UN-ECE EC AIR/WG6/1998/5)
Reference literature
[45, Sema and Sofres, 1991], [127, UN/ECE, 1998], [323, API, 1997], [268, TWG, 2001]
Description
The fixed-roof tank can give rise to emissions in the following ways:
• filling losses; during tank filling the existing tank vapour space, which is more or less
saturated with vapour, is expelled to the atmosphere, while when a tank is emptied the
incoming air is slowly saturated with vapours, which are then expelled by subsequent filling
and/or breathing. Generally those emissions are greater than standing emissions. A
technique to reduce VOC from those tanks is to increase the storage pressure by blanketing.
• breathing losses; during storage of the liquid there will be vapour emissions from tank
breathing caused by the difference in night and day temperatures and by changes in
atmospheric pressure. To a certain extent, pressure controllers and insulation can prevent
breathing losses
• vapours released during water draining.
The techniques that are considered to reduce emissions from fixed roof tanks are:
• a technique to reduce VOC from those tanks is to blanket them
• installation of an internal floating roof.
Cross-media effects
The installation of an internal floating roof in a fixed roof tank requires the tank to be taken out
of commission and reduces the net storage by 5 - 10 %.
Operational data
Performance data on emissions and other useful information about EFRT are available in [323,
API, 1997]
Applicability
Fixed roof tanks are typically used for low or non volatile products e.g. kerosene, heating oil,
with TVP less than 14 kPa. These tanks can be retrofitted to internal floating roof tanks. But are
not suitable for all products, e.g. bitumen tank with an internal floating roof will not work due to
bitumen solidifying in the flaw between wall and internal floating roof.
Economics
The investment cost of installing internal floating roofs in fixed roof tanks can be 0.2 - 0.4 EUR
million for tanks 20 - 60 meters in diameter.
Reference literature
[107, Janson, 1999], [115, CONCAWE, 1999], [323, API, 1997], [268, TWG, 2001]
Description
External Floating roof tanks (EFRT). Filling and breathing losses are greatly diminished in
comparison with fixed roof tanks, but the following vapour losses are inherent to this type of
tank:
• standing storage emissions from floating roof tanks include rim seal and roof-fitting
emissions, which result from stock vapour pressure changes due to temperature, and
pressure variations but more importantly wind effects as well as openings in the roof. The
influence of wind effects is not a factor on internal floating roof tanks. Standing emissions
on external floating roof tanks are generally much more significant than withdrawal
emissions
• wetting losses; by evaporation of liquid from the wet wall, when the level of the liquid is
lowered by discharge
• vapours released during water draining
• in many cases, for an external floating roof tank, the emissions through fittings can exceed
the rim seal losses especially on tanks with secondary seals. In terms of fitting losses, the
major source is from the slotted stillwell (sample well or dipping well).
304 Mineral Oil and Gas Refineries
Techniques to consider in the determination of BAT Chapter 4
External floating roof tanks are used to store crudes, light products and intermediate stocks
having a vapour pressure of more than 14kPa but less than 86 kPa at normal storage
temperature. More information can be found in Section 2.21. Techniques to minimise emissions
there include:
• installing wipers at the floating deck
• sleeves around the pipe, incorporating still well wipers
• floats with wipers inside the slotted pipe
• land as infrequently as possible the floating roof tanks to avoid unnecesary vapour releases.
Cross-media effects
They have potentially more emissions to water than fixed roof tanks, since it is possible for
rainwater to enter the tank past the roof seal. Any such water needs to be drained before
despatch of the product to a customer since the quality of the product can be seriously
jeopardised.
Operational data
Performance data on emissions and other useful information about EFRT are available in [323,
API, 1997].
Applicability
In retrofit situations where a change in the service of the tank is desired, an acceptable
alternative to an EFRT is to retrofit a fixed roof tank with an internal floating cover.
Economics
Investment cost for the retrofitting of a fixed rood tank to a EFRT is EUR 0.26 million per tank
diameter 20 m. Operator is necessary for tank draining, giving rise to some operating cost.
Reference literature
[45, Sema and Sofres, 1991], [252, CONCAWE, 2000], [258, Manduzio, 2000], [323, API,
1997], [268, TWG, 2001]
Description
Two seals at the floating roof rim provide a double barrier to control VOC emissions from
storage tanks. Fitting secondary roof rim seals is an accepted technology for emission reduction.
Rim-mounted seals (as opposed to shoe-mounted seals) are favoured, since the former offer
emission control if the primary seal fails.
Cross-media effects
Retrofitting seals usually results a loss of operating capacity of about 5 %.
Operational data
Performance data on emissions and other useful informations about rim seals are available in
[323, API, 1997].
Applicability
They can be readily installed in new units (named double seals) and generally retrofitted
(secondary seals).
Economics
Equipping an average tank with a secondary seal system was estimated to cost about $20000
(1991). Investment costs: EUR 0.05 - 0.10 million for tanks 20 - 50 metre diameter.
Operating costs: Replacement likely every 10 years.
Example plants
In common use world-wide.
Reference literature
[45, Sema and Sofres, 1991], [252, CONCAWE, 2000], [323, API, 1997], [268, TWG, 2001]
Description
The need for certain tanks can often be eliminated through improved production planning and
more continuous operations. This technique is closely related to Section 4.15.5. An example is
the implementation of an in-line blending system. See Section 4.21.14.
Applicability
The reduction of number of tanks typically requires a full change in the management of product
and intermediates. As consequence this technique is easier to be applied to new units.
Reference literature
[268, TWG, 2001]
It follows techniques to consider in the determination of BAT for prevention of leakage through
tank bottoms. This topic is well documented in EEMUA pub 183 "Guide for the Prevention of
Bottom Leakage from Vertical, Cylindrical, Steel Storage Tanks”.
Description
Double bottoms can either be retrofitted in existing tanks or incorporated into the design of new
tanks. If retrofitted, the existing tank bottom is normally used as the secondary flooring, and
sand, gravel or concrete can be installed between the new primary and the secondary floors. In
this case, it is general practice to keep the interstitial space to a minimum and therefore the
secondary bottom should slope the same way as the primary bottom. The slopes to the base of
tanks can be either straight, cone-up (sloping from the centre down to the tank perimeter) or
cone-down (sloping downward from the tank perimeter). Nearly all tank floors are made of
carbon steel. If a double bottom is to be installed (either retrofit or new build), there are choices
in material selection for the new floor. A second carbon steel floor can be utilised or a more
corrosion-resistant stainless steel floor can be installed. A third choice is to use a glass fibre-
reinforced epoxy coating over the steel.
The use of double bottom tanks allows a vacuum system to be installed, and in this case the
space between the lower and upper floor is not filled, but retained as an air space using steel
spacers (which can typically be a steel reinforcement mesh). In this more recent system, the
space between the floors is kept under a vacuum that is continuously monitored. Any leak in
the primary or secondary floor will dissipate the vacuum and trigger an alarm. A further test of
the extracted air will indicate an upper floor failure if product or vapours are present, or a lower
floor failure if neither product nor vapour is present (subject to any previous underfloor
contamination).
Cross-media effects
In the case of retrofitting, the application of this technique may prolong the outage time of tanks
during installation of the double bottom. Its application can reduce the capacity of the tank.
Operational data
By installing double bottoms, the time between inside inspections and the the number of the
annual tank cleaning procedures are decreased.
Applicability
Either retrofit or new-build tanks.
Economics
Typical retrofit costs for double bottoms, as quoted from German or Swiss suppliers, are as
follows and include provision of a vacuum leak detection system:
• carbon steel: 110 EUR/m2
• stainless steel: 190 EUR/m2
• glass fibre-reinforced epoxy: 175 EUR/m2
An UK refinery reported that the actual cost of installing a double bottom on a 10340 m3 tank
was 600000 EUR.
Reference literature
[253, MWV, 2000], [112, Foster Wheeler Energy, 1999]
Description
The impervious membrane liner is a continuous leak barrier under the entire bottom surface of
the tanks. It can be an alternative to a double bottom or it can be added as an extra measure of
safety below the double bottom. Like the double bottom, it is primarily intended to arrest the
small but persistent leak rather than address a catastrophic failure of the entire tank. The key to
an effective liner is that the seams need to be liquid-tight against either the steel shell of the tank
or the concrete wall that supports and surrounds the tank. The minimum thickness of the
flexible membrane is 1 mm, although 1.5 to 2 mm thick sheets are commonly used. The
membrane need to be chemically resistant to the product stored in the tank.
Cross-media effects
Prolonged tank outage if liner has to be retrofitted to existing tank.
Applicability
They can be installed either in a new-build design or as a retrofit and they generally include a
leak detection system.
Economics
The cost of retrofitting a liner is slightly higher than installing a double bottom as it involves
jacking-up the existing tank to install the membrane and leak detection system. An indicative
cost is of the order of 200 EUR/m2. As new-build construction, an impervious liner may be
cheaper than a double bottom, but it may have a higher life cycle cost. This is because any
future failures of the liner would require re-lifting the tank or reverting to a new double bottom
solution.
The following are costs per unit for placing the following materials on the ground:
• concrete: 30 EUR/m2
• asphalt: 24 EUR/m2
• HD polyethylene: 23 EUR/m2
• bentonite: 18 EUR/m2
• clay: 11-17 EUR/m2
This costs exclude the cost of lifting the tank for installation. However, costs of placing such
materials in the floors of bunds would be of this order. An additional cost arises from moving
any pipework in the bunds.
For comparison of the installed costs of such a liner, one refinery reported that since 1994, it
had installed a clay bottom liner under each lifted tank during maintenance. A leak detection
system was installed at the same time. The cost of this was about 20000 to 30000 EUR per tank
with the cost of the clay liner being 11 to 17 EUR/m2. The total cost for 14 tanks amounted to
350000 EUR. A similar cost (35000 EUR) was quoted in another installation of an impervious
liner (plus leak detection) for a 12000 m3 concrete tank.
Example plant(s)
Impervious membrane liners are used in lieu of double bottoms in a number of non-European
countries.
Reference literature
[45, Sema and Sofres, 1991], [268, TWG, 2001]
Description
As with sewers, one way to protect against soil and groundwater contamination is by detecting
leaks at an early stage. Leaks through a tank bottom can be detected by a leak detection system.
Conventional systems include inspection ports, inventory control and inspection wells. More
advanced systems include electronic sensing probes or energy pulse cables whereby product that
comes into contact with the probe or cable will alter its impedance and trip an alarm. In
addition, it is common practice to use a variety of inspection procedures on tanks at intervals to
prove their integrity. See also Section 4.23.6.1 LDAR programme. Some techniques to consider
are:
• Equipping storage tanks with overflow alarms and automatic pump shut-offs.
• Installing double bottoms with integrated leak detection systems on tanks where practicable.
Relief valves on pressurised storage should undergo periodic checks for internal leaks. This can
be carried out using portable acoustic monitors or, if venting to atmosphere with an accessible
open end, tested with a hydrocarbon analyser as part of an LDAR programme.
Applicability
Probes and cables have to be packed pretty densely if the leak to be detected is small.
Consequently, inspection, where possible, may in some cases be more reliable than probes.
Economics
One refinery reported that the installation of a leak detection system on a group of four 12 m
diameter tanks cost a total of 55000 EUR with an operating cost of 4000 EUR per year. The
routine inspection of tanks at another refinery was quoted as 2000 EUR/tank/year.
Reference literature
[253, MWV, 2000]
Description
To avoid corrosion on the underside of the tank bottom, tanks can be equipped with cathodic
corrosion protection.
Cross-media effects
Electricity is needed.
Reference literature
[253, MWV, 2000]
Description
Whereas double bottoms or impervious liners protect against the small but incessant leak, an
impermeable tank farm bund is designed to contain large spills (for safety as well as
environmental reasons), such as caused by a shell rupture or a large spill from overfilling. The
bund consists of a wall or dike around the outside of the tank to contain all or part of the tank
contents in the event of a spill, and (in some cases) an impermeable ground barrier between the
tank and the dike to prevent infiltration of the product into the ground. The dike is typically
constructed of well-compacted earth or reinforced concrete. The height is normally sized to
accommodate the maximum contents of the largest tank within the volume enclosed by the dike.
However this philosophy of total capture is flawed if the ground between the tank and dike is
permeable. Oil can permeate downward and under the dike in these situations. Asphalt, concrete
surface or high density polyethylene (HDPE) liner are some of the possibilities.
Cross-media effects
This technique might call for the soil to be compacted, which may be seen as negative impact. A
study in the United States has concluded that the effectiveness of liners in protecting the
environment is limited because of their unreliability and the difficulty in inspecting or testing
their integrity. Also, there are few releases that would be contained by such liners, and
moreover, they are expensive to install. Because of this, it was concluded that other
preventative measures are more effective in protecting the environment and are more cost-
effective in the long run.
Applicability
In some cases, retrofitting may not be possible.
Economics
Some refineries have quoted the total cost for installing concrete paving under 6 big tanks at
EUR 1.3 millions (0.22 per tank), the cost of sealing bund dike walls with asphalt (around 10
tanks) at EUR 0.8 million, and the cost of concrete paving bunds at 70 - 140 EUR/m2. One
refinery quoted the cost of installing 200m of impermeable HDPE barrier along the site
boundary at 150000 EUR (750 EUR/m).
Reference literature
[147, HMIP UK, 1995]
Description
Minimization of tank bottoms can be carried out through careful separation of the oil and water
remaining in the tank bottom. Filters and centrifuges can also be used to recover the oil for
recycling. Other techniques to consider are the installation of side entry or jet mixers on tanks or
the use of chemicals. This means that basic sediment and water is passed on to the receiving
refineries.
Cross-media effects
Passing the sediments and water from the crude oil tanks to the refinery means that they will
most probably show up in the desalter.
Reference literature
[147, HMIP UK, 1995]
Description
For routine internal tank inspections and for the purpose of repair, crude oil and product tanks
have to be emptied, cleaned and rendered gas-free. A technique to clean tank bottoms include to
dissolve the majority of the tank bottoms (>90%) with hot diesel fraction at temperatures of
around 50 ºC where it disolve most of the tank bottoms and they can be blended after filtration
to the crude tanks.
Cross-media effects
If the refinery operates its own sludge incineration plant, cleaning residues can supplied to it.
Operational data
The emission arising during the cleaning of crude oil tanks and the reduction of these emissions
are exhaustively described in [302, UBA Germany, 2000]. The use of hot diesel to clean up the
tank bottoms require heating.
Applicability
Fully applicable.
Example plants
Examples exist of applications in many EU+ refineries and tank farms.
Reference literature
[117, VDI, 2000], [316, TWG, 2000], [302, UBA Germany, 2000], [122, REPSOL, 2001]
Description
It is preferable to paint tanks containing volatile materials in a light colour to prevent a)
increasing the evaporation because of increased product temperature and b) increasing the
breathing rate of fixed roof tanks.
Cross-media effects
Painting in light colours can result in tanks being more "visible" which can have negative visual
effect.
Applicability
Painting of the roof and the upper part of the shell of a tank in the middle of a tank farm can be
as effective as painting the entire tank.
Reference literature
[262, Jansson, 2000], [268, TWG, 2001], [302, UBA Germany, 2000]
Description
Proper material handling and storage minimises the possibility of spills, leaks and other losses
which result in waste, air emissions and emissions to water. Some good storage practices are:
• use larger containers instead of drums. Larger containers are reusable when equipped
for top and bottom discharge, whereas drums have to be recycled or disposed of as
waste. Bulk storage can minimize the chances of leaks and spills as compared to drums.
Applicability: Safe disposal of un-refillable large containers can be a problem.
• reduce generation of empty oil drums. Bulk purchasing (via tank truck) of frequently
used oils and filling tote bins as intermediate storage. Personnel could then transfer the
oils from the tote bin to reusable drums, buckets, or other containers. This would reduce
the generation of empty drums and their associated handling costs
• storage of drums off the floor to prevent corrosion through spills or concrete 'sweating'
• keeping containers closed except when removing material
• practise corrosion monitoring, prevention and control in underground piping and tank
bottoms (Related with 4.21.8 and 4.21.22)
• tanks for storage of ballast water may cause large VOC emissions. They can be
therefore equipped with a floating roof. Those tanks are also relevant as equalizing
tanks for the waste water treatment system
• vents from sulphur storage tanks to pass to sour gas or other arrestment systems
• vent collection and ducting from tank farms to central abatement sytems.
• install self-sealing hose connections or implement line draining procedures
• install barriers and/or interlock systems to prevent damage to equipment from the
accidental movement or driving away of vehicles (road or rail tank cars) during loading
operations
• implement procedures to ensure that arms are not operated until inserted fully into the
container to avoid splashing where top loading arms are used
• apply instrumentation or procedures to prevent overfilling of tanks
• install level alarms independent of normal tank gauging system
Reference literature
[127, UN/ECE, 1998], [268, TWG, 2001], [252, CONCAWE, 2000], [19, Irish EPA, 1993],
[262, Jansson, 2000]
Description
Figure 4.8 shows a scheme of a in-line blending system for gasoils.
PRODUCT
PRODUCT BLENDING MANIFOLD 1
TO STORAGE
: Storage
ANALYSERS:
FC FC FC A FC
-FLASH
-SULPHUR
-CLOUD
KERO
from CDU
LGO
from CDU
HGO BLENDING
from CDU COMPUTER ADDITIVES
SIGNAL TO FC's
LVGO
from HVU
LCO
from FCCU
VSBGO
from VBU ANALYSERS:
FC FC FC FC FC A -FLASH FC
-SULPHUR
-CLOUD
-DENSITY
-VISCOSITY
PRODUCT BLENDING MANIFOLD 2 HEATING OIL
FC FC FC FC FC FC TO STORAGE
HYDROFINER
HYDROFINER BLENDING MANIFOLD
Figure 4.8: Simplified scheme of an in-line blending system for gasoils (automotive diesel and
heating oil)
Cross-media effects
In-line blending systems contain many flange connections and valves, which could be sources
of leakage, particularly during maintenance.
Operational data
Extreme care has to be paid to on-line quality analysers to assure quality of the blended product.
Applicability
Optimising the blend ratios to meet all critical specifications is to a certain extent a trial-and-
error procedure which is most economically accomplished with the use of a computer. The large
number of variables usually permits a number of comparable solutions that give roughly
equivalent total overall cost or profit. Optimisation programmes allow the computer to provide
the optimum blend to minimise cost and maximise profit.
Example plants
In-line blending systems are normally applied for high volume feed and/or product streams.
Reference literature
[118, VROM, 1999]
Description
See Section 2.21
Applicability
The reasons for batch blending are the minimum storage requirements for strategic reasons,
fiscal and tax control, operating flexibility and tank dimensions.
Example plants
For a variety of reasons given under applicability, batch blending of feed and product is still
done to a certain extent.
Vapour recovery/destruction systems and vapour return during loading applied to storage are
included in Section 4.23.6.2.
Description
Vapour recovery systems, used for the reduction of hydrocarbon emissions from storage and
loading facilities of gasoline and other highly volatile products. Displaced vapour containing
air/gas from loading of volatile products (such as gasoline or products of similar vapour
pressure) into vessels and barges should ideally be recycled or routed through a vapour recovery
unit. The pressure swing adsorption technique is one of a number of techniques to recover
chemical compounds from vapour streams. Examples of these vapour streams are process vents,
refinery fuel gas, flare or incinerator feed streams. The units can recover organic compounds
from gaseous streams. More information in Section 4.23.6.
Table 4.35: Emissions values for vapour recovery plants during the loading of motor gasolines
Cross-media effects
Where explosive mixtures can occur, it is important that safeguards are implemented to limit the
risk of ignitions and ignition propagation. Need to implement safeguards for the handling of
explosive mixtures, especially from mixed chemical vapour streams. Two stage units are high
energy consumers. The energy consumption of these plants is about twice as high as that of
plant with a residual emissions of 5 g/m3, which correlates with higher CO2 emissions.
Operational data
Energy consumption (for cooling, pumping, heating, vacuum), waste (adsorbent / membrane
replacement), water effluent (i.e. condensates from steam regeneration of adsorbent, defrost
water from condensation units).
Applicability
It can be applied to crude loading (with the exception of adsorption if a pretreatment as sulphur
stripper is not used, because of adsorbent fouling) having smaller effectiveness than those
mentioned above because methane and ethane are collected in the vapour stream, despatch
product stations and ship loading of products. These systems are not applicable in unloading
processes when the receiving tank is equipped with an external floating roof. Vapour recovery
units are not considered applicable when recovery of small quantity of chemicals is under
discussion.
Economics
Next table shows some examples of the cost of vapour recovery units
Example plants
Many examples are found in European refineries. Gothenburg oil harbour is going to install a
VRU for ship loading.
Reference literature
[107, Janson, 1999], [181, HP, 1998], [211, Ecker, 1999], [45, Sema and Sofres, 1991], [117,
VDI, 2000], [316, TWG, 2000], [247, UBA Austria, 1998], [268, TWG, 2001]
Description
Vapour destruction is also a technically viable option, which may include energy recovery. It
also includes the destruction of vented products in process heaters, special incinerators or flares.
The heat generated in the combustion can be used to produce low-pressure steam or to heat
water or air.
Cross-media effects
CO2 is produced in the combustion process. CO and NOx may be also produced. Additional fuel
is consumed in the combustion of streams with low concentrations and for preheating of
catalysts. Experience with existing units shows 19 - 82 kg/h of gasoil. Continuous ignition
source within vapour system is a major safety concern that requires implementation of systems
to prevent ignition propagation.
Applicability
Applicable to reactor and process vents, tank vents, loading facilities, ovens and dryers,
remediation systems and other manufacturing operations. Applicability limited to very light
hydrocarbons, methane, ethane and propylene.
Economics
Investment Costs (EUR million) 2 - 25 and 3.2 - 16 for marine loading (0.1 - 0.22 EUR million
/ship for conversion).
Operating Costs: 0.02 - 1.1 EUR million
Retrofit costs are very site-specific.
Example plants
Catalytic combution processes have been applied in the chemical industry.
Reference literature
[181, HP, 1998], [118, VROM, 1999], [268, TWG, 2001]
Description
In order to prevent emissions to atmosphere from loading operations, several options are
available. Where loading takes place from fixed roof tanks, a balance line can be used. The
expelled mixture is then returned to the liquid supply tank and so replaces the pumped-out
volume. Vapours expelled during loading operations may be returned to the loading tank, if it is
of the fixed roof type, where it can be stored prior to vapour recovery or destruction. This
system can also be used for vessels and barges.
Cross-media effects
Due to evaporation in the receiving vessel during transfer (vapour evolution by splashing), there
is usually a surplus volume of vapours compared to the displaced liquid volume. The balancing
line is not considered an efficient means of VOC-abatement for the most volatile liquids.
Operational data
Where explosive mixtures can occur, it is important that safeguards are implemented to limit the
risk of ignitions and ignition propagation. Tanks should remain closed to prevent emissions, so
they may operate under a low pressure and ullaging and sampling should not be undertaken
through open dipatches. Detonation arresters require regular cleaning when vapours contain
particulates (e.g. soot from badly operated cargo tank inerting systems).
Applicability
Not all vapours can be collected. May affect loading rates and operational flexibility. Tanks
which may contain incompatible vapours cannot be linked.
Economics
Investment costs are 0.08 EUR million per tank and the operating costs are low.
Example plant(s)
LPG loading. Subsequent loading of transport containers is carried out by means such as closed
loop systems or by venting and release to the refinery fuel gas system.
Reference literature
[115, CONCAWE, 1999], [268, TWG, 2001]
Description
The loading/unloading pipe is flange-connected to a nozzle situated at the lowest point of the
tank. A vent pipe on the tank can be connected to a gas balancing line, VRU or to a vent. In the
last case VOC emissions can occur. The flange connection in the filling line has a special design
("dry connection") which enables it to be disconnected with minimum spillage/emissions.
Reference literature
[80, March Consulting Group, 1991], [268, TWG, 2001]
Description
Handling of materials used in the refinery may result in spills that contaminate soil or rainwater.
This technique paves and kerbs the area where materials are handled, for the collection of the
possible spilled material.
Example plants
Many examples can be found in European refineries.
Reference literature
[80, March Consulting Group, 1991]
Description
The design and operation of odorising facilities should be to a very high standard which
minimises the risk of odorant leaks or spillages. To an extent, the operational aspects of this are
facilitated by the ease of detection of even the smallest odorant release, but this need to be used
as an aid to vigilance, not as a reason for lax control.
Odorising plants should be designed to minimise the potential for leaks, e.g. by having the
minimum number of pumps/valves/filters/tank connections, etc, by using welded, not flanged,
connections wherever possible and by protecting the plant from possible impact damage. All
such items used need to be designed to a very high standard of sealing efficiency. Devices such
as automatic self-sealing couplings for loading lines are preferred.
Plants should be designed to deal with the high vapour pressure of odorants and should use inert
or pressure controlled natural gas blanketing of the vapour spaces above stored odorant. During
deliveries into tanks any displaced vapour need to be returned to the delivery vehicle, adsorbed
on activated carbon or incinerated. The use of demountable semi-bulk containers avoids the
generation of displaced vapours and is preferred where appropriate.
Reference literature
[18, Irish EPA, 1992]
Description
In overground pipelines is easier to detect leakings
Applicability
Its implementation is easier in new installations.
Economics
The replacement of existing underground pipelines is typically very expensive.
Reference literature
[19, Irish EPA, 1993]
4.22 Visbreaking
4.22.1 Deep thermal conversion
Description
Deep thermal conversion bridges the gap between visbreaking and coking. It maximizes yields
of stable distillates and produces a stable residue, called liquid coke.
Cross-media effects
Increased energy consumption.
Economics
The investment amounts to 9120 to 11300 EUR/m3 installed, excluding treatment facilities and
depending on the capacity and configuration (basis: 1998)
Reference literature
[212, Hydrocarbon processing, 1998], [316, TWG, 2000]
4.22.2 Hydrovisbreaking
Economics
Investment (basis: 750 kt/yr of topped heavy Canadian crude; battery limits including desalting,
topping and engineering; 1994 USA$ Gulf Coast)
Example plants
This technology was tested in a 2 Mt/yr visbreaker in a refinery in Curaçao.
Reference literature
[212, Hydrocarbon processing, 1998], [250, Winter, 2000]
Description
In this process, a soaker drum is added after the furnace. So the cracking takes place at lower
furnace outlet temperatures, and longer residence time.
times before shutting down to remove coke from the furnace tubes (run times of 6 - 18 months
versus 3 - 6 months in coil cracking).
Cross-media effects
The above apparent advantage to reduce the number of clean-ups for soaker visbreakers is at
least partially balanced by the greater difficulty in cleaning the soaking drum.
Operational data
Run times of 3 - 6 months are common for furnace visbreakers and 6 - 18 months for soaker
visbreakers. Fuel consumption is around 11 kg FOE/t. Power and steam consumption is similar
to the coil one. Operating temperature is 400 - 420 ºC.
Example plants
Some soaker visbreakers exists in European refineries.
Reference literature
[297, Italy, 2000], [268, TWG, 2001]
Description
The gas generated in the visbreaking may be sour and contain sulphur compounds, in which
case a gas sweetening operation, typically amine scrubbing, should be applied before the gas is
recovered into products or used as refinery fuel gas. Visbreaking gas, apart from being amine
scrubbed to remove H2S, is further treated to remove approximately 400 to 600 mg/Nm3 of
mercaptan sulphur (depending on feed) in order to meet the refinery fuel gas sulphur-content
specification.
The aqueous condensate is usually sour, and should pass to the sour water stripper in an
enclosed system.
Cross-media effects
Need for chemicals and utilities to run the processes.
Example plants
Those techniques are applied in many refineries.
Reference literature
[316, TWG, 2000]
Description
During the thermal cracking some coke formation is produced and deposits in the furnace tubes.
The coke should be cleaned when necessary. It exist in the market additives to control the
sodium content within the feedstock. Also the control of the addition of caustics to the feedstock
upstream can be also used.
Reference literature
[268, TWG, 2001]
Description
CO boilers and catalyst reduction of CO (and NOx). Good primary measures for the CO
reduction are:
• good operation control
• constant delivery of liquid fuel in the secondary heating
• good mixing of the exhaust gases
• catalytic afterburning
• catalysts with oxidation promoters.
Applicability
FCC and in heavy residues fired. (Section 4.10.4.9 and 4.5.5)
Reference literature
[316, TWG, 2000]
Description
Unlike SO2, NOx or particulates treatment of flue gases, as a feasible abatement technology for
CO2 is not available. CO2 separation techniques are available but the problem is the storage and
the recycling of the CO2. Options for the refiner to reduce CO2 emissions are:
I. effective energy management (topic discussed in Section 4.10.1) including:
• improving heat exchange between refinery streams
• integration of refinery processes to avoid intermediate cooling of components (e.g.
4.19.1 Progressive distillation unit)
• recovery of waste gases and their use as fuels (e.g. flare gas recovery)
• use of the heat content of flue gases.
II. use of fuels with high hydrogen contents (topic discussed in Section 4.10.2)
III. effective energy production techniques (see Section 4.10.3). It means the highest
possible recovery of energy from fuel combustion.
Cross-media effects
Use of fuels with high hydrogen contents reduce CO2 emissions from the refineries but overall
will not reduce CO2 as these fuels will not then be available for other uses.
Operational data
Rational energy use needs good operation to maximise heat recovery and process control (e.g.:
O2 excess, heat balances between reflux, product temperature to storage, equipment survey and
cleaning). To get optimum results repeated operator training and clear instructions are
necessary.
Reference literature
[252, CONCAWE, 2000], [268, TWG, 2001]
The NOx emission level of an oil refinery or a natural gas plant is influenced by the refining
scheme, the type of fuels used and the abatement techniques implemented. As the number of
emission sources and their contribution to the overall emission level may vary greatly, one of
the first techniques to consider is the precise quantification and characterisation of the emission
sources in each specific case. For more information see Section 3.26.
Description
The low-temperature oxidation process injects ozone into a flue gas stream at optimal
temperatures below 200 ºC, to oxidise insoluble NO and NO2 to highly soluble N2O5. The N2O5
is removed in a wet scrubber by forming a dilute nitric acid waste water that can be used in
plant processes or neutralised for release. General information about this technique can be found
in Waste water and wastegas BREF [312, EIPPCB, 2001]
Cross-media effects
Ozone should be produced on site, on demand, from stored O2. Risk of ozone slip. The system
can increase nitrate concentration in the waste water.
Operational data
The use of ozone and the low optimal temperatures of the process provide stable treatment
conditions. Energy consumption for the production of ozone range from 7 - 10 MJ/kg (2 – 2.8
kWh/kg) of ozone produced with a concentration of 1-3 % w/w with dry oxygen feed.
Temperature should be below 160 ºC to minimise ozone decomposition. Heavy particulate
laden fuels may require additional equipment.
Applicability
It can be used as a stand-alone treatment system or can follow other combustion modifications
and post-combustion treatment systems, such as low-NOx burners, SCR or SOx removal, as a
final polish step. The process can be used either as a “stand alone”treatment system or as a
retrofit to existing plant. It can also be used as a polishing system to other NOx removal
technologies and for the treatment of ammonia slips.
Economics
It requires low maintenance and low operator interface. The relative capital cost and operating
costs given by technology providers claim that are equal to, or less than SCR-type systems.
Example plants
The process is in use on USA commercial installation for sectors as acid pickling, steam boilers
and coal fired boilers. A demonstration is currently planned on a refinery FCCU.
Reference literature
[181, HP, 1998], [344, Crowther, 2001], [268, TWG, 2001]
Description
SNCR is a non-catalytic process for removing oxides of nitrogen from the flue gas by gas phase
reaction of ammonia or urea at high temperature. General information about this technique can
be found in Waste water and wastegas BREF [312, EIPPCB, 2001]
Cross-media effects
Ammonia injection produces significantly less N2O and CO emissions than other reagents.
When the temperature lower it is needed to reheat-up the flue gas with the consequent use of
energy. The systems also need the equipment necessary for the storage of ammonia or urea.
Disadvantages are also NH3 slip (5 - 40 mg/Nm3) and possible side reactions (N2O). The U.S.
EPA identified ammonia as the single largest precursor of fine particulate (2.5 microns in size
and smaller) air pollution in the country.
Operational data
SNCR process needs to achieve temperatures of 800 ºC to 1200 ºC for efficient conversion.
With the injection of hydrogen along with the NH3 and carrier, optimum temperatures can be
shifted to as low as 700 ºC. To achieve good mixing, the small amount of reactant is injected
along with a carrier gas, usually air or steam. The ammonia produced in the sour water stripping
(see Section 4.24.2) can be used as denox agent. The effectiveness of this process is very
temperature dependent.
Applicability
It is typically applied to flue gases from heaters and boilers. No or very small space
requirements. Space requirements are mainly restricted to NH3 storage.
Economics
Cost considerations include the initial capital costs for modifying the furnace or boiler, piping to
inject the reactant, the reactant supply system and a recurring cost for ammonia or urea to react
with the NOx. Cost estimates for the use of an SNCR after a power plant are reported in the
Section 4.10.4.6. The following table gives cost effectiveness of the use of SNCR under
different conditions
Example plants
Applied to FCCs and boilers
Reference literature
[211, Ecker, 1999], [302, UBA Germany, 2000], [115, CONCAWE, 1999], [181, HP, 1998],
[247, UBA Austria, 1998], [268, TWG, 2001]
Description
A further deNOx technique is known as catalytic deNOx. The ammonia vapour is mixed with the
flue gas through an injection grid before being passed through a catalyst to complete the
reaction. Various catalyst formulations are available for different temperature ranges: zeolites
for 300 - 500 °C, traditional base metals employed between 200 - 400 °C, for low temperature
applications 150 - 300 °C metals and activated carbon are used. More information about this
technique can be found in the Waste gas BREF [312, EIPPCB, 2001]
Cross-media effects
SCR has some cross-media effects as NH3 slip of less than <2 – 20 mg/Nm3 depending on the
application and requirements or sulphate aerosols are controllable to very low levels. SCR
operation at the end of its catalyst life may also entail bigger slip of NH3. This means a
constraint in conversion efficiency. The U.S. EPA identified ammonia as the single largest
precursor of fine particulate (2.5 microns in size and smaller) air pollution in the country. N2O
generation can occur with some catalysts. At the end of its life, some components of the catalyst
can be recycled and the other are dispose.
Operational data
The ammonia produced in the sour water stripping (see Section 4.24.2) can be used as denox
agent. The catalytic deNOx, process operates at low temperatures (250 ºC to 450 ºC). Operator
and maintenance requirements are low for clean gas applications. More attention is required for
applications where fouling substances like SO3 and soot or dust are present in the off-gas. The
presence of SO3 will lead to formation of ammonium sulphates, which have a detrimental effect
on the activity of the catalyst and lead to fouling of downstream heat exchangers. However
sulphate deactivation can be avoided with a relevant temperature control. Reactivation of the
catalyst can be executed by heating at 400°C, to counteract fouling a water washing step is
generally applied. Soot has to be removed by blowing of the catalyst with steam or inert gas
(e.g. nitrogen), while a ‘dummy’ layer can optionally be applied for extra protection of the
catalyst. If the flue gas contain particulates containing metals, metal deactivation may occur.
The NH3 consumption is directly related to the NOx concentration in the flue gas and amounts to
about 0.4 tonne NH3 per tonne of NOx removed. The amount of ammonia is normally dosify in
a 10 % more than the stoquiometric needs. The catalyst has a lifetime of 4 to 7 years for oil
firing, for gas firing 7 to 10 years. The pressure drop of the catalyst bed will lead to a minor
amount of additional power consumption from fans in the system or loss of efficiency in case of
a gasturbine application (equivalent to 0.5 - 1 % of the heat-input). Additional fuel in the event
of reheating the flue gas is necessary, corresponding to a loss of energy efficiency of 1 - 2 %.
Common safety precautions need to be taken when using pure NH3. To avoid elaborate safety
measures (required for larger volumes of NH3) the trend is towards using aqueous NH3 or urea.
The NH3 storage also varies in size allowing 1 - 3 week operational hold-up.
When a so-called tail-end configuration is preferred (e.g. where the SCR is projected behind a
wet scrubber, or at the end of a gasfired unit where the flue gas has reached a temperature of
less than 150°C), it may be necessary to reheat the flue gases to have an adequate temperature
for the reduction reaction. For these applications in-line burners are required. If the required
temperature rise is larger than 70 - 100°C heat recuperation by means of a gas-to-gas heat
exchanger could be cost-effective.
Applicability
Selective catalytic reduction (SCR) has been applied to combustion flue gases or process off-
gases (e.g. FCC). The introduction of an SCR-system into an existing installation is a challenge
due to space, pressure and temperature problems. Creative solutions can often be found to
reduce the retrofit costs. Higher operating temperatures reduce the catalyst size and costs but
introduce retrofit complexity. Lower operating temperatures increase the required catalyst
volume and costs but often allow a simple retrofit.
Sofar, predominantly operating temperatures between 200 and 450°C prevail. These
temperatures are normally available before the economiser section or the air preheater of a
boiler. Gas fired refinery furnaces generally have a stack temperature of 150 - 300°C.
Dependent on the flue gas sulphur content a medium (MT) or low temperature (LT) catalysts
can be applied in tail-end configuration. Zeolite type catalysts have reached the market place. In
the USA a number of gas turbines have been fitted with these catalysts, some other applications
also exist (the DESONOX process, a combined technique for SO2 and NOx removal, has such
catalyst). As for oil firing, only MT catalysts can be applied, because of the presence of sulphur
and particulates. MT catalysts have been widely applied in coal fired power plants
predominantly at conditions where the flue gas still contain all the fly-ash and SO2 from the
boiler. Two other configurations are low dust / high SO2 level and low dust / low SO2 level (tail-
end configuration). There is little experience with the use of SCR for units applying vacuum
residue as fuel. However, the power plant of the refinery Mider in Germany uses vacuum
residue, visbreaker residue and FCC-slurry as fuel in its three oil-fired heaters. A SCR-plant in
high dust configuration is installed after the heaters, cleaning the gas of NOx to concentration
less than 150 mg/Nm3. The particular difference oil firing represent, lies in the nature of its fly
ash. In contrast to coal, the fly ash is extremely fine (> 90 % fall in the category of PM2.5),
though the quantity is much less. The composition is also quite different. Coal fly ash
predominantly consists of silica compounds, with a minimum of unburnt HCs. Oil fly ash
contains metal oxides, soot and coke. Unabated fly-ash concentrations for oil firing are in the
range of 100 - 600 mg/Nm3 (with highest values for vacuum residues). SCR applied under these
conditions may suffer from clogging by fly ash and sulphates (at coal firing the fly ash by its
sandblasting effect keeps the catalyst clean). The potential for sulphate precipitation is generally
higher with vacuum residue, due to its high sulphur content (2.5 - 4 %).
Pressure drop can be an important consideration as to whether SCR can be applied to a flue gas
system. For this reason natural draft furnaces could probably not be fitted with it.
See Section 4.23.8 for combined flue gas desulphurisation and deNOxing.
GAS TURBINE FLUE GAS STACK CANNED FURNACE WITH WEAK STRUCTURE
PC
BOILER HOUSE
ESP OF CATCRACKER
ECONOMISER
ECONOMISER
ESP ESP
Figure 4.9: Some examples of SCR retrofit situations in refineries (clean gas situations)
Economics
See applications in each Section (Energy, catcracker and coker). The investment costs of new
SCR-system depend largely on the flue gas volume, its sulphur and dust content and the retrofit
complexity. The total erection costs for an SCR system (including all material costs - i.e. reactor
system + first catalyst charge, NH3 dosage & storage, piping & instrumentation -, engineering
and construction costs) for a flue gas flow of 100000 Nm3 are EUR 1 - 3 million for gas firing
and between EUR 3 - 6 million for liquid fuel firing. For existing installations the possible
location of the catalyst bed is often constrained by space limitations, which will lead to
additional retrofit costs.
Operation costs comprise costs of energy, NH3 and catalyst replacement. Specific costs for NH3
and the catalyst charge are of the order of EUR 250 per tonne NH3 respectively EUR 1 per Nm3
of flue gas when firing gas and EUR 1.5 per Nm3 of flue gas when firing oil.
The following table gives cost effectiveness of the use of SCR under different conditions
Example plants
It has been applied to FCC off-gases, gas turbines, process boilers, process heaters. SCR has
been applied successfully on a large variety of applications: coal and oil fired power plants,
waste incineration plants, diesel and gas engines, gas turbine plants, steam boilers and refinery
furnaces (such us naphtha reformers, steam reformers, crude and vacuum distillations units,
thermal cracking and hydroprocessing units) and FCC plants. Some examples are shown in
Figure 4.9. Less experience is available with flue gas from firing high-sulphur residual oil.
SCR is used extensively on power plant in Japan, Germany, Austria and on gas turbine plant in
Netherlands, California in the USA- SCR are also widely used in waste incineration plants. To
date, SCR has been sucessfully applied worldwide in refinery processes as power plants and the
FCCs. For example, at refineries in Japan, SCR applications are common. In European
refineries, more than 6 applications have been found (Austria, the Netherlands and Sweden).
Reference literature
[302, UBA Germany, 2000], [211, Ecker, 1999], [118, VROM, 1999], [257, Gilbert, 2000],
[175, Constructors, 1998], [181, HP, 1998], [115, CONCAWE, 1999], [268, TWG, 2001]
4.23.4 Particulates
PM emissions from refinery units entail particles present in the flue gas from furnaces,
particularly soot, catalyst fines emitted from FCC regeneration units and other catalyst based
processes, the handling of coke and coke fines and ash generated during incineration of sludges.
Particulates generated in the refinery contain metals. Then, a reduction in the particules content
reduces the metal emissions from the refinery. The particles may range in size from large
molecules, measuring a few tenths of a nanometer, to the coarse dusts arising from the attrition
of catalyst. Distinction is usually made between aerosols, < 1 - 3 µm and larger dust particles.
The avoidance of fine particulates < 10 µm (PM10) is particularly important for health reasons.
The available dust removal techniques can be subdivided into dry and wet techniques or a
combination of the two. The usually at refineries applied techniques for dust emission reduction
are briefly discussed below. Dry techniques are cyclones, electrofilters and filtration and wet
techniques are wet scrubbers and washers. More information can be found in the waste gas
BREF [312, EIPPCB, 2001].
4.23.4.1 Cyclones
Description
The principle of cyclone separation is based on centrifugal force, whereby the particle is
separated from the carrier gas. More information can be found in the waste gas BREF [312,
EIPPCB, 2001].
Cross-media effects
A reduction of the environmental impact is best achieved if a useful outlet is found for the
collected solid dust material. Dust collection is essentially a shift of an air emission problem to a
waste problem.
Operational data
Cyclones can be designed for high temperature and pressure operation. Dust collection
equipment is normally simple to operate and fully automated. For the dry separation no utilities
are needed. For dust removal normally no additives are used. Sometimes the collected dust
needs rewetting to avoid spreading of dust during handling.
Applicability
Multi-cyclones as employed in FCC-units, cokers are not able to remove fine particulate matter
(PM10 = dust <10 µm) and therefore they are mainly used as pre-separation step.
Example plants
In refineries multi-cyclones and ESPs are employed in FCC-, heavy oil and residue cracker
units.
Reference literature
[250, Winter, 2000], [118, VROM, 1999]
Description
The basic principle of operation of electrofilters (Electro-Static Precipitator, ESP) is simple. The
gas is ionised in passing between a high-voltage electrode and an earthed (grounded) electrode;
the dust particles become charged and are attracted to the earthed electrode. The precipitated
dust is removed from the electrodes mechanically, usually by vibration, or by washing in so
called wet electrofilters.
Cross-media effects
High voltage in an ESP introduces a new safety risk in refineries. A reduction of the
environmental impact is best achieved if a useful outlet is found for the collected solid dust
material. Dust collection is essentially a shift of an air emission problem to a waste problem.
Operational data
An ESP uses some electricity. For dust removal normally no additives are used. Sometimes the
collected dust needs rewetting to avoid spreading of dust during handling.
Applicability
Application of ESPs can be found in FCC-units, FGD processes, power plants and incinerators.
It may not be applicable for some particulates with high electric resistance. They typically can
be installed in new and existing plants.
Economics
Abatement Plant size range Approximate capital Approximate operating
cost (EUR million cost per year
installed) (EUR million)
Electrostatic precipitator Flue gas flows 0.16
150000 Nm3/h 1.9
650000 Nm3/h 4.4
They are battery limit costs based on 1998 prices and include such items as equipment, licence fees,
foundations, contruction and commissioning. They are an order of magnitude only. Site-specific factors
such as layout, available space and necessary modifications to existing plant could have a significant
impact. In some cases these factors might be expected to increase the costs by some 50 %.
Operating costs are in the order of EUR0.5 to 10 per 1000 Nm3 treated flue gas. Inclusion of
fines disposal costs gives figures 2.5 - 3 times higher than the ones specified in the above table
and 10 times higher the operating cost.
Example plants
In refineries multi-cyclones and ESPs are employed in FCC-, heavy oil and residue cracker
units.
Reference literature
[250, Winter, 2000], [118, VROM, 1999]
4.23.4.3 Filtration
Description
Fabric filters
Cross-media effects
A reduction of the environmental impact is best achieved if a useful outlet is found for the
collected solid dust material. Dust collection is essentially a shift of an air emission problem to a
waste problem. Filter material life is limited (1 - 2 years) and may add a disposal problem.
Operational data
Dust collection equipment is normally simple to operate and fully automated. For the dry
separation only limited utilities are needed.
Applicability
Fabric filters are effective, except in applications with sticky dust or at of temperatures over
240 ºC. Filters are typically used for cleaning of flue gases with a volume < 50000 Nm3/h.
Reference literature
[250, Winter, 2000], [118, VROM, 1999]
Description
In wet scrubbing the dust is removed by counter-current washing with a liquid, usually water,
and the solids are removed as slurry. Venturi and orifice scrubbers are simple forms of wet
scrubbers. The Electro-dynamic venturi washer (EDV), is a recent development which reduces
the dust emission to 5 mg/Nm3. The technique combines venturi with electrostatic dust
separation. EDV is sometimes used in the flue gas treatment of combustion installations and
incinerators.
Cross-media effects
A reduction of the environmental impact is best achieved if a useful outlet is found for the
collected solid dust material. Dust collection is essentially a shift of an air emission problem to a
waste problem. Scrubbers for dust removal, may also be effective for SO2 reduction.
Operational data
Dust collection equipment is normally simple to operate and fully automated. Wet scrubbing
requires pumping energy as well as water and alkali.
Applicability
Spray towers have a low pressure drop but are not suitable for removing particles below 10 µm.
Venturi and packed bed washers have been installed for sludge incinerators.
Economics
Operating costs are in the order of EUR0.5 to 10 per 1000 Nm3 treated flue gas.
Example plants
Some FCC-units are equipped with scrubbers.
Reference literature
[250, Winter, 2000], [118, VROM, 1999]
4.23.4.5 Washers
Description
Ventury washers.
Centrifugal washers combine the cyclone principle and an intensive contact with water, like the
venturi washer.
Operational data
Dust collection equipment is normally simple to operate and fully automated.
Venturi washers, need to be operated at sufficient pressure and at full water saturation of the gas
phase to reduce the dust emission.
Applicability
Wash columns, or absorbers with packing are applied in various processes.
Example plants
Ventury washers: they are mostly applied for removal of a combination of dust, acidic
components (HCl and HF) at for example cokers and incinerators. For SO2 removal scrubbing at
pH ± 6 with caustic or lime is necessary.
Reference literature
[118, VROM, 1999]
Sulphur is an inherent component of crude oil. Part of this sulphur leaves the refinery in its
products, part of it is emitted to the atmosphere (already described in all the processes) and part
of it is recovered by some processes located in the refinery for this purpose (more information
about the sulphur split in a refinery in Section 1.4.1). Sulphur is an environmental issue that
cannot be addressed in an integrated manner if sulphur coming out with the products is not
considered. In other words, decreasing emissions from a refinery may result in the production of
fuels (products) that may later be burned in a way that is not environmentally efficient, thereby
jeopardising the environmental effort made in refineries.
Because it is such an integrated issue, proper sulphur management (See Annex VI) should also
consider other topics that are not within the objectives of this document. For example, a global
environmental impact assessment of sulphur dioxide emissions should contain:
1. Atmospheric emissions from processes generating SOx (furnaces, boilers, FCC, etc.). Those
issues are in fact included in this document in each of the process sections.
In principle refiners have the following options and combinations thereof to reduce the SO2
emissions of the refinery:
• increase the use of non sulphur containing gas (LPG, natural gas, etc.) discussed in
Section 4.10 Energy system
• to lower the sulphur content of the fuels used (reduce the H2S content in the refinery fuel
gas with amine treating Section 4.23.5.1, gasification of heavy fuels Section 4.10,
hydrotreatment of fuels in Section 4.13.2 )
• to use low sulphur crudes discussed in Sections 4.10 Energy system
• increase the efficiency of the SRU discussed in Section 4.23.5.2 Sulphur recovery units
(SRU)
• to apply end-of-pipe techniques to capture the SO2 from the fumes (namely flue gas
desulphurisation(FGD)) (those techniques are the ones discussed in Section 4.23.5.4and
have been considered in Sections 4.5 Catalytic cracking and 4.10 Energy system).
• reduce SO2 emission from typically small contributors whey they become significant
part of the total emission. For example in the case of gas refineries, the overall picture of
the sulphur emissions change because small contributors in liquid-fired refineries
become significant.
In this section, only those processes dealing with the recovery of sulphur is going to be
considered. Between 17 to 53 % (average 36 %) of the sulphur is recovered in European
refineries. H2S formed at various processes such as hydrotreating, cracking, and coking in a
refinery, finally ends up as contaminant in refinery fuel gas and treat gas streams. In addition to
H2S these gases also contain NH3 and to a lesser extent CO2 and traces of COS/CS2. H2S
removal from these gases is achieved by extraction with an amine solvent. After regeneration of
the solvent the H2S is released and sent to a Sulphur Recovery Unit (SRU).
Description
Before elemental sulphur can be recovered in the SRU, the fuel gases (primarily methane and
ethane) need to be separated from the hydrogen sulphide. This is typically accomplished by
dissolving the hydrogen sulphide in a chemical solvent (absorption). Solvents most commonly
used are amines. Dry adsorbents such as molecular sieves, activated carbon, iron sponge and
zinc oxide may also be used. In the amine solvent processes, amine solvent is pumped to an
absorption tower where the gases are contacted and hydrogen sulphide is dissolved in the
solution. The fuel gases are removed for use as fuel in process furnaces in other refinery
operations. The amine-hydrogen sulphide solution is then heated and steam stripped to remove
the hydrogen sulphide gas. In Figure 4.10 a simplified process flow diagram of an amine
treating unit is shown.
ACID GAS
FUEL GAS
TO SULPHUR
RECOVERY
WATER
MAKE-UP
ABSORBER REGENERATOR
LEAN
AMINE
FLASH DRUM
OPTIONAL
The main solvents used are MEA (Mono ethanol amine), DEA (diethanol amine), DGA
(diglycol amine), DIPA (Di-isopropanol amine), MDEA (Methyl diethanol amine) and a
number of proprietary formulations comprising mixtures of amines with various additives. One
important issue concerning the selection of the type of amine is the selectivity concerning H2S
and CO2.
1. MEA has had a wide spread use, as it is inexpensive and highly reactive. However, it is
irreversibly degraded by impurities such as COS, CS2 and O2, and therefore is not
recommended to use when gases from cracking units are present.
2. DEA is more expensive than MEA but is resistant to degradation by COS and CS2 and
has obtained a wide spread use.
3. DGA is also resistant to degradation by COS and CS2 but is more expensive than DEA
and has the disadvantage to absorb also hydrocarbons.
4. DIPA, which is used in the ADIP process, licensed by Shell. It can be used for selective
H2S removal in the presence of CO2 and is also effective in removing COS and CS2.
5. MDEA is nowadays most widely used, MDEA has a similar characteristic as DIPA, i.e. it
has a high selectivity to H2S, but not to CO2. As MDEA is used as a 40 - 50 % solution
(activated MDEA) in water, this has also potential energy savings. Because of the low
selectivity for CO2 absorption DIPA and MDEA are very suitable for use in Claus Tailgas
amine absorbers, as these do not tend to recycle CO2 over the Claus unit. MDEA is
applied as a single solvent or as in proprietary formulation comprising mixtures.
Cross-media effects
Operational data
The use of selective amines should be considered, e.g. for a stream containing carbon dioxide.
Measures should be taken to minimise hydrocarbons entering the sulphur recovery system;
operation of regenerator feed drums should be controlled to prevent hydrocarbon accumulation
in, and sudden release from, the amine regenerator as this is likely to lead to an emergency shut-
down of the SRU.
Utility consumption per tonne of H2S removed in an amine treating unit is approximately:
Usually a fresh solvent make up rate of 10 - 50 t/yr is required to maintain solvent strength for a
5 Mt/yr refinery.
Amine solutions should be re-used wherever possible and where necessary, suitably treated
before disposal which should not be to land. Recycling of monoethanoleamine-solutions:
Corrosive salts, which concentrate during recycling, can be removed via ion exchange
techniques. Some proprietary solutions may be biodegradable under suitable conditions.
It is important also that the amine processes have sufficient capacity to allow maintenance
activities and upsets. This sufficient capacity can be achieved by having redundancy equipment,
apply load shedding, emergency amine scrubbers or multiple scrubber systems.
Applicability
Process off-gas streams from the coker, catalytic cracking unit, hydrotreating units and
hydroprocessing units can contain high concentrations of hydrogen sulphide mixed with light
refinery fuel gases. Emergency H2S scrubbers are also important.
Economics
The cost of upgrade the refinery amine treatment system (2 %) to meet 0.01 to 0.02 % v/v of
H2S in fuel gas is around 3.75 to 4.5 million EUR. This cost is battery limit costs based on 1998
prices and include such items as equipment, licence fees, foundations, erection, tie-ins to
existing plant and commissioning. They are an order of magnitude only. Site-specific factors
such as layout, available space and necessary modifications to existing plant could have a
significant impact. In some cases these factors might be expected to increase the costs by some
50 %.
Example plant(s)
Common technology used all over the world
Reference literature
[118, VROM, 1999], [211, Ecker, 1999], [19, Irish EPA, 1993], [268, TWG, 2001]
H2S-rich gas streams from Amine Treating Units (see above section) and Sour Water Strippers
(see Section 4.24.2) are treated in a Sulphur Recovery Unit (SRU) normally a Claus process for
bulk sulphur removal and subsequently in a Tail Gas Clean-up Unit (TGCU, see later in this
section) for trace H2S removal. Other components entering the SRU include NH3, CO2 and to a
minor extent various hydrocarbons.
Description
The Claus process consists of partial combustion of the hydrogen sulphide-rich gas stream (with
one-third the stoichiometric quantity of air) and then reacting the resulting sulphur dioxide and
unburned hydrogen sulphide in the presence of a activated alumina catalyst to produce
elemental sulphur.
OFF-GAS
STEAM TO STACK
ACID GAS LINE BURNER
REACTORS INCINERATOR
REACTOR
MAIN BURNER
LINE BURNERS
230-300°C 200-220°C
AIR STEAM 0.2 barg TAIL GAS
0.3 barg
TO TGT
BFW
BFW STEAM
Figure 4.11: Simplified process flow diagram of a sulphur recovery unit (CLAUS) unit
The capacity of the Claus plants can be increased with the use of oxygen instead of air
(OxyClaus process) however this has not any beneficial effect in the efficiency of the Claus
plant. Use of this process increase capacity up to 200 % in existing Claus sulphur recovery
units, or for a more economical design of Claus sulphur units.
Cross-media effects
The reduction of SO2 leads to an increase of the CO2 emission. For example for a 100 t/d
sulphur claus plant, the application of three reactors would lead to an emission of 4.8 tonnes of
sulphur per day at a cost of 8.5 tonnes of CO2 per day.
Operational data
• Feed/air ratio control, temperature control of the furnace, reactors and condensers and good
demisting of liquid sulphur, especially from the final condenser exit gas stream are
important parameters in obtaining maximum sulphur recovery. Good control and
availability is crucial as a technique, to deliver any design targets. In this line, the use of
state-of-the-art control and monitoring systems can be seen as an important technique. Use
of a tail gas analyser linked to the process control system (feedback control) will aid
optimum conversion during all plant operating conditions, including changes to sulphur
throughput
• To have a SRU configuration with sufficient capacity for the H2S feed to the unit including
the sourest crude oil to be used is important. The duplication of the SRU capacity is
important to consider to obtain low sulphur emissions. This enough capacity also should
consider to allow the scheduled maintenance activity to proceed every two years, without a
significant increase of sulphur emissions.
• To have utilisation factors close to 100% increase how efficient the units are used. Those
capacity factors should plan also major turnaround maintenance.
• use a good furnace burning-zone design and effective furnace temperature and oxygen
control systems where sour water stripper off-gases are a feed stream, because the process
must also be designed and operated to complete the destruction of ammonia. Ammonia
breakthrough may lead to deposition and blockages of catalyst beds by ammonium salts (eg
carbonate/sulphate) and these SRUs need to be monitored for evidence of this.
In some cases, the SRU need a pilot flame when the H2S concentration is so low that a stable
flame cannot be achieved.
Applicability
Fully applicable
Economics
Abatement Plant size Approximate Approximate
range capital cost operating cost per
(EUR million year (EUR million)
installed)
Upgrade SRU with O2 enrichment to 100 t/d 2.1 - 5.3 1.6 (costs are for
increase throughput from100 t/d to 170 t/d. oxygen)
They are battery limit costs based on 1998 prices and include such items as equipment, licence fees,
foundations, erection, tie-ins to existing plant and commissioning. They are an order of magnitude only.
Site-specific factors such as layout, available space and necessary modifications to existing plant could
have a significant impact. In some cases these factors might be expected to increase the costs by some
50 %.
Example plants
In the market, it exists more than 5 licensors of this process. The Claus process is public domain
and virtually applied at any refinery. Two stages Claus process is the most common in Europe.
More than 30 Oxyclaus systems are in operation in the world.
Reference literature
[250, Winter, 2000], [258, Manduzio, 2000], [115, CONCAWE, 1999], [45, Sema and Sofres,
1991], [181, HP, 1998], [114, Ademe, 1999]
Description
Current methods for removing sulphur from the hydrogen sulphide gas streams are typically a
combination of two processes: the Claus process (See section above) followed by a tail gas
clean-up or treatment unit. Since the Claus process by itself removes about 96 % (2 stages) of
the hydrogen sulphide in the gas stream, the TGTU processes are often used to further recover
sulphur.
More than 20 processes for TGTU have been developed in order to enhance the recovery of
Sulphur compounds from natural gas and/or refinery sources. TGTU processes can be broadly
divided according to the principles applied:
- Dry bed processes, where the main process step is achieved on a solid catalyst. Two paths
have been followed within this group: a) Extend Claus reaction on a solid bed, b)Oxidise
sulphur compounds to SO2 prior to absorption, or reaction.
- Liquid Phase Sub-DewPoint processes, consisting of extending the Claus reaction under
sub-dewpoint conditions in liquid phase.
- Liquid scrubbing processes. There are two main categories, H2S scrubbing processes and
SO2 scrubbing processes. In the most commonly applied configurations, H2S or SO2 are
recycled to the upstream Claus Unit.
The first and third categories can further be divided in sub-categories depending on the sulphur
recovery method used. It should be noted that a strict distinction between dry beds and liquid
scrubbing processes may become uneasy as some arrangement combine the capabilities of both
types of processes. Some processes belonging to the four groups above-mentioned are further
explained below; this list is not intended to be exhaustive:
The H2S Scrubbing process is by far the most widely applied. The concept underlying H2S
scrubbing processes are:
- Hydrogenation and hydrolysis of all sulphur compounds to H2S passing it through a cobalt-
molybdenum catalyst with the addition of a reducing gas
- Absorption of H2S by an amine solution (generic amine or specialty amine)
- Regeneration of the amine solution and recycle of the H2S to the upfront Claus reaction
furnace.
Several Licensor currently propose variations on the H2S scrubbing process, using solvents
available on the market place, or in some instances proprietary solvents.
OFF-GAS
TO INCINERATOR
TAIL GAS
FROM SRU RETURN ACID
SCOT BURNER GAS TO SRU
NATURAL GAS QUENCH
WATER COOLER
AIR
WATER
MAKE-UP
ABSORBER REGENERATOR
REDUCTION
REACTOR
STEAM
STEAM BFW
SOUR WATER
TO SWS
Figure 4.12: Simplified process flow diagram of a tail gas Claus unit (SCOT) unit
The Sulfreen process is a dry-bed, sub-dew point absorption process based on the extension of
the Claus reaction, i.e. catalytic oxidation of H2S to S. Basically consists of two (occasionally
three for large capacities) Sulfreen reactors in series with the Claus reactors. Activated Alumina
is used as a catalyst. Regeneration is needed since the sulphur accumulates on the catalyst
decreasing its activity. Sulphur from the hot regeneration stream is condensed in a dedicated
condenser. Two variations are used: Hydrosulfreen and DoxoSulfreen.
The HydroSulfreen adds a conversion step upstream of the first Sulfreen reactor, to perform the
hydrolysis of COS and CS2 to H2S with the help of a activated Titanium oxide Claus catalyst.
The Claus reaction takes place in the HydroSulfreen reactor and produced sulphur is condensed
in a dedicated condenser
The DoxoSulfreen concept is based on two ideas: the upstream units are operated to get a slight
excess of H2S, compared to the quantity necessary to maintain the Claus ratio, therefore a nearly
total SO2 conversion takes place on the conventional Sulfreen catalyst; then the remaining H2S
is directly oxidised to elemental Sulphur
In the Beaven process, the hydrogen sulphide in the relatively low concentration gas stream
from the Claus process can be almost completely removed by absorption in a quinone solution.
The dissolved hydrogen sulphide is oxidized to form a mixture of elemental sulphur and hydro-
quinone. The solution is injected with air or oxygen to oxidize the hydro-quinone back to
quinone. The solution is then filtered or centrifuged to remove the sulphur and the quinone is
then re-used. The Beaven process is also effective in removing small amounts of sulphur
dioxide, carbonyl sulphide, and carbon disulphide that are not affected by the Claus process.
These compounds are first converted to hydrogen sulphide at elevated temperatures in a cobalt
molybdate catalyst prior to being fed to the Beaven unit.
The CBA (cold bed absorption), process is very similar to the Sulfreen process except in the
fact that the CBA process uses a hot process stream indigenous to the Claus process to
accomplish regeneration of the sulphur loaded catalyst bed. The hot process stream is part of the
effluent of the first Claus reactor. Several configurations are available depending on the number
of Claus converters.
The Clauspol is a process where the tail gas is put in contact with a solvent (polyethylene
glycol) and the reaction of H2S and SO2 is catalysed by a dissolved catalysts. (sodium salt of an
inorganic acid) which is a solvent for H2S and SO2, but not for liquid sulphur.The Claus reaction
can therefore proceed at low temperature (120 °C) and is shifted further to the right as the
produced sulphur is removed from the reaction medium, as it is not soluble and separates.
The LO-CAT process. Absorption and regeneration are performed in a single vessel divided in
two sections: the centerwell and the outer space where aeration with air is performed. The
purpose of the Centerwell is to separate the sulphite ions from air in order to minimize by-
product formation (e.g. thiosulphate). The difference in aeration (and therefore of density)
between the centerwell and the outer space give sufficient driving force for solution circulation
between the absorption and the regeneration zones without the need of a specific pump. The last
type of processing scheme is called the “aerobic unit” and is used to treat air contaminated with
H2S. All reactions take place in the same vessel, at the expense of increased by-product
formation, but with the advantage of a reduced capital cost.
The SO2 abatement from the Claus plant is a process that uses a physical scrubbing mechanism
to remove SO2 from the incinerated tail gas of a Claus plant. The recovered SO2 is recycled to
the inlet of the Claus plant unit.
emits around 5 t/d of sulphur. If a tail gas clean-up process is included in such a refinery the
emissions of sulphur may be reduced to 0.5 t/d, representing that a reduction of 90 % of the
sulphur emissions from the sulphur recovery units. Next table shows the expected overall
sulphur recovery yield, the resulting additional recovered sulphur and the dry basis sulphur
emission (in the form of SO2 specie) after incineration, of the tail gas treatments considered in
this Section.
Table 4.37: Expected overall sulphur recovery yield, the resulting additional recovered sulphur
and the SO2 emissions (dry basis) after incineration
Cross-media effects
The reduction of SO2 leads to an increase of the CO2 emission. For instance the application of a
tailing gas treatment would lead to a SO2 reduction of 96 % (if compared with the three reactor
option), however at an increase for CO2 of 110 %. For example for a 100 t/d sulphur claus plant
with three reactors, the application of a TGTU would reduce the emissions of SO2 to 0.1 t/d but
at a price of increasing the CO2 emissions to 18 t/d.
Operational data
Good control and availability is crucial as a technique, to deliver any design targets. The
estimate of operating costs, including sulphur produced, utilities and chemicals as well as
additional manpower expense are given in the following table:
Applicability
Applicable to both new and existing plants. Capacities range from 2 to more than 2000 tonnes
of sulphur per day from the combined Claus/tail gas treatment units.
Economics
The cost of the SRU depends strongly on the type of tail gas treatment. Following several tables
show some examples of economics of the TGTUs.
Usual practice is to relate the capital cost of the TGCU to the one of the up front Claus unit. The
following table gives estimate of such ratio, for a 100 t/d Claus unit (including catalyst) in a
refinery environment.
The reference to the upstream sulphur unit is indicative and corresponds to the way this kind of
comparison is usually presented in the literature. This comparison should be taken with care
when comparing with other studies, as the capital cost of a sulphur unit may vary greatly.
A particular example of the cost of a three stages Claus plant plus TGTU superclaus process is
shown in the following table:
EXPENSE
Licensing Fee 0.5 2
Subtotal 0.5 2
Final Total 29.4 99
Table 4.40: Economics of tail gas treatment units of the sulphur recovery units
Another example of the cost of a TGTU unit reports that for a Clauspol unit treating a typical
Claus unit tail gas, combined production of 100 tonnes of sulphur per day (ISBL 1998 Gulf
Coast location), the investment (excluding engineering and license fees) came up to 3 million
US$.
The following table shows specific costs data for SO2 abatement under the assumptions outlined
in Annex IV.
Example plants
Reference literature
[195, The world refining association, 1999], [112, Foster Wheeler Energy, 1999], [309,
Kerkhof, 2000], [257, Gilbert, 2000], [115, CONCAWE, 1999], [107, Janson, 1999], [181, HP,
1998], [114, Ademe, 1999], [45, Sema and Sofres, 1991], [346, France, 2001]
Description
In order to reduce the emissions of H2S from the storage and transport of liquid sulphur, the
amount of H2S and polysulphides in the sulphur can be reduced to <10 ppm by oxidation or
treatment with a suitable additive.
Reference literature
[268, TWG, 2001]
Description
The system works with a fixed-bed or batch-type granular reactant.
Applicability
Applications to waste water sytems, land oil tanker vent, oil storage and transportation, and
bitumen plants.
Example plants
More than 1000 applications worldwide.
Reference literature
[181, HP, 1998]
Flue gas desulphurisation is a technique by which SO2 is removed from flue gases or other
waste gases. The process often involves an alkaline sorbent which captures SO2 and transforms
it to a solid product. SO2 in refinery waste gases may have unabated concentration levels of
1500 - 7500 mg/m3. Various FGD methods exist with varying SO2 removal efficiencies. The
FGD market is dominated by the so-called wet lime/limestone process followed by spray dry
scrubbers, and the application of sorbent injection and regenerable processes. More information
can be found in the Waste gas waste water BREF [312, EIPPCB, 2001].
Systems either of the regenerable or the non-regenerable type, do exist for SOx removal only,
but also for simultaneous removal of dust and NOx. Most of these combined processes are still
in the development phase, but some are commercially available already. They are or may
become competitive with systems consisting of separate units for SO2 elimination (i.e. wet
scrubbers) and NOx removal (i.e. SCR).
Description
Additive injection (AI) and Spray Dry Absorber (SD) are scrubbing processes that remove SO2
according to the same principle as WS (reacting with a Ca-based sorbent) however, without the
complicating requirements for producing a high-grade gypsum by-product (i.e. prescrubbing
and oxidising). The by-product is a mixture of sulphites, sulphates and fly-ash for which there
are few or no useful applications.
The AI-process provides moderate SO2 removal at relative low sulphur loads. A dry sorbent is
injected in the furnace. The sorbent used is limestone or hydrated lime (for small boilers, more
active sodium bicarbonate (NaHCO3) is injected in the flue gas duct).
The SNOX process combines high removal of SO2, NOx and particulates. More information see
Section 4.23.8
The seawater scrubbing process uses the natural alkaline - bicarbonates - of seawater to remove
SO2. This implies a potential for high removal efficiencies. The discharge will contain sulphate
and chlorine ions - which are a natural constituent of seawater.
In the Walther (WA)-process, SO2 is absorbed by spray injection of aqueous ammonia, yielding
ammonium sulphite. The sulphite is subsequently oxidised to sulphate. The ammonium salt
solution from the scrubbing section is concentrated in an evaporation unit and granulated. The
end-product is a marketable fertiliser.
The Wellman Lord (WL)-process is the most widely used regenerable process. The process is
based on sodium sulphite/bisulphite equilibrium.
Present-day Wet Limestone Scrubber (WS)-systems are very much improved and less complex
than the earlier systems. Generally a limestone/water slurry is used as the sorbent. Gypsum is
produced by oxidation (aeration) in the sump of the absorber.
Cross-media effects
Operational data
Applicability
Economics
Example plants
Reference literature
[250, Winter, 2000], [257, Gilbert, 2000], [181, HP, 1998], [258, Manduzio, 2000]
When transferring liquids to vessels at atmospheric pressure, the existing mixture of vapour and
gas (often air, but also inerts) in the receiving vessel is often emitted to atmosphere. Such
loading operation is recognised as having an impact on the environment due to the presence of
VOC, a precursor of ozone. EU Stage 1 Directive 94/63/EC requires that Member States
prescribe the specific emission reduction measures is laid down in the annexes of the Directive.
Also the application of VRUs to prevent escape of these vapours to the atmosphere is specified
in the directive. VRUs aim for recovery of hydrocarbons for re-use. In some cases recovery is
not economic, and preference will be given to vapour destruction units (VDU). A more general
term covering both options is Vapour Handling Systems (VHS).
Description
Control of fugitive emissions involves minimising leaks and spills through equipment changes,
procedure changes, and improved monitoring, good housekeeping and maintenance practices.
The aim in all refinery processes should be to prevent or minimise the release of VOCS.
Because of the size, scope and nature of hydrocarbon processing on refineries, this presents a
major challenge, which requires an overall strategy that also devolves down into individual
action at a process unit and plant item level. Most VOCs are released through fugitive losses
from sources including valves, flanges, pump seals and equipment vents. Even a small simple
refinery may have over 10000 potential sources and the problem is magnified proportionally on
the complete complex refineries.
In the case of process component fugitive releases the only real option, is the implementation of
a permanent on-going Leak Detection and Repair (LDAR) programme. This should be
developed and tailored to suit the situation concerned, using appropriate techniques, frequencies
and priorities. It should provide estimates of fugitive VOC releases for monitoring returns and
enable action to be taken to minimise releases. A LDAR contains the following elements:
§ type of measurement (e.g. detection limit of 500 ppm for valves and flanges, against the
interface of the flange)
§ frequency (e.g. twice a year)
§ type of components to be checked (e.g. pumps, control valves, heat exchangers, connectors,
flanges)
§ type of compound lines (e.g. exclude lines that contain liquids with a vapour pressure above
13 kPa)
§ what leaks should be repaired and how fast the action should be taken
The principal areas of fugitive loss are well known and their minimising has been the subject of
much investigation and action on refineries world-wide, mainly led by operators subject to
extremely tight regulation. Some techniques to consider can be:
• an essential first step of any programme is to establish a fugitive release inventory for the
refinery. This normally involves a combination of sampling, measurements, environmental
monitoring, dispersion modeling and estimates based on emission factors
• identify all potential sources of VOC releases, by establishing population counts of
equipment components in line with up-to-date P & I drawings for processes. This survey
should cover gas, vapour and light liquid duties
• quantifying of the VOC releases, initially as "baseline" estimates, and subsequently to more
refined levels. Suitable protocols for this include the “1995 Protocol for Equipment leak
Emission Estimates” (USAEPA-453/R-95-017) for process component losses and API
methods(9) for tankage losses. Some major companies have developed their own techniques
and protocols. USAEPA method defines an emission estimate of 10000 ppm and is more
correct for a plant that until today, has not implemented LDAR programme. A recent report
shown that from 1000 valves analysed by type found that the average leakage measured on
regulating control valves was over 70000 ppm. Some TWG members question that high
emission leakage limit and propose to set the limit to 500 - 1000 ppm and specially with
small chemicals are handled.
• using appropriate dispersion modeling techniques, predict atmospheric mass flux and
concentrations
• employing environmental monitoring techniques, compare the predicted situation with the
measured one
• identification of the processes that have higher fugitive releases
• VOC releases need also to be minimised from refinery utility and ancillary systems and
during plant maintenance and cleaning operations. Hydrocarbons should be minimised in
cooling water and adequate separation facilities provided to deal with incidents of
contamination. After depressurising off-line process vessels to RFG/flare, they need
normally to be steamed out to flare and residual liquids returned to oil slops plants, not run
to drain before venting to air, when personnel entry is required
• vapour recovery/incineration/use
• a strategy to reduce VOC emissions may include a complete inventory and quantification by
a DIAL LIDAR technique (differential absorption light detection and ranging) (see
Section 3.26 for more information).
• emissions that occur through relief valves are to be routed to flare or dedicated incineration
systems, and in exceptional cases vented to a safe location
• fugitive VOC emissions (sometimes containing H2S) can be minimised through double
mechanical seals on pumps, compressors or agitators. For example dual pressurized sealing
systems on pumps virtually eliminate leakage of the process fluid into the environment and
typically have emissions approaching zero, usually described as ‘not measurable with
existing instrument technology’
• use of low emission valve stem packing (500 ppm) on critical valves, e.g. rising stem gate
type control valves in continuous operation, particularly on gas/light liquid high
pressure/temperature duties. Valves is the type of equipment that causes most of the leaks
(40 - 65 %) Valves with rising stems and in particular control valves are an important
source of leaks and may account for as much as 64% of the leaking losses in a plant. It is
therefore recommended that a high integrity low emission packing be installed on these
valves that can provide emissions performance below 500 ppm and at the same time is fire
safe. To ensure low term leakage performance valve live loading spring assemblies mounted
on the gland bolts will ensure that the low emission packing is constantly loaded even as the
material relaxes over time. Control Valve solutions with valve live loading exist as
Maximum Achievable Control Technology (MACT) and emissions performance of 3-5
years is achievable at below 500 ppm. Low emission packing should be specified, and
should be certified by an independent test by a reputable testing house. Packing also be fire
safe. Rising Stem control valves even with low emission packings can leak excessively after
the packing has relaxed in the stuffing box. Studies carried out by a French research
institute have shown average leakage concentrations on control valves to be as much as
70000 ppm. (Power Point Slide Enclosed) In the chemical, petrochemical and refinery
industry the use of packed valves incorporating live loading spring assemblies mounted on
the gland bolts compensates for normal packing relaxation or the effects of thermal cycling
and vibration. Valve Live Loading systems offer considerable improvements in long-term
sealability on rising stem gate valves, globe valves and regulating control valves. On
frequently operated rising stem valves (most problematic) Live Loading offers a
considerable improvement in long-term emissions performance (< 500 ppm for 3-5 years).
Valves have been identified as the greatest source of fugitive emissions in refineries and
rising stem valves particularly frequently operated valves represent the greatest sources of
fugitive emissions.
• use of alternative proven types of low-release valves where gate valves are not essential,
e.g. quarter turn and sleeved plug valves, both of which have two independent seals
• use of balanced bellows type relief valves to minimise valve leakage outside the design lift
range and piping of reliefs to RFG or flare gas, normally via phase separation, without
header back pressure
• minimising the number of flanged connections on pipelines and the use of high
specification jointing materials. Flanges should be minimized and high integrity sealing
materials should be used and should be fire safe materials. In some critical services,
particularly where thermal cycling or possible vibration may exist, flange disc springs can
be mounted under the flange bolts to compensate for the effects of thermal cycling and
gasket relaxation. Such solutions combined with high performance sealing materials can
provide emissions performance of <500 ppm for periods of 3-5 years. Often heat
exchangers are not given a separate consideration in the discussion of fugitive emissions.
On a typical refinery there can be several hundred heat exchangers. Due to the size of these
flanged connections, the large no. of bolts involved, the non uniform temperature on the
flange and the thermal cycling conditions under which many operate, it is recommended
that flange disc springs are fitted under the nuts of the bolts to maintain a constant gland
load on the sealing material. Heat exchangers have a far greater tendency to leak than
normal pipe flanges due to their dimensions and particularly in thermal cycling conditions.
High integrity sealing materials are recommended and not the standard typical metal
jacketed gaskets as indicated in the design codes.
• use of canned pumps or double seals on conventional pumps. Study the use of non-leaking
pumps as diaphragm pumps, bellow pumps, canned rotor pumps, or pumps with magnetic
clutches. Magnetic pumps cannot be used for clean liquids at low temperature. Emissions
from centrifugal pumps can be reduced to 33 % by replacing packed seals by mechanical
seals. Turbo-compressors can be sealed with labyrinth glands, with rotating rings or floating
rings in liquid seals. An escape of the transported medium through the seals should be
prevented e.g. by inert gas or liquid sealing medium at elevated pressure. Furthermore,
magnetic drive pumps and canned motor pumps have limited applicability and care must be
taken when using such pumps in contaminated process streams, process streams containing
particulate and higher pressures. It should also be noted that magnetic pumps typically have
lower efficiency compared to conventional pumps, drawing more power for the same
service. In end effect this transfers emissions from the pump ( as VOC emissions) to the
stack ( as CO2, SO2, NO2 emissions).
• piping of compressor seals, vent and purge lines to RFG or flare systems
• use of end caps or plugs on open ended lines and closed loop flush on liquid sampling
points (no applicable with circulating fluid (pump discharge) or in tanks). Minimising the
releases to air from process hydrocarbon analysers, by optimising sampling
volume/frequency and venting to RFG or flare systems
• install a maintenance drain-out system to eliminate open discharges from drains
• using totally closed-loop in all routine samplers. To follow product quality, inspection
requirements, environmental standards, etc. a sampling schedule is typically set up. With
routine sampling is understood those that are frequently used or needed. For example, by
routine samplers is understood, the samplers of effluent water quality, daily samples of
processed crude fuel fired, fuel gas, certain feedstocks (i.e. FCCU), that are taken to analyse
the sulphur concentration. and samplers from intermediates, final products, tanks, etc.
About 93 % of the sources of fugitive emission are accessible. The achievable emission
reduction depends on the current conditions of the components, with typical reduction rates of at
least 50 to 75 % related to average emission factors. The estimated efficiency for quarterly
inspection and maintenance is 80 to 90 %. Higher efficiencies may be reached, when more
intensive inspection and maintenance programmes are implemented.
A good inspection and maintenance programme for valves and flanges is a very cost-effective
way to reduce NMVOC emissions in a refinery. Savings may reach 0.19 EUR/kg of
hydrocarbon reduced.
Applicability
Fully applicable. See descriptions for applicability of each of the techniques.
Economics
A system for monitoring approximately 3000 refinery components (primarily pumps seals,
valves inhigh liquid vapour pressure or gas services, and road/rail loading) is estimated to costs
about EUR87500 excluding modification of poorly performing equipment and cost of labour.
For a reduction efficiency of 50 % by annual inspection for fugitive emissions, OECD estimated
no additional costs, but a cost benefit due to hydrocarbon savings. A reduction of 80 % by
quarterly inspection and maintenance will cost about 193 EUR/t NMVOC abated. A simple
Inspection and Maintenance programme at a small plant may result in an annual cost of EUR
44000, whereas at a large plant with stringent requirements some EUR 875000 may have to be
accounted.
The cost of leak detection and repair programme for pumps and compressors is 1.75 - 2.5
EUR/kg of hydrocarbon emissions reduced.
Example plants
LDAR programmes have been used successfully world-wide to reduce VOC releases. Typical
survey results show that leaks from glands on valves and pumps are responsible for 90 % or
more of estimated fugitive releases and that a small proportion of valves, virtually all on gas or
high temperature light material streams, contribute almost all of the total.
Reference literature
[107, Janson, 1999], [45, Sema and Sofres, 1991], [112, Foster Wheeler Energy, 1999], [79,
API, 1993], [127, UN/ECE, 1998], [19, Irish EPA, 1993], [260, Sandgrind, 2000], [350,
European Sealing Association, 2001]
Description
Vapour recovery units (VRUs) are installations designed for the emission reduction of volatile
organic compounds (VOC) which are emitted during loading and unloading operations of light
products. As VOC emission abatement by VRUs is only one aspect of total VOC control in a
refinery, this section should be considered in combination with storage, handling and integrated
Refinery management. Several commercial techniques are available for the recovery of VOC.
Those techniques can be divided in two large groups according to the type of separation. One
group includes techniques where the separation of the VOC from the air due to pressure swing
adsorption on activated carbon, absorption by lean oil washing (kerosene), selective membrane
separation or condensation by cooling or compression (this is a special case because separation
and re-condensation are both achieved in a single process). The other group integrates those
techniques that the VOC are separated by condensation to liquid state. It includes re-absorption
into the gasoline or crude oil, condensation and compression. Follows a short description of
those techniques.
• absorption: The vapour molecules dissolve in a suitable absorption liquid (water, lyes,
glycols or mineral oil fractions such as reformate)
• adsorption: The vapour molecules adhere physically to activate sites on the surface of solid
materials, like activated carbon (AC) or zeolite
• hybrid systems. Nowadays, combinations of the VRUs are on the market which are able to
meet very low emissions standards. Examples are cooling/absorption and
compression/absorption/membrane separation.
MEMBRANE
VAPOURS FROM
BARGE, SHIPS, ABSORPTION
STORAGE SECTION VENT OR STACK
COMPRESSOR RECOVERED
LIQUIDS IN
REFORMATE
FRESH
REFORMATE CIRCULATING
REFORMATE
REFORMATE WITH
RECOVERED LIQUIDS
• membrane gas separation: The vapour molecules dissolve into the membrane, move by
diffusion to the other side and desorb into the support material, driven by a pressure
difference. The concentration is higher at the desorption side of the membrane, as the
vapour molecules are preferentially absorbed into the membrane compared to the gas
molecules.
• refrigeration/condensation: By cooling of the vapour/gas mixture the vapour molecules
condense and are separated as a liquid.
Emissions of the various systems are directly related to the reduction efficiency as reported
above and can be as low as 10 mg/Nm3 (without methane). An efficiency of 99.9 %
concentrations of 150 mg/Nm3 (without methane) or 2500 mg/Nm3 (with methane) can be
achieved.
Cross-media effects
Effluent is usually only condensate and is usually negligible.
VRU technique
Absorption If water is used regeneration of the absorption liquid is not necessary, as
water can be treated in a waste water treatment plant. Regeneration more
than doubles the investment + energy costs. The only waste generated is
the exhausted liquid which need to be replaced once in many years.
Adsorption Waste is only relevant in case absorption bed.
Operational data
The handling of VOC involves always safety measures in view of explosion risks (flame
arresters) and the presence of toxic compounds such as benzene. VRUs are compact and require
very little energy and process materials to operate. Usually operating times are longer than two
years.
VRU technique
Absorption If reformate is used recycling to blending tanks is done.
Hybrid systems Due to the complexity of operation it is difficult to maintain the high
performance
Applicability
Vapour recovery units (VRUs) are installations designed for the emission reduction of volatile
organic compounds (VOC) which are emitted during loading and unloading operations of light
products. For a refinery this is particularly relevant to gasoline storage and loading as well as for
products with equivalent volatility characteristics such as naphtha and BTEX.
VRU technique
Adsorption Due to heat of adsorption this method cannot handle high inlet
concentrations (auto-ignition). At higher inlet concentrations the AC is
quickly saturated, so regeneration is necessary. Normally, adsorption
VRUs applies therefore to a two-stage system. The only waste generated
is the exhausted activated carbons which need to be replaced once in
many years.
VRUs occupy limited space. Usually they are preassembled and delivered skid mounted.
Commercial VRU capacities range from 500 - 2000 Nm3/h.
Scores of VRUs at gasoline (un)loading installations for VOC emission reduction have been
built in W-Europe following Stage-1 legislation. Hybrid systems are popular due to simplicity,
good operability and high performance.
100
INLET CONCENTRATION (g/m3)
CONDENSATION
CATALYTIC
10 GAS SCRUBBER
THERMAL
OXIDATION
REGENERABLE
CRYOGENIC
ADSORPTION
CONDENSATION
(REGENERABLE)
MEMBRANES
BIOSCRUBBER ROTATING
ADSORPTION BIOFILTRATION ADSORBER
(NON GENERABLE)
0.1
1 10 100 1000 10000 100000
Figure 4.14: Applicability different techniques for end-of-pipe gas treatment for VOC removal
Economics
A typical VRU with a capacity of 1000 Nm3/h involves a capital cost of EUR 2 million with an
installation factor of 1.5 (as package unit) to 5 (in exceptional cases). Capital investment
strongly depends on site-specific factors, such as the number of loading berths connected to the
system, the distance between berth and the emission control facility (cost of ducting), the need
for blowers, safety systems (explosion and flame arresters) systems. The capital costs can range
from EUR 4 - 20 million for a VRU of 2000 Nm3/h capacity. The investment costs can range
from EUR 2 - 25 million for an efficiency of 99.2 % implying a operating cost from EUR
0.02 - 1 million applied to loading operations (slurry, road, rail and internal refnery
movements). The cost for a VRU unit of 2000 Nm3/h capacity and applying utilities is shown as
follows:
DESCRIPTION k EUR %
INDIRECT COSTS
230 Detailed Engineering 275 7
213 Field Supervision 137 3
213 Inspection 83 2
260 PMT - NPQC - Owner 424 10
Subtotal 919 22
DIRECT COSTS - EQUIPMENT
311 Heat Exchangers 0 0
314 Towers 0 0
315 Drums 0 0
316 Reactors 943 22
324 Pumps and Drivers 13 <1
326 Tanks and Spheres 0 0
347 Cooling Towers 0 0
Subtotal 956 22
DIRECT COSTS - NON EQUIPMENT
208 Scaffolding 25 <1
209 Cleanup Construction 8 <1
216 Construction Equipment 16 <1
307 Excavation & Cable Trenches 142 3
308 Concrete Work and Paving 216 5
309 Piling 33 1
313 Piping 1179 28
318 Structural Steel 77 2
322 Instrumentation 285 7
328 Roads, Walks, Fences 15 <1
337 Electric Power and Light 107 3
345 Communication Equipment 0 0
348 Insulation 3 <1
349 Paint 167 4
380 Standby Equipment (Cap. Spare) 0 0
Subtotal 2273 54
Total Capital 4148 98
EXPENSE
101 Dismantling 17 <1
102 Site Clearance 0 0
103 Relocation 7 <1
104 Reconditioning 0 0
105 Temporary Bypasses 0 0
111 Gas Freeing & Cleaning 8 41
165 Owner Expense Charges 81 2
Subtotal 113 2
Example plants
VRU technique
Membrane gas separation This new technology is now used at a number of refineries.
Reference literature
[258, Manduzio, 2000], [181, HP, 1998], [115, CONCAWE, 1999]
Description
Two systems are relevant in this respect:
• oxidation: The vapour molecules are converted to CO2 and H2O either by thermal oxidation
at high temperatures or by catalytic oxidation at lower temperatures. The stage 1 directive
only allows oxidation in special situations, e.g. when energy is recovered by a gas motor
• biofiltration: Decomposition to CO2 and H2O is achieved at temperatures slightly above
ambient by micro-organisms located in a solid humidified mass.
Cross-media effects
Thermal oxidation can give unwanted combustion products such as NOx, requiring extra
treatment. Catalytic oxidation requires less energy to arrive at combustion temperatures, and can
be competitive with thermal oxidation at low inlet concentrations. Thermal oxidation requires
good primary and/or secondary safety measures to prevent explosions, while the efficiency of
catalytic oxidation may be reduced by catalyst poisoning and ageing. Incineration of VOC also
generates CO2. Combustion of streams with low concentrations and for preheating of catalysts
additional fuel is consumed.
Waste is only relevant in case biofilters are exhausted. No secondary pollutants or wastes are
created.
Operational data
Usually operating times are longer than two years. Biofilters: Input air should be 5 - 55 ºC and
moist.
Applicability
Biofilters: Any gas that can burn in air will oxidise in biofilters. Installation for the treatment of
flows of 17 m3/h up to 135000 m3/h are found in the literature.
Bioxidation: The method is highly suited for treatment of continuous constant-composition air
streams with low concentrations of organic pollutants. The method is not suitable for the direct
treatment of the vapour/air mixtures often encountered in transshipments, because such
mixtures have mostly higher vapour concentrations (>1 % v/v) and appear as sudden peak flows
during the rather infrequent unloading operations. Biotreatment facilities are certainly sensitive
to poisoning by presence of unexpected compounds in the incoming vapour stream.
Consequently most of those systems needs continuous monitoring to prevent the entrance of
undesired compounds.
Economics
Thermal oxidation of VOCs. The economics of a thermal oxidation system with heat recovery
will depend on many factors, including the calorific value of the waste stream. If a thermal
oxidiser is furnished with a gas-to-gas heat exchanger that has an efficiency of 60 % and a gas
flow of 4720 l/s, the payback for heat recovery will be impressive. Assume natural gas, used as
supplementary fuel, costs $20/ million of kcal and electric power cost $0.08/kWh. If a system
runs 24 h/day, 350 d/yr, the payback for the 200000$ additional capital investment for a plant-
type heat exchanger would be less than five months.
Biofilters: Biofiltration costs significantly cost less than the other air pollution control
techniques. Capital costs vary with flowrate and the destruction/removal efficiency. Capital
costs start at about $15/m3/h. Operation and maintenance costs are extremely low because no
fuel or chemicals are required.
Example plants
Thermal incineration: More than 107 units operaring in the world and over 76 portables.
Reference literature
[118, VROM, 1999], [181, HP, 1998]
4.23.7 Flares
Description
Flares are used for safety and environmental control of discharges of undesired or excess
combustibles and for surges of gases in emergency situation or upsets. The flare is usually
required to be smokeless for the gas flows that are expected to occur from normal day-to-day
operation. This gas flow is usually designed at 15 to 20 percent of the maximum design flow.
Flares may become an important SO2 emitter when clean fuels are used within the refinery.
Flare systems normally can be divided into two main sections, i.e. the flare collection system
with a flare knock out drum and the flare stack itself. When dealing with large refinery
complexes separate knock-out drums may be installed in different process areas with ‘blocking-
in’ facilities to allow maintenance during shut downs of these areas.
PURGE
AIR
SUPPLY
FLARE
STACK
STEAM FOR
SMOKELESS
BURNING
KNOCKOUT DRUM
SEAL DRUM
STEAM
STEAM
OILYWATER TO WWTP
TO OIL RECOVERY
FACILITIES OR SLOP
Cross-media effects
Flare seal water is typically in need of treatment before release.
Operational data
Since air present in the stack can create a potentially explosive mixture with incoming flare gas
during low-flare gas loads, a continuous stream of purge gas is required. Also in a lot of cases a
molecular water seal is employed which allows using a smaller purging rate.
Applicability
Flaring of toxic gases (never via a ground flare) requires special considerations. To ensure safe
operation during periods when the flare may not have a flame present, ground level
concentration calculations for hazardous components should be performed assuming the flare as
a vent only. Other safeguards may be necessary to mitigate groundlevel exposure hazards.
Reliable continuous pilot monitoring is considered critical when flaring toxic gases.
There are basically two types of flares, an elevated flare and a ground flare. Ground flares are
used when it is necessary to conceal the flare (for various reasons), otherwise the elevated flare
is commonly chosen, because it can handle larger flow releases more economically. Sometimes
a refinery has both a ground flare and an elevated flare. In that case the ground flare is used to
combust small amounts of continuous vents and small relief loads.
The elevated flare is always a vital part of the refinery complex, its primary purpose being
safety. This flare system is designed to conduct easily ignitable and toxic vapours from the
process area to a remote elevated location. The elevated flare system includes a seal drum to
prevent the flame from flashing back into the process units, and an elevated stack with pilot-
ignition and steam nozzles at the tip. Due to different burning characteristics of the gases, a
separate sour gas flare is usually provided; this flare could be equipped with different burners
than the hydrocarbon flare to allow more efficient combustion of sour gases (H2S).
Flare-gas recovery systems often are installed to comply with local regulatory limits on flare
operation and, therefore must be sized to conform to any such limits.
Example plants
Flares are common in refineries
Reference literature
[101, World Bank, 1998], [19, Irish EPA, 1993], [117, VDI, 2000], [118, VROM, 1999]
Description
The SNOX plant removes SO2, NOx and particulate from the flue gas produced by the coal-fired
boilers at the Gela refinery power plant.
The SNOX plant is based on catalytic processes where the only additional material necessary is
the ammonia used for the NOx removal. Furthermore, natural gas and water are needed, besides
small quantities of silicone oil for the acid mist control unit.
The process produces 94 - 95 % pure sulphuric acid (H2SO4) for sale. The system employs a
catalytic converter to oxidise SO2 to SO3 at a temperature of 400 - 420 °C. This temperature
level allows a deNOx, operating at 380 °C, to be well fitted in the process. High NOx removal at
high NH3 slip is possible with no risk of precipitation of ammonium sulphates, as the reactor
temperatures are above the decomposition temperature (350 °C) and any NH3 slip is destroyed
in the SO2 / SO3 oxidiser. The process produces no waste water or waste products, nor does it
consume any chemical apart from ammonia for NOx control. The high removal of dust is
demanded by the H2SO4 production step. Dedusting with an efficiency consistently at 99.9 % is
required as to avoid frequent clean-up of the SO2 / SO3 converter and to maintain good product
quality. The flue gas train subsequently comprises an air preheater, a high-temperature ESP, the
cold side of the regenerative heat exchanger, the deNOx reactor, a heat supply, the SO2/ SO3
converter, hot-side regeneration heat-exchanger, the acid condenser (operating between
240 - 100 °C, hydrating SO3 and condensing the resulting acid product) which heat is used as
the first step in preheating combustion air. The recovered heat, produced by the conversion
process, is substantial and make up for the power demand when the sulphur content in the fuel
(oil or coal) is 2 - 3 %. Areas relating to dust that require attention in operation are the HTEP,
the SO2 / SO3 converter and the acid falling-film condenser (made of borosilicate glass tubes).
Condensation of H2SO4
The SO3 rich flue gas from the SO2 converter is cooled from approximately 422°C down to
260°C in the gas/gas exchanger, by heat exchange with the cold unconverted flue gas.
Then the SO3 gas is condensed in the WSA heat exchanger.
When the SO3 rich gas is cooled, the SO3 is hydrated to vapours sulphuric acid according to the
following reaction scheme:
1) SO3 + H2O ----> H2SO4 + 24.1 kcal/gmol SO3
The cleaned gas is discharged into the atmosphere through the stack at 106°C.
Cross-media effects
H2SO4 production (95 % w/w H2SO4 concentration) 13 tonne/h with a 5.5 % w/w sulphur
pet-coke
Applicability
The Gela Refinery SNOX plant was designed for cleaning the flue gas produced in the boilers at
power plant burning a mixture of high sulphur Fuel Oil and high sulphur Petroleum Coke.
Economics
A SNOX plant designed for a flue gas load of 1.0 million Nm3/h at the exit of the existing
induced draught fans had a cost of EUR 100 million.
Example plants
GELA AgipPetroli Refinery Start up September 1999
Reference literature
[297, Italy, 2000]
Description
Scrubber with hypochlorite ion is used for odour/low-level VOC abatement.
Description
Flares, compressors, pumps, turbines and air coolers require particular attention as regards
sources of noise. Abatement measures in refineries usually focus on these types of equipment.
Refinery waste water streams that require purification prior to discharge comprises mainly two
streams. The first effluent is process water generated in various refinery units as a consequence
of steam injection and/or washing hydrocarbon fractions with water. The direct contact with
hydrocarbon fractions usually results in a high concentration of dissolved organics. Process
water is treated in a sour water stripper (SWS) when it contains hydrogen sulphide and
ammonia and is subsequently re-used as wash water for the desalter. The resulting effluent is
high in oil and solids, has a high chemical oxygen demand (COD 300 - >600 mg/l) and might
contain also specific contaminants for which legal limits apply. The second significant effluent
stream that requires treatment is contaminated rainwater, referred to as surface water run-off,
which usually requires only removal of free oil and solids as COD is usually below 100 mg/l.
Prior to effluent treatment proper segregation and process water integration should be
considered. Other waste water streams that appear in a refinery are the cooling water blowdown,
tank water bottoms and ballast water. This section only includes the end-of-pipe processes that
appear in treatment of waste water in refinery. Deeper information about the techniques
included in chapter can be found in the Common waste water and wastegas treatments BREF.
Waste water treatment plants are advanced environmental protection systems for the control of
surface water pollution. The purpose of waste water treatment is to remove floating and
dispersed oil, suspended solids, dissolved oil, notably BTEX and phenol and other specific
compounds such as sulphides, cyanides, heavy metals, phosphate (when polymerisation is
present) and nitrogen compounds and other COD. Additionally, the treated effluent can be
rendered suitable for re-use in the refinery process. The principle of waste water purification is
based on (1) dedicated separation techniques for the dispersed and floating oil and suspended
solids from the waste water and (2) removal of dissolved compounds by biological treatment or
further treatments.
This section try to give some indication on what can be done in a refinery with the management
of the waste water. In other words, it is better to combine waste waters from different processes
or is better to process them independtly in some independent waste water plant. Topics related
with the storage of waste water is also included here. This section deals with process waste
water, cooling water and sanitary waste water, ballast water, cleaning water, accidentally oil
contaminated water, continuous oil contaminated water, etc.
Process water integration can be applied at any industrial site, be it a refinery or a combined
refinery and petrochemical complex. In this document the concept and practical tools for
preparing a process water integration scheme for a refinery is presented in Section 4.15.7.1.
Reference is made to a EU BREF document on waste water treatment.
Description
Some techniques to consider in that respect are:
1. use of sour water can be used to the maximum extend possible as desalter wash water or as
wash water in overhead FCC main column
2. process water from the dimerisation process should be treated properly because the high
content on phosphate that waste water from this process typically have
3. use of equalising tanks for the storage of waste water
4. tanks for storage of ballast water may cause large VOC emissions. A way to prevent those
emissions is to cover the equalising tanks of the waste water treatment systems with floating
roofs. Technique to consider also included in the Section 4.21.13
5. the temperature of the waste water is controlled in order to reduce volatilisation and to
secure the performance of the biological treatment
6. rainwater from polluted plant areas (storm water) should be collected and routed to a
treatment plant. For collection a “first flush” scheme may be used. Depending on the initial
degree of pollution (mainly oil), the contaminated water should be treated in oil/water/solids
separation systems (API, flotation units, sandfilters) and/or biotreatment units. Non-
contaminated water may be discharged directly or re-used in order to save costs. It is
necessary that the WWTP is able to handle high quantities of rainwater. For this purpose
sufficient buffer capacity need to be installed (storm water tank). In locations with low
rainfall segregation of streams is less relevant
7. control of surfactants in waste water - Surfactants entering the refinery waste water streams
will increase the amount of emulsions and sludges generated. Surfactants can enter the
system from a number of sources including: washing unit pads with detergents; treating
gasolines with an end point over 400 degrees (F) thereby producing spent caustics; cleaning
tank truck tank interiors; and using soaps and cleaners for miscellaneous tasks. In addition,
the overuse and mixing of the organic polymers used to separate oil, water and solids in the
waste water treatment plant can actually stabilise emulsions. The use of surfactants need to
be minimised by educating operators, routing surfactant sources to a point downstream of
the DAF unit and by using dry cleaning, high pressure water or steam to clean oil surfaces
of oil and dirt
8. install high pressure power washer - Chlorinated solvent vapour degreasers can be replaced
with high pressure power washers which do not generate spent solvent hazardous wastes
9. use non-hazardous degreasers - Spent conventional degreaser solvents can be reduced or
eliminated through substitution with less toxic and/or biodegradable products.
Other data from an EU+ country with many refineries gives the following ranges depending
on the type of refinery:
Other data from another EU+ country gave a range from 0.18 – 0.21 m3/ tonne throughput
for three refineries in his country.
Reference literature
[118, VROM, 1999], [197, Hellenic Petroleum, 1999], [107, Janson, 1999], [262, Jansson,
2000], [268, TWG, 2001]
Sour water from the various refinery units is for the main part stripped in a SWS and can
normally be re-used together with crude distillation unit overhead wash water as desalter wash
water. This is the main process water source in refineries.
Description
Most sour water strippers are single stage, requiring one stripping column. Figure 4.16 shows a
typical single stage sour water stripper. The sour water streams from the process units are
collected in a sour water collection vessel. This provides both hold-up for feed and acts as a
settler, where oil separation takes place. From this vessel sour water is pumped via a
feed/effluent exchanger, to the top of the stripper column. The sour water is counter-currently
stripped in the column by steam, either injected live or generated in a reboiler. This column is
usually refluxed to reduce the water content in the sour gas. Operating pressure in the column
varies from 0.5 to 1.2 barg depending on the destination of the off-gas. Often pH control is
applied to maximise either H2S or NH3 removal.
Sour off-gases from a stripper unit can be routed either to a sulphur recovery unit (SRU), to an
incinerator, or to the sour flare. Since off-gases routed to an incinerator or a flare contribute to
the release of SO2 and NOx, routing to a SRU unit is preferred and now commonly practised.
Normally the off-gas from the sour water stripper, leaving the reflux drum contains
approximately 30 mol% water.
Two stage stripping:. The two-stage sour water stripper differs from the single stage stripper by
operating the first column at a lower pH (6), removing H2S over the top and NH3/water via the
bottom, and the second stage at a higher pH (10), removing the NH3 over the top and a stripped
water stream over the bottom. These results in much lower H2S and NH3 concentration in the
stripped water to be discharged to the effluent treatment plant.
AMMONIA
SOUR WATER
STRIPPER
SETTLER
STEAM STEAM
SLOPS OIL
2nd STAGE
WATER
TO WWTP
Figure 4.16: Simplified process flow diagram of a sour water stripping unit (SWS)
Note: The section enclosed in the dotted box is the second stage of a two stage SWS
Stripped sour water can be routed to the process units for re-use or to the waste water treatment
plant, after appropriate cooling if necessary. Normally stripped sour water can be used as
desalter wash water, provided its contaminants levels are adequate (NH3 less than 150 ppm and
H2S less than 20 ppm). These limits are required to avoid corrosion in downstream units (i.e.:
CDU overhead system).
Two stage stripping: In the first one, H2S is separated, whereas in the second one NH3 is
removed from water and concentrated to a solution containing 10 % NH3, which can be re-used
to reduce NOx emissions. Exemplary data of this two stage stripping are given in the following
table.
The two stage SWS process can reach an overall H2S and NH3 recovery, of 98 % and 95 %
respectively. Hydrogen sulphide: 0.1 - 1.0 mg/l; Ammonia: 1 - 10 mg/l. Ammonia generated in
the two stage SWS process can be used within the refinery.
Other benefits are that the reduction of ammonia and sulphur content in the waste water by
SWS. Reduction of ammonia entering the waste water system reduce the need for a nitrification
/ denitrification process.
The addition of a sour water tank homogenises the different water streams, it further removes oil
that may cause plugging in the stripper and it helps to produce acid gas of constant composition
for the SRU. Because less hydrocarbon enters to the SRU, less coking of catalyst results.
Cross-media effects
Routing the off-gases from the stripper unit can influence negatively the efficiency and the
running conditions of the SRU mainly due to the NH3 content in the gas
Combination of 2 stage sour water stripping and re-use of NH3-rich effluent for reduction of
NOx from CO-boiler achieves reduction of NOx by 180 t/yr, NH4-N by 250 t/yr and solid waste
from WWTP by 10 % at Holborn refinery.
Operational data
Most commonly SWSs have a surge drum to remove entrained hydrocarbons, that can cause
upsets in the downstream SRU. The SWS feed is normally preheated by feed/effluent
exchangers to a tower inlet temperature of 100°C to save stripping steam. A higher feed
temperature than 100°C is not advisable since flashing of the feed on entry to the stripper tower
is to be avoided.
Decreasing the water content in the off-gas below 30 mol% is not practised as problems with
salt deposition from the vapour phase can occur. This salt deposition particularly occurs when
CO2 is present in the sour gas, and the level of corrosive ammonium bisulphide (NH4HS) in the
condensate refluxed to the stripper column increases beyond the level acceptable from a
materials and corrosion perspective.
The use of a second stripping column consume extra chemicals for pH control (acid, caustic)
and extra energy.
Applicability
Two stage stripping: In case the SWS bottoms are not re-used but sent to biotreating it may still
contain too much NH3. In order to solve this in the SWS unit, the Sour Water Stripper Column
can be either equipped with a larger number of stages or a two stage stripper can be installed
(Section enclosed in dotted box in Figure 4.16). The more or less pure ammonia stream from the
second stripper top might be sent to the CO boiler of the FCCU for deNOxing purposes or to the
hot flue gas of a furnace.
Economics
The cost for new steam strippers depends on the flow rate and range from about EUR
0.525 - 0.700 millions.
The investment costs for expansion of existing sour water stripping in the NH3 stripper were
about EUR 3.3 million. Annual operating costs (steam, energy, cooling water, sulphuric acid,
caustic lye, fuel gas) for the entire plant is approximately EUR 0.45 million. In contrast to the
annual operating cost for operation of only H2S strippers (EUR 0.35 million), the annual
operating cost are approximately 75000 EUR because the expansion of the NH3 stripper.
Other data shows the following costs for different sour water strippers:
Example plants
Two-stage stripping is seldomly applied in refineries. In the Holborn refinery in Germany a two
stage SWS is integrated as an alternative to an effluent denitrification stage in the WWTP. Also
in the new Mider refinery a two-stage SWS unit has been installed.
Reference literature
[19, Irish EPA, 1993], [257, Gilbert, 2000], [118, VROM, 1999], [211, Ecker, 1999], [127,
UN/ECE, 1998], [302, UBA Germany, 2000], [181, HP, 1998], [316, TWG, 2000], [115,
CONCAWE, 1999]
Description
Benzene, phenols and hydrocarbons in general in waste water can often be treated more easily
and effectively at the point it is generated rather than at the waste water treatment plant after it is
mixed with other waste water. Consequently the identification of hydrocarbon sources is the
first measure to consider.
1. Nitrogen or air stripping for benzene recovery from waste water. Nitrogen stripping can be
used to strip benzene and other low aromatic compounds from waste water. The mixture is
treated by activated carbon beds that capture the organic allowing cleaned nitrogen to
recycle to the waste water stripper. Peridiocally, the carbon bed is insitu regenerated with
live steam: desorbed organic vapours are carried by the steam to a condenser and
subsequently decanted into organic and water layers. Organics are returned to refinery as
valuable feed
2. Liquid-Liquid extraction from waste waters for phenol extraction from waste water
3. Reduction of hydrocarbons and aromatic compounds
4. High-pressure wet air oxidation. Water is intensively mixed with air and organic
compounds are oxidised in the presence of a catalyst at high temperature and high pressure
(250 ºC, 7 MPa). Sulphur containing substances are oxidised to sulphates; amines and
nitriles are converted to molecular nitrogen
5. Low pressure oxidation. Persistent organic compounds are treated with oxygen and are
mineralised in a biological treatment plant. (BOC Gases)
6. Supercritical water oxidation process. Supercritical water (373 ºC, 221 bar) is used to
dissolve organic compounds, which are oxidised in a reactor by injection of oxygen
7. Energy absorption technology. Sources of contamination with hydrocarbons are desalters
(40 %), storage tanks (20 %), slop-systems (15 %) and other processes (25 %). It is possible
to recognise type and extent of water contamination directly at the source by use of
electromagnetically high frequencies (EA – Energy Absorption Technology).
Operational data
1. Nitrogen stripping has several advantages over air. No oxygen reduces biological fouling of
the stripper and reduces the risk of refinery upsets creating explosive mixtures in the recovery
unit.
2. Utilities: typical per m3 water feed
Electricity 159 kWh
Steam, 20.7 barg, 15.6 kg
Steam, 2.07 barg, 103 kg
Water tempered at 45 ºC, (∆T = 19 ºC) 5.6 m3
cooling at 29 ºC, (∆T = 11 ºC) 2.5 m3
Applicability
1. This technology is used for the treatment of desalting water and waste water from BTEX-
plants (Texaco Development Co; AMCEC, Inc.).
2. They can be designed to handle waste water streams with phenol levels from several hundred
ppm to saturation (approx. 7 %) and above.
3. This method is not applied for high flows.
Economics
1. Economics: The design and equipment supply cost of about $1250000. Annual utility costs
are around $85000.
2. Economics: Cost effective for levels of phenols higher than 1 %. A base case example for a
waste water stream of 27.2 m3/h containing 6 % phenol was treated with 4.3 m3/h solvent in a
four-column system. The overall phenol recovery was 99.3 %.
Investment, extractor only, $1.32/m3
complete system, $3.43/m3
Recovered value $3.96/m3
3. Method very expensive
Example plants
1. More than 15 systems ranging from 800 to 12000 l/min are now operating in various USA
refineries. Air stripping for the removal of MTBE has been used succesfully at least in one
European refinery.
2. This system is typically applied when phenol concentrations are high.
Reference literature
[181, HP, 1998], [211, Ecker, 1999], [316, TWG, 2000], [321, Helm, Spencer et al., 1998]
Process water coming from the SWS is the main process water source in refineries. Moreover,
some incompatible process effluents and drains from off-sites (flares and tanks) add to the total
flow of process water. This effluent has to pass first an oil/water separator (CPI, PPI or API) for
free oil and solids removal and an equalisation tank where sometimes further oil skimming can
be accomplished.
Description
Emissions to air of VOC (including benzene) and odorous components (hydrogen sulphides and
mercaptans), however, are not uncommon and cannot always be controlled fully by up-stream
measures. Therefore APIs, CPIs can be provided with covers as an emission reduction measure,
sometimes with off-gas treatment (biofilter or re-injected into the aeration basin). Safety aspects
(explosiveness of VOC-air mixtures) are to be attended.
The next level of control is to install water seals (traps) on sewers and drains and gas tight
covers on junction boxes in the system. The use of covers on oil/water separators with good oil
removal facilities will prevent or reduce evaporation of liquid hydrocarbons from exposed
surfaces. Alternatively, incineration of the vapours coming from the API could be achieved
from a covered API separator.
The HC emissions from the waste water systems can be determined by calculation from the
exposed surface area of the oil-contaminated untreated water tank (API separator) and an
empirical oil evaporation factor of [117, VDI, 2000]:
• 20 g/m2 per hour for open oil separator
• 2 g/m2 per hour for covered oil separator
Cross-media effects
When API or PPI separators are covered the flammability/explosion limits may easily be
reached. Consequently safety and environmental protection should be considered.
Operational data
Covers present problems with skimming of free oil.
Applicability
Those systems are fully applicable. CPI are easier to cover than API.
Economics
Costs are estimated with about 460 EUR per kt of HC emissions reduction by provision of
floating covers for waste water separator bays. The investment for the installation of two 200 m3
floating slab-covers is about EUR 75000 and the total annual operating costs about EUR 42800
per year.
Example plants
Well proven technologies. In some European refineries, the API and PPI separators have been
covered.
Reference literature
[107, Janson, 1999], [115, CONCAWE, 1999], [258, Manduzio, 2000], [127, UN/ECE, 1998],
[247, UBA Austria, 1998]
The next step after the primary treatments waste water is processed for the removal of dispersed
oil and solids by air flotation where with the aid of polyelectrolytes flocs are formed which
capture oil and suspended solids (sometimes a sandfilter is used for this duty). The thus formed
sludge is brought to the water surface by air flotation whereby tiny air bubbles are also captured
by the sludge flocs. The sludge is skimmed off and the water is routed to a biotreater.
Description
Emissions to air of VOC (including benzene) and odorous components (hydrogen sulphides and
mercaptans), however, are not uncommon and cannot always be controlled fully by up-stream
measures. Therefore DAF units can be provided with covers as an emission reduction measure,
sometimes with off-gas treatment (biofilter or re-injected into the aeration basin). Safety aspects
(explosiveness of VOC-air mixtures) are to be attended.
Operational data
Chemicals utilisation includes acid and/or alkali for pH correction, polyelectrolyte or FeSO4 or
FeCl2 flocculant requirements of the floculation units. In practice pH correction and polymer
dosing to the flocculation flotation unit needs daily attention and fine tuning if needed. The
energy requirement for the WWTP is relatively low, the aeration compressor being the largest
consumer.
Applicability
Fully applicable
Economics
Year built Type Design flow rate Capital cost Operating cost/yr
(m³/h) (EUR x 106) (EUR x 103)
1995 DAF 80 0.2 18
1994 DAF 300 1.4 20
1989 IAF 400 2.4 47
1993 DAF 350 8.0 683
1996 IAF 818 0.4 112
1996 DAF 50 3.1 –
1996 DAF 800 1.5 –
Example plants
Well proven technologies
Reference literature
[113, Noyes, 1993], [115, CONCAWE, 1999]
Description
After the sludge is skimmed off in the flotation process, the water is routed to a biotreater,
usually an activated sludge unit or to a trickling filter. By means of bacteria nearly all dissolved
hydrocarbons and other organics are removed. In case deep nitrogen removal is required,
denitrification in a biotreater is an option. Oxygen using (aerobic) bacteria is able to convert
ammonia into nitrate and in a so-called anoxic stage (without aeration) in the biotreater other
bacteria are able to convert nitrate into atmospheric nitrogen, which escape as bubbles. The
generated biomass or biosludge is allowed to settle in a clarifier and for the main part recycled
to the biotreater. Primary sludge from the flocculation-unit and excess biosludge are usually
sent to a thickener prior to dewatering and disposal (Figure 4.17). Air Flocculation Flotation
processes can be operated with Dissolved or Induced Air (DAF or IAF resp.).
Cross-media effects
Energy consumption, activated carbon wastes when GAC or PAC are appplied and methanol
consumption when denitrification is applied. Waste water treatment generate sludge. If waste
water treatment is not properly operated the amount of sludge can be increased.
Operational data
If a denitrification biotreater is applied pumping energy is relatively high. If a tertiary treatment
unit is used for ammonia removal by biological nitrification, methanol can be used as in a
subsequent denitrification stage as an oxygen receptor for denitrifiying bacteria. In some
refinery biotreaters powdered activated carbon is dosed for meeting discharge specs. The
sensitivity of a biotreater against shock loads (peak discharges) of toxic compounds (e.g.
sulfolane, MTBE, phenol, cyanide, sulphide) is relatively large and should be avoided by spill
prevention measures, good housekeeping and equalisation in a buffer tank.
Applicability
Fully applicable. Toxic streams should not enter a biotreater.
Economics
A complete API, equalisation tank, DAF, DNB biotreater, clarifier train for 125 m3/h requires
an investment cost of about EUR 15 million provided the up-stream drainage infrastructure is
available. Operating costs are in the order of EUR 1.5/m3.
Floating cover: Capacity of the process 800 m3/h of oily water may emit up to 1000 t/yr
Efficiency: 90 % of recovery of the VOC
Investment cost EUR 0.6 million
Operating cost: EUR 0.03 million per year
Example plants
Well proven technologies. Nitrification/denitrification biotreater installations are operational at
Harburg, Godorf and Gothenburg refineries.
Reference literature
[181, HP, 1998], [45, Sema and Sofres, 1991]
Description
Techniques for the reduction of salt contents of the waste water: Ion exchange, Membrane
processes or Osmosis.
Metals may be separated by Precipitation, Flotation, Extraction, Ion exchange or Vacuum
distillation.
Cross-media effects
Energy consumption, wastes from exhausted activated carbon, membranes and metal sludges.
Operational data
Activated carbon is needed if AC filtration appears in the system.
Economics
When the WWTP is extended with SF and AC, the operating costs of the resulting WWTP is
doubled in reference with a WWTP without those systems (base case). With UF and RO both
the investment costs and the operating costs will approximately triple the base case.
Example plants
Sandfilters, ultrafiltration, activated carbon and reverse osmosis are proven technologies.
Reference literature
[181, HP, 1998].
This section contains what emission values can be achieved by an overall waste water treatment
of refineries. The waste water system considered contains primary, secondary and tertiary
treatment.
VOC
VOC DAF-unit VOC
waste water
from API Equalisation
refinery tank
processes
blower
API sludge primary sludge
recovered oil
VOC air
clarifier
aeration
Aeration blower Water
Aeration Aeration
basin 1 basin 3 discharge
basin 2
or anoxic zone
for DNB
configuration 80% effluent recycle in
case of DNB
excess
sludge recycle sludge
optional post-
denitrification
unit recycle for
reuse
sludge brine
Figure 4.17: Simplified process flow diagram of a typical refinery waste water treatment plant,
including denitrification/nitrification biotreater.
Achievable levels
The following ranges of release levels are considered to be achievable for key water parameters
typically used to characterise a refinery release to water using a good combination of techniques
described in this section.
Table 4.50: Emission concentration and loads found in good operated WWTP
Effluent treatment units are designed such that the most important key parameters (suspended
solids, TOC, COD and BOD) can be monitored automatically leading to limited operator
attention and even less if sufficient up-stream water management measures are being taken to
ensure a constant quality and flow of water and adequate buffering. TOC and COD can be
monitored continuously whereas BOD need several days for the analysis.
Applicability
WWTP usually occupy a significant space at refineries, particularly the biotreater due to the
relative slow biodegradation process. In order to save space it is advisable to integrate water
management principles into the design leading to a compact. A full train WWTP occupies a
space of 1 ha, not including an observation basin which many refineries have installed as a last
line of defense. The amount of rainwater depends of course on the local climatic conditions and
the size and layout of the refinery.
Residual refinery wastes are typically in the form of sludges, spent process catalysts, filter clay,
and incinerator ash. Other waste fractions are waste from flue gas desulphurisation, fly ash,
bottom ash, spent activated carbon, filter dust, inorganic salts as ammonium sulphate and lime
from water pretreatment, oil-contaminated soil, bitumen, sweepings, spent acid and caustic
solutions, chemicals, among others. Treatment of these wastes includes incineration, land
treating off-site, land filling onsite, land filling off-site, chemical fixation, neutralization, and
other treatment methods.
Description
The establishment of an Environmental Management System (Section 4.15.1) should contain
elements to prevent waste generation and some pollution prevention techniques that help to
prevent pollution of soil and groundwater. Some techniques that may be included are:
• implementation of a sludge master plan with the aim to reduce the sludge generated.
• provide enclosed sampling loops
• clean and assembly only in especially constructed and dedicated areas
• provide dedicated drainage systems
• physical barriers such as clay walls or plastic membranes can be installed around the site
boundary. To be effective at containing oil pollution, these have to extend below the water
table. Monitoring wells will also be required to ensure that if oil does build up behind the
barrier, it can be removed before it has a chance to escape under the barrier. A second form
of barrier is a ditch also extending to below the groundwater table. Any oil leaving the site
can be observed floating on the water in the ditch and recovered. The outer face of the ditch
can be sealed with an impervious layer such as concrete, plastic, clay, steel piling etc
• groundwater, like surface water basically flows downhill. Therefore, groundwater can be
prevented from leaving a site by lowering the level inside the site by pumping so that it is
lower than the level outside. Water will then flow into the site rather than outwards. The
pumped water has obviously to be disposed of. It may be possible to use this water for
supply purposes on site. If it is to be discharged, then monitoring will be required to ensure
that it is not contaminated. If it is, then it will have to be treated in which case, this method
will be limited by the capacity of the effluent treatment system
• minimisation of underground piping. Underground piping can be a source of undetected
releases to the soil and groundwater. Inspecting, repairing or replacing underground piping
with surface piping can reduce or eliminate these potential sources. Retrofitting can be very
expensive
• perform a risk analysis to rank in order of significance cases where accidental leak may
occur (elements to consider are the product in the tanks/pipes, the age of the equipment, the
nature of the soil and groundwater that would be affected). Prioritise areas where
impermeable floors are needed most. Produce a multiyear master plan to program necessary
steps
• periodically screening of leaks in sewers and pipes.
• correct conditioning of a catalyst during a run extends catalytic life
Reference literature
[195, The world refining association, 1999], [316, TWG, 2000], [115, CONCAWE, 1999]
Description
Sludges are defined as emulsions of oil in water, stabilised by the presence of solids. In
refineries a number of different types of sludges are generated at the following sources: crude
and product tanks (bottoms), API separation units, flocculation and flotation units, DAF,
contaminated soil. Biological sludges represent an important different category of sludges in
terms of oil content and dewaterability. According to CONCAWE, in 1993, 44 % of the
European refinery sludges were incinerated, 9 % was land farmed and 30 % land filled. It is
foreseen that land filling and land farming of sludges will be increasingly prohibited by
forthcoming EU legislation, which means that the scope for sludge prevention reduction but
also for incineration at third parties will increase in the future.
The purpose of sludge treatment by dewatering, drying and/or incineration is to reduce the
volume and the residual hydrocarbon content in order to save costs in subsequent processing or
disposal. The principle of mechanical dewatering by decanters is based on centrifugal forces and
on the density difference between water, oil and solids. The principle of the thermal treatment
steps is based on a combination of evaporation by indirect heating and/or destruction of the
organic constituents by thermal oxidation (incineration).
Decanter centrifuges are the most widely used for dewatering both oily and for biosludges.
Steam dryers are almost exclusively applied for biosludge and function often as a pretreatment
step for incineration. Landfarming of sludge cakes is still being practised but is increasingly
restricted in view of emissions and soil contamination risks.
Decanter centrifuges are widely used in sludge dewatering and deoiling applications throughout
the (refining) industry, either as fixed facilities or as a mobile service rendered by contractors.
Dewatered biological and oily sludges can be further processed using drying and/or incineration
techniques resulting into virtually oil free residues for which useful applications are available.
Oily sludge dewatering is only applied at refineries that dispose of sludge cakes outside their
premises in order to reduce the volume and associated disposal cost. Disposal to cement kilns,
coal fired power plants, dedicated sludge incinerators, municipal and hazardous waste
incinerators are all practised. Drying is currently hardly employed by refineries due to safety
risks. de-oiling/dewatering of sludges gives small volume of solid, low solvent waste,
(centrifuging or filtration). De-oiling/dewatering of sludges gives small volume of solid, low
solvent waste, (centrifuging or filtration).
In Figure 4.18 a simplified process flow diagram is shown of a decanter combined with a
fluidised bed incinerator (FBI) being the most appropriate technique for sludge incineration.
The feeding of sludge cake into the sand bed of the FBI can be accomplished with a positive
displacement pump. The oxygen for incineration is provided by air injection into the bed via a
wind box and a perforated floor. An air ventilator enables the sand in the bed to fluidise. The
burned out ash particulates leave the bed and are removed by flue gas treatment. This can
consist of a waste heat boiler (WHB), an electrostatic precipitator (ESP), a scrubber section, a
flue gas heater (to avoid a visible steam plume), a flue gas ventilator and a stack. The ash is
transported by a chain/bucket conveyor combination or pneumatically to a hopper, and it
usually requires dedusting by water spray and agglomeration in a double paddle mixer. At
smaller incinerators a water quench is used to cool the flue gas from 850°C to about 150 °C
prior to fabric (bag) filtration.
Decantate
back to
decanters biotreater
oily effluent to
oil/water separator
MP-steam
Flue gas
water injection ash
heat exchanger
hopper
flue gas
from ESP stack
blower water
effluent to direct
discharge or to
treatment/recycle ash to disposal
Figure 4.18: Simplified process flow diagram of sludge handling and incineration
Operating temperatures of the bed can be controlled at respectively 800 – 850 °C and 850 –
950 °C. The WHB operates between 250 and 450 °C, the ESP at 220 °C and the scrubber
between 200 and 50 °C. In a FBI all ash is carried over the top with the flue gas, which provides
a large dust loading (typically 50000 mg/Nm3) to the WHB and ESP. Some 50 % of the dust is
collected in the WHB, which can be regarded as a settling chamber. The catchment efficiency of
an ESP is typically 99 %, leaving the (venturi) scrubber system to eliminate ± 500 mg/Nm3 to
the stringent CEC limit of 5 mg/Nm3. The ash might find a useful outlet in concrete or asphalt
dependent on the quality.
It should be mentioned that the emissions of SO2, NOx, CO, organic compound, PAH, heavy
metals generated by those systems should be properly controlled by suitable abatement
techniques.
Cross-media effects
Scrubbing might require caustic if acidic (SO2 and NOx) components are present in the flue gas
in concentrations which need to be removed to legal limits. The FBI with flue gas treatment and
ash handling is a closed system, with fully automated control. The system is kept under slight
negative pressure. The described system is regarded state-of-the-art for sludge incineration and
is able to fulfil all legal stack emission requirements. Dioxins (PCDD/F) could be an issue at
high chlorine loading combined with sub-optimum operation conditions. In flue gas treatment
systems of domestic sludge incinerators impregnated activated carbon filters (ACF) or zeolites
are installed for mercury/dioxin control. Effluent from the scrubber system can be substantial.
Recirculation of scrubber water is only a cost effective proposition if treatment is required prior
to discharge.
Operational data
Safety issues are associated with start-up and (planned or emergency) shut-down, for which a
flame control interlock and a nitrogen purging system is required. Anaerobic and pyrophoric
sludges (tank bottoms) can cause safety problems during storage and drying. The energy
requirements of sludge drying and incineration depend strongly on the sludge cake
characteristics (water and residual oil contents). Energy integration with FBI is two fold: air
preheating of the primary fluidisation air is standard, and steam generation can be economically
accomplished with systems able to generate >8 tonnes of MP steam/hour. An intrinsic
disadvantage of FBI is the relatively large excess of air utilised in order to keep the bed
fluidised.
Applicability
Because its high cost, options as reduction/recycling of waste are more typical waste
minimisation options to be applied before waste incineration. Fluidised bed incinerators may be
an option for third-party waste industry receiving waste streams.
Economics
For a FBI with a capacity of 4 t/h sludge cake (20 % dry solids) a footprint of about 50x100 m is
required including tankage and complete flue gas and ash handling. The installation is normally
12 - 15 m high (tanks, incinerator, WHB, ESP, ash hopper) and the stack height is usually at
least 40 m, dependent on ancillary installations. The above-described system would be suitable
for a large 20 Mt/yr refinery and requires a capital investment of about EUR 37.5 million
(including installation). The operational costs could amount to EUR 500 - 700/tonne dry solids.
If large scale domestic sludge incinerators are available, which are entitled to take industrial
sludges, the installation of a dedicated sludge incineration system is not economically
justifiable. This certainly would be the case for smaller refineries, indicating that contract
dewatering, drying and incineration are competitive to own processing.
A combination of decanters with an advanced drying system for oily sludge occupies only
10 - 15 % of the above indicated space, and would involve a capital investment of EUR 5
million due to the fact that expensive flue gas treatment is avoided.
Example plants
Decanter centrifuges are reliable, advanced and proven technology with minimum emissions. At
refinery locations permanent decanter centrifuges have for example been installed at refineries
in Godorf, Gothenburg and Stanlow. A few times per year contractors are hired in either for on-
site sludge processing or for collection and off-site processing (decanting, drying, incineration
in cement kilns, power plants or in industrial/domestic waste or dedicated sludge incinerators).
Sludge incineration plants using the fluidised bed system are state-of-the-art but require
advanced design and process control. At several refineries Mobil, Wilhelmshafen, Shell Pernis
and Godorf, Esso Botlek these systems were already built in the 1970s. Some of them have been
dismantled due to the current availability of more economic options as compared with necessary
investment and operation in additional flue gas cleaning equipment. For oily sludges mixing
with other wastes and use as secondary fuel in cement kilns and/or power plants represents an
attractive disposal route if those plants use appropriate abatement techniques. International
sludge contractors employ either mobile decanter and drying systems for oil recovery (Impex,
EPMS, Soil Recovery A/S) or fixed oily sludge treatment systems (ATM).
A BREF on Non-Ferrous metals has been already published. In that BREF the techniques used
for the recovery of metals from secondary materials (e.g. refinery catalysts) is extensivelly
treated.
The use of catalytic processes in refineries has increased significantly in the last 20 years. This
increase is mainly due to the introduction of catalytic residue conversion processes such as
heavy oil residue cracking, hydrocracking and residue hydroconversion, hydrodemetallisation
and hydrofinishing and also Hydrogen production. Since 1980 a significant expansion has taken
place of the hydrotreating and hydrodesulphurisation capacity as well as of sulphur recovery
units and associated tail gas treatment, which processes also utilise catalysts. Traditional
catalytic processes such as fluid catalytic cracking, catalytic reforming and isomerisation are
also generators of spent catalysts.
Description
The purpose of spent catalyst management is to minimise environmental and health impacts. To
achieve this goal spent catalysts are carefully handled, safely removed, carefully packed and
sent for reactivation or metal reclamation. Metal recovery aims at conversion of spent catalysts
into useful products, which can be recycled and re-used with minimum impact on the
environment.
The principle of spent catalyst management is a scheduled, strictly regulated and safe handling
of the materials involved, usually executed by specialised contractors during plant turn-arounds.
Arrangements with the supplier of fresh catalyst can sometimes be made for them to take back
the spent catalyst.
process is hardly used due to the availability of cheaper alternatives. The catalyst carrier
(alumina and/or silica) can sometimes be converted into products or is disposed of otherwise.
Some refineries practise storage of spent catalysts at the refinery site with permission of the
authorities. Spent catalysts are distinguished by type, process, composition and recyclability. A
summary is given in Table 3.78.
Ni/Mo catalysts are typically used in hydrotreaters and hydrocracking units. Regeneration and
reclamation capacity is available.
Ni/W catalysts are used in lube oil hydrofinishing. This category experiences limitations to
dispose off in view of the high Tungsten content (24 % w/w).
FCC spent catalysts, also including heavy oil and residue cracking spent catalysts (RCC), are
the largest catalyst waste category in refineries. Outlets in road building are becoming available.
Reformer and Isomerisation catalysts are exclusively reprocessed by the suppliers of the fresh
catalysts. Replacement contracts have been concluded since the introduction of these processes
due to the very expensive noble Pt metal involved.
Hydrodemetallisation catalysts typically have high vanadium contents (10 - 20 %) and are
currently alumina based (used to be silica). Direct disposal to the steel industry is possibly the
most cost-effective option.
Zn containing beds from H2 plants are typically recycled to the Zn-industry where ZnS ores are
processed. Amounts are some 50 t/yr.
The reclamation plants (more information in Non-Ferrous BREF) are typically operated in batch
mode and contain many different unit operations. From the total spent catalysts only some 5 %
pure products are produced. The remainder is either ferro-alloy feed or ceramic raw material.
The plant usually processes relatively small batches. Spent hydroprocessing catalysts can vary
considerably in composition. Hence the plants have limited process automation.
Cross-media effects
The management of spent catalysts is an area of environmental concern because prolonged
uncontrolled storage may lead to soil and groundwater contamination with heavy metals.
If the refinery has taken proper catalyst management measures the environmental concerns
focus usually on the spent catalyst regeneration facilities. Most of these plants currently operate
with proper flue gas and effluent treatment facilities and comply with present day emission
standards. This is an important requirement for being acceptable for oil companies to make use
of these spent catalyst reclamation contractors. Covering spent catalyst reception facilities is
operational or being considered.
Operational data
Safe handling of chemicals is an important issue for spent catalyst metal reclaimers as toxic and
pyrophoric materials are being handled. For the refinery utilisation of energy and process
materials during spent catalyst handling are not particularly relevant.
Economics
Cost for processing and metal recovery depend strongly on the composition of the spent catalyst
concerned. For high metal waste even a recovery fee will be paid to the waste generator. Current
processing costs for an average HDS catalyst could be in the order of EUR 500/tonne. Recent
stringent internationally accepted legal procedures, amongst which special packaging (for rent
2 m3 containers at EUR 5/day), labelling and acceptance requirements as well as transportation
costs could add considerably to these costs. On-site processing of spent catalysts is not
economic for a refinery.
Spent FCC catalysts with limited V/Ni concentrations are accepted with consent of authorities,
by raw material suppliers to the road building industry. The material is sometimes used in
cement and as asphalt filler.
Example plants
The regeneration of spent catalysts started in the early 1980s. All processors use pyro- and
hydrometallurgical processes with varying degrees of sophistication, recovery capabilities and
environmental performance. The pyro- and hydrometallurgical processes currently applied are
considered acceptable and proven technology.
Reference literature
[118, VROM, 1999; 122, REPSOL, 2001]
Heavy residues generated by refineries are the heaviest fractions from the different units
(distillation, conversion) and that have not application for products and are typically re-used
within the refinery. Those residues have a calorific value that can be exploited. Follow a list of
treatments that may be applied to reduce the amount of those residues. All these techniques are
already analysed in other sections but they are put together here to help the reader to understand
what refineries may do with their heavy residues.
• Catalytic cracking (included within Section 4.5) as Reduced crude cracking (RCC),
Heavy oil Cracking, VEBA combi cracking (VCC), Deep catalytic cracking (DCC).
• Non catalytic cracking (included within Section 4.7 and 4.22) as Delayed Coking,
Fluid Coking, Flexicoking, LR-Coking, Deasphalting: DEMEX, Rose-Technology,
Visbreaking, Thermal Cracking, Partial Oxidation.
Because oily sludges make up a large portion of refinery solid wastes, any improvement in the
recovery of oil from the sludges can significantly reduce the volume of waste. There are a
number of technologies currently in use to mechanically separate oil, water and solids,
including: belt filter presses, recessed chamber pressure filters, rotary vacuum filters, scroll
centrifuges, disc centrifuges, shakers, thermal dryers and centrifuge-dryer combinations.
Clay from refinery filters must periodically be replaced. Spent clay often contains significant
amounts of entrained hydrocarbons and, therefore, must be designated as hazardous waste.
Techniques that may be applied are:
• Back washing spent clay with water or steam can reduce the hydrocarbon content to levels
so that it can be re-used or handled as a nonhazardous waste.
• Another method used to regenerate clay is to wash the clay with naphtha, dry it by steam
heating and then feed it to a burning kiln for regeneration.
• In some cases clay filtration can be replaced entirely with hydrotreating (See
Section 4.20.4).
Streams from water treatment units (e.g. oil/water mixtures from interceptors) can be routed to
the wet slops tank. Oil from sludge thickeners (centrifuges/decanters) can be routed to the slops
tanks. In that way also the oil in sludges from DAF units can be recovered [259, Dekkers,
2000].
Some options to reduce the waste generation is the recycle or re-use of them. Follows some
examples that can be considered as good environmental practices:
• some catalysts for metal recovery (reforming, desulphurisation) (See Section 4.25.3)
• waste lubes: re-refining (See Section 4.25.4.6)
• drums/containers: reconditioning
• spent caustic can be used: (See Section 4.20.2)
• alkylation process: CaF2
§ for HF production
§ as a fluxing agent (steel industry)
• resale of polymerisation plant catalyst as a fertiliser (4.18.2)
• sale of gypsum or sulphuric acid from flue gas desulphurisation units
• dust: In the regenerative flue gas desulphurisation process according to Welmann Lord NH3
is injected to prevent the formation of SO3. Therefore up to 80 % of the ash consists of
(NH4)2SO4, which can be used as a fertiliser or as a basic material for NH3 production [250,
Winter, 2000].
• paper, wood, glass, scrap metal
• construction/demolition debris:
§ concrete to a crusher, for use in road building etc.
§ asphalt scrapings re-used, e.g. in road building
It has to be mentioned that re-use of waste fractions/residues outside the company is only an
option, if these waste fractions meet certain criteria (such as concentration of pollutants in a
fertiliser) and do not alter the original product characteristics (e.g. when using gypsum in the
cement industry).
Description
Waste lubes can be re-used within the refinery as fuel component or as feedstock for re-refining.
Controlled incineration of waste lube oils is another alternative.
Cross-media effects
Waste lubricating oils are generally undefined and may contain all kind of additives and
contaminants.(even PCB’s). Inclusion of contaminated waste lubes into commercial fuel oil
would expose the customer to high risks. Inclusion of this waste in refinery fuel poses a safety
risk.
Applicability
Some applicability problems have been detected. For example, waste lubricating oil from
garages may contain organic chlorides used as degreasers; in hydrotreaters they will be
converted into HCl and cause severe corrosion in this high pressure unit filled with hydrogen.
Only by application of well-controlled pretreatment techniques, the treated waste oil can be
upgraded to a fuel component. Such a pretreatment activity is normally not compatible with
refinery operations and is done outside the refinery by specialized companies, which also do the
collection of the waste oils. The only exception to the use of waste lubes as fuel component is
for waste lube oil that is generated in the refinery itself and of which its composition is beyond
any doubt.
Reference literature
[259, Dekkers, 2000].
This section contains methods for the biodegradation of refinery wastes that may be used
specifically within refineries on-site. This section does not contain methods for the remediation
of contaminated soils. As it was stated in the scope of this document, soil remediation
techniques are considered out of the scope of this document.
Description
Many hazardous chemicals present in refinery waste can be converted by microbiological
methods to harmless compounds such as water and carbon dioxide. In general, the
microbiological degradation of contaminants in soil is very slow in nature, because process
conditions for such degradation are seldom favourable. To accelerate degradation a number of
conditions have to be fulfilled.
Current techniques for biological decontamination are based on optimisation of the process
conditions for microbiological degradation. The appropriate micro-organisms for
microbiological degradation may already be present in the waste to be treated or may have to be
added. The latter is necessary if special micro-organisms are required. These special micro-
organisms can be obtained by selection and adaptation.
Operational data
The most important factors are the availability of temperature control, sufficient oxygen,
nutrients, and the appropriate micro-organisms. Also important are the concentration level of the
contaminants and the variation in concentration. The presence of toxic compounds can disturb
the degradation process. Sometimes the presence of natural organic compounds has a positive
influence on the biodegradation process.
In summary the following conditions have to be met in order to optimise the degradation rate of
a waste generated by a refinery:
Reference literature
[115, CONCAWE, 1999]
Description
The following processes are commonly used:
Solidification – A process in which materials are added to the waste to produce a solid. It may
involve a solidifying agent that physically surrounds the contaminant (i.e. cement or lime), or it
may utilise a chemical fixation process (i.e. sorbents). The resulting waste is usually an easily
handled solid with low leachability. Several processes exists:
• Cement-based processes. In this process the slurried waste is mixed with cement and during
the hardening process is incorporated in the rigid concrete matrices. In the same line, spent
catalytic cracker unit (FCCU) catalyst may be used as an additive in cement manufacturing.
When the cement is used, the catalyst component forms insoluble hydrates with the chalk
present in the cement mixture, which also gives beneficial fixation of heavy metals.
• Thermoplastic techniques. In general the use of thermoplastic solidification techniques is
restricted to dry solid materials.
• Mixing with bitumen. The processing of waste with bitumen is applied in the oil
manufacturing industry as a disposal method. This process allows treatment of soils with
high levels (up to 10 %) of high boiling range hydrocarbons.
Stabilisation – The conversion of a waste to a chemically stable form that resists leaching. This
may be accomplished by a pH adjustment. Stabilisation also generally results in a solidification
of some sort (monolith or dry granular solid).
• Chemical Stabilisation. These processes are based on the reaction of lime with waste
materials and water to form a chemically stable product. When compacted, the porosity to
water is very low. This reduces the risk of leaching.
The micro-encapsulation techniques are based on the reduction to surface-to-volume ratio of the
waste by formation of monolithic, hard mass with a very low permeability.
Macro-encapsulation is the enclosing of a relatively large quantity of waste, such as an entire
waste container. Wastes are macro-encapsulated by surrounding them with a stiff, weight-
supporting matrix, and a seam-free jacket.
Applicability
Cement-based processes: The process is especially effective when the waste contains metals
because at the high pH of the cement mixture most metal compounds are converted into
insoluble metal hydroxides. In the case of spent catalyst, most metal compounds are present as
hydroxides which as such may also increase the strength and stability of the waste containing
concrete. On the other hand, the presence of organic impurities may act as interfering agents to
the curing of the concrete and this limits the application of this disposal route.
Mixing with bitumen. It is applied as a disposal method for spent FCCU catalyst and used for
the production of a filler component in road asphalt where the catalyst is present as a minor
component. The catalyst particles are completely encapsulated which makes aqueous extraction
unlikely. This process allows treatment of soils with high levels (up to 10 %) of high boiling
range hydrocarbons. The soil is mixed with bitumen to produce a stable end-product suitable for
use in road construction.
Encapsulation is suitable for on-site treatment of disposal sites of accumulated spent acid tars
and oily sludges which are difficult to transport and to dispose of by other means. A
disadvantage is that the treated product occupies a larger volume than the original sludge.
Because it can be applied on-site, the encapsulation process may be considered for single
applications like rehabilitating refinery sites after decommissioning or cleaning up an oil-
polluted site after a spill. The decision to apply the process depends on the future use of the site
and local legislation. The process is less attractive for the treatment of regularly produced
sludges, because of the increased mass generated for disposal.
Reference literature
[115, CONCAWE, 1999]
• identification of the key environmental issues for the Mineral Oil refinery and Natural
Gas Plant sectors
• examination of the techniques most relevant to address those key issues
• identification of the best environmental performance levels, on the basis of the available
data in the European Union and world-wide
• examination of the conditions under which these performance levels were achieved;
such as costs, cross-media effects, main driving forces involved in implementation of
these techniques
• selection of the best available techniques (BAT) and the associated emission and/or
consumption levels for this sector in a general sense all according to Article 2(11) and
Annex IV of the Directive.
Expert judgement by the European IPPC Bureau and the relevant Technical Working Group
(TWG) has played a key role in each of these steps and in the way in which the information is
presented here.
On the basis of this assessment, techniques, and as far as possible emission and consumption
levels associated with the use of BAT, are presented in this chapter that are considered to be
appropriate to the sector as a whole and in many cases reflect current performance of some
installations within the sector. Where emission or consumption levels “associated with best
available techniques” are presented, this is to be understood as meaning that those levels
represent the environmental performance that could be anticipated as a result of the application,
in this sector, of the techniques described, bearing in mind the balance of costs and advantages
inherent within the definition of BAT. However, they are neither emission nor consumption
limit values and should not be understood as such. In some cases it may be technically possible
to achieve better emission or consumption levels but due to the costs involved or cross-media
considerations, they are not considered to be appropriate as BAT for the sector as a whole.
However, such levels may be considered to be justified in more specific cases where there are
special driving forces.
The emission and consumption levels associated with the use of BAT have to be seen together
with any specified reference conditions (e.g. averaging periods).
The concept of “levels associated with BAT” described above is to be distinguished from the
term “achievable level” used elsewhere in this document. Where a level is described as
“achievable” using a particular technique or combination of techniques, this should be
understood to mean that the level may be expected to be achieved over a substantial period of
time in a well maintained and operated installation or process using those techniques.
Where available, data concerning costs have been given together with the description of the
techniques presented in the previous chapters. These give a rough indication about the
magnitude of costs involved. However, the actual cost of applying a technique will depend
strongly on the specific situation regarding, for example, taxes, fees, and the technical
characteristics of the installation concerned. It is not possible to evaluate such site-specific
factors fully in this document. In the absence of data concerning costs, conclusions on
economic viability of techniques are drawn from observations on existing installations.
It is intended that the general BAT in this chapter are a reference point against which to judge
the current performance of an existing installation or to judge a proposal for a new installation.
Mineral Oil and Gas Refineries 391
Chapter 5 Best available techniques
In this way they will assist in the determination of appropriate "BAT-based" conditions for the
installation or in the establishment of general binding rules under Article 9(8). It is foreseen that
new installations can be designed to perform at or even better than the general BAT levels
presented here. It is also considered that existing installations could move towards the general
BAT levels or do better, subject to the technical and economic applicability of the techniques in
each case.
While the BREFs do not set legally binding standards, they are meant to give information for
the guidance of industry, Member States and the public on achievable emission and
consumption levels when using specified techniques. The appropriate limit values for any
specific case will need to be determined taking into account the objectives of the IPPC Directive
and the local considerations.
In order to complement what has been said above, the approach followed in this document for
the BAT assessment can be summarised in the following steps:
As mentioned in the above paragraphs and in the Preface, Chapter 5 contains, where possible,
the range of emission or performance levels associated with the use of a technique or group of
techniques considered to be BAT. All emission levels associated with BAT given in this chapter
are daily averages unless stated otherwise. Gaseous concentrations are corrected to 3% O2 and
in dry conditions unless stated otherwise. Where BAT is currently represented by only one
technique, the emission or performance level cannot be defined as a single number because of
differences in practical situations of feedstock, degree of integration, process variations or other
circumstances. Where BAT is a group or combination of techniques, each single technique
within that group has its own range of associated levels. However, the ranges appropriate for
each technique are combined in a single range, typically wider than those related to a single
technique. In some cases, some of the extreme values of the range have only been reported to be
reached by a combination of at least two techniques.
The ranges given contain the lowest level of emissions found in Chapter 4, unless the lowest
level has been achieved in special local circumstances. The upper end of the range does not
generally correspond with the highest value found in Chapter 4. This upper value is determined
from the better environmental performance levels reported but also takes account of the cost of
reducing emissions and the process requirements constraints. In this respect, the proposed range
is in accordance with the definition of BAT. For certain techniques, reduction percentages are
given as well as associated emission levels. The reduction percentage (based on data in previous
chapters) relates to the potential of the technique; it is acknowledged that, if a more
contaminated feed is processed, a higher reduction percentage needs to be applied to reach the
BAT associated emission values. Each range has its own purpose and importance and can be of
help during the permit process.
In the determination of the upper value of the associated emission range in this chapter, a split
view has been expressed by industry and two Member States. The split view is that they
consider that the upper value in Chapter 5 should correspond with the upper value found in
Chapter 4. Their rationale is that if a technique considered BAT is already applied in an
installation and achieves a certain value, that value should be considered in the range of
associated emission values.
• Due to the complex nature of refineries, the diversity in different refinery configurations
and the high degree of process integration, it is strongly advised to read Chapter 5 in
conjunction with Chapter 4 when determining the most appropriate solution for an
individual refinery. To help the reader in this subject, references to Chapter 4 have been
included in Chapter 5.
• Since the likelihood of new facilities being built in Europe is low, environmental
improvement in the sector is normally achieved by applying BAT to existing installations.
This retrofitting is often referred to as “revamping”. This document takes account of the
differences between new and existing installations, where relevant and quantifiable.
• Certain prevention techniques (e.g replacing an alkylation process by another) that appear
in the following sections may be very difficult to implement in existing installations. The
reason is that the change from current practice to a more environmentally friendly
technique has certain environmental and economic costs (e.g. decommissioning) which
may outweigh the environmental and economic advantages of applying that technique.
Thus, application is only justified in major rebuildings or changes and new installations.
Existing installations may have factors such as space or height limitations that prevent
full adoption of some of those techniques. A proper assessment can only be done at
local/site level.
• Without prejudice to the preceding paragraph, the control and abatement techniques that
appear in this chapter are widely applicable to both new and existing plants in the refinery
sector.
• Many factors influence decisions about whether or not a refinery should have a certain
process technique or pollution abatement technique. Factors such as the refinery’s
products, type of crude oil processed and type of refinery will need to be taken into
account when using this document at local level. Similarly, actual costs for a technique in
one location/site may differ from those in another; even in the same location the cost for
an identical technique may differ widely depending on specific situations. (e.g.
differences in furnace layout, in space availability). Independent assessments, by e.g.
engineering contractors, may resolve issues when they arise and can only be done
properly at local level.
• For an existing plant where the actual environmental performance is close to a BAT
performance, it may be appropriate to consider the marginal cost-effectiveness of
implementing a new measure in order to reach the BAT level. (See below for more
information on cost-effectiveness analysis.)
• In addition to the BAT referenced in this chapter, BAT for a refinery contains elements
from other IPPC documents and international regulations. In this context, special
attention is drawn to the Storage and Handling BREF, Industrial Cooling BREF,
Monitoring BREF and BREF on Large Combustion Plants when commercial fuels (e.g.
natural gas, commercial fuel oil, gasoil) are used in the refinery.
Cost- effectiveness
In judging whether a technique included in Chapter 4 and 5 is “available” in general, the
concept of cost-effectiveness - taking account of investment and operating costs - is a useful
tool for the BAT assessment of the respective technique as compared to alternative techniques
that can achieve comparable levels of pollution prevention or control. Levels of cost-
effectiveness (e.g. the cost to reduce one tonne per year of SO2 emission) can be useful in
setting so-called “reference values for cost-effectiveness”. Such indicators of cost-effectiveness
are sometimes used in other industries. Some data provided by the TWG in this respect are
found in some techniques in Chapter 4 as well as in Annex IV.
The discussion among the experts of the TWG reflected differences of opinion that go beyond
purely techno-economic considerations. Basically they may be considered to reflect differences
in legislative and administrative procedures in Member States, which in turn are basically the
result of differences in administrative and legislative cultures. The differences in approach
should therefore not be classified as right or wrong within the IPPC legal structure. Both
approaches should be respected as having their own merits in the permitting procedure and can
be used in a complementary rather than contradictory manner. However, several important
conclusions were drawn from the discussion on both approaches:
• The bottom-up approach has the advantage that the assessment of the emission values
associated with the application of individual techniques on a separate process unit is more
straightforward and more in line with the approach used in other BREFs. It provides the
licensing authority with a clear picture of what it is considered to be the best
environmental performance of a single refinery process/activity and how the goal can be
achieved. This approach subsequently requires that the aspects of integration of the
refinery as whole should be built from the individual processes/activities. If this is not
performed correctly, it may result in a sub-optimal solution. This approach recognises
that this integration aspect and the calculation of the bubble level can only be assessed
properly at site level.
• The top-down approach can be used as an effective management tool to prioritise the
application of environmental techniques in installations without becoming entangled in
the complex technical-economic interrelationships in the refinery. However, it does not
provide the licensing authority with guidance on what environmental goals may be
achieved in certain processes/activies or on how a certain environmental goal can be
achieved. The flexibility on how to achieve a goal is largely given to the operator with
this approach.
• Both approaches recognise that the overall reduction of emissions is typically achieved in
practice by a combined effort focused on the performance of the ‘processes/activities’
(e.g. improved SRU efficiency, application of low NOx techniques) and on the
performance of the ‘integrated/whole’ installation (e.g. energy efficiency, fuel
management, sulphur balance). As a result of the high variability between refineries, the
set of measures that leads to the lowest overall emissions at acceptable costs may also
vary significantly from one refinery to the other.
• A certain direct relationship has been recognised between the two approaches. The more
environmental measures that are applied, the lower the bubble level can be. Annex V
gives some examples of calculation of the bubble figure considering BAT associated
emission values for single processes/activities.
environmental performance. They provide the framework for ensuring the identification,
adoption of and adherence to BAT options that, whilst often down-to-earth, are important.
These good housekeeping/management techniques/tools often prevent emissions.
BAT is to:
• implement and adhere to an Environmental Management System (EMS) (See
Section 4.15.1). A good EMS could include:
§ the preparation and publication of an annual environmental performance report. Such
a report will also enable the dissemination of performance improvements to others,
and will be a vehicle for information exchange (Art. 16 of the Directive). External
verifications may enhance the credibility of the report.
§ the delivery to stakeholders on an annual basis of an environmental performance
improvement plan. Continuous improvement is assured by such a plan.
§ the practice of benchmarking on a continuous basis, including energy efficiency and
energy conservation activities, emissions to air (SO2, NOx, VOC, and particulates),
discharges to water and generation of waste. Benchmarking for energy efficiency
should involve an internal system of energy efficiency improvements, or intra- and
inter-company energy efficiency benchmarking exercises, aiming for continuous
improvements and learning lessons.
§ an annual report of the mass balance data on sulphur input and output via emissions
and products (including low-grade and off-spec products and further use and fate).
(See Section 4.23.5)
• improve stability of unit operation by applying advanced process control and limiting
plant upsets, thereby minimising times with elevated emissions (e.g. shutdowns and start-
ups) (see Section 4.15.5)
• apply good practices for maintenance and cleaning (More information in 4.15.3).
• implement environmental awareness and include it in training programmes (More
information in Section 4.15.4).
• implement a monitoring system that allows adequate processing and emission control.
For more information on monitoring, refer to Section 3.26 and the Monitoring BREF.
Some elements of a monitoring system could include:
§ continuous monitoring of pollutants for high volume flows with a high variability in
pollutant concentrations
§ periodic monitoring or the use of emission relevant parameters for flows with a low
variability
§ regular calibration of measurement equipment
§ periodic verification of measurement by simultaneous comparative measurements.
BAT is to:
• improve the energy efficiency (reduction of all air pollutants generated by combustion)
by enhancing heat integration and recovery throughout the refinery, applying energy
• use clean RFG and, if necessary to supply the rest of the refinery energy demand, liquid
fuel combined with control and abatement techniques (see Section 5.2.10) or other fuel
gases such as natural gas or LPG. Where gaseous fuels replace liquid fuels, the reduction
of SO2 and NOx emissions as a result of the fuel switch is not complicated to calculate for
single processing units and for the refinery as a whole (see Section 4.15.2). Information
on cross-media effects and applicability restrictions of a full gas switch is in Section
4.10.2.1
The TWG has not been able to identify a single range of emissions associated with the
application of BAT under the bubble concept (see reasons within the introduction to BAT
for Reduction of Emissions to Air). However, several benchmarks were identified (See
Section 4.15.2):
For the concentration bubble approach (all in mg/Nm3 and @ 3% O2):
§ One Member State proposed that the full implementation of BAT results in a
bubble of 60 - 200 (daily average).
§ Two Member States proposed that the implementation of BAT results in a bubble
of 100 - 600 (monthly average) based on the calculations in Annex V.
§ One Member State proposed a bubble value of 850 (daily average)
§ Two Member States proposed a bubble range of 800 - 1200 (monthly average)
based on current practice.
§ Industry proposed a bubble range of 1000 – 1400 (yearly average) based on current
European refinery performance.
For the load bubble approach (all in t SO2/Mt of throughput):
§ One Member State proposed that the full implementation of BAT result in a bubble
range of 50 – 230 (yearly average)
§ One Member State proposed a bubble range of 50 – 210 (yearly average) based on
the upper quartile of specific emissions from 40 existing EU refineries.
The above data demonstrates the wide variety of proposed ranges and the inconsistency
when daily-monthly-yearly values are compared (yearly values should be the lowest
values).
One Member State does not agree with the above approach of dealing with the sulphur
dioxide emissions in the refinery as a whole. Their proposal is to follow their national
methodology for BAT determination and implementation.
The TWG has not been able to identify a single range of emissions associated with the
application of BAT under the bubble concept (see reasons within the introduction to BAT
for Reduction of Emissions to Air). However, several benchmarks were identified (See
Section 4.15.2):
For the concentration bubble approach (all in mg/Nm3 and @ 3% O2):
§ One Member State proposed that the full implementation of BAT results in a
bubble of 70 - 150 (daily average).
§ One Member State proposed that the implementation of BAT results in a bubble of
100 - 200 (monthly average) based on calculation in Annex V.
§ One Member State proposed that the implementation of BAT results in a bubble of
150 (monthly average) and 200 daily.
§ Two Member States proposed a bubble range of 250 – 450 (monthly average)
based on current practice.
§ Industry proposed a bubble range of 200 – 500 (yearly average) based on current
European refinery performance.
For the load bubble approach (all in t NOx/Mt of throughput):
§ One Member State proposed a bubble range of 20 – 150 (yearly average) based on
the upper quartile of specific emissions from 40 existing EU refineries.
§ One Member State proposed that the full implementation of BAT results in a
bubble range of 80 – 170 (daily average)
The above data demonstrate the wide variety of proposed ranges and the inconsistency
when daily-monthly-yearly values are compared (yearly values should be the lowest
values).
One Member State does not agree with the above approach of dealing with the nitrogen
oxides emissions in the refinery as a whole. Their proposal is to follow their national
methodology for BAT determination and implementation.
The TWG has not shared so much information on values on particulate emission under
the bubble concept. The reason is because the bubble concept it is less applied for
particulates emissions than for the case of for SOx and NOx emissions. For this reason no
figures appear here. However, several benchmarks were included in Section 4.15.2.
• achieve the following water parameters in the WWTP effluent (see Table 4.50 in
Section 4.24.8)
§ combination of waste water from several processes with comparable qualities for
pretreatment (e.g. treatment of sour water from primary distillation unit, catalytic
cracker, coking and from other sour water sources by stripping; see Section 4.24.1).
BAT is to:
• use triple effect evaporation systems in the solvent recovery sections of deasphalting,
extraction and dewaxing process units. In some instances, for technical reasons (increase
in temperature and pressure) it is not feasible to go from double to triple effect. Triple
effect systems are typically used in non-fouling feedstocks (e.g. wax). (see 4.3.1)
• use N-methyl pyrrolidone (NMP) as solvent in the aromatic extraction. In some cases, the
switch from furfural to NMP may not be justified environmentally or technically
especially when producing lower boiling point base oils (e.g transformer oil distillates).
Because a solvent switch typically requires different temperature, pressure and solvent
volumes, they are typically very costly (See 4.3.2). Industry claims that considering the
information within this document, both NMP and furfural are equally viable solvent
candidates. In Industry’s opinion, no clear case has been made in the BREF to arrive at
one preference.
• use hydrotreating for clean-up of the base oil streams and wax finishing if final clean-up
is required. Clay treatment may be justified in some cases because product quality cannot
be achieved by hydrotreating, but it produces solid waste. (see 4.3.4,5)
• consider the application of a common hot oil system for these solvent recovery systems to
reduce the number of furnaces (see 4.3.10) In stand-alone lube refineries, this will
facilitate application of flue gas treatment where liquid fuel is used (See 5.2.10)
• apply prevention techniques for VOC emissions from solvent-containing systems (e.g.
storage). Practise benchmarking of solvent consumption (See 4.3.6-7)
• apply BAT on sulphur recovery units in stand-alone lube refineries (see 5.2.23)
• review the possibility of waste water stripping from aromatic extraction before re-use (see
4.3.9)
• consider the effect of solvents when designing and operating the WWTP. This BAT can
be seen as a part of a water management scheme (See Section on generic BAT and 4.3.2)
• apply leakage prevention measures for solvent-containing systems, to prevent soil and
groundwater contamination with the water-soluble solvents employed. This can be seen
as a part of the waste management system (See Section on generic BAT and
Section 4.3.7)
5. Catalytic Cracking
A catalytic cracking unit is usually part of a processing complex that includes a gas plant, amine
treating of the light (incl.C3/C4) gases and treatment of various product streams. BAT
determinations for these associated activities are dealt with in the respective parts (gas
separation processes, hydrogen-consuming processes and waste gas treatments) of this section.
For the catcracker itself (reactor, regenerator, fractionator, catalyst receipt, storage/loading, wet
gas compressor, absorber and debutaniser) the main specific BAT determinations are given
above. Solutions and measures should be considered in an integrated way with measures and
solutions in other units.
Three split views from Member States appeared on the range of the NOx emissions. One
Member State claims that the application of an SCR is always possible and that the upper
value should be therefore 100. One Member State claims that the range should be 300 –
450 mg/Nm3 because the FCC technology is chosen according to the crude oil processed,
the refinery configuration and the product demand. So NOx emissions cannot drive the
production operations. One Member State specifies that the range should be 10 – 450
because SCR and SNCR are not applicable to all existing units.
• reduce SO2 emission to 10 - 350 mg/Nm3 (lower end of the range only achievable when
both low sulphur feedstock and FGD are used) by a suitable combination of:
§ hydrotreatment of the feedstock if it is economically and technically viable (see
4.5.4). This technique is easier to implement when amine scrubbing and Claus
capacities and hydrogen are available (see 4.5.4).
§ using DeSOx catalyst additive (see 4.5.10.1)
§ using FGD (techniques mentioned in Sections 4.5.10.2-6) of the regenerator gas with
95 – 99 % efficiency (emission target depends on uncontrolled level) especially if
hydrotreatment of feedstock is not applicable.
One Member State maintains that FGD is always applicable and that the range should
therefore be changed to 10 – 100.
One Member State does not agree with the approach followed in this section for the SOx
and NOx emissions. Their proposal is to follow their national methodology for BAT
determination and implementation.
• apply good desalting practices (described in Section 4.9.1-3) that result in optimum
downstream processing (minimal corrosion, meeting product specifications and reducing
catalyst contamination) and waste water quality (oil and N content)
• maximise the use of already-used refinery water as desalting water (see 4.9.4).
BAT is to:
• adopt an Energy Management System as part of the Environmental Management System
(described in Generic BAT section) (see 4.10.1.2). The aim is to increase the refinery
energy efficiency (see below). As elements mentioned in the EMS section, a good Energy
Management System could include:
§ reporting the energy efficiency of the refinery and the plan to increase it (as part of
environmental performance report described in EMS)
§ an energy consumption reduction plan (as part of environmental performance report
described in EMS)
§ participation in ranking/benchmarking activities in energy consumption (as part of
benchmarking described in EMS)
• improve the refinery energy efficiency (see Generic BAT). Techniques that help to
improve the overall efficiency are:
§ application of efficient energy production techniques such as use of gas turbines,
combined cycle power generation/co-generation plants (CHP), IGCC, efficiently
designed and operated furnaces and boilers and replacement of inefficient boilers
and heaters. In the implementation phase of the replacement of inefficient boilers
and heaters, consideration should be given tothe retrofitability, size, actual amount of
emissions, age and remaining life in order to evaluate the cost-effectiveness and the
timing for the action. (see 4.10.3)
§ application of combustion improvement campaigns (see 4.10.1.2)
§ matching energy production and consumption with computerised control systems
(see 4.10.1.3)
§ optimised use of steam in the stripping processes and the use of steam traps (see
4.10.1.4)
§ enhancing the heat integration through processes/activities by energy optimisation
analysis (see 4.10.1.3)
§ enhancing the recovery of heat and power within the refinery (see 4.10.1.3)
§ using of waste heat boilers to reduce the use of fuel for the production of steam (see
4.10.1.3)
§ identifying and using, if possible, opportunities for synergy outside the refinery
fence (e.g. district/industrial heating, power generation) (See 4.10.1.3)
• use clean RFG and, if necessary to supply the rest of the refinery energy demand, liquid
fuel combined with control and abatement techniques (see Section 5.2.10) or other fuel
gases such as natural gas or LPG. Where gaseous fuels replace liquid fuels, the reduction
of SO2 and NOx emissions as a result of the fuel switch is not complicated to calculate for
single processing units and for the refinery as a whole (see Section 4.15.2). Information
on cross-media effects and applicability restrictions of a full gas switch is in Section
4.10.2.1
• increase the proportion of clean fuels used. This can achieved by a suitable combination
of:
§ maximisation of the use of refinery fuel gas (RFG) with low H2S content (20 - 150
mg/Nm3 by amine treating) (see 4.10.2.1-2 and 4.23.5.1). One Member State claim
that for existing refineries the range should be <500 – 1000 ppm because the H2S
content in RFG is negligible if the bubble limit is respected.
§ balance and control of the RFG system between suitable pressure limits to give
system flexibility, with make-up available from sulphur-free sources such as LPG or
imported gas (see 4.10.2.1)
§ use state-of-the-art controls to optimise the performance of the RFG system (see
4.10.11.3 and 4.10.12.1)
§ use flaring of RFG only during start-up/shutdown/upset/emergency conditions.
During normal operation give a constructive use to the RFG, including sale (related
to flare section at the end of this chapter) (see 4.23.7)
§ upgrading and cleaning heavy fuel oil used in the refinery to low-sulphur liquid fuel.
As mentioned in the SO2 emission section below, the abatement of emissions after
firing liquid fuels is also considered BAT (see 4.10.2.3).
• reduce CO2 emissions by:
§ increasing the refinery energy efficiency (see above and Generic BAT section)
§ increasing the use of gaseous fuels that have higher hydrogen-to-carbon ratios (see
4.10.2.1)
• reduce CO emissions by application of efficient combustion techniques
• reduce NOx emissions:
§ by reducing fuel consumption (increase the energy efficiency, see above)
§ by replacing existing burners with low-NOx types during major outages. The lowest-
NOx burners suitable for individual applications should be used (see 4.10.4.1.2)
§ from boilers and heaters using gas fuel to 20 - 150 mg/Nm3 (lower levels for natural gas
and higher ones for small heaters with primary measures. Two Member States claim
that the upper value should be 100 because it can be reached with the implementation of
primary measures and SCR) by applying a suitable combination of:
§ high thermal efficiency heater/boiler designs with good control systems (e.g. oxygen
trim) (see 4.10.3.1)
§ low-NOx burner techniques (see 4.10.4.1-2)
§ flue gas circulation in boilers (see 4.10.4.3)
§ reburning technique (see 4.10.4.5)
§ SCR/SNCR. Ammonia slip considered associated with the use of SCR is
2 - 5 mg/Nm3. The lower value is achievable with new catalysts and ammonia slip
typically increases with catalyst life. (see 4.10.4.6-7)
§ from boilers and heaters using liquid fuel to 55 - 300 mg/Nm3 (the lower levels relates
only to boilers with SCR and the higher levels only to small heaters with primary
measures. One Member State claims that small heaters (<50MW) can reach 200 and
with the big heaters and boilers (>50MW) the installation of SCR is justified and
consequenly values of less than 100 can be reached. One Member State claims a level
from 200 – 400 because of the nitrogen content in the fuel) by applying a suitable
combination of:
§ fuels with low nitrogen content (related with low sulphur content) (see 4.10.2.3)
§ low-NOx burner techniques (see 4.10.4.1-2)
§ flue gas circulation in boilers (see 4.10.4.3)
§ reburning technique (see 4.10.4.5)
§ SCR/SNCR to liquid fuels heavier than gasoil type. Ammonia slip considered
associated with BAT is 2 - 5 mg/Nm3. The lower value is achievable with new
catalysts and ammonia slip typically increases with catalyst life. (see 4.10.4.6-7)
§ from gas turbines to 20 – 75 mg/Nm3@ 15 % O2 (lower levels for natural gas and
higher levels for small gas turbines and RFG. One Member State claims that the upper
value should be 35 based on primary measures and SCR) by applying a suitable
combination of:
§ diluent injection (see 4.10.4.4)
§ dry low-NOx combustors (see 4.10.4.2)
§ SCR. Ammonia slip considered associated to BAT is 2 – 5 mg/Nm3. The lower value
is achievable with new catalysts and ammonia slip typically increases with catalyst
life. (see 4.10.4.6-7)
• reduce particulate emissions (particulates from liquid firing contain Ni,V) to 5 - 20
mg/Nm3 by applying a suitable combination of:
§ reducing the fuel consumption (increasing the energy efficiency. See above)
§ maximising the use of gas and low ash content liquid fuels (see 4.10.5.1)
§ steam atomisation on the liquid fuels (see 4.10.5.2)
§ ESP or filters in the flue gas of furnaces and boilers when heavy liquid fuel is used
(see 4.10.5.3-4)
One Member State claims that the range for particulates should be 30 – 50 because those
values are consistent with a 95% abatement. Industry rationale for the range 5 to 50
mg/Nm3 is that this is consistent with the whole range reported in Chapter 4.
• reduce sulphur dioxide emissions:
§ by reducing the fuel consumption (increasing the energy efficiency)
§ from combustion processes (boilers, heaters and gas turbines):
§ by increasing the proportion of clean fuels used (to low-sulphur residual fuel, to
gasoil, ultimately to gas) (See BAT on clean fuels at the beginning of
Section 5.2.10)
§ to 5 – 20 mg SO2/Nm3 when using fuel gas by cleaning refinery fuel gas (20 to
150 mg H2S /Nm3) including monitoring of the sulphur content of the refinery
fuel gas. Refer to split views given on the H2S concentration in BAT for
refinery fuel gas.
§ to achieve an average emission value of 50 – 850 mg SO2/Nm3 (lower end of
the range is for the implementation of FGD and deep hydrodesulphurisation to
all liquid fuels) for the total refinery liquid fuel pool by applying a suitable
combination of:
§ hydrodesulphurising the necessary amount of liquid fuel (See
section 4.10.2.3)
§ applying flue gas desulphurisation (See Section 4.5.10 and 4.23.5.4). This
technique is more cost-effective in large furnaces and boilers.
One Member State claims that the use of FGD is always possible and that the
upper value should therefore be 200. One Member State and Industry claim that
the limit should be no lower than 1700 mg SO2/Nm3 which is equivalent to a
1 % sulphur level in fuel oil with no abatement.
The Commission has noted the divergent views of the TWG concerning the
average sulphur dioxide emission levels when burning liquid fuels, associated
with the use of BAT. The Commission further notes that Council Directive
1999/32/EC on the sulphur content of certain liquid fuels prescribes a
maximum emission limit value of 1700 mg/Nm3, which equates to 1 % sulphur
in heavy fuel oil, as a monthly mean value averaged over all plants in the
refinery from 1 January 2003. In addition, the more recently adopted Directive
2001/80/EC on large combustion plants provides for emission limit values in
the range of 200 to 1700 mg/Nm3 depending on the characteristics of plants
covered by that directive.
In this perspective, the Commission believes the range of 50 to 850 mg/Nm3, as
average sulphur dioxide emission levels when burning liquid fuels to be
consistent with BAT. In many cases, achieving the lower end of this range
would incur costs and cause other environmental effects which outweigh the
environmental benefit of the lower sulphur dioxide emission (reference in
Section 4.10.2.3). A driver towards the lower end could be the national
emission ceiling for sulphur dioxide as fixed in Directive 2001/81/EC on
national emission ceilings for centain atmospheric pollutants or if the
installation is located in a sulphur sensitive area.
• reduce water use by:
§ re-using condensate water as de-areator feed water. Where the condensate and the
de-areator are located far apart, their integration is not always cost effective. (see
4.10.3.2)
§ preheating the boiler feed water with waste heat. Where the BFW and the available
waste heat are located far apart, their integration is sometimes too costly compared
with the environmental benefit (see 4.10.3.2)
One Member State does not agree with the approach followed in this section for the SOx
and NOx emissions. Their proposal is to follow their national methodology for BAT
determination and implementation. Another Member State does not agree with the
approach followed in this Section because they claim that the bubble approach should
drive the Energy BAT Section.
• use as fuel preferably gas that is of saleable quality (typically less than 5 mg of H2S/Nm3)
(see 4.17.1)
• consider, particularly for large carbon dioxide flows, alternatives to direct releases of
CO2. (See Section 4.17.3)
• dispose of mercury recovered from raw natural gas (if present) in an environmentally
acceptable way (Refer to Section 4.17.7)
BAT is to:
• maximise heat integration by selecting between:
§ design highly integrated units (e.g. progressive distillation) (see 4.19.1)
§ increase heat integration between the atmospheric crude distiller with the vacuum
unit or with other refinery process units (4.19.2-3). Some techniques that could be
used are:
§ application of energy optimisation analysis to the crude preheat train
§ increasing crude distillation column pumparounds. Reboiling sidestrippers with
a heat transfer oil rather than by steam stripping.
• maximise the use of liquid ring vacuum pump and surface condensers in place of some
stage steam jet ejectors on the vacuum tower overhead. Especially applicable to the last
vacuum stage which gives the most environmental benefit avoiding contamination of
water. Minimise waste water flow/transfer of hazardous substances from vacuum pumps
applying water recycling/waste water-free technique. See applicability restriction and
more information in Section 4.19.4.
• apply advanced process control to optimise energy utilisation (see 4.19.2-3)
• use crude distillation units as an alternative (other is in 5.2.7 third bullet) to reprocess
slop. This technique may generate problems in the desalter or may foul up the heat
exchangers (see 4.19.8).
BAT is to:
• apply the BAT determined in the BREF on Storage
• ensure the liquids and gases stored are in appropriate tanks or vessels based upon the true
vapour pressure of the stored material (see Storage BREF).
• implement containment BAT (see storage BREF)
• use high-efficiency seals in floating roof tanks (see storage BREF)
• bund all stored chemicals, with separate bunding for incompatibles (see storage BREF)
• apply emission reduction measures during tank cleaning (See 4.21.10-11)
• apply concepts of good housekeeping and environmental management (see Section 5.1
and 4.15.3)
• minimise number of tanks and volume by a suitable combination of: application of in-line
blending, integration of processing units, co-operation with partners in industry. This
technique is much easier to apply on new facilities. (see 4.21.7, 4.21.14, 4.15.5)
• enhance vapour balancing and back venting during loading/unloading processes, e.g. by
vapour balance lines that transfer the displaced vapour from the container being filled to
the one being emptied. Incompatibility of tank vapours and applicability to external
floating roofs tank are some examples of restrictions of application. Applicability needs
to reflect economics, the type and size of vessel to be used (e.g. tank, truck, railcar, ship),
type of hydrocarbon fraction and frequency of use of the tank. Because this technique is
related to the next one, both should be evaluated together when implementing on a
specific site. (see 4.21.18)
• apply vapour recovery (not applicable to non-volatile products) on tanks, vehicles, ships
etc. in stationary use and during loading/unloading. Achieved emission levels are very
dependent on the application, but recoveries of 95 - >99 % are considered BAT. If VRUs
are not considered appropriate for certain streams, vapour destruction units are considered
BAT. Properties of streams, such as type of substance, compatibility of substances or
quantity need to be considered in the applicability of this BAT. Applicability needs to
reflect economics, the type and size of vessel to be used (e.g. tank, truck, railcar, ship),
type of hydrocarbon fraction and frequency of use of the tank. Because this technique is
related to the above one, both should be evaluated together when implementing on a
specific site (see 4.21.16 and 4.23.6.2)
• reduce (risk of) soil contamination by the implementation of an inspection and
maintenance programme that could include implementing good housekeeping measures,
double-bottom tanks, impervious liners, good housekeeping practices (draining,
sampling, tank bottoms) (as part of the EMS) (see 4.21.8 and 4.21.13)
• install self-sealing hose connections or implement line draining procedures (see 4.21.13)
• install barriers and/or interlock systems to prevent damage to equipment from the
accidental movement or driving away of vehicles (road or rail tank cars) during loading
operations (see 4.21.13)
• implement procedures to ensure that arms are not operated until inserted fully into the
container to avoid splashing where top loading arms are used (see 4.21.13)
• apply instrumentation or procedures to prevent overfilling (see 4.21.13)
• install level alarms independent of normal tank gauging system (see 4.21.13).
Air pollutant Section of this chapter where specific BAT-related air pollutants are
mentioned
CO Catcracker, energy system
CO2 Energy system
NOx Generic, catcracker, energy system
PM Generic, catcracker, reformer, coker, energy system
SOx Generic, catcracker, coker, energy system, product treatments,
visbreaker
VOC Generic, bitumen, gas separation processes, NG plants, storage
6 EMERGING TECHNIQUES
Emerging technique is understood in this document as a novel technique that has not yet been
applied in any industrial sector on a commercial basis. This chapter contains those techniques
that may appear in the near future and that may be applicable to the refinery sector.
1. Large oil companies are cutting down on their R&D budgets, and are more and more
relying on third parties for new developments in refinery technology and catalytic
processes. These third parties develop techniques for licensing income or for income from
catalyst sales. Developments are reported in the technical literature (Hydrocarbon
Processing, Chemical Engineering Progress, Oil and Gas Journal, Erdöl, Gas und Kohle,
Petroleum Technology Review) and during seminars and conferences (World Petroleum
Congress, WEFA, Hart’s Fuel Conference, European Refining Technology Conference,
NPRA and API specialist meetings) for the dissemination of those techniques.
2. Technological developments are concentrating on optimising current systems for higher
yields (e.g. more selective catalysts and solvents), higher energy efficiency (e.g. improved
reactor design and better heat integration) and shorter downtimes (e.g. scavenging of
impurities, automatic cleaning systems) rather than novel processes;
3. The current tool box of conversion, separation, treatment and environmental technologies
seems adequate and sufficient to meet any desired product slate and product specifications
for the coming decade as well as meeting stringent regulatory requirements;
Meanwhile, the upgrading process of the refinery sector to implement existing techniques for
meeting new product specifications will continue. The rationalisation process in the refining
industry will also continue and low margins will force refineries to look for cost savings.
Alkylation
Most of the safety and environmental concerns are related to the potential for a large release of
HF or H2SO4. A solid acid catalyst would potentially overcome most of the drawbacks
associated with liquid acid catalyst systems and open up a new area of marketing for alkylation.
Numerous companies are putting a large R&D effort into the development of a new solid
catalyst for the alkylation process.
Catalyst Technology
BF3/Al2O3 Grignard
SbF5/SiO2 Circulating bed + washing
Zeolite H-b Grignard
CF3SO3H fixed in a porous carrier (silica) Haldor Topsoe A/S
[61, Decroocq, 1997], [330, Hommeltoft, 2000]
Technology providers claim that these techniques will be ready in the market in one to two
years.
Catalytic cracking
Some promising lines of investigation for the improvement of the environmental performance
of catcrackers are:
Catalytic reforming
The current practice of application of continuously improved catalysts (supplied by catalyst
manufacturers) is expected to continue. [259, Dekkers, 2000]
Energy system
Some promising lines of investigation for the improvement of the environmental performance
of the energy systems are CO2 abatement techniques (more information within the waste gas
treatments). Others are the heat integration. The search for further energy improvements is
continuing, with the current focus on attractive cogeneration opportunities and more complex
heat integration.
Hydrogen production
Some promising lines of investigation in hydrogen production technologies are:
• the hydrocarb process, in which the residual oil is essentially cracked to carbon and
hydrogen. This process can be seen as an internal source of natural gas for a refinery. The
process produces carbon, hydrogen and methanol. It has been calculated in a refinery of
4.98 t/yr that this process can increase by 40 % the total gasoline production, 1150 m3/d of
methanol and 795 m3/d of C/H2O slurry [12, Steinberg and Tung, 1992].
• methane pyrolysis, which takes advantage of the thermal decomposition of natural gas and
the direct production of hydrogen while sequestering the carbon or using the carbon for
other commodity purposes. Consequently the CO2 generation is completely eliminated [12,
Steinberg and Tung, 1992], [281, Steinberg, 2000].
Hydrogen-consuming processes
Some promising lines of investigation for the improvement of the environmental performance
of energy systems are:
• residue hydrotreating and hydroconversion processes (e.g. slurry bed technology). This
process has only been demonstrated at semi-commercial scale and no commercial plants are
in operation yet.
• gasoline deep desulphurisation techniques with a lower hydrogen consumption are currently
under development. Parameters are not yet available.
Product treatments
A development to be mentioned is the biodesulphurisation of gas oil and even crude oil [259,
Dekkers, 2000]
Assuming refinery utilisation gradually increases and that product specification and abatement
requirements continue to be more stringent, then CO2 releases, if unabated, are likely to
continue to rise, since measures to meet these requirements will require energy. In principle CO2
emissions can be reduced by separation, collection and finding a useful application. Injection of
CO2 aiming at secondary and tertiary oil recovery is a potential possibility. Injection into sub-
soil formations for storage or supply to greenhouses as a gaseous fertiliser product has also been
suggested. However, considering the large quantities of CO2 involved these projects will be
very costly indeed. Moreover, these types of solutions can only partly alleviate the CO2
emission issue.
Disposal of CO2
Unlike the abatement of other pollutants no feasible technology exists for the removal of carbon
dioxide from flue gases. A number of disposal options are, however. under scientific
consideration. For technical, ecological and economic reasons, a viable solution is not yet
available, but this option is currently being investigated by certain major operators and the
International Energy Agency (IEA).
7 CONCLUDING REMARKS
Some general concluding remarks
The environmental situation of European refineries varies greatly across the European
Community, so the starting point for each case is very different. Different environmental
perceptions and priorities are also evident.
Sources of information
More than 350 items of information were used for the preparation of this document. Eight
reports from authorities and industry were used as the main sources of information.
Complementary to that information, 17 site visits, sub-group meetings and much other
information from suppliers and general literature were collected. In the later stages, comments
on draft documents have represented the main source of information, complementing,
qualifying or adding to the information into the BREF.
Most of the information on techniques submitted by TWG was focused on abatement techniques
in general. Very little information was made available on the performance of production
techniques to be considered in the determination of BAT, and especially on achieved emission
and consumption levels that those production techniques may achieve. Most commonly, this
type of data came from technology providers or from general literature.
Level of consensus
The refineries sector is a large and complex one, spread across all Member States except
Luxembourg. This size and complexity are reflected in the number of processes / activities
addressed in the BREF and the number (200+) of BAT it contains. The fact that agreement has
been reached on all but 27 of those 200+ BAT is a measure of the broad commitment of TWG
members to the conclusions reached. One corresponds to the general introduction to Chapter 5,
eleven to the Generic BAT and fifthteen to Specific BAT. All of them are listed below.
Very few data were exchanged on production techniques and related good environmental
practices (e.g. current emission and consumption levels, performance, consumption benchmarks
or economics for improvement). That is the main reason why few production techniques appear
in Chapter 5. For future BREF reviews, all TWG members and interested parties should
continue, or start, to collect these data for all production processes. The data should be
expressed as concentrations as well as loads and, where appropriate, in relation to the amount of
input and/or output material.
TWG recognises that technology providers or European technology providers associations (e.g.
European Sealing Association) can give excellent input to the BREF, especially to give their
expertise in new applications in the refinery sector. Consequently it is recommended that they
be formally involved in the next revision of the BREF. The TWG advice is to start the review of
this document in 2006.
For fugitive VOC emissions, a guide at the EU level covering equipment performances and
good practices would be useful. It can be compiled as a preparatory step for the next BREF
revision in partnership between member states, European environmental bodies, suppliers and
industry representatives.
Information on plants and actual performance data are scarce; for the revision of this document
the missing information should be provided. Apart from data missing in general, as mentioned
above, the following comments relate to specific areas where data and information are missing:
a collect and analyse more data on split views with the intention to solve them
b try to achieve consensus in the definition of the bubble and try to collect technical
support for the BAT associated values under the bubble concept for the main air
pollutants.
c energy efficiency has been recognised as one of the most important issues in the BREF.
The reason is because refinery processes are very energy intensive. In the next revision,
a further analysis of the methods and quantification of energy efficiency needs to be
done. The development and promotion of a transparent methodology for the calculation
of energy efficiency should be promoted.
d provide data on decommissioning of processes
e provide data on mass balance of the metals contained within the crude oil. Mass balance
on sulphur is typically done in refineries, but no information on mass balance of metals
has been reported. Volatile metals such as mercury, contained within the crude oil are an
issue that needs to be explored. Where do they go and how can they be abated?.
f data on characteristics of particulates such as unburnt hydrocarbons, metals (e.g. Ni, V)
and PM10 need to be qualified
g very little information has been provided about noise levels or techniques to reduce
them
h no information has been provided on odour abatement. It is recognised that odour
abatement is linked to VOC abatement (typically sulphur compounds), but there is no
identification of compounds and where they may come from. In other words, analysis of
the VOC needs to be done.
i collect data on dioxin emissions and on the effect of catalyst regeneration processes and
some clean-up facilities
j an important issue when considering refinery waste is that there are still many differing
definitions between countries, which makes comparisons of waste difficult. Make a
catalogue of types of wastes produced by refineries to make comparison easier and more
accurate.
k collect information on NOx and SOx emissions from the flaring system
l collect information on NOx emissions from coking processes
m very few information have been exchanged on natural gas plants. This can partially be
justified because the number of those plants are much smaller than Mineral Oil
Refineries. However, some concrete data on how those plants can be improved
environmentally need to be exchanged.
It has been recognised that the cross-media issues that may be generated by some process
techniques are very difficult to assess. For example, alkylation productions raises the difficult
question of which is the better process: HF alkylation (more energy efficient but very toxic
catalyst) or sulphuric alkylation (less energy efficient and the waste generated needs to recycled,
but less toxic). It may be worthwhile to develop some methodology to compare two or more
technologies.
A study to compare energy efficiency in the refinery sector world wide. This needs to consider
the different types of refineries. During this study, a transparent methodology needs to be
defined. The study should also include an analysis of the techniques already used within the
refinery sector to improve their energy efficiency and how they affect the efficiency, cost and
environmental benefit.
No much information about the physical and chemical characteristics of dust (e.g. content of
unburnt hydrocarbons, content of metals and content of PM10) has been found during the work.
Because those analyses are difficult to do, they can be supported by an R&D project to
standardise them.
A study of the fugitive emissions from refineries. This should include the collection of fugitive
VOC emission data from the refinery sector. This may be based on the information gathered by
the Commission under Article 15 (European Pollutant Emission Register, EPER). This should
also include an analysis of the monitoring methods and an analysis of techniques to reduce those
VOC emissions in order to quantify emission reduction capabilities and cost. This programme
can also focus on the different types of equipments that could be used, their performances and
their cost, or on the need to develop new equipment presenting a better trade-off between
performance and cost or allowing fugitive VOC emissions to be identified and characterised at a
lower cost than the actual LDAR programmes.
Refineries handle harmful products like aromatics (benzine, etc…). Some studies should be
promoted to better evaluate the emissions of these products, their impact on the environment
and the different ways to solve these problems. This is to some extent connected to fugitive
VOC emissions, except that here there is a need to characterise these emissions.
There is a need expressed by the permitters to get more precise technical information in this
area. A lot of recommendations are made concerning the seals of different equipment. It could
be particularly useful to study more precisely their efficiency and their implementation
conditions and to set up good practices and methodologies concerning the type of seals to
recommend, the way to check their efficiency, their utilisation conditions and the relevant
replacement interval.
The flameless oxidation mode has begun to be used more widely in sectors such as iron and
steel and ceramics. Studies and experiments should be promoted to encourage similar work in
the refinery sector. This technique seems to offer higher energy efficiency and low NOx
emissions.
It has been very difficult to express environmental performances for this sector in mass flow or
specific mass flow that are highly representative of the environmental impact. This is because
the energy consumption of a refinery increases with its complexity and the degree of conversion
of the products. In parallel, the rate of sulphur recovered increases with the complexity and the
energy consumption. A methodology to facilitate a benchmarking in specific mass flow,
considering for example several kinds of refineries, would be useful.
The EC is launching and supporting, through its RTD programmes, a series of projects dealing
with clean technologies, emerging effluent treatment and recycling technologies and
management strategies. Potentially these projects could provide a useful contribution to future
BREF reviews. Readers are therefore invited to inform the EIPPCB of any research results
which are relevant to the scope of this document (see also the preface of this document).
8 LIST OF REFERENCES
12. Steinberg, M. and Y. Tung, (1992). "Residual Oil refining by hydrocarb process"
Brookhaven National Laboratory, Upton, NY 11973, USA (516)282-3036 (1 author);
Hydrocarb Corporation, New York, NY, USA (2 author), 8 pgs.
18. Irish EPA, (1992). "BATNEEC Guidance Note. Class 9.2. Crude Petroleum Handling
and storage. Draft 3" Irish Environmental Protection Agency, 16 pgs.
19. Irish EPA, (1993). "BATNEEC Guidance Note. Class 9.3. Refining petroleum or gas.
Draft 3" Irish Environmental Protection Agency, 29 pgs.
34. Italy, (1999). "The italian refining industry" Ministero dell'Ambiente, Servizio
inquinamento atmosferico e acustico e le industrie a rischio., 38 pgs.
43. Dekkers, C. and R. Daane, (1999). “Metal contents in crudes much lower than
expected" Oil & Gas Journal, (March 1, 1999) pp. 44-51.
45. Sema and Sofres, (1991). "Technical note on the best available technologies to reduce
emissions of pollutants into the air from the refining industry. Application of articles 7
& 13 of the Directive 84/360 EEC" Report made for European Commission, 135 pgs.
49. CONCAWE, (1998). "Sulphur dioxide emissions from oil refineries and combustion of
oil products in western Europe and Hungary (1995)" Concawe, 30 pgs.
60. Balik, J. A. and S. M. Koraido, (1991). “Identifying pollution prevention options for a
petroleum refinery" Pollution prevention review, (summer 1991) pp. 273-293.
61. Decroocq, D., (1997). “Major scientific and technical challenges about development of
new processes in refining and petrochemistry" Revue de l'institut francais du petrole, 52
(5) pp. 469-489.
73. Radler, M., (1998). “1998 & 2000 worldwide refining survey" Oil & Gas Journal,
Special (Dec. 21, 1998) pp. 49-92.
79. API, (1993). "Environmental design considerations for petroleum refining crude
processing units" American Petroleum Institute and Health and Environment Affairs
Department, 133 pgs.
80. March Consulting Group, (1991). "Pollution Control for Petroleum Processes" March
Consulting Group commissioned research for the Department of Environment (HMIP),
93 pgs.
82. CONCAWE, (1995). "Oil refinery waste disposal methods, quantities and costs. 1993
survey", 32 pgs.
83. CONCAWE, (1990). "A field guide on reduction and disposal of waste from oil
refineries and marketing installations" Concawe, 34 pgs.
101. World Bank, (1998). "Pollution prevention & abatement handbook. Petroleum
refining", 469-472 pgs.
107. Janson, B., (1999). "Swedish BAT notes on refineries" Swedish environmental
protection agency, 18 pgs.
108. USAEPA, (1995). "Profile of the Petroleum Refining Industry (EPA Office of
compliance Sector Notebook Project)" USA Environmental Protection Agency, 150
pgs.
110. HMIP UK (1993). "Guidance on effective flaring in the gas, petroleum, petrochemical
and associated industries (Technical guidance note (abatement) A1)" HM Inspectorate
of pollution in UK, London.
112. Foster Wheeler Energy, (1999). "A strategic review of the petroleum refinery industry
sector" Foster Wheeler Energy Ltd. work for Environment Agency of UK, 141 pgs.
113. Noyes, R., (1993). "Petroleum refining" Pollution Prevention Technology Handbook,
Mill Road, Park Ridge, New Jersey, Noyes Publications.
115. CONCAWE, (1999). "Best available techniques to reduce emissions from refineries"
Concawe, 200 pgs.
117. VDI, (2000). "Emission control from mineral oil refineries" VDI / UBA, 80 pgs.
118. VROM, (1999). "Dutch notes on BAT for the mineral oil refineries" Ministry of
Housing, Spatial Planning and the Environment (VROM)- Directorate for Air and
Energy-Raytheon Engineers & Constructors, 50 pgs.
119. Bloemkolk, J. W. and R. J. van der Schaaf, (1996). “Design alternatives for the use of
cooling water in the process industry: minimization of the environmental impact from
cooling systems" Journal for Cleaner Production, 4 (1) pp. 21-27.
127. UN/ECE, (1998). "VOC task force on emission reduction for the oil and gas refining
industry" DFIU-IFARE, 80 pgs.
136. MRI, (1997). "Emission factors" Midwest research institute, 130 pgs.
144. HMIP UK, (1997). "Natural Gas Processing" UK environmental protection agency, 17
pgs.
147. HMIP UK, (1995). "Petroleum Processes: Oil Refining and Associated Processes"
HMSO, 86 pgs.
175. Constructors, R. E., (1998). "Validation of inventaries of NOx measures in the refinery
sector (Validatie van de inventarisatie van NOx maatregelen in de raffinage sector)"
Raytheon Engineers & Constructors for Ministry of VROM, Netherlands, 200 pgs.
181. HP, (1998). “Environmental processes '98" Hydrocarbon processing, (August 1998) pp.
71-118.
197. Hellenic Petroleum, (1999). "Contribution of Hellenic Petroleum to the refinery BREF"
Hellenic Petroleum, 25 pgs.
198. (Hellas), M. O., (1999). "Contribution of Motor Oil (Hellas) to the BREF" Motor Oil
(Hellas), pgs.
199. Petrola Hellas, (1999). "Contribution of Petrola Hellas to the BREF" Petrola Hellas,
pgs.
211. Ecker, A., (1999). "State of the art of refineries focused on the IPPC directive"
Umweltbundesamt, Austrian Environmental Agency, 160 pgs.
215. Jansson, B., (1999). "Preem and Scanraff refineries in Sweden. Technical information"
Swedish environmental protection agency, 200 pgs.
222. Shell Pernis, (1999). "Environmental report 1998, Technical Information of the
refinery" Shell, 10 pgs.
229. Smithers, B., (1995). "VOC emissions from external floating roof tanks: Comparison of
remote measurements by laser with calculation methods" CONCAWE, 50 pgs.
242. CONCAWE, (1998). "European downstream Oil Industry Safety Performance 1997",
30 pgs.
244. USAEPA, (1992). "Workbook for estimating fugitive emissions from petroleum
production operations", 150 pgs.
245. API, (1983/1989/1990). "Evaporative loss from external and internal floating roof-
tanks" American Petroleum Institute, 150 pgs.
247. UBA Austria, (1998). "State of the art in the refining industry with regard to the IPPC-
directive (Jahrbuch der Europäischen Erdölindustrie)" Federal Environmental Agency
of Austria, 200 pgs.
250. Winter, B., (2000). "Comments from Austrian TWG member to the first draft".
251. Krause, B., (2000). "Comments from German TWG member to first draft".
253. MWV, (2000). "Comments from German refinery association to first draft".
256. Lameranta, J., (2000). "Comments from Finish TWG member to first draft".
257. Gilbert, D., (2000). "Comments from French TWG member to first draft".
258. Manduzio, L., (2000). "Comments from Italian TWG member to first draft".
259. Dekkers, C., (2000). "Comments from Dutch TWG member to first draft".
260. Sandgrind, S., (2000). "Comments from Norwegian TWG member to first draft".
261. Canales, C., (2000). "Comments from Spain TWG member to the first draft".
262. Jansson, B., (2000). "Comments from Swedish TWG to first draft".
268. TWG, (2001). "Comments from TWG to second draft of Refineries BREF".
269. Confuorto, N., (2000). "Controlling FCCU flue gas emissions with scrubbers" Belco, 23
pgs.
271. Martinez del Pozo, J. L., (2000). "Comments to delayed coking chapter of first draft".
272. Shawcross, P., (2000). "Information on coking, calcining and etherification processes"
Conoco, 2 pgs.
278. Alstom Power, (2000). "Flakt-Hydro Process" ABB Alstom Power, 20 pgs.
281. Steinberg, M., (2000). “A new process for ammonia production with reduced CO2
emission or coproduct carbon production" , Draft pp. 12.
282. Conoco, (2000). "Material from the Conoco Humber refinery in the UK" Conoco, 50
pgs.
287. Johnston, D., (1996). “Complexity index indicates refinery capability value" Oil & Gas
Journal, (March 18, 1996) pp. 74.
290. Statoil, (2000). "Information from Statoil Karsto natural gas plant" Statoil-Norway, 11
pgs.
292. HMIP UK, (2000). “UK refineries "lag behind"on pollution abatement" ENDS Report,
302 pp. 11-12.
293. France, (2000). "Material collected during the visit to France in April 2000", 500 pgs.
296. IFP, (2000). "Reports from the IFP" Institute Francais du Petrole, 150 pgs.
302. UBA Germany, (2000). "German Notes on BAT in the Refinery Industry. The German
Refinery Industry" German Environmental Agency, 100 pgs.
310. Swain, E. J., (2000). “Processed-crude quality in US continues downward trend" Oil &
Gas Journal, March 13, 2000 pp. 44-47.
312. EIPPCB (2001). "BREF on common waste water and waste gas treatment/management
systems in the chemical sector", Sevilla.
315. USAEPA, (2000). "Exposure and human health reassesment of 2,3,7,8-TCDD and
related compounds" USA Environmental Protection Agency, pgs.
316. TWG, (2000). "Comments from TWG members to the second draft of chapter 4 and
chapter 5.1", pgs.
319. Sequeira, A. (1998). "Lubricant base oil and wax processing" Marcel De KKer, New
York.
320. Italy, (1996). “Italian IGCC project sets pace for new refining area" Oil & gas journal,
Dec. 9 1996 .
321. Helm, L., C. Spencer, et al., (1998). Hydrocarbon Processing, 1998 pp. 91-100.
322. HMIP UK, (1995). "Chief Inspectors Guidance Note - Processes Subject to Integrated
Pollution Control - Combustion Processes: Large Boilers and Furnaces 50 MWth and
over" HM Inspectorate of pollution, 120 pgs.
324. Sicelub, (2001). "Lubrification systems for pumps, air coolers and compressors"
Sicelub, 400 pgs.
330. Hommeltoft, S. I., (2000). "Information about solid catalyst alkylation" Haldor Topsoe
A/S, 100 pgs.
337. Journal, O. G., (2000). “Special report on refining catalyst demand" Oil & Gas Journal,
2000 .
344. Crowther, A., (2001). "Low temperature NOx oxidation" BOC gases, 7 pgs.
345. Molero de Blas, L. J. (2000). "Pollutant formation and interaction in the combustion of
heavy liquid fuels" Department of Chemical Engineering University Collage London,
London.
348. Ashworth Leininger Group, (2001). "Petroleum Refining Sector. NOx controls report .
Technical advice on emission control technology applie in the California" Ministry of
Environment. Air&Energy Directorate, 20 pgs.
350. European Sealing Association, (2001). "Comments to the Second Draft" European
Sealing Association, 5 pgs.
9 GLOSSARY
Symbols
~ Around, more or less
∆T Increase of temperature, temperature rise
€ EUR (EU currency)
A
AC Alternating Current
AC Activated Carbon
Amylenes Pentenes
AOC Accidentally Oil Contaminated
ºAPI Density unit (specific density = 141.5/(ºAPI+131.5))
API American Petroleum Institute
API separator Oil/water/sludge separator (developed by American Petroleum Institute)
ASU Activated Sludge Unit
atm Atmosphere. Pressure unit equal to 101.3 kPa
Auto-Oil I,II Co-operative programmes between the EU and the oil and motor industries
to find the most cost-effective way to improve air quality in Europe. These
have led to Directives on fuel properties and vehicle emissions.
B
bbl US barrel (159 litres)
bar Pressure unit equal 100 kPa (1 bar = 0.987 atmospheres)
barg Bar gauge (relative pressure over the atmospheric pressure)
BBU Bitumen Blowing Unit
BFO Bunker Fuel Oil
BFW Boiler Feed Water to produce steam
BOD Biological Oxygen Demand
bpcd Barrels per calendar day (average flow rates based on operating 365 day per
year)
bpsd Barrels per stream day (flow rates based on actual on-stream time of a unit)
BREF Best Available Techniques reference document (Documents developed by
the EIPPCB on different IPPC sectors)
BTEX Benzene, Toluene, Ethylbenzene, Xylene
BTX Benzene, Toluene, Xylene
C
cat catalytic
catcracker catalytic cracker (typically refers to fluid catalytic cracker)
CCGT Combined Cycle Gas Turbine
CCR Conradson Carbon Residue (See under Concarbon more information)
CDD/CDF Chlorodibenzoparadioxins/furans
CDU Atmospheric Crude Distillation Unit
CEC Comission of the European Community
CHP Combined Heat and Power (co-generation)
COC Continuously Oil Contaminated
COD Chemical Oxygen Demand
Concarbon Conradson carbon = measurement of carbon residue (expressed in % w/w).
Measurement of the tendency of a hydrocarbon to form coke.
CONCAWE European Refinery Association for Environment, Health and Safety
Corinair European Air Emissions inventory
CPI Corrugated Plated Interceptor
Cx Hydrocarbons with x number of carbons
CW Cooling Water
Currencies http://fxtop.com/en/adv.htm
D
d day
DAF Dissolved Air Flotation
dB(A) Noise unit
DC Direct Current
DCU Delayed Coker Unit
DEA Diethanol Amine
DG Direction General from the European Commission
DGA Diglycolamine
DIAL Differential absorption light detection and ranging
DIPA Di-isopropanol amine
DNB Nitrification/denitrification biotreater
DS Dissolved Solids
E
EC European Community
EFRT External Floating Roof Tanks
EIA Environmental Impact Assessmente
EII Energy intensity index (Solomon index. See Section 3.10.1)
EIPPCB European IPPC Bureau
EMS Environmental Management System (see Section 4.15.1)
EOP End-of-pipe
EP European Parliament
ETBE Ethyl tert-butyl ether
EU+ EU countries plus Norway and Switzerland
EUR Euro: EU currency
F
FBI Fluidised Bed Incinerator
FC Flue Control
FCC Fluidised-bed Catalytic Cracking
FCCU Fluidised-bed Catalytic Cracking Unit
FFU Flocculation/flotation unit
FGD Flue Gas Desulphurisation
FGR Flue Gas Recirculation
FOE Fuel oil equivalent (1 tonne FOE = 4.25·1010 J)
G
GAC Granular activated carbon
GO Gasoil
H
h hour
HC (1) Hydrocarbon
HC (2) Hydrocracker/ing
HCU Hydrocracker Unit
HDM Hydrodemetallisation
HDS Hydrodesulphurisation
HFO Heavy Fuel Oil
HGO Heavy Gasoil
Horizontal BREF BAT reference documents prepared to deal with topics common to more
than one IPPC industrial sector (namely storage, cooling, waste water and
waste gas and monitoring)
HORC Heavy Oil and Residue Catalytic cracking
HP High Pressure
HSE Health, Safey and Environment
HT High Temperature
I
I-TEQDF International toxic equivalent of dioxins/furans
IAF Induced Air Flotation
IFRT Internal Floating Roof Tank
IGCC Integrated Gasification Combined Cycle
IMO International Maritime Organisation
K
k-- kilo
kero kerosene
KjN Kjendahl nitrogen (Measure of the total nitrogen content)
kt kilotonne (106 kg)
L
LCO Light Coking Oil
LCP Large Combustion Plant BREF
LDAR Leakage Detection and Repair Programme (see Section 4.23.6.1)
LGO Light Gasoil
LP Low Pressure
LPG Liquefied Petroleum Gas
LT Low Temperature
LVGO Light Vacuum Gasoil
LVOC Large Volume Organic Chemicals BREF
M
M--- Million
MAH Mono Aromatic Hydrocarbon
MDEA Mono diethatnolamine
MEA Mono ethanolamine
MP Medium Pressure
MS EU Member State
MTBE Methyl Tert-buthyl Ether
Mt/yr Million of tonnes per year
MWe MegaWatts electric (energy)
MWth MegaWatts thermal (energy)
N
n- normal, linear organic compound
N--- Normal (refers to volume of gases under normal conditions (temperature
0 ºC and pressure of 1 atmosphere (101.3 kPa)))
n/a not applicable
n.a. not available (typically for data not available within the document).
Highlighted within the document to help next revisions.
N-Kj Nitrogen analysed by Kjendahl method.
Naphthenic The naphthenes contain one or more saturated rings of 5 or 6 carbon atoms
in their molecules to which paraffinic-type branches are attached.
NOC Non-oil contaminated
NOx Nitrogen Oxides (NO + NO2, normally expressed as NO2)
P
Pa Pascal (N/m2). SI unit for pressure: 1 atm=101.3 kPa
PAC Powdered activated carbon
PAH Polyaromatic Hydrocarbons
PC Pressure Control
PCDD/F Polychlorodibenzoparadioxins/furans
Platformer Type of catalytic reformer
PM Particulate Matter (any finely divided solid or liquid material emitted into
the air)
PM10 Particulate matter of size less than 10 µm
Poly-unit Polymerisation unit
PPI Parallel Plate Interceptor
ppm parts per million (weight)
ppm v/v parts per million in volume
PPS Pressurised cross flow plate separation of the desalting unit
PSA Pressure-swing adsorption used for purification of hydrogen
psi pounds per inch (British pressure unit 1 bar = 14.5 psi)
psia pounds per inch (absolute pressure)
R
RCC Residue Catalytic Cracker
Residue This term is used in this document as the heaviest fractions of the distillation
and conversion units. These fractions are used as feedstock in other
processes or as liquid refinery fuel.
RFG Refinery Fuel Gas
RO Reversed Osmosis
RON Research Octane Number
RSH Mercaptan
RVP Reid Vapour Pressure
S
S Sulphur
SF Sand Filter
SOx Sulphur oxides (SO2 and SO3)
SRU Sulphur Recovery Unit
SS Suspended Solids
SW Sour Water
SWS Sour Water Stripper
T
t Tonne (1000 kg)
t/yr Tonnes per year (annum)
t/d Tonnes per day
TAME Tertiary Amyl Methyl Ether
TCDD/F Tetrachlorodibenzoparadioxins/furans (toxicity reference for dioxins)
TEL Tethra ethyl lead
TGT Tail Gas Treatment of the sulphur recovery unit
TML Tethra methyl lead
TOC Total Organic Carbon
TSS Total Suspended Solids (water)
TWG European Technical Working Group on refineries
U
U Unit (used together with names of processes)
UF Ultrafiltration
Ultraformer Type of reformer
USAEPA Environmental Protection Agency of the USA
V
v/v volume by volume
V.I. Viscosity Index (see Section 2.3)
VBU Visbreaking Unit
W
w/w weight by weight. Ratio expressed in weight
WHB Waste Heat Boiler
WWTP Waste Water Treatment Plant
Y
yr year
10 ANNEXES
Six annexes have been prepared to complement the information provided in the document. The
information provided is the following:
Annex I Environmental legislation and emission limit values applied to refineries.
Annex II Refinery configurations.
Annex III Refinery feedstock, intermediates and products.
Annex IV Submitted examples of cost-effectiveness and implementation of environmental
technologies.
Annex V Background information from TWG members for the different proposals of SO2
and NOx emission under the bubble concept.
Annex VI Proposals from some Member States on the implementation of BAT in the
refinery sector
Refineries generate emission to the air, discharges to water and also waste. Most pollutants are
emitted to streams that are transboundary in nature. As consequence many of the environmental
policies and regulations with respect to oil refining are to a large extent influenced by
international developments (EU-directives and other international regulations). These
regulations focus on product quality, refinery emissions and, more recently, on environmental
reporting. The issue of emergency procedures is normally included in the external safety report
and in the permit for the location.
Table 10.1 gives a summary of some of the main EU legislation that it is applied to refineries.
The specifications for heavy fuel oils are regulated in Council Directive 1999/32/EC relating to
the reduction in the sulphur content of certain liquid fuels and amending 93/12/EC. OJ L121/13,
11 may 1999 EU B. For heating oil, the Directive entails a reduction of the sulphur content to
0.1 % in 2008 and for landtrade fuel oils a limitation of the sulphur content to 1 % from 2003
onwards. For seagoing vessels the regulations of Annex VI of the MARPOL treaty, once
ratified, imply that as of 2003, in so-called SOx emission control areas, the use of fuel oils may
be restricted to fuel oils with a maximum sulphur content of 1.5 %.
Legislation Concerning
National Emission Ceiling Directive Important Directive for the reduction of CO2, SO2,
NOx and VOC
Air Pollution Protocol (Gothenburg Protocol, 1 Important protocol for the reduction of SO2, NOx
December 1999). and VOC
VOC prevention during transfer of liquids When transferring liquids to vessels at atmospheric
pressure, the vapour phase (often air, but also
inerts) in the receiving vessel is often emitted to
atmosphere. Such loading operation is recognised
as having an impact on the environment due to the
presence of VOC. EU Stage 1 Directive 94/63/EC
prescribes emission reduction measures and the
application of Vapour Recovery Units (VRU) or
Vapour Recovery Systems (VRS) to prevent escape
of vapours to the atmosphere.
Air Quality Directive (AQD) The Air Quality Directive gives limit values. PM
emissions has been agreed upon by the Council of
Ministers in second reading and will be published
shortly.
Legislation Concerning
EC Large Combustion Plant Directive New units (ie units which came into existence on
or after 1 July 1987) with individual furnaces
greater than 50MWth or multiple heaters where
aggregate thermal input is greater than 50MW,
come within the scope of the EC Large
Combustion Plant Directive, and must comply with
its requirements. However the Directive does not
cover direct refinery processes, e.g. FCCU
regenerators, coking processes or SRUS, nor gas
turbines. The Directive requires Member States to
introduce programmes for a successive reduction of
sulphur dioxide and oxides of nitrogen emissions
from large combustion plants operating, or
licensed, before 1 July 1987.
Compound Legislation
International Post-Kyoto Protocol UN/Greenhouse gas convention
Carbon dioxide EC Directives 93/76/EEC
93/389/EEC
International EU/UN-ECE
EC Directives Large Combustion Plant Directive (LCPD),
Particulate Integrated Pollution Prevention Control Directive(IPPC)
Air Quality Directives (AQFD)
Other: UN-WHO
USAEPA
International EU/UN-ECE, Acidification Strategy
EC Directives Large Combustion Plant Directive (LCPD),
Sulphur in Liquid Fuels Directive(SLFD)
Integrated Pollution Prevention and Control Directive (IPPC).
Sulphur oxides Air Quality Sulphur protocol (incl. Bunkers)
Directives Air Quality Framework Directive (AQFD)
National Ceilings Directives
Other UN-ECE/- World Health Organisation (WHO)
USAEPA/ 1990 Clean Air Act
International EU/UN-ECE
Acidification Strategy
EC Directives LCPD
IPPC
Nitrogen oxides Air Quality Directives: AQFD
N-protocol (93/361/EEC)
National Ceilings Directives
Other UN-ECE/-WHO
USAEPA/ 1990 Clean Air Act
International EU/UN-ECE
UN-ECE/VOC Protocol Stage I (Marine VR), Stage II
MARPOL Convention Annex VI
IMO-MARPOL 73/78
EC Directives Air Quality Directives AQFD
Volatile organic 94/63/EC
compounds National Ceiling Directives
Ozone Strategy
Other: UN-WHO
USAEPA/1990
Clean Air Act
Draft EC-Long range transboundary air pollution
Compound Legislation
International UN/Greenhouse gas convention
Carbon monoxide EC directives IPPC
Other UN-WHO
Methane International Post-Kyoto Protocol UN/Greenhouse gas convention
EC directive (93/389/EEC)
Halon and CFC International UN Montreal Protocol (94/84/EC)
Nickel International UN-ECE protocols on Heavy Metal
EC directives Air Quality Directives, AQFD
IPPC
Benzene EC directives Air Quality Directives
AQFD
IPPC
PAH EC directives Air Quality Directives
AQFD
IPPC
Others UN-ECE POP protocol
There is an OSPARCOM Decision (89/5) which sets an oil limit for refinery discharges.
External safety
On the basis of the post-Seveso guideline of the EU (82/501) refineries are nowadays obliged to
prepare an external safety report. The Seveso-1 guideline was implemented in 1999, and
involves further requirements on safety reporting. It applies to the refinery as a whole.
Compound
The EC directive on large combustion plants regulating the sulphur content in liquid
fuels sets an overall limit of 1700 mg/Nm3 (bubble) for existing plants in the refinery.
For new refinery plants the directive on the reduction of emissions from large
SO2 combustion plants limits the SO2 emission to 1000 mg/Nm3 flue gas. Moreover, the
proposed revision of this directive includes a proposal to lower the SO2 emission limit
for new refinery installations to 450 mg/Nm3.
Particulates The EC directive for large combustion plants applies a limit to total emissions of
particulate matter (PM) of 50 mg/Nm3.
The remaining tables in this section give a brief summary of the environmental legislation
and/or emission limit values affecting refineries in some of the EU+ countries.
Austria
Belgium
Air (Vlarem), water and ground water (Vlarem), waste and soil (Vlarem, Vlarebo and Vlaria)
and noise (Vlarem).
Cr 0.5 mg Cr/l
P 2 mg P/l
Pb 0.05 mg Pb/l
Air SO2 1300 mg/Nm³ bubble emissions for total of fuel and
(ELV at process emissions, including
3% O2) emissions from cogeneration plants;
ELV and scope currently under
revision
1700 mg/Nm³ new large combustion installations,
fuel oil, 50 to 300 MWth
1700 to 400 mg/Nm³ new large combustion installations,
fuel oil, 300 to 500 MWth
400 mg/Nm³ new large combustion installations,
fuel oil, >= 500 MWth
35 mg/Nm³ new large combustion installations,
fuel gas
NOx 450 mg/Nm³ bubble emissions for total of fuel and
process emissions, including
emissions from cogeneration plants;
ELV and scope currently under
revision
450 mg/Nm³ new large combustion installations,
fuel oil
350 mg/Nm³ new large combustion installations,
fuel gas
dust 150 mg/Nm³ bubble emissions for total of fuel and
process emissions, including
emissions of cogeneration plants
50 mg/Nm³ new large combustion installations,
fuel oil
5 mg/Nm³ new large combustion installations,
fuel gas
50 mg/Nm³ fluid bed catalytic cracking
regeneration (before 01/01/2005: 300
mg/Nm³)
CO 150 mg/Nm³ bubble emissions for total of fuel and
process emissions, including
emissions of cogeneration plants
Ni 2 mg/Nm³ bubble emissions for total of fuel and
process emissions, including
emissions of cogeneration plants
Denmark
Denmark has set SO2 emission limit values of 1000 mg/Nm3 for residual product fuels, with a
maximum of 1 % S in the fuel oil. SO2 limits for gaseous and LPG fuels have been set at much
lower levels (35 mg/Nm3 to 5 mg/Nm3). NOx levels have been set at 225 mg/Nm3 for both
liquid and gaseous fuels.
Finland
Table 10.7: Summary of environmental legislation and emission limit values in Finland applied to
refineries
Note: VNp = Decision of Council of State
France
Air
The maximum allowable SO2 refinery emission concentration (as bubble) permitted since 2000
is 1700 mg/Nm3.
Water emissions.
According to the different refinery categories, the mass flow limits per tonne of product
processed (crude, residues, etc…) are the following:
Type of refinery 1 2 3
Mass flow limit (monthly
average)
Water flow (in m3/t) 0.25 0.65 1
Total suspended solids (in g/t) 6 15 25
COD (in g/t) 25 65 100
BOD5 (in g/t) 6 15 25
Total nitrogen (in g/t) 5 12.5 20
Hydrocarbons (in g/t) 1.2 3 4
Phenols (in g/t) 0.06 0.15 0.25
Formula for the noise level requirement at the plant limit: 45dBA + ct (takes into account the
different periods of the day) + cz (take into account the type of area)
Germany
Important regulations relevant for industrial installations in Germany are laid down in the
Federal Immission Control Act [Bundes-Immissionsschutzgesetz - BImSchG], the Federal
Water Act [Wasserhaushaltsgesetz - WHG] and the Federal Recycling and Waste Management
Act [Kreislaufwirtschafts- und Abfallgesetz - KrW-/AbfG]. Germany uses a segregated media
permitting system for different environmental media, but the final decision on an application is
reached by the assessment of environmental impacts on all media by the local authorities. Also
noise requirements are considered in the licensing procedure, whereby Germany aims at
favouring pollution prevention. The “precautionary principle” has a legal status which permits
the settings of standards. Legal standards are not subject to any negotiation in the licensing
process in Germany. The cited regulations are currently amended according to the IPPC
Directive.
In compliance with the federal structure of Germany, the implementation of environmental laws
and decrees is under the responsibility of the federal states (Bundesländer), which may
implement the administrative procedure differently. For new plants, that are regarded as relevant
with respect to emissions and releases into the environment, also an environmental impact
assessment is required during the licensing procedure (cf. Gesetz über die
Umweltverträglichkeitsprüfung [UVPG]).
Table 10.9 gives an overview of the current German legal basis and regulations for
environmental protection in German industry alongside the process chain. In the following, the
most important acts and regulations are presented.
Table 10.9 Legal basis and regulations alongside the process chain
The emitted substances are split into 3 classes (vaporous or gaseous inorganic substances: 4
classes), whereby class I substances are the most toxic and Class III or IV are the least harmful.
The emission limit values contained in the TA Luft represent the State of the art for technical
measures for reducing emissions (last amended in 1986, currently being amended). These
quality values were developed referring to scientific findings and research taking into account
toxicological, bioaccumulative and epidemiological aspects. The required concentration limits
are given in mass of emitted substances related to the volume of emitted gas under standard
conditions (0°C, 1013 mbar) after subtraction of the water vapour content.
If inorganic dust particles of several classes are present, the mass concentration in the emitted
gas should not exceed a total of 1 mg/m3 for coinciding class I and II substances as well as a
total of 5 mg/m3 for coinciding class I and III or class II and III substances.
If organic substances in several classes are present, the mass concentration in the emitted gas
should not exceed a total of 0.15 g/m3 with a total mass flow of 3 kg/h or more.
No daily mean value of the respective emitted substances should exceed the required emission
limits, 97 % of all half-hourly means should not exceed six fifths of the required emission
limits, and all half-hourly means should not exceed the required emission limits by more than
twice.
Because state of the art technology significantly proceeded since 1986, the local authorities in
some cases demand stricter emission limit values than laid down in the TA Luft.
Issued by the Federal Government in consent with the Federal States (Länder), these minimum
requirements are binding for the authorities responsible for licensing and governmental control
of the discharges. Depending on the local conditions, even more stringent requirements can be
established. The minimum requirements are based on the ‘emission principle’ and the
precautionary principle, i.e. application of stringent, technically derived emission standards
irrespective of the loading reserves of the receiving water bodies resources or the potential
effects of the various substances discharged. Additionally, the Federal Ministry of Environment
publishes explanations and comments on the Waste Water Ordinance. Refining of petroleum is
dealt with in Annex 45 of this regulation. Table 10.11 gives the main restrictions established in
Annex 45 of the AbwV.
Parameter *)
Limit value
[mg/l]
Phenol index after distillation and dye extraction 0.15**)
Total hydrocarbons 2
Total phosphorus 1.5
Cyanide, easily released 0.1**)
Sulphide sulphur and mercaptan sulphur 0.6**)
AOX 0.1**)
5-day biochemical oxygen demand 25
COD 80
Total nitrogen as the sum of ammonia, nitrite and nitrate nitrogen 40
*)Qualified random sample or 2 hours composite sample
**) Requirements apply to the waste water prior to blending with other waste water
Source: AbwV, Annex 45 „Petroleum Refining“
Table 10.11 Limit values for discharges into water for petroleum refining plants
The required loads shall be production-specific loads in relation to the production capacity on
which the water discharge licence is based. This should be determined by a qualified random
sample or a 2 hour composite sample.
A qualified random sample should refer to a composite sample of at least five random samples
taken over a maximum period of two hours at intervals of no less than two minutes, which are
then blended. A composite sample should refer to a sample which is taken continuously over a
given period, or a sample consisting of several samples taken either continuously or
continuously over a given period and then blended. A random sample should refer to a single
sample taken from a waste water flow.
Waste water from cooling systems for the indirect cooling of industrial processes is excluded
from this regulation. Indirect cooling water is subject to the provisions laid down in Annex 31,
AbwV. Table 10.12 gives the relevant requirements for discharges from cooling circuits in
Annex 31. If the stated values are not reached, approval for the discharge of waste water will be
denied.
Parameter *)
Minimum Requirements
COD 40 mg/l
Phosphor compounds, given as P 3 mg/l
Zinc 4 mg/l
AOX 0.15 mg/l
Available residual chlorine 0.3 mg/l
Chromium compounds must not be contained
Mercury compounds must not be contained
Nitrite must not be contained
Metal organic Compounds (Metal-Carbon-Compound) must not be contained
*)Qualified random sample or 2 hours composite sample
Source: Annex 31, Abwasserverordnung
Table 10.12: Requirements for discharges from cooling circuits of industrial processes
German regulations concerning the waste management and disposal of hazardous materials
Regulations concerning the waste management and disposal of hazardous materials are laid
down in the Federal Recycling and Waste Management Act [Kreislaufwirtschafts- und
Abfallgesetz - KrW-/AbfG] and the Federal Immission Control Act [BImSchG]. According to
the BImSchG, an operator of a waste generating installation which is subject to licensing, is
obliged to avoid waste production or to make sure that the waste is environmentally soundly
recovered. If that is technically or economically not reasonable, the waste has to be disposed of
without any harmful effects. The federal States’ working group on immission control
(Länderausschuss für Immissionsschutz, LAI) has published examplary administrative
regulations for particular industrial sectors containing measures for avoidance and recovery of
wastes which are considered as technically and economically reasonable.
The KrW/AbfG states that installations generating more than 2 tonnes of hazardous waste or
more than 2000 tonnes of non-hazardous waste (per waste key), must produce a waste
management concept and yearly waste balances.
Ireland
Table 10.14: Emission limit values for air emissions applied in Ireland
Table 10.15: Emission limit values for discharges to water applied in Ireland
Italy
The maximum allowable SO2 refinery emission concentration (as bubble) permitted in 2000 will
be 1700 mg/Nm3.
Norway
Emission standards are not generally applied in Norway. The emission limits for the refineries
are set according to each case. Norway has introduced a special tax duty to minimise emissions
of CO2 and SO2. Heavy fuel oil with more than 1.0 % sulphur is prohibited in the twelve
southern and southwestern counties of Norway. In the northern parts of Norway heavy fuel oil
is allowed with a maximum 2.5 % sulphur. But due to high sulphur taxes, heavy fuel oil with
up to 2.5 % sulphur content is more expensive than the 1.0 % oil, and is therefore hardly used.
Noise is regulated to a limit of 45 dB(A).
TOC 1) - tonnes/year 1)
Metals 1) - kg/day 1)
P (total) 1) - tonnes/year 1)
NMVOC 1) - tonnes/year 1)
Methane 1) - tonnes/year 1)
The permit values vary from refinery to refinery due to design, age and local factors. The
permits are planned to be under revision in 2001 and 2002.
Sweden
Air
The maximum allowable SO2 refinery emission concentrations (as bubble) permitted in 2000
will be 800 mg/Nm3. Use of high sulphur fuels (>0.5 %) is prohibited throughout the country.
Sulphur recovery must be in excess of 99 % including the flaring of residual tail gas. NOx and
particulates must be reduced to the minimum feasible level.
The total permitted release levels for the Scanraff Refinery are 2000 t/yr SO2 and 1000 t/yr NOx.
The absolute limit of particulates from the FCCU is 75 mg/Nm3 (the crude capacity of the
refinery is 10 Mt/yr))
The Netherlands
A characteristic of the Dutch regulations and permitting is “self-regulation within boundaries”.
This characteristic stems from a broad support for achieving consensus, and includes as well the
environmental policy, both in the way it was brought into existence and in the way it has been
implemented.
Air emissions
Regulation Affecting
Combustion Plant Emission Limits Decree (BEES) SO2, NOx, particulates
Emission Guidelines for Air (NeR) Sulphur Recovery Plants
Hydrocarbons 2000 programme (KWS2000) VOC
The National Environmental Policy Plan (NEPP, 1997) states as a policy objective that by 2010
the refineries will be gas-fired, or that installations that are still oil-fired will be equipped with
FGD, SCR and ESP. The KWS2000 measures comprise preventive maintenance and spills and
leaks abatement. Furthermore, vapour recovery for inland barge loading operations of gasoline
and double mechanical seals on floating roof tanks are specified as well as improved sealings of
pumps, compressors, valves and flanges and covering of open basins. Measures for vapour
recovery are also applicable for crude loading and unloading.
Table 10.16 summaries the emission limit values applied to different air contaminants.
Permit values vary from location to location due to differences in design, age and other local
factors. The authority aims at uniformity for future permit revisions. A number of permit
parameters and typical compliance values for Dutch refineries are (1997): Oil 140 t/yr, Benzene
9 t/yr, Phenol 5 t/yr and N-total 540 t/yr. The parameters and compliance values refer to typical
refinery effluent composition (ranges), having passed primary effluent treatment facilities but
prior to secondary and/or tertiary treatments.
Waste
Hazardous waste in refineries mainly involves the management of oily sludges, spent catalysts
and spent caustic. For these waste categories increased use is made of third-party waste
contractors for off-site recovery or disposal. The Environmental Management Act (Wm)
governs legal disposal and two EU-Directives cover the classification, packing and labelling of
hazardous substances. As regards the export of hazardous waste a similar procedure exists as
laid down in the Basle Convention on the Control of Transboundary Movements of Hazardous
Waste and their Disposal.
Noise
The Netherlands has an integrated policy for noise, and aims for a proper inventory of noise
sources and a planned abatement programme. Germany, Switzerland, Austria, France and the
Scandinavian countries are developing similar policies on noise.
Nowadays noise level contour values and monitoring obligations are included in permit
requirements of the refineries. Typical for industrial noise requirements in urban areas in most
countries are maximum allowable limits between 55 and 73 dB(A) during daytime and 45-66
dB(A) during the night.
United Kingdom
The UK does not have standard emission limit values for refineries, but their guidance notes
provide expected BAT levels for new plant units. Existing unit operations should meet
BATNEEC standarts (Best Available Techniques Not Entailing Excessive Cost).
Japan
There are three regulation levels: the State, the Regional Prefecture and the municipality.
Japanish refineries are forced to use as liquid fuel an industrial gas oil (so called A-fuel)
containing less than 0.1 % S.
State regulations on NOx are: Refinery furnaces, 100 ppm; boilers, 150 ppm; hydrogen
production unit, 150 ppm; and FCC, 250 ppm. Municipality emission limit value for NOx is
85 Nm3/h (equal to 1396 t/yr) [248, Ademe, 2001]
USA
Air
The US Environmental Protection Agency has set a limit for particulate matter in the FCC
generator exhaust gas of 1.0 kg/1000 kg of coke burnt off in the catalyst regenerator (equates
approximately to 75 mg/Nm3). The coke burn-off rate is determined from the volumetric flow
rate of regenerator exhaust gas and the content of CO2, CO and O2. The SO2 emission limit for
New Source Performance Standards is 700 mg/Nm3.
Water
Strict limits apply to toxic and conventional pollutants for waste water, BOD, COD, TOC,
ammonia, sulphides and phenolics.
World Bank
The following guidelines present emission levels normally acceptable to the World Bank Group
in making decisions regarding provision of World Bank Group assistance; any deviations from
these levels must be described in the World Bank Group project documentation. The guidelines
are expressed as concentrations to facilitate monitoring. Dilution of air emissions or effluents to
achieve these guidelines is unacceptable. All the maximum levels should be achieved for at least
95 % of the time that the plant or unit is operating, to be calculated as a proportion of annual
operating hours. [101, World Bank, 1998]
Air Emissions
Parameter Maximum value (mg/Nm3)
Particulate matter (PM) 50
Nitrogen oxide (NOx)* (as NO2) 460
Sulphur oxide (SOx) (as SO2) 150 for sulphur recovery units and 500 for other units
Nickel and vanadium (combined) 2
Hydrogen sulphide 15
* Excluding NOx emissions from catalytic units.
Liquid effluents
Parameter Maximum value milligrams per litre (mg/l)
pH 6-9
BOD5 30
COD 150
Total suspended solids 30
Oil and grease 10
Chromium (hexavalent) 0.1
Chromium (total) 0.5
Lead 0.1
Phenol 0.5
Benzene 0.05
Benzo(a)pyrene 0.05
Sulphide 1
Nitrogen (total) (1) 10
∆T (ºC) Less than or equal to 3 (2)
(1) The maximum effluent concentration of nitrogen (total) may be up to 40 mg/l in processes that
include hydrogenation
(2) The effluent should result in a temperature increase of no more than 3 ºC at the edge of the
zone where initial mixing and dilution takes place. Where the zone is not defined, use 100
meters from the point of discharge provided there are no sensitive ecosystems within this
range.
Note: Effluent requirements are for direct discharge to surface waters. Discharge to an off-site
waste water treatment plant should meet applicable pretreatment requirements.
Ambient Noise
Noise abatement measures should achieve either the following levels or a maximum increase in
background levels of 3 dB(A). Measurements are to be taken at noise receptors located outside
the project property boundary.
The emission requirements given here can be consistently achieved by well designed, well
operated and well maintained pollution control systems.
In this section four of the most common refinery schemes or configurations will be discussed in
detail. The classification of refineries into these configurations is slightly arbitrary. Other
approaches do exist, but the main purpose is to illustrate the great diversity of refinery types in
existence.
The overall block flow diagrams of the above four configurations are shown in Figure 10.1 to
Figure 10.4.
Another approach is to express the conversion (of residual material into distillates) of each
conversion unit into its “catcracking equivalency”, and than to define the refinery complexity as
the sum of all the catcracking equivalencies. In comparing refineries, both the configuration
(which units are in) and the complexity (what is the capacity of these units) are important to
consider.
Energy requirements will generally increase from configuration 1 to 4 and with complexity.
Similar generalisations can be made regarding e.g. process water requirements, sulphur
production (at same sulphur intake) etc., but qualifications often need to be made. In the
schemes that represent the configurations in the following pages, processing units are
abbreviated; for more details on these process units, see the Glossary.
This is the simplest type of refinery, producing fuels by straightforward operations (see Figure
10.1). This type of refinery has a very rigid product distribution pattern; the produced fuels are
almost entirely fixed by the type of crude being processed. The production cannot be greatly
influenced by changing the operating modes of the various processing units.
Many of the Hydroskimming refineries were built in the fifties and sixties when the demand for
all fuels increased significantly, the cost of crude was relatively low and the demand for heavy
fuel oil was relatively high. In the seventies and eighties the majority of the Hydroskimming
refineries were expanded to include a cracking complex, but quite a lot of Hydroskimming
refineries dating from that period are still in operation. In the Crude Oil Distillation Unit the
crude oil is fractionated into straight-run naphtha overhead product, kerosene, gas oil side-
stream products and atmospheric residue product from the bottom of the column. The straight-
run unstabilised naphtha is passed to the Naphtha Hydrotreating Unit to make it suitable for
Catalytic Reforming. The hydrotreated naphtha stream is split into a light and a heavy naphtha
fraction.
PROPANE 1 1 NAPHTHA
REFINERY CRACKER
FUEL GAS GAS 2 FEEDSTOCK
2 PLANT BUTANE 3 1
LIGHT NAPHTHA 4
3 C5/C6
3 C3/C4 5 ISOMERATE
1
7 3 C3/C4 ISOMERI-
STABILIZED SATION 1
STRAIGHT RUN HEAVY
NAPHTHA MINUS NAPHTHA NAPHTHA NAPHTHA NAPHTHA CAT. 5 MOGAS
HDS
27 SPLITTER
20 REFORM. 17 REFORMATE 17 POOL
30 24
40
REFINERY FUEL OIL
1 2
FUEL
OIL
39 POOL
41
In the past part of the light naphtha fraction was used as mogas (motor gasoline) blend
component, and the balance was sold as feedstock to Naphtha Crackers. More recently, as a
result of the lead (TEL) phase-out programme, many Isomerisation Units are included in the
Hydroskimming refineries to isomerise the light naphtha stream. The isomerate produced has a
roughly 20 - 25 higher octane number than its feed and compensates for the loss of octane
resulting from the lead phase-out. This avoids the complete light naphtha stream being sold as
low value feedstock to Naphtha Crackers. The heavy naphtha fraction from the Naphtha
Hydrotreater is upgraded in the Catalytic Reformer to a high-octane gasoline-blending
component. The hydrogen from the Catalytic Reformer is used to desulphurise the gasoil and
naphtha. Normally two grades of gasoline are produced in a refinery, regular and premium
gasoline. These grades of gasoline are produced by blending different ratios of isomerate,
reformate and butane. Saturated light hydrocarbons from the CDU Naphtha Hydrotreater,
Isomerisation Unit and Catalytic Reformer Unit are sent to the Gas Plant. Propane from the Gas
Plant is either sold directly as propane product or blended with butane to be sold as LPG. A
proportion of the butanes produced in the Gas Plant is used as blend stock for the mogas pool,
the remainder of the butanes are either blended with propane to be sold as LPG or directly sold
as butane product. The components lighter than propane are sent to the refinery fuel gas system.
Straight-run kerosene is often hydrotreated in the Kero Hydrotreating Unit (or kero HDS) and
run down as jet fuel. Normally a portion of the hydrotreated kerosene is used as blending
component for the automotive diesel oil pool.
Straight-run gas oil is hydrotreated in the Gas Oil Hydrotreating Unit to reduce its sulphur
content. Automotive diesel oil is produced by blending desulphurised kerosene with
desulphurised gas oil. Heating oil and marine diesel oil (MDO) are produced by blending
desulphurised gasoil with straight-run gasoil. Both products have higher maximum allowable
sulphur contents than diesel oil. Heavy fuel oil is produced from atmospheric residue with
adjustment for viscosity and sulphur content by addition of desulphurised gas oil. In some cases
the atmospheric residue is thermally cracked. This is not indicated in the configuration scheme.
The sour off-gases from the Naphtha Hydrotreater, the Kerosene and Gas Oil Hydrotreating
units (or Gasoil HDS) are treated in the Amine Treating Unit to remove H2S and other sour
components present, prior to being sent to the refinery fuel gas system. The sour LPG stream
from the Naphtha Hydrotreater is also treated in the Amine Unit to reduce its sulphur content
before combining with the sweet LPG stream from the Catalytic Reformers. Sour waste water
from all process units is stripped, prior to utilisation as wash water in the Desalter and final
purification in Effluent Water Treating facilities. The sour off-gas from the Sour Water Stripper
is combined with the H2S rich-off gas from the Amine Unit and sent to a Sulphur Recovery Unit
(SRU), where the H2S is converted to elemental sulphur, and a nearly sulphur-free off-gas is
vented to the atmosphere.
30 ISOMERI-
3
2 2 36
SATION 1
6
iC4 C4
HEAVY CC
1 ALKYLA-
TION
NAPHTHA 3
CDU
KERO
2 KERO 22 MeOH MTBE 20 KERO
CRUDE OIL MTBE
HDS POOL
20 1 22
100 2
GASOIL 6 2
10
GASOIL 20 MEROX MEROX
GASOIL
HDS
40 VGO
5 LIGHT CC 4 4 18 POOL
20
5 NAPHTHA
REFINERY C4
1 FUEL GAS
2 TO KERO HDS
FCCU NAPHTHA
2 2
VD
FCCU GAS
NAPHTHA 4 MEROX
15 SPLITTER HEAVY CC
HVU 15 PLANT 2
LCO NAPHTHA
4 1
2 FUEL
1 VBLGO 2 OIL
VR
VBHGO POOL
VBU
20 14 VB RESIDUE
11 16
REFINERY
2 3 3
FUEL OIL
5
Many Catcracker refineries in Europe include a Visbreaker Unit to reduce the heavy fuel oil
production. If no VBU is applied, significant quantities of high value gasoil components will
have to be blended to the vacuum residue product, mainly to meet the viscosity specifications of
heavy fuel oil. The VBU converts the vacuum residue into a residue which has a much lower
viscosity and also produces some naphtha and gas. This type of refinery has a flexible product
distribution pattern. The fuel slate can be significantly influenced by changing the operating
modes of the various processing units and by the blending of products. This type of refinery
configuration is predominant in Europe. Most of these FCC refineries were built in the fifties
and sixties as a simple Hydroskimming refinery. In the seventies and eighties, a Catcracker
complex was added to these refineries.
Scheme 3 is a Hydroskimming refinery expanded with an HVU, Hydrocracker Unit (HCU) and
e.g. a Delayed Coker Unit (DCU, see Figure 10.3.). In this configuration an even larger part of
the atmospheric residue is converted into lighter fuel components together with reduced
production of residual fuel oil. The addition of a Coker allows this refinery to eliminate the
production of residual fuel oil completely. HCU units are specifically used to maximise the
production of gasoline and middle distillates. A limited number of Hydrocracker refineries in
Europe include a Delayed Coker Unit (DCU) to reduce heavy fuel production and maximise
light fuels production. The DCU converts heavy residues in lighter hydrocarbon fractions and
petroleum coke. Dependent on the quality, coke is used in the cement and steel or in the
aluminium industry. The hydrogen necessary for the HCU is supplied by a Hydrogen Plant. A
Catalytic Reformer does not normally produce enough hydrogen for the HCU. Alternatively, the
supplemental hydrogen requirement can be provided via partial oxidation of heavy
hydrocarbons followed by hydrogen separation.
NAPHTHA MINUS
7
CDU
20 LIGHT NAPHTHA
KERO KERO KERO
CRUDE OIL HDS NAT.GAS
20 POOL
20
100 25
C3/C4 2
GASOIL
10
GASOIL HYDROGEN 28 GASOIL
40 HDS PLANT 20
5 20 20 POOL
22
5 8
H2
3
8
VD 10 HYDRO KERO 7 FUEL
CRACKER OIL
HVU GASOIL
BLEED POOL
8 2 10
5
NAPHTHA
2 MINUS
25 VR LCGO
DELAYED REFINERY
20 COKER HCGO 8 FUEL OIL 3
5 5 COKE
This type of refinery has a higher degree of flexibility with respect to either maximum gasoline
or maximum middle distillate production, whereas a FCC configuration is primarily focused on
increased gasoline production. About 15 % of the existing refinery complexes in Europe have
already been extended with a Hydrocracker. It should be noted that such extensions require a
relatively high capital investment and high energy consumption compared to the installation of a
Catcracker.
Scheme 4 is a refinery with a Hydrocracker and additional “deep conversion”. With a Residue
Hydrocracker and/or a Gasification Unit (IGCC), so that even more conversion capacity for
heavier feed stocks is introduced (see Figure 10.4). This is an example of a large complex
refinery, and it includes some options to produce high value petrochemical feedstocks. This
extended range of heavy residue conversion techniques can in principle eliminate the need for
mixing heavy residue to the fuel oil pool or can process more heavy crude oils. The Residue
Hydrocracker, i.e. Hycon, H-oil Unit or other types of Residue Hydrocrackers, converts vacuum
residues to high-value transportation fuels e.g. gasolines, kerosene and diesel. The addition of a
Residue Hydrocracker to a refinery provides considerable flexibility and increases production of
gasoline and middle distillate fuels at the expense of fuel oil. The IGCC converts the visbreaker
residue to power, steam, hydrogen, and some waste streams. The main advantage of this
arrangement is the production of hydrogen by use of low-value vacuum residue instead of the
more valuable light naphtha, other light hydrocarbon streams or natural gas. Moreover, the fuel
oil production is reduced and the production of LPG and diesel fuels is increased.
C3 EX FCCU 1 NAPHTHA
REFINERY PROPANE 2 CRACKER
FUEL GAS SATURATED
GAS
BUTANE 3 FEED
3 PLANT 1 STOCK 6
7
LIGHT NAPHTHA 2
5 C5/C6
3 ISOMERI- ISOMERATE 5 1
STRAIGHT RUN C3/C4 5 C3/C4 5 SATION
NAPHTHA MINUS
STABILIZED HEAVY
20 NAPHTHA 7 REFORMATE
MOGAS
1 NAPHTHA NAPHTHA NAPHTHA CAT.
VBU HDS SPLITTER REFORM. ALKYLATE POOL
NAPHTA 19 12
1 C4
1 20
NAPHTHA MINUS ISOMERI- 1 CC NAPHTHA
SATION
4
2
iC4 1 C4
1 ALKYLA-
HEAVY CC TION
NAPHTHA
CDU
KERO 2 KERO 27 MeOH MTBE
25 KERO
CRUDE OIL HDS
MTBE 27 POOL
100 25 1 27
2 2
2
1 2
GASOIL 2 2
H2
15 LIGHT CC
LVGO GASOIL 22 KERO NAPHTHA LCC NAPHTHA
HYDRO
MEROX MEROX
22 GASOIL
5 VBLGO 2 HDS 5 CRACKER GASOIL 3 20 POOL
FROM 2 24
HVGO C3 TO C4
VBU
5 BLEED 1 PETROCHEM.
2
FEEDSTOCK TO KERO HDS 1
40 2 2
FCCU HCC NAPHTHA
NAPHTHA NAPHTHA 4 2
HVU FCCU GAS MEROX 1
12 PLANT 2 SPLITTER HEAVY CC
LCO 2 NAPHTHA 2
20 HVGO 10 2 FUEL
5 OIL
VBLGO 3 1 POOL
TO GASOIL HDS STEAM
VR RESIDUE NAPHTHA 2
15 8 VBU TO NAPHTHA HDS IGCC
POWER 8 7
HYDRO VBLGO 3
H2 CRACKER 6 SOOT TO SARU
13
VB RESIDUE 6 SLAG
5 1
H2 TO HYDRO
CRACKERS REFINERY
FUEL OIL
The visbroken residue is sent to the IGCC where it is converted into syngas (H2/CO). The
produced hydrogen can partly be used in the Hydrocracker and the Residue Hydrocracker.
Remaining syngas may be combusted in a gas turbine followed by a steam turbine for electricity
production and a boiler for steam production. This can be an energy efficient alternative to the
classic heavy fuel-fired boilers. Waste streams from the IGCC are slag and soot recovered from
the soot rework unit. The quantity and quality of the soot depends on the visbroken residue feed
quality and ultimately on the crude oil processed.
The hydrocarbons in crude oil are a mixture of three chemical groups: paraffins (straight and
branched chains are called normal- and iso-paraffins), naphthenes (saturated rings or
cycloparaffins) and aromatics (one or more unsaturated rings). The most used rough distinction
between crude oil types is sweet or sour. Sweet crude is normally low in sulphur and lightly
paraffinic. Sour crude is usually high in sulphur and heavily naphthenic.
The composition of the crude is the most important parameter in establishing the range and
quality of products that may be produced from a refinery. The impurities of the crude, which
usually make up 1 - 5 % of the total, are also very important in establishing the value of the
crude and the difficulties in converting it into marketable products. The most important impurity
of crude oil is sulphur, which is present largely in the form of organic compounds such as
mercaptans and sulphides. Some elemental sulphur, H2S, and FeS may also be present but only
in small quantities. The total sulphur content may be as low as 0.04 % w/w or as high as
5 % w/w. Crudes containing more than 0.5 % w/w S are commonly referred to as ‘sour’ and the
others as being ‘sweet’. In general, the sulphur content increases in the higher boiling fractions.
Examples of crude types processed in European refineries are shown in Table 10.17. A distinct
shift in crude oils processed in European refineries took place between 1993 and 1997, from
Middle East crude oils to “North Sea” crudes. Market forces mainly caused this and this trend
could therefore easily reverse. CONCAWE has reported that the average sulphur content
processed in European refineries decreased from 1.45 % in 1979 to approx. 1.0 - 1.1 % from
1985 onwards. The availability and marketing of low sulphur crude oils has helped the
European refineries to reduce SO2 emissions and to some extent also the emissions of NOx.
The next two tables show the metal content of crude oil. Nickel and vanadium are present as
porphyrin-systems in the crude oil. In order to determine heavy metals in crude, sampling is of
crucial importance. More information about that can be found in [43, Dekkers and Daane,
1999].
A recent report [43, Dekkers and Daane, 1999] has shown that the quantity of Cadmium, Zinc,
Chromium, Copper and Arsenic in the crudes are indeed much lower than traditionally assumed.
Reasons for this include improper analytical techniques and contamination during sampling.
The following table shows the results achieved in the above-mentioned report.
Table 10.18: Content of certain metals in some crude compared to other published data
Crude oil is a mixture of many compounds and, apart from the lightest components, the
characterisation by a complete chemical analysis is almost impossible. As a result, the general
physical properties are expressed in terms of simple parameters, which can be quickly
estimated. The underlying nature of the crude is then derived from this information by
comparison with corresponding parameters from known feedstocks. Most of these parameters
are not limited to crudes, but may be used for most petroleum products as well. One of the
simplest crude properties to measure is the specific gravity, normally expressed in terms of the
°API. Crudes are often classified as being ‘heavy’ or ‘light’ according to their °API, the heavy
being less than about 28 - 32°API and the light over 32 – 35 °API.
There are over a hundred of refinery products, including gaseous and liquid refinery fuels for
domestic and industrial use, fuel for most forms of transport, lubricating oils for all types of
machinery and the basic raw materials for the petrochemical industry. Individual refineries do
not usually supply the full range of possible products, but try to select those most suited to the
quality of the crude feed, the available processing equipment and the local market requirements.
The composition of a product will vary between refineries, being a function of crude type and
refining processes used, but the overall quality will meet the product specifications prevailing
for a given market. The major refinery product groups are shown in Table 2.1 by refinery
process unit. A short description of the product groups follows.
LPG
Sometimes C3 and C4 components are also used for refinery fuel, but most of the C3 and C4
components will be sold as liquefied petroleum gas (LPG), which is used as a fuel for numerous
applications. LPG is sold either as a mixed LPG fraction or as separated C3 and C4 fractions.
The liquid propane normally contains a minimum of 95 %w/w C3 compounds, the remainder
being C2 and C4. Butanes and butenes are also used as a petrochemical feedstock for the
manufacture of MTBE, acetic acid, solvents, polybutylenes, and rubber.
Gasoline
Gasoline, the fuel for motor cars and light aircraft, represents the highest volume and one of the
more valuable refinery products. Automotive gasoline, which is by far the most important
gasoline type, consists of a complex mixture of hydrocarbons ranging from C4 to C10, boiling
468 Mineral Oil and Gas Refineries
Annexes Chapter 10
between 38°C and 205°C. Most refiners produce gasoline in three or four octane grades, the
principal difference being the anti-knock performance. Mogas is a term normally used for motor
gasoline.
Fuel oils
Fuel oils cover a very wide range of applications and are produced in a variety of grades.
Gasoils are sometimes referred to as fuel oils, but in Europe the term is usually used to describe
heavy fuel oils (atmospheric residue) which are used for electricity etc. Heavy fuel oil are used
for electricity or power generation by utility companies, or sold as bunker fuel oil for seagoing
vessels. Refineries use heavier fractions from crude oil distillation, from vacuum distillation or
from visbreaking for generation of heat, electricity and steam in their furnaces and boilers.
Heavy fuel oils consist largely of the residue remaining from the distillation of crudes. These
residues consist of large hydrocarbon structures containing components that require additional
processing to convert them into more valuable, lighter products for use in gasoline and diesel
engines. These residues generally have a high sulphur and ash content, are mostly very viscous
and will therefore be blended with lighter gasoils in order to sell them as a commercial fuel for
ships and utilities.
Bitumen
Bitumen (referred to as asphalt in the USA) is used mainly for paving roads but also for a
number of other applications, such as materials for roofing. The basic source of bitumen is the
residue remaining after vacuum distillation of crude. The residue can also be treated by blowing
with air at high temperature (bitumen blowing process) to increase the asphaltene content.
Petroleum coke
Petroleum coke is the residue remaining after destructive distillation of petroleum residue in a
delayed coking unit. It is widely used as fuel in the cement and steel industry. It can also be
used as a fuel for power plants if the sulphur content is low enough. Coke also has nonfuel
applications as a raw material for many carbon and graphite products.
Sulphur
Sulphur is basically a by-product of refining resulting from the removal of sulphur compounds
from the main hydrocarbon product streams. The quantity of sulphur produced by a refinery
depends on the amount present in the crude and the desulphurisation and sulphur recovery
capacity installed. It is a valuable product which is mainly used as the raw material for sulphuric
acid manufacture.
In general, it is found that the heavier the crude, the higher the quantities of sulphur, Conradson
carbon and heavy metals, and the lower its hydrogen content. Heavy crudes are also more
difficult to process. Heavy crude normally results in high fuel-oil production while light crude is
more appropriate for a high gasoline and middle distillate production. Other simple property
indicators are sulphur content and pour point.
In order to evaluate the properties of the various likely products such as gasoline octane rating,
the aromatics content of kerosene, etc., it is necessary to distil the crude and analyse the
fractions boiling within different boiling ranges. This will give information on the process units
that may be required for quality improvement, such as desulphurisation facilities, a reformer
unit or an aromatics extraction plant. The residue fractions can be similarly analysed to give
more detailed information on the conversion units that may be appropriate.
It follows some examples submitted by the TWG on the calculation of some cost-effectiveness
of the implementation of environmental technologies in the refinery sector. One feasibility study
(Section 10.4.4) provided by a TWG member has shown that the cost and the technical solution
for the achievement of the same environmental goal (e.g. bubble concept) depends on the local
situation.
Table 10.19: Example of abatement technology cost for a new installation, emission reduction and
emissions remaining after the application of the technology.
Due to local regulations and/or corporate policies there are many situations where certain levels
of control already exist within specific sites. In these cases the cost for achieving a given %
emission reduction target is significantly increased over the initial cost effectiveness values.
This needs to be accounted for in determining the cost effectiveness of a technology/technique.
As can be seen in Table 10.19, if technology C is already in place at a location, the additional
emissions reduction for going to technology G would be only 49%. The cost of implementing
technology G therefore becomes 2.45 kEUR/tonne for the incremental emissions reductions,
rather than the 1.21 kEUR/tonne when going to 99% control from 0%. If one were to go to
technology G in a location where technology E was in place, the incremental cost would be 13.3
kEUR/tonne. In all of these instances the final situation is to be at 99% control of emissions.
The actual cost for going to the 99% level in these situations, if done incrementally, would
effectively be the cost of implementing Technology C plus E, plus G, to get the 99% reduction.
Table 10.20 shows the resultant cost of applying various technologies starting from different
levels of existing control.
Table 10.20: Cost effectiveness (kEUR/tonne abated) starting from different percentages of
reduction and considering that the previous technology should be replaced 100%
Next table gives the data on cost-effectiveness of the application of serveral sulphur recovery
unit techniques.
The following graph shows the cost to abate a tonne of NOx in industry sectors as well as some
reference values used in some countries [248, Ademe, 2001],
abated)
10000
9000
x
8000
n e
xa 3)
de vel
e
e
ed fer fere )
en M mg )
C th+ 3
G R ( -1 t R ( l)
(1 gl .5 3)
I)
ss C R 03 us R
D Du ple ple
x
Sw h re re CR
rn at alu
C ish enc nc
lu
a
/m
CR
C P 8 ry 6 Fr e
W Fr e s e /m
te uro o 0 /m
N (A C
S C ev
tra le
va
te am m
S
c tch /S
ifo tax e v
g
g
(
W ss ( 0.7 0.5
ia io
LC fin m3
em 00 5
W
l a 1- o
re g/
a
t
n
se m
as x
tS
2
ne 00
H
SC (
as E
pa s 2
ut
al
Ja rie
la S
G s
e
s
in
la
ef
G
R
Figure 10.5: Some values on the cost for the abatement of NOx in different industrial sectors
The following tables show data concerning cost-effectiveness for NOx abatement measures.
NOx Control for Fired Heaters and Boilers Firing Refinery Blend Gas
Basis: 100 Giga joules/hr installation
Retrofit of existing unit
Refinery blend gas firing
Uncontrolled NOx emissions of 150 ppm at 3% oxygen (300 mg/Nm3)
Flue Gas Ultra Low NOx SNCR SCR Ultra Low NOx
Recirculation Burners Burner
plus Low NOx plus SCR
Burner (6)
NOx Control for Fired Heaters / Boilers Firing Residual Fuel Oil
Basis: 100 Giga joules/hr installation
Retrofit of existing unit
Residual Oil firing
Uncontrolled NOx emissions of 250 ppm at 3% oxygen (500 mg/Nm3)
Boilers Heaters
NOx Control for Gas Turbines Firing Natural or Refinery Blend Gas
Basis: 85 MW output turbine (representative of a GE Frame 7 size unit) (electrical output)
-tural gas or refinery blend gas firing
Uncontrolled NOx emissions of 250 ppm at 15% oxygen (350 g/GJ)
Cost Effectiveness of some already applied NOx abatement techniques in USA refineries.
For both sites, all the major existing units contributing globally for more than 80 % of the NOx
emissions from the site have been considered: furnace, boilers, regenerator, etc. The BAT
candidates listed in the [115, CONCAWE, 1999] have been considered for each item at the
technical, economic and maintenance levels.
Summary graphs are presented showing the variable efficiency of considered and validated
techniques on different units: low-NOx burners, flue gas recirculation, SNCR, SCR, reburning,
etc. An estimated cost (about 30 %) is expressed for implementation of each BAT candidate in
terms of cost per reduced NOx tonne per year. A large variation of cost is observed on different
units, for various techniques and with different impacts on the global NOx emissions reductions
of the two sites respectively.
90000
80000
70000
Cost en FF/Tons NOX
60000
50000
40000
30000
20000
10000
0
0 200 400 600 800 1000 1200
Reduction NOx Tons/Yr
Figure 10.6: Variable efficiency and cost of retrofitting of existing units (FRF/tonne NOx reduced)
Note: Using feasibility of NOx BAT candidates study on two refinery sites (1 Euro = 6.56 FRF)
Finally, combined scenarios of the potential applicability of these BAT are presented for the
both sites and for similar environmental objectives based on a global NOx bubble concentration.
Cost impact is different for each existing site to be retrofitted and is very expensive according to
the environmental objectives to be reached.
This technical feasibility and economic considerations have to be remembered in the context of
the BAT associated emissions levels to be proposed, taking into account the complexity of
retrofitting due to the difference in the current European existing refinery sites.
§ Increased energy efficiency of the various processes integrated with the refinery fuel
management system;
§ Decrease use of liquid fuels by increased use of gaseous fuels from within the refinery or,
where applicable, from additional processing and conversion in the refinery, FCC, IGCC,
coker, etc.;
§ Lower the S-content of liquid fuels (e.g. hydrotreatment) used in the refinery;
§ Use of clean gaseous fuels (-tural gas or cleaned refinery gas);
§ Use of liquid fuels combined with control techniques to clean the flue gases, thereby
reducing the emissions of SO2 and NOx;
§ Apply BAT to single installations, especially to the sulphur recovery unit (SRU).
It follows some calculations, justifications for the proposals given by some TWG members.
In order to illustrate the use of the bubble concept in a refinery, a number of cases and their
relevant emissions are calculated along the lines outlined below. The following assumptions are
used.
Gas-fired refinery:
There are a few full conversion refineries in Europe that are fully gas-fired. The main source of
energy in these cases is cleaned refinery gas. In some cases also natural gas is used as
supplementary fuel.
Assumptions for bubble calculation: Full conversion, fully gas fired refinery, processing 10
million tonne of crude oil with an annual fuel use of 700000 tonnes (7% on intake). SRU
applies BAT with 99.8% efficiency and has an annual production of 100000 tonnes of
elementary sulphur. The FCC has an annual throughput of 1.5 million tonne (consuming some
12% of the total refinery fuel consumption). The 100% gas case is presented in Tables 10.21-24.
Gas and Liquid Fired Refinery using low sulphur oil (0.5% S)
Most European refineries use a combination of gaseous and liquid fuels from within the
refinery. The calculation of the technical achievable emission values in such a refinery is based
on a) existence of a combined fuel management system in the refinery, and b) emission values
of the combined fuel system with fuels that are either inherently clean (certain refinery gases or
natural gas) or refinery gaseous or liquid fuels that can be cleaned relatively easy. It may also
cover the situation of the cleaning of the flue gases from fuels that “untreated” will cause
pollution to air.
Assumptions for the calculation: Medium conversion, partially gas fired refinery, processing 10
million tonne of crude oil with an annual fuel use of 400000 tonne (4% on intake). SRU applies
BAT with 99.8% efficiency and has an annual production of 50000 tonnes of elementary
sulphur. The FCC has an annual throughput of 1.5 million tonne (consuming some 20% of the
total refinery fuel consumption). Several cases are shown (70% gas, 50%gas and 30% gas).
Assumptions for the calculation: Medium conversion, partially gas fired refinery, processing 10
million tonne of crude oil with an annual fuel use of 400000 tonne (4% on intake). SRU applies
BAT with 99.8% efficiency and has an annual production of 50000 tonnes of elementary
sulphur. The FCC has an annual throughput of 1.5 million tonne (consuming some 20% of the
total refinery fuel consumption). Heavy residue (3.5 wt% S) is fired on the power plant, which
is equipped with flue gas desulphurisation. Two desulphurisation efficiency cases are presented,
90% DeSOx and 95% DeSOx.
The results from the calculations are summarised in Table 10.21. The cases are elaborated in
more detail in Tables 10.22-24. It should be explicitly noted that these bubbles are calculated
from a technical basis and by taking the ‘bottom up’ approach. It does not take account of
process upsets or outages.
Table 10.21: Summary of calculated SOx and NOx bubbles from the cases described in this section
and presented in following tables
It should be noted that the 100% gas case is the only case that is a full conversion refinery. The
other cases refer to medium conversion refineries, which has a considerable effect on the fuel
consumption and thus also on the load bubbles.
Description Unit 100% gas 70% gas 50% gas 30% gas DeSOx-90 DeSOx-95
Refinery type conversion Full medium medium medium medium medium
Crude throughput million tonnes/year 10 10 10 10 10 10
Fuel consumption % of feedstock input 7 4 4 4 4 4
Fuel consumption tonnes per year 700000 400000 400000 400000 400000 400000
Flue gas from FCC* 103 m3, ISO, dry, 3% O2 1008000 960000 960000 960000 960000 960000
Flue gas from gas firing* 103 m3, ISO, dry, 3% O2 7392000 2688000 1920000 1152000 1420800 1420800
Flue gas from oil firing* 103 m3, ISO, dry, 3% O2 0 1152000 1920000 2688000 2419200 2419200
S-content fuel oil wt. % 0.5 0.5 0.5 0.5 3.5 3.5
Flue gas desulphurisation % 0 0 0 0 90 95
Table 10.22: Description of 6 different cases of refinery firing and flue gas treatment
* The flue gas flow is calculated from the assumption that the combustion of 1 tonne of oil or 1000 m3 of gas yields a flue gas flow of 12000 Nm3 (at 3% O2)
SOx* FCC Gas firing Oil firing FCC Gas firing Oil firing SRU Total Concentration bubble Load bubble
Description mg/Nm3 mg/Nm3 mg/Nm3 tonne/yr tonne/yr tonne/yr tonne/yr
tonne/yr mg/Nm3 tonne/Mtonne crude
100% gas 50 – 400 5 – 15 - 50 – 403 37 – 111 0 400 487 - 914 58 – 109 49 – 91
70% gas 50 – 400 5 – 15 850 48 – 384 13 – 40 979 200 1241 – 258 – 334 124 – 160
1604
50% gas 50 – 400 5 – 15 850 48 – 384 10 – 29 1632 200 1890 – 394 – 468 189 – 224
2245
30% gas 50 – 400 5 – 15 850 48 – 384 6 – 17 2285 200 2539 – 529 – 601 254 – 289
2886
DeSOx-90 50 – 400 5 – 15 595 48 – 384 7 – 21 1439 200 1695 – 353 – 426 169 – 204
2045
DeSOx-95 50 – 400 5 – 15 298 48 – 384 7 – 21 720 200 975 – 1325 203 – 276 97 – 133
* Concentrations are assumed from the application of BAT. Loads and bubbles are calculated accordingly.
Table 10.23: Assumed and calculated SO2 emissions and bubbles for the different cases described in Table 10.22
NOx* FCC Gas firing Oil firing FCC Gas firing Oil firing Total Concentration bubble Load bubble
Description mg/Nm3 mg/Nm3 mg/Nm3 tonne/yr tonne/yr tonne/yr tonne/yr mg/Nm3 Tonne/Mtonne
crude
100% gas 100 – 150 100 – 150 – 101 – 151 739 – 1101 0 840 – 1262 100 – 150 84 – 126
70% gas 100 – 150 100 – 150 120 – 200 96 – 144 269 – 403 138 – 230 503 – 778 105 – 162 50 – 78
50% gas 100 – 150 100 – 150 120 – 200 96 – 144 192 – 288 230 – 384 518 – 816 108 – 170 52 – 82
30% gas 100 – 150 100 – 150 120 – 200 96 – 144 115 – 173 323 – 538 534 – 854 111 – 178 53 – 85
DeSOx-90 100 – 150 100 – 150 120 – 200 96 – 144 142 – 213 290 – 484 528 – 841 110 – 175 53 – 84
DeSOx-95 100 – 150 100 – 150 120 – 200 96 – 144 142 – 213 290 – 484 528 – 841 110 – 175 53 – 84
* Concentrations are assumed from the application of BAT. Loads and bubbles are calculated accordingly.
Table 10.24: Assumed and calculated NOx emissions and bubbles for the different cases described in Table 10.22
Hypothesis
The following calculations are based on a complex refinery firing heavy residues in the power
plant, which is equipped with flue gas desulphurisation. To allow some flexibility a
gaseous/liquid fuel ratio between 1 and 2 is assumed, which means that between 50 % and 33 %
of the total energy used by the refinery is derived from combustion of liquid fuels (there are
refineries existing with 100 % gas firing). Average S-content of the liquid fuels is assumed to be
in the range of 2.8 – 3.5 %. Measures to increase energy efficieny are not included in these
calculations.
Calculations
Assuming the above given parameters, calculated bubble (including all installations, but without
flares) concentrations for SO2 are in the range of 109 mg/Nm3 (case a; 95 % efficiency of FGD)
and 338 mg/Nm3 (case b; 90 % eff. of FGD). Case b together with an efficiency of FGD of
95 % results in a bubble concentration of 175 mg/Nm3 (corresponding to 235 t/Mt crude
throughput).
Corresponding loads are from 126 t SO2/Mt to 455 t SO2/Mt crude throughput.
Refinery characteristics
A complex refinery processing 15 Mt/yr of crude oil. 15% fuel oil at an average sulphur content
of 1.3 %. 160000 tonnes of fuel oil and 840000 tonnes of gas (RFG + natural gas)
BAT assumptions
50% reduction of fuel emissions by switching fuel oil to low S fuel oil or by installing FGD on
stacks with highest SO2 fuel emissions.
50% reduction of catcracker emissions by reducint sulphur input to catcracker and/or by
installing FGD on stacks of catcracker
High efficient SRU >99.5% sulphur recovery
Conclusions
Concentration bubble of 200 mg SO2 /Nm3
Refinery characteristics
A complex refinery processing around 12 Mt/yr of crude oil.
Fuel consumption: 156000 t/yr liquid at 3 % S
400000 t/yr of refinery fuel gas + natural gas
SRU at 99% SO2 concentration 20000 mg/Nm3
BAT assumptions
Improve the SRU to BAT levels
DeNOx in the main stacks
Conclusions
SOx bubble concentration of 50 mg/Nm3
NOx bubble concentration of <100 mg/Nm3.
10.5.4 Italian proposals and justification for the given bubble figures
With respect to environmental benefits, Italy and other mediterranean countries are
characterised by a nearly full compliance with objectives for acidification and very marginal
transboundary pollution. In other words we don’t need very stringent emission ranges for SO2
and NOx.
From the technical and economic point of view, we should consider that the refining sector in
the Mediterranean area is characterised as follows:
• 70 % of the European fuel oil market demand is located in the Mediterranean area;
• the current configuration of the existing refineries in this area derives from the actual
petroleum products demand that is very different from the North Europe demand;
• the crude oil slate processed is mainly derived from Middle East crude oils that represent
the best choice to fit both the market demand and the refinery configuration. The heavy fuel
oil produced under these conditions needs to be used internally because any other options
(deep conversion, desulphurisation, gasification, etc) are not economic sustainable
In conclusion we believe that the level associated with BAT should take into account the cost
and the advantages associated to different emission ranges and, moreover, BAT must be
techniques generally applicable otherwise the competitiveness inside the refining sector in
Europe will be distorted.
On the basis of these considerations we propose the following bubble emission ranges:
The ranges here proposed for SO2 and NOx are derived from the emission values associated to
the BAT considered for the revision of the Large Combustion Plants Directive, recently
approved at European level. The BATs adopted for this revision are the most stringent
techniques applicable to the new and existing large combustion plants.
In the revision of the Large Combustion Plants Directive the SO2 emission limit values for new
and existing refineries are respectively 600 and 1000 mg/Nm3. Having considered that the
bubble emission levels in the BREF are referred to all plants, these emission limit values are
slightly increased to include the SO2 emission also from Claus and Catalytic Cracking plants.
The emission limit values for NOx, in the revision of the Large Combustion Plants Directive are
defined in the same range we proposed for refinery bubble and this seems enough to justify our
proposal.
The above emission values ranges are coherent with the definition of general BAT and so the
requirement of more stringent ranges do not have justification.
By the way the techniques that need to be adopted to meet the more stringent ranges will
increase the refinery energy consumption and the associated CO2 emissions.
Table 10.25: Variation of the Bubble Results from Use of Fuels with Varying Sulphur Content with
and without FGD
In terms of the regulatory point of view or operating conditions of a single site, the imposed or
reachable bubble limit has to take into account the huge variable conditions of operations of a
refinery site on a daily basis, particularly on the averaging time considerations.
Figure 2 shows the reported daily variation of a calculated SO2 global bubble of a representative
refinery, all units included, and the corresponding daily use of potential fuels within the site.
The annual global average SO2 bubble is 1200 mg/Nm3, ranging from 500 to 2000 mg/Nm3, the
daily variation is given for each type of fuels use (low sulphur fuel oil, high-sulphur fuel oil,
vacuum residue, refinery fuel gas).
In order to avoid making the graph over complex, it does not reflect the fuels sulphur content
which is also varying according to the initial sulphur content in crude feedstock, we observe
respectively:
• for LS FO, average S% of 0.7 %, ranging from 0.4 to 1 % by month
• for HS FO, average S% of 1.87 %, ranging from 1.1 to 2.3 % by month
• RSV, average S% of 1.94 %, ranging from 0.9 to 2.7 % by month
• fuel gas (average S% of 0.05 %)
1500,0
1000,0
500,0
0,0
103
120
137
154
171
188
205
222
239
256
273
290
307
324
341
358
18
35
52
69
86
1
days
Figure 10.7: Daily SO2 bubble variation and fuels breakdown – real refinery case (global annual
average of 1200 mg/Nm3)
It is recommended, to consider yearly averaged value of the SO2 bubble in order to maintain the
flexibility of the process.
We would believe that very challenging global and annual bubbles for SO2 in the future, based
on local conditions and considering the SO2 emissions of all the units, could range from 1000 to
1400 mg/Nm3. This leaves flexibility to still respond to market constraints and opportunities on
the crude oil, the supply/demand situation for oil products and to operate effectively in a very
competitive environment. Individual refineries that have the local opportunity to burn gas only
and still remain competitive can go to lower values than this.
In order to avoid any negative local impact on neighbouring communities, this daily variation
can be locally controlled by alert procedures based on unfavourable weather forecast conditions,
resulting in temporary specific operating constraints for the sites.
CONCAWE does not have survey data available on the NOx emissions. These values are not
only dependent on the fuels management but also from the operating conditions (excess air,
temperature, etc.) and processes.
a) REAL NEW REFINERY CASE – Daily Variation of NOx single stack ELV
As observed for an SO2 bubble, similar large variations are observed on the short-term basis for
single stack: the variation is very sensitive, even on a new built refinery, where feedstock is
however not so varyiable.
We would believe that NOx annual bubble for a good performing existing refinery (mixed
gas/liquid fired) could be 350 to 500 (based on low NOx burners on fuel oil accounting for
combustion air preheat). NOx for a new refinery (gas fired) possibly 200 can be reached.
Table 10.27: Variation of the Bubble Results from Use of Fuels with Varying Sulphur Content with
and without FGD
2. Quantification of the sulphur emissions of the different refinery sources being part of S2
and S4 in order to clarify the bubble concept and to identify the main emitters in each
specific case,
3. Selection, among the different technical options indicated below (described more precisely
in the specific BAT), the technical measures allowing to achieve the following global
indicative reference values (bubble benchmark)
600 mg/Nm3 (Monthly average),
850 mg/Nm3 (Daily average),
considering the possible cross-media (extra energy consumption, production of waste
and water effluents, atmospheric emissions outside the refinery, increase of residue
production, etc...) and the cost effectiveness.
Possible technical measures which could be implemented to fulfill the global
environmental goal include the following:
Due to the specific characteristics of the sulphur recovery units for flue gases, it is additionaly
considered that BAT is also to apply the specific BAT defined for this type of unit (see
Section 5.2.23).
The overall sulphur strategy should also lead to minimise S7 by increasing the conversion of
products without specifications with processes like cokers, hydrocracking, gasification, etc…,
by improving the characteristics of these products or by having an external use ensuring that the
environmental impact is minimised (cement plants, conversion in another refinery for example).
This kind of external use has also to be taken into account in the OPSR calculation.
Primary measures have generally the highest cost effectiveness but also limited performances
especially for liquid fuels. These measures can easily be implemented for new units but may
raise important integration problems with existing units. Given the limited cost of this type of
measure its promotion may be encouraged where possible.
The limited performance of the primary measures makes necessary to consider secondary
measures, whose cost effectiveness considerations are of great importance. Its implementation
should be restricted to the main sources, e.g. those representing about 80% of the emissions of
the whole refinery. Secondary measures includes three main abatement techniques: Reburning,
SNCR (Selective non catalytic reduction) and SCR (Selective catalytic reduction) with a range
of efficiencies (50 to 60%, 50 to 70 % and 80 to 90 %) and costs per tone of pollutant abated.
BAT can be considered an average abatement rate of 70 % for the sources representing 80% of
the whole emission of a refinery. According to the principles described above, the combination
of primary and secondary measures allows to reduce the global rough emission level of a
refinery by 60-80 % and to achieve average emission levels in the range of 100-150 mg/Nm3 on
a long-term period.
Other techniques, such as catalytic combustion, or oxyfiring, may need also consideration and
could represent relevant technical measures to achieve the global BAT emission level above-
mentioned.
Other technical options have been implemented in some world areas to achieve lower emission
levels. In a refinery in California there was a switch from liquid to gaseous fuels and
implementation of secondary measures on gas fired units, achieving 20-30 mg/Nm3. In Japan it
has been applied hydrotreatment of liquid fuels and feeds combined with the implementation on
an important number of SCR units, achieving 60-65 mg/Nm3.
However, although this type of environmental goal may be locally justified, it is not considered
that these experiences represent at the time of writing a general level of BAT in the EU
framework because of the direct cost of these measures and the other investments required to
manage the great amount of residues produced or to improve their characteristics.
1. a CO-furnace boiler for partial oxidation conditions. The associated emission levels are less
than 50 mg/Nm3 for CO and between 100-300 mg/Nm3 for NOx (see Section 4.5.3),
2. a monitoring and control of the oxygen excess (around 2%) for the full combustion mode.
The associated emission levels are between 50-100 mg/Nm3 for CO and between 300 - 600
mg/Nm3 for NOx (see Section 4.5.1),
3. a reduction of particulates emission in the range 10 - 30 mg/Nm3 (metals (Ni, V, etc...) less
than 5 mg/Nm3) by a suitable combination of the following techniques:
In addition, the FCC unit may be a major source of NOx and SOx. In this case, several technical
options described below have to be considered and evaluated alone or in combination:
• the hydro-treatment of the feedstock (economical and technical viability are described in the
Section 4.5.4). The main purpose of this technique is to reduce the sulphur content of the
feed but this option has also a positive influence on NOx, metals and on the product quality.
Most units around the world (California for example) are designed to achieve 300 to 500
ppm in the feedstock,
• flue gas desulphurisation; depending on the process considered and the running conditions,
the abatement rate may vary between 50 and 90 %. These techniques may produce large
amounts of residues and water effluents,
• DeSOx catalyst. This technique may have a limited efficiency and involves an important
consumption of catalyst. This option is probably more suitable to solve peak pollution
problems,
• DeNOx measures such as SCR (abatement rate between 80-90%) and SNCR (abatement
rate between 50 and 70 %) (these techniques have an associated emission of ammonia
which can be limited to 10 mg/Nm3 and it has to be pointed out that SCR is much more
common for FCC units than SNCR) or modifications concerning the design of the unit
which can induce a trade-off between NOx and CO emissions.
These technical options allow thus to achieve ranges as low as 50 - 200 mg/Nm3 for SOx as SO2
and 50 - 100 mg/Nm3 for NOx as NO2. Their combination has to be optimised concerning
performances in regard of the different parameters, energy consumption, waste and water
effluents and cost effectiveness. They may not be justified in all refinery frameworks but have
to be considered to fulfil the whole environmental goal of a refinery.
The energy system is the main source of air pollution of a refinery, however, due to the diversity
of the units used (boilers, furnaces, turbines), their characteristics (size, age, type of burners,
purpose) and their running conditions (air preheat, fuels used, temperature requirements, oxygen
excess and type of control), the strategy to reduce the environmental impact of these units can
only be given by the general guidelines described below.
Firstly, it is necessary to have an overview of the fuels used by the refinery, types of combustion
units, and characteristics of the air emissions (concentration and mass flow) for the main
pollutants (SOx, NOx, dust, metals, carbone oxides).
On these bases, the first step is to consider technical options which seems totally justified from
an economical point of view. This category includes specific BATs such as:
• the use of clean refinery gaseous fuels (H2S content between 20 and 100 mg/Nm3),
• the reduction of CO emissions below 50 mg/Nm3 with efficient combustion techniques, the
reduction of dust emissions in the range 10 - 20 mg/Nm3 (metal content (Ni, V, etc...below
5 mg/Nm3) applying the suitable following combination of techniques (improvement of fuel
characteristics, steam atomisation for liquid fuels, ESP or bag filter),
• the use of primary DeNOx measures for new units, existing ones when these measures can
be implemented quite easily and in the other cases during the major rebuilding of these
units. The achievable emission levels with primary DeNOx techniques are the following:
Gaseous fuels (30 to 100 mg/Nm3 for boilers and heaters, 50 to 100 mg/Nm3 at 15% O2 for
gas turbines - the lowest values are more likely achieved with natural gas and the highest
with refinery fuel gas but a lot of other parameters may influence the performance level),
liquid fuels (200 to 400 mg/Nm3 for boilers and furnaces -the performance level is mainly
influenced by the nitrogen content of the fuel even if other parameters need also to be
considered).
• It is also considered in this category to (add Section 101).
The second step consists of analysing the relevancy of the following basket of techniques in the
global framework of a refinery concerning NOx and SOx for the main sources (see SOx and NOx
sections of the general BAT):
These last two techniques induce NH3 emissions which can be kept below 10 mg/Nm3.
The application of these techniques or of a combination of these techniques may allow to
achieve emission levels lower than 50 mg/Nm3 for NOx (as NO2) and 100 mg/Nm3 for SOx (as
SO2) on any kind of the energy system sources, but the partial or complete implementation of
these options on a specific unit has to be examined in the global framework of the refinery.
The technico-economical analysis may lead to select a less ambitious option concerning
performances allowing nevertheless to achieve the global environmental target at the refinery
level. On the contrary, better performances may be proposed through technical options like
catalytic combustion for turbines (NOx level lower than 2.5 ppm)
First proposal is to define an upper limit (volume and pressure) to store a liquid in another way
than a pressurised tank designed to prevent any direct organic vapour loss to the atmosphere
(connection to a vapour recovery unit with an efficiency of more than 95%). This limit could be
70 kPa concerning the pressure under storage conditions and 150 m3 for the volume.
Then, a second limit (volume: 150 m3 and pressure under storage conditions: 3.5 kPa) could be
indicated concerning the following technical choices: pressurised tank designed to prevent any
direct organic vapour loss to the atmosphere (connection to a vapour recovery unit with an
efficiency of more than 95%) or external floating roof with two seals (one above the other) or
for new fixed roof storages connection to a vapour recovery unit with an efficiency of more than
95% and for existing fixed roof storages installation of an internal floating- type cover with two
seals or connection to a vapour recovery unit with an efficiency of more than 95%.
For storages non connected to a vapour recovery unit, BAT is also to paint the external walls
and the roof with a material having a radiant heat reflection index of at least 70%. This BAT
could apply to facilities that load organic liquids with a vapour pressure higher than 10500 Pa
under loading conditions. Continuous monitoring concerning the vapour recovery unit
emissions should also be considered.
Possible
Furnaces Atmospheric
Raw materials Abatement
Boilers emissions
(crude, other feeds, techniques
Turbines S2
etc…) Oil refinery Etc…
(S)
FCC
Etc..
.
Other units producing
Products H 2S Sulphur Sulphur S3
recovery units (inert form)
Atmospheric
emissions
S4
Potential
Atmospheric Inert sulphur in Atmospheric Water effluents Waste
Emission products Emissions S8 S9
Outside the S6 Outside the refinery
refinery S7
S5
With : S = S2 + S3 + S4 + S5 + S6 + S7 + S8 +S9