Water Cannons

Demonstration of Clyde Bergemann
Water Cannons at Alabama Power

Company's Plant Miller Unit 1
Technical Report
12762100
12762100
Demonstration of Clyde Bergemann
Water Cannons at Alabama Power
Company’s Plant Miller Unit 1
1011120
Final Report, November 2004
Cosponsor
Southern Company Services, Inc.
600 North 18th Street
Birmingham, Alabama 35202
EPRI Project Manager

J. Stallings
EPRI • 3412 Hillview Avenue, Palo Alto, California 94304 • PO Box 10412, Palo Alto, California 94303 • USA
800.313.3774 • 650.855.2121 • [email protected] • www.epri.com
12762100
DISCLAIMER OF WARRANTIES AND LIMITATION OF LIABILITIES
THIS DOCUMENT WAS PREPARED BY THE ORGANIZATION(S) NAMED BELOW AS AN
ACCOUNT OF WORK SPONSORED OR COSPONSORED BY THE ELECTRIC POWER RESEARCH
INSTITUTE, INC. (EPRI). NEITHER EPRI, ANY MEMBER OF EPRI, ANY COSPONSOR, THE
ORGANIZATION(S) BELOW, NOR ANY PERSON ACTING ON BEHALF OF ANY OF THEM:
(A) MAKES ANY WARRANTY OR REPRESENTATION WHATSOEVER, EXPRESS OR IMPLIED, (I)

WITH RESPECT TO THE USE OF ANY INFORMATION, APPARATUS, METHOD, PROCESS, OR
SIMILAR ITEM DISCLOSED IN THIS DOCUMENT, INCLUDING MERCHANTABILITY AND FITNESS
FOR A PARTICULAR PURPOSE, OR (II) THAT SUCH USE DOES NOT INFRINGE ON OR
INTERFERE WITH PRIVATELY OWNED RIGHTS, INCLUDING ANY PARTY'S INTELLECTUAL
PROPERTY, OR (III) THAT THIS DOCUMENT IS SUITABLE TO ANY PARTICULAR USER'S
CIRCUMSTANCE; OR
(B) ASSUMES RESPONSIBILITY FOR ANY DAMAGES OR OTHER LIABILITY WHATSOEVER

(INCLUDING ANY CONSEQUENTIAL DAMAGES, EVEN IF EPRI OR ANY EPRI REPRESENTATIVE
HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGES) RESULTING FROM YOUR
SELECTION OR USE OF THIS DOCUMENT OR ANY INFORMATION, APPARATUS, METHOD,
PROCESS, OR SIMILAR ITEM DISCLOSED IN THIS DOCUMENT.
ORGANIZATION(S) THAT PREPARED THIS DOCUMENT
Alabama Power Company, Inc.
ORDERING INFORMATION
Requests for copies of this report should be directed to EPRI Orders and Conferences, 1355 Willow
Way, Suite 278, Concord, CA 94520, (800) 313-3774, press 2 or internally x5379, (925) 609-9169,
(925) 609-1310 (fax).
Electric Power Research Institute and EPRI are registered service marks of the Electric Power
Research Institute, Inc. EPRI. ELECTRIFY THE WORLD is a service mark of the Electric Power
Research Institute, Inc.
Copyright © 2004 Electric Power Research Institute, Inc. All rights reserved.
12762100
CITATIONS
This report was prepared by

Birmingham, AL 35202
Principal Investigators
C. Boohaker
B. Mead
J. Sorge
Alabama Power Company, Inc.

Birmingham, AL 35202
Principal Investigator
M. Carlisle
This report describes research sponsored by EPRI and Southern Company Services, Inc.
The report is a corporate document that should be cited in the literature in the following manner:
Demonstration of Clyde Bergemann Water Cannons at Alabama Power Company’s Plant Miller
Unit 1, EPRI, Palo Alto, CA, and Southern Company Services, Inc., Birmingham, AL: 2004.
1011120.
iii
12762100
12762100
REPORT SUMMARY
This report documents the findings of a demonstration of Clyde Bergemann Water Cannons at
Alabama Power Company’s Plant Miller Unit 1.
Background
Alabama Power Company's Miller Unit 1 is a 700-MWe unit equipped with a Babcock and
Wilcox opposed wall-fired boiler. The unit, originally designed to burn eastern bituminous coal,
switched to Powder River Basin (PRB) coal during the mid-1990s. During fourth quarter 2001,
Plant Miller contracted with Clyde Bergemann to install their SmartCannon waterwall cleaning
system. The primary motivation for the installation was to reduce condensate usage and mitigate
quench cracking of the waterwall tubes while maintaining at least the de-slagging performance
of the current water lance system. The installation of the system was completed during first
quarter 2002.
Objectives
To obtain a better understanding of the Clyde Bergemann SmartCannon waterwall cleaning
system and its potential benefits.
Approach
EPRI and Southern Company co-sponsored a test program to document the SmartCannon
waterwall cleaning system’s impact on plant operation and performance metrics such as
efficiency, emissions, and reliability.
Results
Key findings of the program are as follows:
Reliability—From installation through the end of 2003, the Clyde Bergemann water cannons and
heat flux sensors were very reliable and used almost exclusively for furnace wall cleaning.
Maintenance requirements were much reduced compared to the water lances also installed on
this unit.
Condensate Usage—The water cannons use strained water instead of condensate water as the
cleaning medium. It is estimated that makeup water requirements were reduced by
approximately 10% or nearly 10 million gallons/year (38 ML/year).
Waterwall Tube Damage—For 2003, the total number of cleaning events with the water cannons
was much reduced (approximately 50%) over what had been normal plant practice with the water
lances. Based on data collected during the test program, tube damage produced by an individual
water cannon cleaning cycle was not materially different than that observed with the water
lances. Tube failure mechanisms may take several years to manifest to an actual failure, and
12762100
direct measurement is difficult over a short period. However, based on the frequency of cleaning
and magnitude of temperature swings, it may be inferred that the tube failure rate was less
following the installation of the water cannons. It is the authors' opinion that the availability of
reliable waterwall heat flux and surface temperature measurements are of great importance to the
successful application of intelligent sootblowing technologies, particularly if water is used as the
cleaning medium. The damage incurred is much more highly dependent on temperature change
than the frequency of cleaning and, as such, should be taken into consideration when developing
the cleaning strategy. At Miller for 2003, the nominal surface temperature change was 71°F
(39°C), but it varied greatly with location and, over the course of the year, with several
temperature excursions greater than 400°F (222°C). The damage incurred as a result of the latter
was several orders of magnitude greater than that of transients with temperature changes less
than 100°F (56°C).
Heat Transfer Distribution—Furnace heat transfer was more uniform following installation of
the water cannons. This improvement can, in large part, be attributed to better understanding of
the furnace slagging condition provided by the heat flux sensors and not the result of the water
cannons. Similar results would likely be obtained using the water lances if heat flux sensors were
used to manage cleaning cycles.
Efficiency—The use of water cannons did not have a measurable impact on boiler efficiency.
Although it may be coincidental, steam temperatures, particularly at lower loads, improved
following installation of the water cannons.
Emissions—NOx emissions were not materially affected with the installation of the water
cannons.
Unit Controllability—The water cannons did not have an adverse impact on plant controllability
or stability.
EPRI Perspective
Overall, the Clyde Bergemann water cannon installation at Miller Unit 1 was considered a
significant benefit to plant operation primarily by reducing maintenance costs and tube failure
rates. EPRI’s Heat Rate and Cost Optimization (71.005) has published numerous reports from
other intelligent sootblowing demonstrations, including Texas Genco’s W.A. Parish (1004113
and 1004114) and TVA’s Bull Run (1004115). Substantial efforts are currently underway at six
locations.
Keywords
Boiler efficiency
Furnace
Heat flux
Heat transfer
Intelligent sootblowing
Surface temperature
Tube failures
Water cannons
12762100
ABSTRACT
Alabama Power Company's Miller Unit 1 is a 700 MWe unit equipped with a Babcock and
Wilcox, opposed wall-fired boiler. The unit, originally designed to burn eastern bituminous coal,
quarter 2002. To obtain a better understanding of the technology and its potential benefits, EPRI
and Southern Company co-sponsored a test program to document its impact on plant operation
and performance metrics such as efficiency, emissions, and reliability. This report describes the
findings of this project.
vii
12762100
12762100
ACKNOWLEDGMENTS
The authors express their gratitude to Alabama Power Company Plant Miller staff for sponsoring
and supporting this test program, especially John Banger, Planning & Engineering Manager at
the site. Many others at Miller graciously took time from their normal duties to provide support
and guidance to the project.
We also wish to thank those from Clyde Bergemann, particularly Charlie Breeding, for bearing
with our many requests for data and documentation. The successful installation of the water
cannons at Miller was in no small part a result of their efforts.
We would like to thank Dr. Mark Barkey from the University of Alabama for his time and
patience in assisting us with the understanding of the material data analysis presented in this
paper.
Lastly, many thanks to Jeff Stallings of EPRI for sponsoring this test program and being very
understanding of the many delays in this (and other) test program.
ix
12762100
12762100
ABBREVIATIONS AND ACRONYMS
°C Degrees Celsius
°F Degrees Fahrenheit
AH Air heater
APC Alabama Power Company
ASME American Society of Mechanical Engineers
ASTM American Society Testing and Materials
B&W Babcock and Wilcox
Btu British thermal units
CBI Clyde Bergemann Incorporated
d Day
DCS Distributed Control System
EPRI Electric Power Research Institute
FEGT Furnace Exit Gas Temperature, usually at the nose of the furnace
g Gram
GE General Electric
gpm Gallons per minute
h hour
HP High pressure
HVT High-velocity thermocouples
J Joule
L Liter
lb Pound
LH, LHt Left hand
m Meter
MPa Mega Pascal
xi
12762100
ms Milli-second
MW, MWe, MWh Megawatt, Megawatt – electrical, Megawatt-hour
NOx Nitrogen oxides (NO and NO2)
Pa Pascal
PI OSI PI plant historian software
PRB Powder River Basin
psi Pounds per square inch
RH Right hand
RH, RHt Reheat
s second
SCS Southern Company Services
SH, SHt Superheat
S-N Stress-cycles
VWO Valves wide open
W Watts
inHg Inches of mercury
ft feet
xii
12762100
CONVERSION FACTORS
From To Multiply by
British thermal units (Btu) Joules (J) 1055.056
Btu/lb MJ/kg 0.002326
Cubic meters (m3) Gallons (gal) 264.1721
Degrees Celsius (°F) Degrees Fahrenheit (°C) 1.8*t+32
Degrees Celsius (°F) (interval) Degrees Fahrenheit (°C) 1.8
Degrees Fahrenheit (°F) Degrees Celsius (°C) (t – 32)/1.8
Degrees Fahrenheit (°F) (interval) Degrees Celsius (°C) 0.55556
Feet (ft) Meters (m) 0.3028
Gallons (gal) Cubic meters (m3) 0.003785412
Gallons (gal) Liter (L) 3.785412
Gallons per minute (gpm) Cubic meters/hour (m3/h) 0.2271247
Inches of mercury Kilo-Pascal (kPa) 3.386388
kBtu/hr/ft2 kW/m2 3.15
Meters (m) Feet (ft) 3.28083
NOx concentration (ppm) NOx concentration 1.912 x 106
(ng/scm)
Pounds (lb) Kilograms (kg) 0.4535924
Pounds / square inch (psi) Mega-Pascal (MPa) 0.006894757
SO2 concentration (ppm) SO2 concentration 2.66 x 106
(ng/scm)
xiii
12762100
12762100
CONTENTS
1 INTRODUCTION....................................................................................................... 1-1
Background.............................................................................................................. 1-1
Unit Description........................................................................................................ 1-2
Boiler ................................................................................................................... 1-2
Burners ................................................................................................................ 1-2
Condensate System ............................................................................................ 1-3
Water Lances ...................................................................................................... 1-3
Turbine / Generator ............................................................................................. 1-3
2 TECHNOLOGY DESCRIPTION................................................................................ 2-1

Technology Installation and Integration at Miller 1 ................................................... 2-5
3 TEST PROGRAM ..................................................................................................... 3-1

Overview .................................................................................................................. 3-1
Characterization of Water Cleaning Events ............................................................. 3-1
Frequency Characteristics ................................................................................... 3-1
Temperature Characteristics ............................................................................... 3-2
Heat Flux Characteristics..................................................................................... 3-3
Waterwall Tube Life ............................................................................................... 3-25
Field Testing and Data Collection ...................................................................... 3-26
Thermal Impacts ................................................................................................ 3-26
Model Test Case Application ............................................................................. 3-27
Model Field Data Application ............................................................................. 3-28
Damage Index from Long-Term Data ................................................................ 3-32
Summary of Waterwall Tube Damage Analysis................................................. 3-33
Condensate Requirements .................................................................................... 3-52
xv
12762100
Furnace Heat Absorption, Boiler Efficiency, and Emissions................................... 3-53
CBI Conducted Performance Tests ................................................................... 3-54
Pre- and Post-Installation Operational Data ...................................................... 3-66
Equipment Reliability ............................................................................................. 3-69
Impact on Unit Stability .......................................................................................... 3-70
4 SUMMARY AND CONCLUSIONS............................................................................ 4-1
5 REFERENCES.......................................................................................................... 5-1
A WATER CANNON RELIABILITY............................................................................ A-1
B CLEANING EVENT CHARACTERIZATION ........................................................... B-1
xvi
12762100
LIST OF FIGURES
Figure 2-1 SmartCannon Cleaning Areas................................................................................ 2-2

Figure 2-2 SmartCannon Mimic Diagram ................................................................................ 2-3
Figure 2-3 SmartCannon Status Screen.................................................................................. 2-4
Figure 2-4 CBI System Layout ................................................................................................ 2-5
Figure 2-5 Water Cannon at Miller .......................................................................................... 2-8
Figure 2-6 Heat Flux Sensor (external view) ........................................................................... 2-9
Figure 2-7 Pump Skid at Miller ...............................................................................................2-10
Figure 2-8 Water Cannon I/O Cabinet....................................................................................2-11
Figure 2-9 SmartCannon Locations and Cleaning Areas at Miller 1 .......................................2-12
Figure 2-10 SmartCannon – Areas Cleaned by Cannon 1 .....................................................2-13
Figure 3-1 Average Cleaning Cycles by Month for 2003 ......................................................... 3-4
Figure 3-2 Cleaning Frequencies by Zone for 2003 ................................................................ 3-5
Figure 3-3 Cleaning Frequencies by Zone for 2003 ................................................................ 3-6
Figure 3-4 Cleaning Frequencies by Zone for December 2003 ............................................... 3-7
Figure 3-5 Example Temperature Transients from Miller......................................................... 3-8
Figure 3-6 Definition of Temperature Droop, Gain, and Change ............................................ 3-9
Figure 3-7 Cleaning Events with Change Greater Than 0°F for 2003.....................................3-10
Figure 3-8 Cleaning Events with Change Greater Than 100°F for 2003.................................3-11
Figure 3-9 Cleaning Events with Change Less Than 50°F for 2003 .......................................3-12
Figure 3-10 Cleaning Events Distribution of Temperature Change for 2003...........................3-13
Figure 3-11 Cleaning Events Surface Temperature Characteristics vs. Sensor......................3-15
Figure 3-12 Cleaning Events Temperature Change by Sensor for 2003 ................................3-16
Figure 3-13 Cleaning Events Temperature Droop vs. Zone for 2003......................................3-17
Figure 3-14 Cleaning Events Temperature Gain vs. Zone for 2003........................................3-18
Figure 3-15 Cleaning Events Temperature Change vs. Zone for 2003...................................3-19
Figure 3-16 Heat Flux vs. Load for 2003 ................................................................................3-20
Figure 3-17 Surface Temperature vs. Load for 2003 ..............................................................3-21
Figure 3-18 Heat Flux vs. Load for 2003 (Scatter)..................................................................3-22
Figure 3-19 Heat Flux vs. Surface Temperature for 2003.......................................................3-23
xvii
12762100
Figure 3-19 Heat Flux vs. Surface Temperature for 2003.......................................................3-23
Figure 3-20 Heat Flux vs. Sensor for 2003.............................................................................3-24
Figure 3-21 Surface Temperature vs. Sensor for 2003...........................................................3-25
Figure 3-22 Test Case Temperature Profiles .........................................................................3-34
Figure 3-23 S-N Curve for Test Cases ...................................................................................3-37
Figure 3-24 Temperature Measurements During Cannon Event – Test ID: Run 1..................3-38
Figure 3-25 Temperature Measurements During Cannon Event – Test ID: Run 4..................3-39
Figure 3-26 Temperatures Measured During Cannon Event – Test ID: Run 0 .......................3-40
Figure 3-27 Temperature Measurements During Water Lance Event – Test ID: Run 2 ..........3-41
Figure 3-28 Temperature Measurements During Water Lance Event – Test ID: Run 3 ..........3-42
Figure 3-29 S-N curve for Temperature Profiles.....................................................................3-43
Figure 3-30 CBI Fast Data Comparison with Agilent 34970A Data.........................................3-44
Figure 3-31 Temperature of Panel Being Cleaned By Water Lance .......................................3-45
Figure 3-32 Temperature Profile of Panel Cleaned Directly by Water Lance..........................3-46
Figure 3-33 Temperature of Panel Being Cleaned by Water Cannon.....................................3-47
Figure 3-34 Waterwall Cleaning Zone Layout Showing Associated TC Sensors (‘A’
designation)....................................................................................................................3-48
Figure 3-35 Temperature During Water Cannon Cleaning Event – Zone 106 ........................3-49
Figure 3-36 Temperature During Water Cannon Cleaning Event – Zone 410 ........................3-50
Figure 3-37 S-N Curve for “double hit” Testing.......................................................................3-51
Figure 3-38 Damage Index as Calculated from Long-Term Data (2003) ................................3-52
Figure 3-39 Condensate Makeup...........................................................................................3-53
Figure 3-40 CBI Conducted Performance Tests – HVT Sample Locations.............................3-55
Figure 3-41 CBI Conducted Performance Tests – HVT Measurement Distributions...............3-59
Figure 3-42 Performance Test 1 - HVT Temperatures ...........................................................3-60
Figure 3-49 Unit Load from January 2002 through April 2002 ................................................3-67
Figure 3-50 SmartCannon System Problems by Category .....................................................3-69
Figure 3-51 Process Response to Water Cannon Events – June 1, 2003 ..............................3-71
Figure 3-54 Process Response to Water Cannon Events – January 1, 2003 .........................3-74
Figure B-1 Cleaning Frequencies by Zone for 2003 ............................................................... B-3
xviii
12762100
Figure B-2 Cleaning Events with Change Greater Than 0°F for 2003 .................................... B-4
Figure B-3 Cleaning Events with Change Greater Than 50°F for 2003 .................................. B-5
Figure B-4 Cleaning Events with Change Greater Than 100°F for 2003 ................................ B-6
Figure B-8 Cleaning Events with Change Less Than 50°F for 2003 ..................................... B-10
Figure B-9 Cleaning Events with Droop Greater Than 0°F for 2003 ..................................... B-11
Figure B-10 Cleaning Events with Droop Greater Than 50°F for 2003 ................................. B-12
Figure B-11 Cleaning Events with Droop Greater Than 100°F for 2003 ............................... B-13
Figure B-15 Cleaning Events with Droop Less Than 50°F for 2003...................................... B-17
Figure B-16 Cleaning Events Distribution of Temperature Droop for 2003 ........................... B-18
Figure B-17 Cleaning Events Distribution of Temperature Gain for 2003 ............................. B-19
Figure B-18 Cleaning Events Distribution of Temperature Change for 2003 ........................ B-20
Figure B-19 Cleaning Events Distribution of Initial Temperature for 2003............................. B-21
Figure B-20 Cleaning Events Temperature Change vs. Sensor for 2003 ............................. B-26
Figure B-21 Cleaning Events Surface Temperature Droop vs. Sensor for 2003 ................... B-27
Figure B-22 Cleaning Events Surface Temperature Gain vs. Sensor for 2003 ..................... B-28
Figure B-23 Cleaning Events Surface Temperature Change vs. Sensor for 2003 ................ B-29
Figure B-24 Cleaning Events Temperature Droop Distribution vs. Sensor for 2003.............. B-30
Figure B-25 Cleaning Events Temperature Gain Distribution vs. Sensor for 2003................ B-31
Figure B-26 Cleaning Events Temperature Change Distribution vs. Sensor for 2003........... B-32
xix
12762100
12762100
LIST OF TABLES
Table 2-1 CBI External Inputs ................................................................................................. 2-7

Table 3-1 Cleaning Event Statistics for 2003........................................................................... 3-9
Table 3-2 Cleaning Event Statistics by Sensor for 2003 (Means)...........................................3-14
Table 3-3 Normalized vs. Actual Damage Indices for Modeled Cases ...................................3-36
Table 3-4 Normalized Damage Indices from Field Data .........................................................3-43
Table 3-5 Normalized Damage Indices from Water Lance vs. Water Cannon Comparison ....3-48
Table 3-6 Normalized Damage Indices from Testing with Water Cannon Sootblowing...........3-51
Table 3-7 CBI Conducted Performance Tests Summary........................................................3-55
Table 3-8 CBI Conducted Performance Tests Results ...........................................................3-56
Table 3-9 CBI Conducted Performance Tests – Heat Flux by Sensor ....................................3-57
Table 3-10 CBI Conducted Performance Tests – HVT Results ..............................................3-58
Table 3-11 Pre- and Post-Installation Performance Parameters (650 – 750 MW) ..................3-68
Table B-1 Cleaning Frequencies by Zone for 2003 ................................................................ B-2
Table B-2 Cleaning Event Means by Sensor for 2003 .......................................................... B-22
Table B-3 Cleaning Event Sensor vs. Temperature Droop for 2003 ..................................... B-23
Table B-4 Cleaning Event Sensor vs. Temperature Gain for 2003 ....................................... B-24
Table B-5 Cleaning Event Sensor vs. Temperature Change for 2003 .................................. B-25
xxi
12762100
12762100
EXECUTIVE SUMMARY
quarter 2002.
Objectives
To obtain a better understanding of the technology and its potential benefits, EPRI and Southern
Company co-sponsored a test program to document its impact on plant operation and
performance metrics such as efficiency, emissions, and reliability.
Results & Findings
Reliability – From installation through the end of 2003, the CBI water cannons and heat flux
sensors have been very reliable and have been used almost exclusively for furnace wall cleaning.
Maintenance requirements are much reduced as compared to the water lances also installed on
this unit.
Condensate Usage – The water cannons use strained water instead of condensate water as the
Waterwall Tube Damage – For 2003, the total number of cleaning events with the water cannons
is much reduced (approximately 50%) over what had been normal plant practice with the water
lances. Based on data collected during the test program, the tube damage produced by an
individual water cannon cleaning cycle is not materially different than that observed with the
water lances. The tube failure mechanisms may take several years to manifest to an actual
failure and direct measurement is difficult over a short period. However, based on the frequency
of cleaning and magnitude of the temperature swings, it may be inferred that the tube failure rate
is less following the installation of the water cannons. It is the authors' opinion that the
availability of reliable waterwall heat flux and surface temperature measurements are of great
importance to the successful application of intelligent sootblowing technologies, particularly if
xxiii
12762100
water is used as the cleaning medium. The damage incurred is much more highly dependent on
temperature change than the frequency of cleaning and, as such, should be taken into
consideration when developing the cleaning strategy. At Miller for 2003, the nominal surface
temperature change was 71°F (39°C) but it varied greatly with location and over the course of
the year with several temperature excursions greater than 400°F (222°C). The damage incurred
as a result of the latter is several orders of magnitude greater than that of transients with
temperature changes less than 100°F (56°C).
Heat Transfer Distribution – Furnace heat transfer was more uniform following the installation
of the water cannons. This improvement can in large part be attributed to better understanding of
the furnace slagging condition as provided by the heat flux sensors and not the result of the water
cannons. Similar results would likely be obtained using the water lances if heat flux sensors
were used to manage cleaning cycles.
Efficiency – The use of water cannons did not have a measurable impact on boiler efficiency.
following the installation of the water cannons.
Emissions – NOx emissions were not materially affected with the installation of the water
cannons.
Unit Controllability - The water cannons did not have an adverse impact on plant controllability
or stability.
significant benefit to plant operation primarily through reducing maintenance costs and tube
failure rate.
xxiv
12762100
1
INTRODUCTION
quarter 2002. To obtain a better understanding of the technology and its potential benefits, EPRI
and Southern Company co-sponsored a test program to document its impact on plant operation
and performance metrics such as efficiency, emissions, and reliability. This report describes the
findings of this project.
Background
switched to Powder River Basin (PRB) coal during the mid-1990s. Since the steam sootblowers
were not effective in removing the ash derived from the PRB coal, the plant installed water
lances to remove ash from the waterwall tubes. These water lance sootblowers clean a circular
area around their location. Although effective at furnace cleaning, as with many plants at the
time, standard plant practice was for the water lances to be continuously cycled. Indicators of
cleanliness, visual or otherwise, were not used in the decision to sootblow in the furnace.
Several factors led the plant to investigate the installation of alternate methods of furnace
cleaning. The greatest concern was damage to the furnace waterwall tubes as a result of
unnecessary cleaning. The spraying of clean tubes can result in thermal quench cracking of the
tubes, greatly reducing their useful life. This phenomenon has resulted in more frequent than
expected replacement of tube panel sections at Miller and also other utility units that use water as
the cleaning medium [4][5][6].
Another problem that Plant Miller had experienced was a shortage of condensate makeup. Plant
Miller’s water plant was designed for the four units burning eastern bituminous coal instead of
PRB coal. The higher sootblowing requirements of the latter resulted in periodic shortages of
condensate makeup at the site.
Maintenance of the water lances was another significant concern. Although it was felt that the
water lances adequately cleaned the furnace walls, their large numbers (28 per unit) and frequent
1-1
12762100
Introduction
use (approximately every 3 hours when the unit was online) has required considerable
maintenance.
To address the above, during fourth quarter 2001, Plant Miller contracted with Clyde Bergemann
Incorporated (CBI) for the installation of the SmartCannon™ system. The term "water cannon"
is another CBI trade name for a waterwall cleaning system utilizing movable nozzles mounted on
the waterwall and using high pressure water to clean the opposing walls. Due to the articulation
of the water cannon and the cleaned surface being on the opposing wall, the area cleaned by one
water cannon is much greater than that of a water lance. Clyde Bergemann has installed water
cannons at a number of other sites (approximately 60 as of fourth quarter 2001). Although the
technology had been installed at many locations, the installation at Miller was novel in that it
utilized CBI developed supervisory controls and heat flux sensors and a variable speed supply
pump.
Unit Description
Alabama Power’s Plant Miller is the largest coal-fired plant in the operating company and is
located approximately 20 miles northwest of Birmingham on the Locust Fork branch of the
Warrior River in Quinton, Alabama. Plant Miller consists of four similar pulverized-coal units
each rated nominally at 700 MWe. Miller Unit 1 began commercial operation on October 12,
1978.
Boiler
The unit is equipped with a Babcock and Wilcox, opposed wall-fired boiler rated at 4,921,000
lb/h (2,232,128 kg/h) of steam at 2400 psi (16.5 MPa) and 1000/1000°F (538°C/538°C). The
boiler was originally designed to burn eastern bituminous coal with higher heating values in the
range of 10,000 to 13,200 Btu/lb (23.3 to 30.7 MJ/kg). During the early 1990s, this unit switched
to a lower specific energy western coal mined from the Powder River Basin (PRB). PRB coal
produces a thin reflective ash on the waterwall sections of the boiler, contributing to higher than
original design superheat and reheat spray flows. To reduce these spray flows, the unit has
undergone surface area modifications that increased economizer surface area and decreased
reheater surface area.
Burners
There are a total of 56 Babcock & Wilcox burners in each unit at Miller, four rows of eight on
the front wall, and three rows of eight on the rear wall. Each burner is equipped with a coal/oil
flame scanner, and a natural gas flame scanner. The wind boxes are compartmentalized with
automated left and right compartment air dampers at each wind box. The burners have
individual, manually adjusted inner and outer air registers as well as spin vanes to control the
flame turbulence. The inner registers control the amount of air that initially mixes with the
incoming primary air and fuel stream. The outer registers provide air around the burner flame,
completing the combustion process. Closing down on register openings gives the secondary air
more swirl and typically produces a wider flame pattern.
1-2
12762100
Introduction
Condensate System
The condensate system at Miller is composed of a high pressure system and a low pressure
system. The low pressure condensate system consists of all piping and equipment from the
discharge of the condenser hotwell, to the low pressure condensate pumps, through the steam
packing exhauster, through the polishing unit, to suction valves of the high pressure condensate
pump. The low pressure condensate pump is a Byron Jackson vertical pump rated at 4650 gpm
(1056 m3/h) and 245 feet (75 m) of head. The high pressure condensate system consists of all
piping, valves and equipment from the suction valves of the high pressure condensate pumps
through the HP pumps, through the low pressure feedwater heaters to the de-aerator. This
system also includes the turbine exhaust hood sprays, the boiler drum fill, and the boiler feed
pump seal injection water. The high pressure condensate pump is a Byron Jackson pump rated at
4400 gpm (999 m3/h) and 630 feet (192 m) of head with a minimum flow requirement of 1500
gpm (340 m3/h).
Water Lances
Before the installation of the water cannons, Miller Unit 1 relied on a system of water lances to
clean the furnace waterwalls. These were Diamond Power Company lances consisting of twenty
(20) IK-4M-WLs/IK-4M-PAs on the furnace walls, two (2) IK-545 Selective Pattern
Waterlances cleaning the upper rear wall and under the nose, and six (6) Selective Pattern
Waterlances cleaning the furnace division wall (3 per side). These water lances were set on a
timed cleaning cycle with no input as to whether the section was clean or dirty. Also, since these
lances cleaned a circular area, the zones cleaned by each lance overlapped to allow for more
complete coverage. This increased the likelihood that certain areas would be cleaned
unnecessarily.
Turbine / Generator
The original Miller Unit 1 turbine generator was a GE tandem compound, four flow steam
turbine with 30 inch (76 cm) last stage buckets, designed for 2400 psi, (16.5 MPa) 1000ºF
(537°C) throttle and 1000ºF (537°C) reheat steam conditions. The nameplate rating for the
turbine was 662,000 kW at an exhaust pressure of 3.5 inch Hg (11.9 kPa) absolute and one-
percent makeup. In 1998, the original HP turbine was replaced with a model from GEC Alstom.
The original GE HP outer shell was reused, but a new inner shell, diaphragms, rotor, and HP
exhaust glands were supplied by GEC. The new HP turbine is throttle controlled (full arc) and
has no nozzle box whereas the original GE high-pressure turbine was a partial arc or nozzle
controlled machine. The new steam path has eleven stages instead of the seven found in the
original design which allows it to have a higher efficiency. The major disadvantage of the new
design is the severe heat rate penalty incurred if the unit has to operate at lower loads. The
current full load (VWO) load level of the unit is approximately 720 MWe.
The GE supplied generator is a 3600-RPM, direct connected, 3-phase, 2-pole, 60-Hertz type
rated at 830,000 kVA at a power factor of 0.85. The principal cooling medium is hydrogen gas
that is contained within the frame.
1-3
12762100
12762100
2
TECHNOLOGY DESCRIPTION
SmartCannon and water cannon are Clyde Bergemann's trade names for a furnace wall cleaning
system utilizing articulated nozzles to direct the flow of water onto opposing furnace walls. The
major elements CBI SmartCannon installed at Miller are:
• Water cannons (four Model 100 cannons are installed at Miller 1)
• Clyde Bergemann developed controls
• Furnace waterwall heat flux sensors (24 installed at Miller 1)
• Variable-speed pump supplying water to the cannons
The water cannons direct water to opposing walls of the furnace. Upon impacting the opposed
wall, the water penetrates the outer layer and evaporates, causing detachment of the slag from the
tube walls. Unlike water lances of the slag deposit that clean full or partial arcs, water cannons
may be tilted such that arbitrarily shaped areas may be cleaned (Figure 2-1). During
commissioning of the system, cleaning zones (generally rectangular) are defined, and each of
these zones are cleaned individually. One water cannon can effectively clean the surface area of
several water lances. The pressure and velocity of the water stream are controlled as is the jet
progression velocity (linear velocity across the waterwall), to minimize the thermal stresses
induced when cleaning the furnace wall. When running, one water cannon uses approximately
200 gpm (45.4 m3/h) of strained water and runs for approximately one minute. For comparison,
a water lance uses condensate makeup water at a nominal flow rate of 40 gpm (9.1 m3/h) and
runs approximately two to three minutes.
This system makes use of heat flux sensors installed in the furnace walls to determine when the
surrounding area is slagged. The heat flux sensors consist of four thermocouples (Type-K,
Chromel-Alumel) that measure the temperature difference across material of known thermal
properties. By knowing the material properties, the heat flux may be calculated. From this
number, the control system determines if the zone is “clean” or “dirty”. These sensors are placed
throughout the boiler to allow for the most coverage by the cannons.
The control system for the water cannons was developed by Clyde Bergemann. This system can
operate in either program mode or automatic mode. In program mode, the zones are cleaned at
preset time intervals. In automatic mode, the zones are cleaned based on feedback from the heat
flux sensors. The system also has an auto-tune module, which adjusts the jet progression
velocity and target heat flux levels in part to reduce temperature transitions when cleaning. The
system runs on a Windows-based personal computer and is implemented using WonderWare®.1
1
Wonderware is a software product and business unit of Invensys plc.
2-1
12762100
Technology Description
The operator interface is through this system. Examples of the operator interface are shown in
Figure 2-2 and Figure 2-3.
Shown in Figure 2-4 is the layout of the CBI equipment. In addition, the CBI computer is
supplied with several 4-20 mA connections to the DCS and other sources which provide the
water cannon system with pertinent unit information, such as current gross load and steam
temperatures.
Water Cannon Water Lances
Figure 2-1
SmartCannon Cleaning Areas
2-2
12762100
Figure 2-2
SmartCannon Mimic Diagram
2-3
12762100
Figure 2-3
SmartCannon Status Screen
2-4
12762100
Figure 2-4
CBI System Layout
Technology Installation and Integration at Miller 1
After project approval in late October 2001, installation of the water cannons began in November
during a previously scheduled unit outage. The mechanical aspect of the installation took
approximately three weeks working ten hour days, six days a week. During this portion of the
installation, the four water cannons along with the bent tube openings (~18 inches x 50 inches, 8
tubes each) were installed as well as the 24 heat flux sensor tube sections
2-5
12762100
Figure 2-5 and Figure 2-6). After the unit came back online in December 2001, most of the
electrical work was done, taking approximately 4 weeks to complete. The cannons began
operating in manual mode the week of February 8 and automatic (feedback) mode the week of
February 18. CBI continued to optimize and debug the system after it came online. In third
quarter 2002, the auto-tuning modules were installed. New aspects for the installation at Miller
were the CBI developed controls, CBI developed heat flux sensors, variable speed/pressure
pump, and an updated water cannon assembly. The Sundyne AnySpeed variable speed/pressure
pump is located on the ground level of the plant, giving it relatively easy access to the ash sluice
water and a positive suction head (Figure 2-7).
The cannons are approximately located at the 415 foot elevation above the burners on the front
and rear walls of the boiler (Figure 2-9). Shown below are the locations of the water cannons as
well as the heat flux sensors. The cleaning areas are also shown on the figure. These areas are
cleaned by the cannon located on the opposite side of the boiler. For instance, the area cleaned
by the cannon located on the right side of the front wall is depicted in green and is on the rear
wall. A breakdown of the individual cleaning zones can be seen in Figure 2-10 through
Figure 2-13.
2-6
12762100
The heat flux sensors, as well as the cannons, are connected to the CBI control computer which
is located in the Unit 1 control room. The entire system is controlled by this computer and is not
routed through the plant’s BOS sootblowing control system. There are 11 inputs into the CBI
computer that are fed via a 4-20 mA copper connection. These inputs are shown in Table 2-1.
CBI can access the computer remotely via a modem connection using pcAnywhere. 2 Utilizing
this connection, CBI is able to update and monitor the software without having to be on-site.
Table 2-1
CBI External Inputs
Feedwater Flow Superheat Spray Flow

Stack – NOx Final Superheat Temperature
Main Steam Temperature at Turbine Superheater Outlet Steam
Reheat Outlet Temperature – Left Reheat Outlet Temperature – Right
A AH Gas Out Temperature B AH Gas Out Temperature
Unit 1 Generated Megawatts
2
A product of Symantec Corporation.
2-7
12762100
Figure 2-5
Water Cannon at Miller
2-8
12762100
Figure 2-6
Heat Flux Sensor (external view)
2-9
12762100
Figure 2-7
Pump Skid at Miller
2-10
12762100
Figure 2-8
Water Cannon I/O Cabinet
2-11
12762100
Figure 2-9
SmartCannon Locations and Cleaning Areas at Miller 1
2-12
12762100
Figure 2-10
SmartCannon – Areas Cleaned by Cannon 1
2-13
12762100
Figure 2-11
2-14
12762100
Figure 2-12
2-15
12762100
Figure 2-13
2-16
12762100
3
TEST PROGRAM
Overview
The primary goal of the test program was to evaluate the potential benefits of installing the water
cannon system vis-à-vis the water lance cleaning system previously installed on the unit. The
primary constraint on the test program was that there was to be minimum impact on plant
operations with the unit to be operated following normal, plant operating procedures including
economic dispatch. Also, to a large degree, to reduce test program cost, plant instrumentation
was to be used exclusively during the test program.
Aspects of plant performance and operations to be addressed during the test program included
the following:
• Waterwall tube life
• Condensate requirements
• Furnace heat absorption and efficiency
• Impact on generation capacity
• Emissions reductions
As discussed previously, the water cannons were first operational in first quarter 2002.
Revisions to the hardware and software continued through most of 2002. The test program at
Miller consisted mostly of monitoring the performance of the cannons and their impact on unit
operations. Ad hoc tests were also conducted as needed. Data collection was mainly through
the plant’s historian (OSI PI system) and CBI system. For the latter, CBI provides several
reports and databases on its system which greatly aids in the analysis of the performance of the
water cannons, including daily logs (important plant and water cannon data continuously
collected at 30 second intervals), fast data files (mainly tube temperatures collected at 100 ms
intervals when a section is being cleaned), and summary tables listing the times and locations of
cleaning.
Characterization of Water Cleaning Events
Frequency Characteristics
A goal of the installation of the water cannons at Miller was to minimize the damage to
waterwalls as a result of cleaning. To a large degree, this damage is mainly a function of the
3-1
12762100
Test Program
frequency of cleaning and the change in tube metal temperatures when cleaning. Prior to the
installation of the water cannons, water lances cycled at approximately 3 hour intervals,
regardless of the slagging condition of the furnace walls, yielding approximately 2500 cleaning
cycles per location per year. Given that the heat flux sensors provide feedback as to when a zone
is dirty, it was anticipated that the cleaning cycles would be reduced with the installation of the
SmartCannon system. During 2003, water cannons were used almost exclusively for furnace
wall cleaning. For that year, there was approximately 1250 cleaning cycles per zone per year
(~100 per month) with an increasing trend through the year (Figure 3-1). The reason for the
increasing trend is not known. Usage varied greatly with the cleaning zone with a minimum of
452 cleaning cycles (Zone 413) and a maximum of 2325 cleaning cycles (Zone 311)
(Figure 3-2). The distribution of the cleaning events in the furnace for all of 2003 and December
2003 is shown in Figure 3-3 and Figure 3-4, respectively. When considering the entire year, the
distribution of the cleaning events was relatively uniform. During December, cleaning events
were considerably less uniform, with most of the cleaning activity on the east furnace wall
toward the rear of the furnace.
Temperature Characteristics
The surface temperature profile during cleaning is another important indicator of the
performance of the cleaning system and the impact on tube life. CBI's goal is to configure the
system such to limit the temperature droop to less than 100°F (56°C).3 At Miller, the
temperature transients varied greatly, examples of which are provided in Figure 3-5. 4 Each of
these charts depicts the cleaning over 24-hours at one sensor location with the time intervals
between cleaning removed to make the charts more legible. As shown, the number of cleaning
cycles per day varies greatly as does the magnitude of the temperature swings. The number of
cleaning cycles is governed in part by the feedback of the heat flux sensors whereas the
temperature transient is influenced by the slagging condition of the waterwall section prior to
cleaning and several configurable parameters of the water cannon system (such as jet progression
velocity, water pressure, and target heat flux). These temperature transients are representative
of those seen at other sites [10]. The characteristics of the temperature transient that are of
interest include (see Figure 3-6 for definition).
• Temperature droop
• Temperature gain
• Temperature change
• Multiple impacts during one cleaning cycle
Temperature droop is indicative of the thermal quenching from water cleaning, temperature gain
is an indicator of the slag removed from the surface, and temperature change relates directly to
3
CBI has referenced the 100°F goal as originating from past EPRI studies. The EPRI report cited is CS-4914. Also,
CBI is only looking at the drop in temperature whereas what is important from a fatigue-stress aspect is the peak-to-
peak change in temperature.
4
The data used in this section is largely from the Fast Data store provided by CBI. The data shown is for the period
from January 1, 2003 through September 30, 2003.
3-2
12762100
Test Program
thermal fatigue. For an ideal cleaning cycle, there would be little or no droop, substantial
temperature gain, and a single impact (no multiple hits per cleaning cycle).
A summary of the thermal impacts of these cleaning events for 2003 are shown in Table 3-2 and
Figure 3-7 through Figure 3-11. The number of cleaning cycles trended upward throughout the
year, but the magnitude of the temperature transients were relatively constant throughout. A
large percentage of the cleaning events produced temperature changes of less than 100°F (56°C)
with the first quartile, median, and third quartile being 25°F (14°C), 52°F (29°C) and 90°F
(50°C), respectively. For the year, there was an increasing trend of cleaning events with
temperature changes of less than 50°F (28°C). Although large changes in temperature are not
desirable from a tube damage aspect, smaller changes may be indicative of ineffective cleaning
either because of the slag not being removed during cleaning or the cleaning of relatively clean
surfaces.
The impact of cleaning also depended greatly on the zone with an average temperature change
ranging from 10°F (6°C) (Sensor 19 – Zones 309, Z310, Z311) to 192°F (107°C)(Sensor 09 –
Zone 205)(Table 3-3). The disparity of the temperature impact of cleaning can be seen clearly in
Figure 3-11 and Figure 3-12 in which the frequency and magnitude of the surface temperature
transients are portrayed. From this chart, it may be inferred that the areas represented by Sensor
9 and 10 have suffered the most damage over this data set since it has the greatest count of
cleaning cycles with the greatest change in surface temperature.5
Heat Flux Characteristics
Heat flux and surface temperatures as collected in the CBI daily log files are shown in
Figure 3-16 though Figure 3-21. At full-load, the sensor average heat flux was approximately
26 kBtu/h/ft2 (81.9 kW/m2) which is consistent with design values for this type unit [12]. The
gross waterwall heat absorption may be calculated using a control volume around the furnace
waterwalls:
m& ⋅ (h2 − h1 )
q& =
A
where:
q& = average heat flux
m& = feedwater flow
h1 = economizer outlet enthalpy
h2 = saturated steam enthalpy at boiler drum pressure
A = waterwall surface area (51,000 ft2)
Using boiler design values yields a heat flux of 28 kBtu/h/ft2 (88.2 kW/m2). As may be
expected, measured heat flux was highly dependent on unit load, varying from lows of
approximately 3 kBtu/h/ft2 (9.4 kW/m2) for loads from 100 to 200 MW to 25 kBtu/h/ft2 (78.8
5
This damage is quantified later in the report.
3-3
12762100
Test Program
kW/m2) above 700 MW. Surface temperature exhibited a similar profile but showed a greater
relative reduction than heat flux at loads below 200 MW. Above that load, the surface
temperature was an excellent predictor of heat flux as measured by the CBI heat flux sensor
(Figure 3-19). A comparison of measured heat flux to design and calculated values is shown in
Figure 3-18. The measured values compared well with the full-load design values but was
significantly lower than the calculated values over the entire load range. This difference may be
due to the following factors:
• The waterwall surface area used in the calculation is not accurate.
• There are no heat flux sensors at or below the burner zones. The lack of sensors in the burner
zone would produce a negative bias in the measured average while a lack of sensors below
the burners would tend to produce a positive bias in the measured average.
Heat flux and surface temperature varied greatly among the sensors. Each were considerably
greater in Sensors 7, 8, 9, and 10 than the others (Figure 3-20). These sensors are located in the
zones immediately above the upper level, rear burners. It is interesting to note that the average
surface temperatures on the front wall just above the burners was, on average, about 100°F lower
than on the rear wall (Figure 3-21).
250
Average Number of Cleaning Events per Zone
200
150
100
50
0
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
2003
Figure 3-1
Average Cleaning Cycles by Month for 2003
3-4
12762100
Test Program
1000 1000 1000 1000

Zone 101 Zone 102 Zone 103 Zone 104
500 500 500 500
0 0 0 0
1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112
1000 1000 1000 1000
500 500 500 500
0 0 0 0
1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112
1000 1000 1000 1000
500 500 500 500
0 0 0 0
1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112
1000 1000 1000 1000
500 500 500 500
0 0 0 0
1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112
1000 1000 1000 1000
500 500 500 500
0 0 0 0
1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112
1000 1000 1000 1000
500 500 500 500
0 0 0 0
1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112
1000 1000 1000 1000
500 500 500 500
0 0 0 0
1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112
1000 1000 1000 1000
500 500 500 500
0 0 0 0
1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112
1000 1000 1000 1000
500 500 500 500
0 0 0 0
1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112
1000 1000 1000 1000
500 500 500 500
0 0 0 0
1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112
1000 1000 1000 1000
500 500 500 500
0 0 0 0
1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112
1000 1000 1000 1000
500 500 500 500
0 0 0 0
1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112
1000 1000 1000 1000
500 500 500 500
0 0 0 0
1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112
Figure 3-2
Cleaning Frequencies by Zone for 2003
3-5
12762100
Test Program
Figure 3-3
3-6
12762100
Test Program
Figure 3-4
Cleaning Frequencies by Zone for December 2003
3-7
12762100
Test Program
A B
C D
E F
Figure 3-5
Example Temperature Transients from Miller
3-8
12762100
Test Program
Gain
Change
Droop
Figure 3-6
Definition of Temperature Droop, Gain, and Change
Table 3-1
Cleaning Event Statistics for 2003
TDroop TGain TMean TChange
Min. : 0.000 Min. : 0.000 Min. :400.0 Min. : 0.00
1st Qu.: 6.575 1st Qu.: 9.841 1st Qu.:795.9 1st Qu.: 24.75
Median : 19.523 Median : 29.453 Median :813.4 Median : 51.76
Mean : 36.672 Mean : 34.066 Mean :821.0 Mean : 70.74
3rd Qu.: 45.446 3rd Qu.: 50.798 3rd Qu.:836.7 3rd Qu.: 89.63
Max. :375.172 Max. :310.558 Max. :993.0 Max. :457.68
Notes: 1. All temperatures in °F.
3-9
12762100
Test Program
Cleaning Events with Change > 0 DegF
250
200
150
Count
100
50
0
20030101 20030215 20030327 20030529 20030706 20030813 20030920
FastFile
Figure 3-7
Cleaning Events with Change Greater Than 0°F for 2003
3-10
12762100
Test Program
250
200
150
Count
100
50
0
20030101 20030215 20030327 20030529 20030706 20030813 20030920
FastFile
Figure 3-8
3-11
12762100
Test Program
Cleaning Events with Change < 50 DegF
250
200
150
Count
100
50
0
20030101 20030215 20030327 20030529 20030706 20030813 20030920
FastFile
Figure 3-9
Cleaning Events with Change Less Than 50°F for 2003
3-12
12762100
Test Program
Temperature Change
1500
1000
Frequency
500
0
0 100 200 300 400 500
Temperature (Deg F)
Notes:
(1) Although it can't been seen on this graph, several water cleaning cycles produced
temperature changes greater than 400°F.
Figure 3-10
Cleaning Events Distribution of Temperature Change for 2003
3-13
12762100
Test Program
Table 3-2
Cleaning Event Statistics by Sensor for 2003 (Means)
TStart TMean TDroop TGain TChange
S01 791.0 810.0 22.167 38.058 60.22
S02 789.5 796.3 25.207 25.898 51.11
S03 799.7 836.3 6.076 54.124 60.20
S04 798.0 828.1 9.223 44.110 53.33
S05 785.9 809.0 9.523 38.235 47.76
S06 772.0 787.6 13.956 29.951 43.91
S07 866.1 887.3 63.392 49.979 113.37
S08 855.9 858.8 28.631 16.649 45.28
S09 861.9 885.7 119.647 72.790 192.44
S10 857.8 895.0 85.882 79.422 165.30
S11 778.4 779.1 64.764 21.442 86.21
S12 801.9 815.3 30.305 36.798 67.10
S13 800.0 815.4 11.365 30.645 42.01
S14 790.9 809.1 6.398 27.618 34.02
S15 802.8 809.4 35.138 26.423 61.56
S16 792.5 802.5 44.650 31.454 76.10
S17 797.5 798.1 31.060 15.253 46.31
S18 790.5 788.6 12.828 4.996 17.82
S19 805.6 805.9 5.840 4.279 10.12
S20 807.7 817.4 38.927 33.910 72.84
S21 787.2 805.0 35.171 34.870 70.04
S22 803.4 820.4 32.696 36.020 68.72
S23 752.1 751.6 24.196 10.904 35.10
S24 777.3 795.7 53.807 41.171 94.98
Notes: 1. All temperatures in °F.
3-14
12762100
Test Program
Temperature Droop by Sensor
300
200
DegF
100
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Sensor
Temperature Gain by Sensor
300
250
200
DegF
150
100
50
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Sensor
Temperature Change by Sensor
400
300
DegF
200
100
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Sensor
Figure 3-11
Cleaning Events Surface Temperature Characteristics vs. Sensor
3-15
12762100
Test Program
Counts of Events with TChange > 0 DegF
1400
1200
1000
Count
800
600
400
200
0
1
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Sensor
1200
1000
Count
800
600
400
200
0
1
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Sensor
30
25
Count
20
15
10
5
0
1
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Sensor

5
4
Count
3
2
1
0
1
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Sensor
Figure 3-12
Cleaning Events Temperature Change by Sensor for 2003
3-16
12762100
Test Program
Figure 3-13
Cleaning Events Temperature Droop vs. Zone for 2003
3-17
12762100
Test Program
Figure 3-14
Cleaning Events Temperature Gain vs. Zone for 2003
3-18
12762100
Test Program
Figure 3-15
Cleaning Events Temperature Change vs. Zone for 2003
3-19
12762100
Test Program
Heat Flux vs. Load
30
25
20
kBtu/h/ft2
15
10
5
0
(100,200] (300,400] (500,600] (700,800]
Load Category, MW
Figure 3-16
Heat Flux vs. Load for 2003
3-20
12762100
Test Program
Surface Temperature vs. Load
800
700
DegF
600
500
400
(100,200] (300,400] (500,600] (700,800]
Load Category, MW
Figure 3-17
Surface Temperature vs. Load for 2003
3-21
12762100
Test Program
Figure 3-18
Heat Flux vs. Load for 2003 (Scatter)
3-22
12762100
Test Program
Figure 3-19
Heat Flux vs. Surface Temperature for 2003
3-23
12762100
Test Program
Figure 3-20
Heat Flux vs. Sensor for 2003
3-24
12762100
Test Program
Figure 3-21
Surface Temperature vs. Sensor for 2003
Waterwall Tube Life
The evaluation of the potential impact of the technology on furnace wall tube life was a primary
goal of this project. Boiler tube temperatures collected from the 24 CBI-installed heat flux
sensors along with the cleaning frequency were used to qualitatively estimate the impact. Also,
the temperatures were used to drive boiler tube failure models to estimate the impact. Since the
damage mechanisms may take several years to actually result in appreciable damage, direct
observation of the impacts over the duration of the project was not feasible (especially with the
anticipated reduced cleaning cycles with the water cannon systems). However, if the opportunity
arose, direct measurement and analysis of tube damage would be performed.
3-25
12762100
Test Program
Field Testing and Data Collection
Field tests were conducted to assist in the water cannon evaluation. The CBI-installed I/O panels
were used to provide access to thermocouple signals from respective heat flux sensor zones.
Leads were connected to the I/O channels of the associated zone being cleaned by the currently
activated water cannon. Temperature profiles for both surface and subsurface (relative to the
fire-side of the tube) were acquired with an Agilent 34970A data acquisition system equipped
with a cold-junction compensated thermocouple card. The data acquisition period was
approximately 10 ms. This was faster than the CBI “fast data” collection scheme on the water
cannon controller module (approximately 100 ms) and was intended to record sub-100 ms
transients that the CBI system did not detect. Due to the scope of the developed impact model,
only temperature data was collected; other dynamic and static stresses6 were not measured, and
were assumed constant and equal for each sensor location in the boiler. Unless otherwise noted,
CBI water cannon events were either anticipated from the unit control room via the CBI queue
panel or manually triggered.
Thermal Impacts
One concern in sootblowing is the impact of thermal fatigue on the waterwall tube material
[9][13]. This thermal fatigue may be associated with quench cracking. Fatigue failure theory
suggests that cyclic temperature changes of the boiler tube material, regardless of the
temperature profile shape, will decrease material life by some amount. By reducing these cyclic
stresses in amplitude, frequency, or a combination thereof, quench cracking effects could be
minimized. Stress-life (S-N) curves are typically used to determine cyclic failure, but are
constructed for simple, constant stress amplitudes. Miner’s Rule is applied for simple cases
where stress amplitudes are constant. When complex stress profiles are encountered, such as
those derived from temperature changes resulting from sootblowing, a more detailed life analysis
method must be employed.
One such method is the rainflow cycle counting fatigue analysis [2][3]. Using this approach, an
index is produced to estimate the composite damage created from complex cyclic stress profiles.
A higher index is indicative of a higher level of damage. For this project, the stress profiles are
calculated from a given temperature history as recorded from the CBI heat flux sensors. Only
temperature effects are considered in the model; all other stresses are assumed equal for
comparison purposes. The stress-strain is assumed to be one-dimensional for simplicity.
Thermal stress is defined by:
σ (t ) = −αE∆T (t )
where
∆T (t ) = T (t ) − Tmax
6
Such as hoop stress resulting from steam/water pressure and stress resulting from weight of tubes.
3-26
12762100
Test Program
Stress levels are assumed to be in the range where high cycle fatigue failure theory is valid.7
Cracking is also assumed to be non-existent so that fracture mechanics are not applied. Dr. Mark
Barkey, a mechanics of materials professor at the University of Alabama, was commissioned to
develop and implement the rainflow cycle calculation in a stand-alone software module.
Temperature changes of less than 3ºF (1.7ºC) are considered noise and hence filtered from the
thermocouple data prior to performing cyclic calculations in the model. Testing with the
inclusion of the 3ºF (1.7ºC) data added insignificant contribution to the calculated damage index;
therefore, the elimination of such data speeds processing.8 Endurance limits are ignored. The
thermal expansion coefficient and the modulus of elasticity for steel, and the ultimate strength
for SA-210-A1 (tube material at Miller) were used in the calculations (α = 6.53x10-6/ºF,
E = 30x103 ksi, UT = 60 ksi).
The stress levels calculated and used in this report (due to thermal expansion and contraction)
may be overestimated, as the model is based upon a fixed-end beam formula. In actuality, there
is some freedom of deformation with the actual amount of constraint being between that of a
completely free and fixed-end model. Both the uneven heating of the furnace and the rapid
cooling from sootblowing induce local thermal stresses to which the simplified model is
applicable. For absolute determination of stress conditions, a more extensive experiment would
need to be conducted. It should also be noted that the tube material in some water cannon spray
zones may be different (such as SA-213-T22) than that which was modeled; however, the focus
of the model is to compare relative impacts and not for total stress determination. It is therefore
important to note that this index, for the scope of this work, is a measure of the relative impact of
thermal cycling, not an absolute damage indication.
Because of the small magnitudes of the damage index, a normalized relative damage index was
calculated. Assuming that a reasonable steady state damage index is not less than 10-30 (a valid
assumption for this study), the actual indices were divided by this best-case baseline. The log of
this normalization was then taken, resulting in an index more palatable for the casual reader (this
also indicates that a single point rise in the normalized index equates to an increase of one
magnitude of the raw index).
 Di 
Dni = log10  −30 
 10 
Model Test Case Application
Several numerical cases were created to demonstrate the relative damage index calculation. All
theoretical test cases were created with equal periods for ease of comparison. Maximum and
minimum temperatures were arbitrarily chosen at 950°F (510ºC) and 650°F (343ºC),
respectively. Case temperature profiles are shown in Figure 3-22.
7
Strain-based methods are recommended for low cycle fatigue analysis. While dependent upon material, generally
it is applied to stress levels in the vicinity of 0.9 Su or greater for steel, which corresponds to a temperature change
of approximately 275ºF. This paper’s analysis assumes this to be an infrequent occurrence regarding model
applicability.
8
The studies were run both with and without the "noise" filter. The results were not appreciably different.
3-27
12762100
Test Program
Cases 1-4 demonstrate similar profiles to those seen in actual data collected from the CBI heat
flux sensor thermocouples at Miller. Relative damage indices were calculated for these cases.
Case 1 and Case 2 produced identical damage indices. This is reasonable since the morphology
of the temperature profiles is similar. Case 3 shows a simulated “double hit”, causing a second
50°F (28ºC) drop in temperature midway through the temperature rise. The index calculated for
this case was very close to those generated in Cases 1 and 2. This is due to the fact that the large
temperature rise, from 650 to 950ºF (343ºC to 510ºC) is dominant in the damage determination,
such that the damage due to the small 50°F (28ºC) intermediate drop is secondary, though
present. Case 4 produced a damage index similar to Cases 1 and 2 as the temperature profile is
also similar.
Cases 5-10 are of academic interest and are not necessarily intended to mimic an actual scenario.
Case 5 represents the estimated profile of a steady-state temperature. The resulting actual
damage index is zero, as might be expected (since the log of zero is undefined, the normalized
index cannot be calculated). Case 6 and Case 7 demonstrate a temperature rise and fall of 50°F
(28ºC), respectively. Normalized damage indices are identical, indicating that the direction of
the temperature change is independent to the amount of damage. Cases 7-9 indicate the relative
damage created by different magnitudes of temperature changes of 50, 100, and 200°F (28ºC,
56ºC, and 112ºC). The calculated results show that a rise of temperature by 50°F (28ºC)
increases thermal impact damage, and the 200°F (111ºC) total rise is significantly worse than the
base 50°F (28ºC) case. Case 10 is a doubled period of the Case 9 temperature profile to
demonstrate the effect on the normalized damage index vs. actual damage index. The
normalized damage index, while slightly higher in Case 10, is not much different than the single
thermal event shown in Case 9; however, the actual damage created from this profile is double
that of Case 9. One should recall that when using the normalized index for comparison purposes,
the log of the actual index is used, such that changes in the normalized indices translate to
exponential changes in the actual damage. A summary of corresponding normalized and actual
damage indices is given in Table 3-4.
Figure 3-23 shows the results of the estimated relative life of the temperature profile theoretical
cases. The average stress amplitude values for the S-N curves in this report are estimated from
the difference between maximum and minimum temperatures of the given profile.
− αE (Tmin − Tmax )
σ amplitude =
2
As expected, higher periodic stresses have lower relative life estimates. It should again be
emphasized that in all of the S-N curves presented in this report, the life cycle value represents
relative life based exclusively on the temperature profiles and can only be used to draw
comparisons among the plotted points. The plots should not be used as design data or as a
definitive measure of material cycle-to-failure analysis.
Model Field Data Application
Field data were collected with the Agilent 34970A and analyzed in the thermal fatigue model.
Steady-state temperatures were recorded to provide baseline measurements with respect to
3-28
12762100
Test Program
calculated damage indices. All temperature data were filtered as necessary to eliminate invalid
values. The results from the thermal fatigue analysis as well as the associated temperature
profiles are discussed. The water cannon system was placed in automatic control, allowing the
control system to determine when each zone was cleaned.
Temperature profiles are shown in Figure 3-24 through Figure 3-28 for five sootblowing events.
These plots show the measured surface and subsurface transient temperatures of the waterwall
tubes from the CBI sensors during a sootblowing event. Either water lances or the CBI water
cannons were used to clean the walls. Fatigue calculations were made on each case for both the
surface and subsurface temperatures for comparison.
Figure 3-24 exhibits a temperature profile that seemed to be common among the temperature
histories acquired during test periods. Prior to sootblowing, the temperature is relatively steady.
When the cleaning cycle for this section is initiated, there is an almost immediate temperature
drop. This is expected as the water stream impacts the walls. Following this drop, there is an
upward ramp during a temperature recovery period. Ideally, this rise would continue until a
higher, slag-free, steady-state temperature is reached. However, there is a second drop in
temperature after the initial hit. This may be occurring because the water jet spray overlaps a
previously cleaned “path”. While this “double hit” is not desirable (as it shortens tube life), the
damage created from the secondary temperature drop is insignificant compared with the damage
created from the initial temperature drop. Adjustments can be made to minimize or eliminate
this effect within the CBI cannon-tracking control system by adjusting the cannon spray patterns
though possibly at the expense of not cleaning some inter-path areas. Figure 3-25 shows a
separate water cannon sootblowing event. The overlapping spray effect is seen in this profile
also. Temperature recovery is only about 10°F (6ºC) on the surface of the tube possibly
suggesting that the tube was not heavily slagged.
In Figure 3-26, the temperature remains fairly constant until the water cannon jet crosses the
zone where the sensors are located. A drop in temperature caused by water to tube impingement
followed by a rapid rise in temperature was expected, as this trend can be seen in other recorded
sootblowing events. Instead, the sharp temperature drop is not present and, whereas the rise in
temperature is evident, the temperature experiences a downturn and nearly returns back to the
original temperature before the sootblowing event is completed. It is unclear why this occurred
since other sootblowing events demonstrate the typical characteristics described formerly. One
thought is that the section was not completely cleaned during this event and, therefore, the water
from the cannon did not completely penetrate the slag to the wall, preventing the sharp drop in
temperature from water impingement on the sensors. This might also account for only the small
recovery in temperatures after sootblowing was complete. Due to the CBI control system
restrictions, the section was not permitted to be automatically cleaned again. Such constraints
prevent failing sensors from triggering a cleaning of bare tubes, which would be detrimental to
tube life.
Figure 3-27 and Figure 3-28 show the temperature responses of water lance cleaning cycles. A
characteristic of water lance cleaning is the cyclic pattern due to the spiraling motion of the lance
during activation. This is evident in both figures, though Figure 3-27 exhibits larger initial
temperature swings whereas Figure 3-28 exhibits more cycles.
3-29
12762100
Test Program
The relative fatigue damage indices for these tests are given in Table 3-5. Both surface and
subsurface relative damage indices were calculated. Figure 3-29 shows the relative life of each
temperature profile. For each profile pair, the lower stress point is associated with the subsurface
temperature profile.
Model results are mixed. Both the highest and lowest damage indices are attributed to water
cannon events (Run 1 and Run 0, respectively). Run 0 data indicates a temperature change,
maximum to minimum of approximately 20°F (11ºC), and recovers to 7°F higher (4ºC) than the
original temperature. Run 1, however, displays a temperature range of about 160°F (89ºC).
There is a double hit recorded in this data as well, which also contributes to its higher relative
damage index. In all tests, as might be expected due to the greater temperature changes, surface
damage indices are higher than subsurface indices. This is assumed to create the most damage
(when comparing damage from surface to its associated subsurface temperature profile), and
therefore, surface thermocouple data can be used for estimation of the extent of localized thermal
damage in further testing.[jns1] This also supports the fact that most cracks initiate at the surface
of a component, and therefore, using these surface temperatures would provide a more accurate
estimation of the relative life of the tube.
Figure 3-30 shows a comparison of the data collected by Southern Company staff to the data
automatically logged by the CBI system. On initial sootblower activation, the CBI system
records “fast data” – temperature sensor data that is logged every 100 ms. There was some
concern that unobserved transients may be occurring in between the fast data logging intervals.
The Agilent 34970A recorded data at approximately 10 ms intervals to determine whether or not
the concern was substantiated. From the data recorded, there was no evidence of significant
transient events occurring between the CBI fast data logging frequency; however, another
phenomenon was detected.
The fast data is recorded for a particular zone as long as the zone is being cleaned. This makes
more efficient use of media storage space. As soon as the zone’s cleaning cycle is completed,
the fast data recording for that section is deactivated and switched to a thirty-second interval data
recording scheme. As shown in Figure 3-30, there are events that were captured in the Agilent
after the CBI fast data had completed its data recording cycle. This data shows there are two
additional temperature drops that occur after the initial sootblowing event. It is thought that
these may be overlaps from adjacent zones being cleaned. Regardless, these events will impact
fatigue life from thermal cycling.
Additional data was collected with the Agilent 34970A at approximately 10 ms intervals. The
unit was at full load (~700 MW). Instead of automatic activation, the cleaning zones were
monitored and the associated water cannons or water lances were manually activated after a zone
reached a heat flux value that indicated the surface was fouled. This was done to estimate the
results of cleaning with water lances versus water cannons if heat flux feedback were present for
both sootblowing methods. Only surface temperature profiles were used in this analysis since
this thermocouple position is much more sensitive to fire-side thermal events (i.e., water
impingement on the fire-side of the tubes).
3-30
12762100
Test Program
Figure 3-31 shows a temperature profile of a panel being directly cleaned by a water lance.
Since the CBI heat flux sensors were installed to support water cannon zones, there is not a direct
correlation for temperature sensors, cleaning zones, and water lances. This water lance was
selected because it was located adjacent to a heat flux sensor directly within in a water cannon
cleaning zone (Zone 410). Automatic water cannon cleaning was disabled to allow the lance to
clean a dirty panel based upon the feedback from the heat flux sensors. The temperature
measurements are taken from the surface temperature sensor in this zone. The characteristic
oscillating temperature is not apparent in Figure 3-31. The waterwall temperature rises
approximately 80ºF (44ºC) after the cleaning cycle.
Figure 3-32 shows another zone being cleaned by a water lance. Once again, instead of allowing
the water cannon to automatically clean this zone, the water cannon system was disabled, and a
water lance in the same vicinity was allowed to clean the section. The characteristic temperature
oscillation is seen in the temperature profile, while the lance successfully cleans the area and
brings the operating surface temperature to approximately 925ºF (496ºC). From this short test, it
is demonstrated that feedback from the heat flux sensors may possibly be used to automate
sootblowing with other types of sootblowing equipment.
Figure 3-33 shows the temperature profile of a panel (Zone 410) being directly cleaned by the
water cannon. The double hit seen here is from pattern overlap of the water jet. Damage indices
were computed for comparison between water lance and water cannon cleaning cycles. These
results are presented in Table 3-6. The water cannon index is higher than the water lance
indices, though all indices were similar.
In a separate test series, temperatures were monitored on the panel being cleaned and selected
adjacent panels. This test effort supported the further investigation of the “double hit”
phenomena as shown in Figure 3-30. These do not include instances attributed to water cannon
spray pattern overlap. Spray pattern adjustments might minimize or eliminate this problem.
Rather, the focus of the investigation was concentrated on those areas where temperature drops
were seen while cleaning other or adjacent waterwall zones. The plant observed this behavior in
day-to-day operations as well. Figure 3-34 shows a map of a subset of water cannon zones and
the associated temperature sensors. Each zone was cleaned before taking data. The unit was at
full load during testing (~700 MW). In an attempt to recreate the double hit and measure its
impact on adjacent panels, Zone 106 and Zone 410 were initially cleaned (from previous testing),
providing clean surfaces and thus maximum sensitivity to temperature changes. Directly
following, Zones 408 and 409 were cleaned while Zones 106 and 410 were monitored for
overspray. The findings are discussed below.
As a reference for comparison, Figure 3-33 shows the profile of Zone 410 being cleaned with the
water cannon prior to the testing. The double hit phenomenon is noticeable here, and is
attributed to overlapping spray patterns.
Figure 3-35 shows the temperature profile from Zone 106 during the test cleaning by the water
cannon on Zone 408 and 409. There appears to be no carryover effect in this zone. Figure 3-36
shows a temperature profile from Zone 410, which is located directly above the area being
cleaned with the water cannon. There are actually two cleaning events captured in this graph.
3-31
12762100
Test Program
Zone 408 was cleaned first. There is a modest deviation from the steady temperature, indicated
by the first slight drop. Zone 409 is then cleaned. The impact from this sootblowing event was
much more dramatic than when cleaning Zone 408. This is seen in the second temperature drop
in Figure 3-36. It is thought that carryover of the water jet by the flue gas may be contacting the
tubes above the area being directly cleaned. This was not visually confirmed. The difference in
magnitude between the two double hits may be due to the relative location of the 408 and 409
sections to Zone 410.
Normalized damage indices for these tests are given in Table 3-7. The lowest value was
calculated from the Zone 106 profile shown in Figure 3-35. This is reasonable, as the
temperature appears steady and no thermal disturbances are recorded. The index calculated from
the (indirect) temperature profile shown in Figure 3-36 (for sections 408 and 409 combined) was
lower than the direct cleaning index, and while it was still relatively high when compared to the
steady-state condition (section 106 indirect), and ranked among the highest compared with other
damage indices from previous tests (see Table 3-5 and Table 3-6), it would take approximately
104 similar waterwall cleanings to be as damaging as the single, direct, cleaning event. The
combined temperature profile (effect on Zone 410 by cleaning Zone 408 plus cleaning Zone 409)
was then split and individual damage calculations were performed on the separated data. The
results, also in Table 3-7, show that the majority of the damage occurs from cleaning Zone 409.
This is reasonable, as Zone 409 is directly above the selected cleaning zone. It should be noted
that the temperature measurements for Zones 408 and 409 during this test are made in a cleaned
state, representing a possible worst-case scenario.
Figure 3-37 is a plot of the relative life calculated from the water lance and cannon comparisons,
and the temperature profiles of the unintended cleaning events as tested. The comparison on this
plot shows that selected cleaning events from both types of sootblowing devices can produce
about the same estimated amount of damage for the stress level calculated. Of particular interest
are the results generated from the study of the temperature profile shown in Figure 3-36.
Relative life from both the composite event (combined from the Zone 408 and Zone 409
cleanings as measured on Zone 410) and each individual sootblowing event was calculated.
Comparing these results to that of the actual contiguous event, the relative damage estimated
from the initial temperature drop (from cleaning Zone 408) is insignificant when compared to the
second temperature drop with the larger temperature span (from cleaning Zone 409). The
location on the plot of the composite event lies on nearly the same point as the second individual
event (Zone 409 cleaning) and since the first individual water cannon event (Zone 408 cleaning)
is lower on the curve, it can be implied that most of the damage is caused by the second event.
Neither the composite nor the individual cases produced as much damage as that of direct water
cannon cleaning event (as shown in Figure 3-33).
Damage Index from Long-Term Data
Using the CBI collected fast data, relative damage indices may be calculated based on the peak-
to-peak temperature change observed during any one cleaning cycle. Although the temperature
cycling produced by multiple hits of the water cannon are ignored in the analysis, the results
should be representative of the total damage since the damage mechanism is dominated by the
largest temperature swings. The method is as follows:
3-32
12762100
Test Program
σ = ε ⋅ E ⋅ ∆T Temperature induced stress
1 / 0.085
 1.62 ⋅ SU 
N =  Cycles to failure
 σ 
1
Di = Di −1 + Damage index
N
Iteratively running the temperature transitions through this set of equations, a cumulative damage
index may be obtained. Using the data collected from 20039, the zones represented by Sensors
10 and 9 (Zones 201 and 205, respectively), accumulated significantly more damage than the
other zones represented in the furnace (Figure 3-38). It is interesting to note that although the
zones represented by Sensor 9 had many more cleaning cycles than Sensor 10 (1400 to 300), the
damage to the zones for Sensor 10 was much greater than that for Sensor 9. This can be
attributed to the few (<10) cleaning cycles where the peak-to-peak temperature change was
greater than 400°F (222ºC). Again, this stresses the need to limit the maximum temperature
changes during any one cleaning cycle. The estimated stress caused by this magnitude of
temperature change places the material into the assumed low-cycle fatigue range, for which the
high-cycle fatigue analysis methods employed in this report are considered invalid. However,
because of the low number (<10) of occurrences of this event, these impacts are neglected for the
purposes of this report.
Summary of Waterwall Tube Damage Analysis
Based on an ASTM method for analyzing complex stress histories, a simple model was built that
produced a damage index for use in comparison of temperature effects on the relative material
life. Several generic test cases were developed to demonstrate the model. Using a high speed
data acquisition unit, actual temperature profiles were recorded from the heat flux sensors
installed as part of the CBI water cannon system. These profiles were then analyzed with the
model to produce damage indices and relative material life curves for comparison. During
testing, it was discovered that the water cannons produced a secondary jet impingement on
sections being cleaned. This may indicate jet spray pattern overlap, and it is thought that it may
be somewhat mitigated by adjusting the water cannon control system, though with a trade-off of
the possibility of not completely cleaning the affected area. Also noted by plant personnel and
during SCS testing, water spray impingement was detected on zones not being cleaned, leading
to unintentional temperature cycles on nearby tube sections. This was seen most noticeably on
zones directly above the cleaning zone. Though not conclusive, this could be attributed to the
water cannon jet being swept along the gas flow path to a higher elevation zone. When
comparing water lances and water cannon effects on tube material relative life, there appeared
that there was no conclusive evidence indicating one was better than the other, all other
parameters being assumed equal, especially with the inclusion of a feedback system to trigger
sootblowing events, such as the heat flux sensors.
9
Using fast data from January 1, 2003 through September 30, 2003.
3-33
12762100
Test Program
1000 1000
950 950
900 900
850 850
800 800
Temperature (F)
Temperature (F)
750 750
700 700
650 650
600 600
550 550
500 500
0 5 10 15 20 25 30 0 5 10 15 20 25 30
Time (dimensionless) Time (dimensionless)
Case 1: Dni=23.9 Case 2: Dni=23.9

1000 1000
950 950
900 900
850 850
800 800
Temperature (F)
Temperature (F)
750 750
700 700
650 650
600 600
550 550
500 500
0 5 10 15 20 25 30 0 5 10 15 20 25 30

1000 1000
950 950
900 900
850 850
800 800
Temperature (F)
Temperature (F)
750 750
700 700
650 650
600 600
550 550
500 500
0 5 10 15 20 25 30 0 5 10 15 20 25 30
Case 5: Dni=0.0* Case 6: Dni=14.8

*The relative index for this case is undefined; however no
damage is expected from a steady-state temperature profile..
Figure 3-22
Test Case Temperature Profiles
3-34
12762100
Test Program
1000 1000
950 950
900 900
850 850
800 800
Temperature (F)
Temperature (F)
750 750
700 700
650 650
600 600
550 550
500 500
0 5 10 15 20 25 30 0 5 10 15 20 25 30

1000 1000
950 950
900 900
850 850
800 800
Temperature (F)
Temperature (F)
750 750
700 700
650 650
600 600
550 550
500 500
0 5 10 15 20 25 30 0 5 10 15 20 25 30

Figure 3-22
Test Case Temperature Profiles (continued)
3-35
12762100
Test Program
Table 3-3
Normalized vs. Actual Damage Indices for Modeled Cases
Test Case Actual Index Normalized Index
Case 1 7.7E-07 23.9

Case 2 7.7E-07 23.9
Case 3 7.7E-07 23.9
Case 4 7.7E-07 23.9
Case 5* 0 0*
Case 6 6.8E-16 14.8
Case 7 6.8E-16 14.8
Case 8 2.1E-12 18.3
Case 9 6.9E-09 21.8
Case 10 1.4E-08 22.1
*The relative index for this case is undefined, as the log of 0 is undefined.
3-36
12762100
Test Program
35
30 Cases 1-4
25
Case 10
20
Stress (ksi)
Case 9
15
Case 7
10
Case 8
Case 6
Note: Case 5 has apparant infinite life
0
1.0E+00 1.0E+01 1.0E+02 1.0E+03 1.0E+04 1.0E+05 1.0E+06 1.0E+07 1.0E+08 1.0E+09 1.0E+10 1.0E+11 1.0E+12 1.0E+13 1.0E+14 1.0E+15 1.0E+16
Relative Life
Figure 3-23
S-N Curve for Test Cases
3-37
12762100
Test Program
900
850
800
Temperature (F)
750
700
650
600
0 20 40 60 80 100 120
Time (s)
Surface Subsurface
Figure 3-24
Temperature Measurements During Cannon Event – Test ID: Run 1
3-38
12762100
Test Program
770
760
750
740
Temperature (F)
730
720
710
700
690
0 20 40 60 80 100 120
Time (s)
Surface Subsurface
Figure 3-25
Temperature Measurements During Cannon Event – Test ID: Run 4
3-39
12762100
Test Program
790
780
770
Temperature (F)
760
750
740
730
720
0 20 40 60 80 100 120 140 160 180
Time (s)
Surface Subsurface
Figure 3-26
Temperatures Measured During Cannon Event – Test ID: Run 0
3-40
12762100
Test Program
780
760
740
Temperature (F)
720
700
680
660
0 20 40 60 80 100 120
Time (s)
Surface Subsurface
Figure 3-27
Temperature Measurements During Water Lance Event – Test ID: Run 2
3-41
12762100
Test Program
790
780
770
760
Temperature (F)
750
740
730
720
710
700
0 20 40 60 80 100 120 140
Time (s)
Surface Subsurface
Figure 3-28
Temperature Measurements During Water Lance Event – Test ID: Run 3
3-42
12762100
Test Program
Table 3-4
Normalized Damage Indices from Field Data
Test ID Normalized Normalized
Surface Subsurface
Index Index
Run 0 11.4 9.8
Run 1 20.4 17.8
Run 2 18.2 15.9
Run 3 15.2 12.6
Run 4 13.3 10.5
16.0
14.0
12.0
Run 1
10.0
Stress (ksi)
8.0
Run 2 Run 3
6.0
Run 0
4.0
2.0
Run 4
0.0
1.0E+06 1.0E+07 1.0E+08 1.0E+09 1.0E+10 1.0E+11 1.0E+12 1.0E+13 1.0E+14 1.0E+15 1.0E+16 1.0E+17 1.0E+18 1.0E+19 1.0E+20 1.0E+21
Relative Life
Figure 3-29
S-N curve for Temperature Profiles
3-43
12762100
Test Program
950
900
850
Temperature (F)
800
750
700
12:53:08 12:53:51 12:54:35 12:55:18 12:56:01 12:56:44 12:57:27
Time
CBI Computer Agilent Surface Agilent Subsurface
Figure 3-30
CBI Fast Data Comparison with Agilent 34970A Data
3-44
12762100
Test Program
950
900
850
Temperature (F)
800
750
700
0 20 40 60 80 100 120 140 160 180 200
Time (s)
Figure 3-31
Temperature of Panel Being Cleaned By Water Lance
3-45
12762100
Test Program
950
900
850
Temperature (F)
800
750
700
0 20 40 60 80 100 120 140 160 180 200
Time (s)
Figure 3-32
Temperature Profile of Panel Cleaned Directly by Water Lance
3-46
12762100
Test Program
950
900
850
Temperature (F)
800
750
700
0 10 20 30 40 50 60 70 80 90 100
Time (s)
Figure 3-33
Temperature of Panel Being Cleaned by Water Cannon
3-47
12762100
Test Program
Table 3-5
Normalized Damage Indices from Water Lance vs. Water Cannon Comparison
Test Case Normalized Index
Water Lance 71 18.8
Water Lance 9 18.5
Water Cannon 19.3
CL Left Side
Front Wall
Zone 106
A06
Zone 410
A10
Zone 409
A08
Zone 408
A08
Figure 3-34
Waterwall Cleaning Zone Layout Showing Associated TC Sensors (‘A’ designation)
3-48
12762100
Test Program
950
900
850
Temperature (F)
800
750
700
0 10 20 30 40 50 60 70 80 90
Time (s)
Figure 3-35
Temperature During Water Cannon Cleaning Event – Zone 106
3-49
12762100
Test Program
950
900
Zone 409 cleaned

Zone 408 cleaned
850
Temperature (F)
800
750
700
0 20 40 60 80 100 120 140 160
Time (s)
Figure 3-36
Temperature During Water Cannon Cleaning Event – Zone 410
3-50
12762100
Test Program
Table 3-6
Normalized Damage Indices from Testing with Water Cannon Sootblowing
Test Case Normalized Index
A10 – Direct – 410 19.3
A10 – Indirect – 106 7.0
A10 – Indirect – 408,409 15.7
A10 – Indirect – 408 (subset) 11.4
A10 – Indirect – 409 (subset) 15.7
14
A10-Direct
12
B7-WL71
A10-WL9
10
8
Stress (ksi)
A10-Indirect-408,409
6
A10-Indirect-409 (subset)
4
A10-Indirect-106
2 A10-Indirect-408 (subset)
0
1.0E+08 1.0E+09 1.0E+10 1.0E+11 1.0E+12 1.0E+13 1.0E+14 1.0E+15 1.0E+16 1.0E+17 1.0E+18 1.0E+19 1.0E+20 1.0E+21 1.0E+22 1.0E+23 1.0E+24 1.0E+25
Relative Life
Figure 3-37
S-N Curve for “double hit” Testing
3-51
12762100
Test Program
4.5E-04
4.0E-04
3.5E-04
3.0E-04
Damage Index
2.5E-04
2.0E-04
1.5E-04
1.0E-04
5.0E-05
0.0E+00
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Sensor
Figure 3-38
Damage Index as Calculated from Long-Term Data (2003)
Condensate Requirements
One of the expected benefits of the technology was the reduction in condensate make-up
requirements for the unit. The condensate usage with the water lances in place was approaching
the capacity (within 15%) of the water treatment plant at Miller. The water treatment plant has a
capacity of approximately 425 million gallons per year (1.6⋅106 m3/y) with each of the four units
at Miller nominally using approximately 100 million gallons per year (0.4⋅106 m3/y). The water
lances use condensate make-up as the cleaning medium while the water cannons use strained
river water.
With the installation of the water cannons, the use of the water lances was virtually eliminated.
Unfortunately, water lance flow rates were not measured on Unit 1 and therefore there was no
direct indication of the reduction in this flow. However, there are circumstantial indications of
the benefits obtained in this regard.
The average daily consumption for the thirteen months prior to the installation was 230,206
gallons (871 m3). This is not including the two months when the unit was down. For the 22
months after the installation, the average daily condensate consumption was 155,740 gallons
(589 m3) (Figure 3-39). Considering generation output, the corresponding average rates are 18
gals/MWh (68 L/MWh) and 12 gals/MWh (45 L/MWh) for the corresponding pre- and post-
installation periods.
3-52
12762100
Test Program
Although the water lance flow was not measured, it can be estimated based on the frequency of
cleaning and the flow of the individual water lances. Standard plant operating practice was for
the water lances to cycle continuously yielding approximately 225 water lance activations per
day with a single activation requiring approximately 120 gallons (40 gpm, 3 minute duration).
Using these figures, the make-up water reduction would be approximately 27,000 gallons/d (102
m3/d) or near 10% of total unit make-up requirements.
In summary, the installation of the water cannons in February 2002 created a substantial
reduction in condensate makeup requirements for the plant.
350000
300000
Condensate Makeup (Avg Gal Per Day)
250000
200000 2001
2002
150000 2003
100000
50000
0
1 2 3 4 5 6 7 8 9 10 11 12
Month
Figure 3-39
Condensate Makeup
Furnace Heat Absorption, Boiler Efficiency, and Emissions
The water cannon installation was expected to have an impact on furnace heat absorption. Also,
CBI speculated that these changes could have a positive impact on boiler efficiency. A goal of
the test program was to evaluate these changes within the test program constraints. Two distinct
approaches were used to make this evaluation. After the system had been installed, CBI
conducted tests to estimate the impact of the water cannons installation on boiler performance
and emissions [11]. Also, pre- and post-installation performance metrics were compared. The
primary sources of data for the analysis included the DCS / PI system, continuous emissions
monitor, and data collected in the CBI system. The findings of these tests are discussed below.
3-53
12762100
Test Program
CBI Conducted Performance Tests
Following startup of the water cannon system but prior to its full optimization, CBI conducted
tests to estimate the impact of the system on cleaning effectiveness, thermal impact, and boiler
efficiency. This testing was conducted from February 18 to 21, 2002. In total, seven tests were
conducted (Table 3-8). Plant instrumentation was used for this testing. This instrumentation,
although considered reliable and appropriate for normal plant use, would not be considered
appropriate without special precautions for use in more exacting ASME Performance Tests [1].
In addition to relying on plant instrumentation, test methods included:
HVTs – Flue gas temperatures were measured at the 468 foot (142 m) elevation in the furnace
(approximately 20 feet above the nose) using high velocity thermocouple (HVT) traverses. This
testing was conducted by Innovative Combustion Technologies. Samples were drawn from the
locations shown in Figure 3-40.
Coal and ash sampling – Coal and ash samples were taken daily for lab analysis. These samples
were taken per normal plant operating procedures. For the coal, samples were obtained in
accordance with ASTM procedures using a cross-belt type mechanical sampling system. Ash
samples were collected from the precipitator hoppers.
Heat flux transmitters – Data was collected from the twenty-four CBI installed heat flux sensors.
This data was collected in the CBI control system.
The results of this testing are shown in Table 3-9 through Table 3-11 and Figure 3-42 through
Figure 3-48. All tests were conducted at full-load and steady-state operation. Boiler
performance parameters, including efficiency, steam temperatures, and gas outlet temperatures,
did not vary substantially amongst the seven tests indicating that the water cannons had little
impact on these parameters as compared to operation with the water lances. This was also the
case for NOx emissions measured at the stack and the nose of the furnace. However, in that the
furnace was relatively clean for all these tests, it not unreasonable to expect little performance
gains. For units that exhibit heavier waterwall slagging, performance benefits may have been
greater.
The water cannons did appear to improve the furnace exit gas temperature (FEGT) distribution
as measured with the HVTs. Mean temperature was reduced by approximately 50°F (28°C) as
compared to the baseline (Test 1). All other factors being equal, this would support the
conclusion that the furnace walls were cleaner when using the water cannons than when using
water lances, though this conjecture is not supported by the mean furnace wall heat flux readings
(Table 3-9). Also, the plant's FEGT measurement, located in the superheat area, did not show
the same reduction in gas temperatures.
3-54
12762100
Test Program
Table 3-7
CBI Conducted Performance Tests Summary
Test No. Date Load Start: Finish: Conditions

1 18-Feb-02 Full 12:30 3:30 Water Lance cleaning over weekend
Load before test
2 19-Feb-02 Full 9:15 11:10 Water Lance and Water Cannon
Load cleaning overnight before test
3 19-Feb-02 Full 13:27 15:30 Water Cannon in Program Mode
Load before and during test
4 20-Feb-02 Full 8:00 9:42 Water Cannon in Automatic mode
Load overnight
Load
Load adjusted JPV & press
Load with automatic heat flux set points, 1
1/2 hour of no cleaning from 10:30 to
Noon. Special large zone test at
12:15.
50
45
Nose
40
35
30
tf Partition Wall
,l
l Sample Location
a 25
W
t
s
a 20
E
15
10
0
0 10 20 30 40 50 6
North Wall, ft
Figure 3-40
CBI Conducted Performance Tests – HVT Sample Locations
3-55
12762100
Test Program
Table 3-8
CBI Conducted Performance Tests Results
Ambient Conditions
Coal Ambient Baro Relative
Test Date Start: Finish: Load, MW Flow, % Temp, °F Pres, inHg Humidity, %
1 18-Feb-02 0:30 15:30 708 85 41 29.5 74.6
2 19-Feb-02 9:15 11:10 708 85 54 29.7 67.8
3 19-Feb-02 13:27 15:30 708 84 62 29.7 68.5
4 20-Feb-02 8:00 9:42 710 84 58 29.5 97.0
5 20-Feb-02 13:00 15:45 709 83 67 29.8 48.1
6 21-Feb-02 8:00 10:00 709 83 49 29.9 70.3
7 21-Feb-02 13:00 14:50 710 82 67 30.0 32.5
Air Heater Outlet Air Heater Air Inlet Air Heater Air Outlet CBI
Gas Temperature, °F Temp, °F Temp, °F Dry Flue Gas Boiler
Test Left Right Left Right Left Right Loss, % Efficiency
1 225 222 56 54 504 521 2.69 87.7%
2 224 222 69 68 508 534 2.61 87.5%
3 224 224 76 75 512 534 2.58 87.6%
4 226 229 73 72 516 537 2.65 87.4%
5 228 229 81 80 515 539 2.64 87.7%
6 218 219 63 63 512 533 2.69 87.5%
7 223 224 80 80 518 541 2.58 87.8%
Plant Economizer Outlet Emissions

Furn Exit Gas Temp, °F Excess Oxygen, % Gas Temperature, °F Stack Stack
Test Left Right Left Right Left Right NOx, lb/Mbtu CO2, %
1 2,179 2,142 2.59 2.80 611 638 0.282 14.3
2 2,116 2,114 2.76 3.03 616 643 0.290 14.3
3 2,133 2,138 2.78 3.02 617 643 0.287 14.0
4 2,138 2,140 2.92 2.89 624 650 0.289 14.1
5 2,146 2,137 3.02 2.80 623 651 0.282 14.0
6 2,151 2,165 2.87 2.91 620 649 0.273 14.7
7 2,159 2,194 2.90 2.93 625 649 0.262 14.5
Feedwater Final SH SH Spray Throttle Final RH RH Spray Hot Reheat

Test Flow, klb/h Temp, °F Flow, klb/h Pres, psig Temp, °F Flow, klb/h Pres., psig
1 4604 1005 238 2406 987 26 503
2 4645 1005 193 2400 989 35 503
3 4588 1004 243 2398 991 36 504
4 4494 1004 293 2400 998 39 505
5 4582 1004 240 2400 993 37 503
6 4588 1004 222 2405 996 35 505
7 4576 1002 234 2405 996 37 505
HVT
Test Temp. CO, ppm NOx, ppm O2, % NOx,
lb/Mbtu
1 1973 244 229 4.8 0.35
2 1905 338 219 5.0 0.34
3 1928 316 190 4.5 0.28
4 1908 476 205 4.7 0.31
5 1935 213 198 4.6 0.30
6 1899 349 193 4.6 0.29
7 1928 418 182 4.7 0.27
Heat Flux, Btu/h/ft^2

Test Avg Max Min Stand. Dev.
1 29.6 93 1 15.2
2 30.2 89 0 14.9
3 28.0 72 4 11.6
4 25.2 52 0 9.1
5 26.4 58 0 9.5
6 27.0 52 0 8.7
7 30.4 86 0 14.9
3-56
12762100
Test Program
Table 3-9
CBI Conducted Performance Tests – Heat Flux by Sensor
Test
Sensor 1 2 3 4 5 6 7 Mean Max Min StdDev
1 22.4 24.2 23.6 23.3 23.2 23.2 23.2 23.3 24.2 22.4 0.55
2 18.2 22.6 23.4 21.6 21.8 21.8 21.8 21.6 23.4 18.2 1.64
3 22.1 24.0 25.2 23.7 23.8 23.8 23.8 23.8 25.2 22.1 0.91
4 25.6 24.2 24.4 25.0 25.5 25.5 25.5 25.1 25.6 24.2 0.59
5 18.3 20.1 20.3 19.2 19.1 19.1 19.1 19.3 20.3 18.3 0.69
6 12.1 11.2 11.0 11.2 11.2 11.2 11.2 11.3 12.1 11.0 0.35
7 43.1 39.9 34.0 37.6 35.5 35.5 35.5 37.3 43.1 34.0 3.18
8 44.2 37.7 39.7 39.9 40.9 40.9 40.9 40.6 44.2 37.7 1.95
9 53.6 58.4 49.4 50.2 44.6 44.6 44.6 49.3 58.4 44.6 5.29
10 35.2 40.3 32.2 35.3 34.0 34.0 34.0 35.0 40.3 32.2 2.54
11 36.7 39.5 34.4 34.6 33.6 33.6 33.6 35.1 39.5 33.6 2.19
12 30.6 24.9 22.2 24.9 23.2 23.2 23.2 24.6 30.6 22.2 2.80
13 28.3 35.1 35.0 33.4 34.8 34.8 34.8 33.8 35.1 28.3 2.49
14 27.2 33.4 32.6 31.0 31.7 31.7 31.7 31.3 33.4 27.2 1.99
15 17.0 16.5 17.3 17.1 17.5 17.5 17.5 17.2 17.5 16.5 0.37
16 33.3 30.5 28.7 30.3 30.1 30.1 30.1 30.5 33.3 28.7 1.37
17 11.9 11.6 11.5 11.4 11.7 11.7 11.7 11.6 11.9 11.4 0.15
18 16.6 17.1 18.2 17.0 17.0 17.0 17.0 17.1 18.2 16.6 0.49
19 34.4 33.7 34.2 33.7 32.6 32.6 32.6 33.4 34.4 32.6 0.77
20 44.0 40.3 32.4 38.1 36.5 36.5 36.5 37.8 44.0 32.4 3.65
21 18.9 22.2 21.6 20.7 21.3 21.3 21.3 21.0 22.2 18.9 1.06
22 76.8 76.0 61.6 67.0 55.2 55.2 55.2 63.9 76.8 55.2 9.63
23 11.6 13.2 14.4 13.0 13.0 13.0 13.0 13.1 14.4 11.6 0.79
24 27.7 28.0 24.8 28.0 28.9 28.9 28.9 27.9 28.9 24.8 1.45
Mean 29.6 30.2 28.0 28.6 27.8 27.8 27.8 28.5 30.2 27.8 0.99
Max 76.8 76.0 61.6 67.0 55.2 55.2 55.2 63.9 76.8 55.2 9.63
Min 11.6 11.2 11.0 11.2 11.2 11.2 11.2 11.2 11.6 11.0 0.19
StdDev 15.2 14.9 11.6 12.8 10.8 10.8 10.8 12.4 15.2 10.8 1.92
Notes: 1. Heat flux in kBtu/h/ft2
3-57
12762100
Test Program
Table 3-10
CBI Conducted Performance Tests – HVT Results
Test 1 - Water lance cleaning over weekend before test (baseline)
TEMPERATURE CO NOx O2
Min. :1430 Min. : 0.0 Min. : 64.0 Min. : 1.150
1st Qu.:1768 1st Qu.: 10.0 1st Qu.:220.0 1st Qu.: 3.650
Median :2010 Median : 50.0 Median :226.0 Median : 4.800
Mean :1973 Mean : 243.8 Mean :228.9 Mean : 4.815
3rd Qu.:2148 3rd Qu.: 250.0 3rd Qu.:240.0 3rd Qu.: 5.650
Max. :2338 Max. :2100.0 Max. :280.0 Max. :15.350
StdDev : 218.903 StdDev : 432.76 StdDev : 26.0544 StdDev : 2.20048
Test 2 - Water lance and water cannon cleaning overnight before test
Min. : 948 Min. : 0.0 Min. :155.0 Min. : 1.550
Max. :2288 Max. :4000.0 Max. :284.0 Max. : 8.150
Test 3 - Water cannon in program mode before and during test
Min. :1326 Min. : 0.0 Min. :169.0 Min. : 1.700
Max. :2270 Max. :4000.0 Max. :218.0 Max. : 7.950
Test 4 - Water cannon in automatic mode overnight
Min. :1290 Min. : 0.0 Min. :149.0 Min. : 1.150
Max. :2295 Max. :4000.0 Max. :247.0 Max. : 7.700
Test 5 - Water cannon in automatic mode
Min. :1508 Min. : 0.0 Min. : 82.0 Min. : 2.200
Max. :2294 Max. :1350.0 Max. :242.0 Max. : 9.250
Test 6 - Water cannon in automatic mode adjusted JPV & press
Min. :1139 Min. : 0.0 Min. :162.0 Min. : 2.650
Max. :2297 Max. :3400.0 Max. :226.0 Max. : 6.700
Test 7 - Water Cannon in automatic mode with modified settings
Min. :1224 Min. : 0.00 Min. :145.0 Min. : 2.400
Max. :2340 Max. :2600.00 Max. :212.0 Max. : 6.900
Notes: 1. Units: CO – ppm, NOx – ppm, O2 - percent
3-58
12762100
Test Program
1000 1400 1800 2200

Temperature, °F
12
O2, %
2 4 6 8
3000
CO, PPM
1000
0
NOx, PPM
200
100
1 2 3 4 5 6 7
Test
Figure 3-41
CBI Conducted Performance Tests – HVT Measurement Distributions
3-59
12762100
Test Program
Figure 3-42
Performance Test 1 - HVT Temperatures
3-60
12762100
Test Program
Figure 3-43
3-61
12762100
Test Program
Figure 3-44
3-62
12762100
Test Program
Figure 3-45
3-63
12762100
Test Program
Figure 3-46
3-64
12762100
Test Program
Figure 3-47
3-65
12762100
Test Program
Figure 3-48
Pre- and Post-Installation Operational Data
The water cannons were installed during the Fall 2001 outage with the unit coming back on-line
on December 21, 2001. The water cannons were not fully operational until mid-February 2002.
During the period from unit startup to the water cannons being operational, water lances were
used per normal plant practice. The load profile for this period is shown in Figure 3-49. As
exhibited in Table 3-12, there were improvements in several important performance parameters
including reheat and main steam temperatures and a reduction in reheat spray flows. Although
this may not be directly attributable to water cannon use as opposed to water lances, it is further
evidence that with the use of the water cannons and heat flux sensors, furnace cleaning was
better managed to improve the thermal performance of the unit. The heat flux sensors were an
important factor in this management. Note that NOx emissions during the two periods were the
same.
3-66
12762100
Test Program
800
700
PI offline
600
500
W
M 400
,
d
a
o
L 300
200
100
0
Jan02 Feb02 Mar02 Apr02
Figure 3-49
Unit Load from January 2002 through April 2002
3-67
12762100
Test Program
Table 3-11
Pre- and Post-Installation Performance Parameters (650 – 750 MW)
Water Lances Water Cannons
Tag Description Units Min Average Max Min Average Max
1_F280 Feedwater Flow klb/h 4223.67 4502.68 4601.03 4118.85 4464.99 4600.77
1_F543 Superheat Spray Flow klb/h 254.17 323.65 416.83 251.30 324.03 404.80
1_F563 Reheat Spray Flow klb/h 3.01 65.25 247.52 0.00 0.74 17.66
1_F585 LH Blr Soot Blowing Steam klb/h 3.94 14.86 18.05 4.28 16.32 20.61
1_F590 RH Blr Soot Blowing Steam klb/h 5.90 15.69 19.53 2.77 15.72 19.14
1_K800 Total Coal Flow % 76.79 83.95 87.72 79.37 84.55 91.00
1_N010 Pct O2 LH Average % 2.71 2.95 3.33 2.79 2.96 3.29
1_N011 Pct O2 RH Average % 2.87 3.07 3.49 2.73 3.01 3.37
1_N035 Stack - NOx lb/Mbtu 0.24 0.29 0.73 0.25 0.29 0.48
1_N040 Stack - CO2 Percentage % 10.03 14.08 14.80 12.80 14.00 14.53
1_P000 Barometric Pressure inHg 29.31 29.68 30.11 29.17 29.63 30.40
1_P551 Throttle Steam Pressure A psig 2316.75 2399.66 2411.07 2389.36 2399.49 2409.17
1_P574 Hot Rht At Turb Press psig 473.08 504.95 513.92 464.61 500.35 506.84
1_R000 Relative Humidity % 39.62 85.83 99.39 21.39 61.30 99.10
1_R628 Avg Ambient Air Temp °F 31.04 53.58 73.54 28.88 52.57 79.75
1_T1044 Final Sh Temp °F 964.97 989.61 1004.63 994.53 999.44 1004.40
1_T582 Rht Out LH °F 960.95 985.12 1000.17 976.57 994.93 1003.74
1_T710 A Stm Coil AH Disch °F 45.99 67.96 87.05 44.18 66.75 93.38
1_T715 B Stm Coil AH Disch °F 45.65 67.04 86.28 43.45 66.15 92.64
1_T720 A Sec AH Disch °F 491.56 505.38 515.63 495.71 507.85 523.39
1_T725 B Sec AH Disch °F 510.97 525.55 536.19 517.33 530.90 544.17
1_T908 Econ Out Gas LH °F 586.30 613.93 624.21 594.16 614.55 625.73
1_T910 Econ Out Gas RH °F 616.83 641.38 651.35 621.66 644.92 657.27
1_T930 A Ah Gas Out °F 196.68 217.63 231.23 204.01 216.77 236.29
1_T935 B Ah Gas Out °F 193.63 217.60 230.97 204.08 217.39 237.43
1_W010 Unit 1 Generated Megawatts MW 657.09 706.70 722.25 650.67 703.05 710.65
Water lance period: Jan 21, 2002 – Jan 29, 2002

Water cannon period: Mar 21, 2002 – Mar 30, 2002
3-68
12762100
Test Program
Equipment Reliability
One of the underlying goals of the project was to determine the reliability of the water cannon
system. In order to track the problems associated with the operation of the water cannon
equipment, a log was kept by plant personnel documenting every encountered instance. The
most common problem encountered was associated with the control system. A majority of these
problems were with the cannons tracking the position on the x-axis and y-axis. Although
persistent, these problems were easily repaired by resetting the cannon or replacing the resolvers.
The most severe problem encountered was the hose coming unscrewed from the cannon. This
was quickly realized due to a high flow alarm and the hose was reattached. The main concern
with this type of failure is one of safety for any personnel that may be in the area; no injuries
resulted from this incident.
The problems are broken down by category in Figure 3-50. The categories are: controls (cannon
control equipment failures such as resolvers); software (PC problems, faulty channels);
pump/valve (replaced seals, faulty valves); cannon (hose coming unscrewed, cannon detaching
from the ball joint); instrumentation (sensor failures, panel failures); upgrades (upgraded to new
balls on all cannons for unknown reason). A complete list of problems are provided in Appendix
A.
Despite these few problems, the overall opinion of plant personnel is positive toward the water
cannons. As compared to the water lances that they replaced, the total maintenance is much less.
System Problems By Category
12
11
10
8
Number of Problems
8
7
6 6
6
2
1
Controls Software Pump/Valve Cannon Instrumentation Upgrades
Figure 3-50
SmartCannon System Problems by Category
3-69
12762100
Test Program
Impact on Unit Stability
Sootblowing operations in the furnace may have a significant impact on plant dynamics,
particularly steam temperatures, steam pressures, and drum level. This effect is the result of
changes in the heat transfer characteristics in the furnace and the resulting changes in furnace
exit gas temperature and energy transferred to the fluid in the furnace tubes. Given that the flow
rate for a water cannon is approximately five times that of a water lance and a water cannon zone
is larger than the area cleaned by an individual water lance, the disturbance injected into the
process would be expected to be greater with the water cannons.
This disturbance injection did not turn out to be an issue at Miller. An example of the plant
response to water cannon events is shown in Figure 3-51 through Figure 3-53. There were 145
cleaning events that day (June 1, 2003) with the water cannon cleaning events indicated by the
flow to the water cannons (FLOW_PV, gpm). As evident, the larger excursions in process
variables (main steam temperature and pressure, reheat temperature, and drum level) were
associated with load changes rather than furnace cleaning. The transients introduced as a result
of water cannon operation cannot be readily observed and were evidently minor. Also, the
events did not appear to exacerbate any splits (east to west) in drum level. Further, the water
cleaning events did not materially affect the primary control variables, again with most of the
variation being the result of load changes (Figure 3-53).
Similar results were found for January 1, 2003 (Figure 3-54). As shown, cleaning cycles – there
were 87 that day - occurred only when the load was above approximately 500 MW (there is a
minimum load permissive which is set at that load point) with most of the events occurring when
the unit transitioned from lower loads to higher load at approximately 9:00 am. As before, the
largest excursions were associated with load changes and not water cannon operation.
The low dynamic impact of these cleaning cycles may be the result of several factors including:
• The zones were not overly slagged when cleaned resulting in a relatively small change in
heat adsorption;
• The zones are sufficiently small in area such that the cleaning of any one zone, even if
heavily slagged, does not produce an observable change in steam temperatures and drum
levels;
• The control system and actuators are sufficiently fast to handle the disturbance injected into
the process during the approximate one-minute cleaning cycle.
3-70
12762100
Test Program
Figure 3-51
Process Response to Water Cannon Events – June 1, 2003
3-71
12762100
Test Program
Figure 3-52
3-72
12762100
Test Program
Figure 3-53
3-73
12762100
Test Program
Figure 3-54
Process Response to Water Cannon Events – January 1, 2003
3-74
12762100
4
SUMMARY AND CONCLUSIONS
Plant Miller contracted with Clyde Bergemann to install the SmartCannon waterwall cleaning
quarter 2002.
To obtain a better understanding of the technology and its potential benefits, EPRI and Southern
Company co-sponsored a test program to document its impact on plant operation and
performance metrics such as efficiency, emissions, and reliability.
Reliability – From installation through the end of 2003, the CBI water cannons and heat flux
sensors have been very reliable and have been used almost exclusively for furnace wall cleaning.
Maintenance requirements are much reduced as compared to the water lances also installed on
this unit.
Condensate Usage – The water cannons use strained water instead of condensate water as the
Waterwall Tube Damage – For 2003, the total number of cleaning events with the water cannons
is much reduced (approximately 50%) over what had been normal plant practice with the water
lances. Based on data collected during the test program, the tube damage produced by an
individual water cannon cleaning cycle is not materially different than that observed with the
water lances. The tube failure mechanisms may take several years to manifest to an actual
failure and direct measurement is difficult over a short period. However, based on the frequency
of cleaning and magnitude of the temperature swings, it may be inferred that the tube failure rate
is less following the installation of the water cannons. It is the authors' opinion that the
availability of reliable waterwall heat flux and surface temperature measurements are of great
importance to the successful application of intelligent sootblowing technologies, particularly if
water is used as the cleaning medium. The damage incurred is much more highly dependent on
temperature change than the frequency of cleaning and, as such, should be taken into
consideration when developing the cleaning strategy. At Miller for 2003, the nominal surface
temperature change was 71°F (39°C) but it varied greatly with location and over the course of
4-1
12762100
Summary and Conclusions
the year with several temperature excursions greater than 400°F (222°C). The damage incurred
as a result of the latter is several orders of magnitude greater than that of transients with
temperature changes less than 100°F (56°C).
Heat Transfer Distribution – Furnace heat transfer was more uniform following the installation
of the water cannons. This improvement can in large part be attributed to better understanding of
the furnace slagging condition as provided by the heat flux sensors and not the result of the water
cannons. Similar results would likely be obtained using the water lances if heat flux sensors
were used to manage cleaning cycles.
Efficiency – The use of water cannons did not have a measurable impact on boiler efficiency.
following the installation of the water cannons.
Emissions – NOx emissions were not materially affected with the installation of the water
cannons.
Unit Controllability - The water cannons did not have an adverse impact on plant controllability
or stability.
significant benefit to plant operation primarily through reducing maintenance costs and tube
failure rate.
4-2
12762100
5
REFERENCES
1. ASME Power Test Codes - Test Code for Steam Generators PTC 4.1, American Society of
Mechanical Engineers, New York, NY: 1985.
2. ASTM E1049-85 Standard Practices for Cycle Counting in Fatigue Analysis, ASTM
International, West Conshohocken, PA: 1997.
3. Barkey, M., A Simple Model for Waterwall Tube Thermal Fatigue, prepared for Southern
Company Services, Birmingham, AL: 2003.
4. Blanton, R., Hemperley, R., Courtney, R., "W.A. Parish Units 7 & 8 ISB Project", Fourth
Intelligent Sootblowing Workshop, EPRI, Palo Alto, CA:2002.
5. Blinka, H, "W.A. Parish Unit 5 ISB Project", Fourth Intelligent Sootblowing Workshop,
EPRI, Palo Alto, CA:2002.
6. Carlisle, M., Mead, B., Sorge, J., "Water Cannon Application at Alabama Power Company's
Plant Miller", Fourth Intelligent Sootblowing Workshop, EPRI, Palo Alto, CA:2002.
7. Intelligent Sootblowing at Texas Genco's W. A. Parish Plant, EPRI, Palo Alto, CA, and
Texas Genco, Houston, TX: 2003. 1004113.
8. Intelligent Sootblowing at TVA's Bull Run Plant, EPRI, Palo Alto, CA, and TVA,
Chattanooga, TN: 2003. 1004115.
9. Kessler, R., "Thermal Fatigue Cracking of Waterwall Tubes from Water Lances and Water
Cannons", Fourth Intelligent Sootblowing Workshop, EPRI, Palo Alto, CA:2002.
10. Sheidler, R, Breeding, C., "Performance Testing of Water Cannons at AEP Gavin", Presented
at PowerGen 2003 conference.
11. SmartCannon Performance Report – Alabama Power, Miller Unit 1, Clyde Bergemann,
Atlanta, GA: 2002.
12. Steam its generation and use, Babcock and Wilcox, Barberton, OH: 1992.
13. Water Blowing of Fireside Deposits in Coal-Fired Utility Boilers, EPRI, Palo Alto,
CA:1986. CS-4914.
5-1
12762100
12762100
A
WATER CANNON RELIABILITY
Problems
DateProblem Description StartDate EndDate
3/28/2002 The B heat flux sensor panel has quit 2 times today. Each time the power was
cycled to restore the system. Also I think the sensor program quit Wednesday.
Rebooted the computer to get the system going.
3/23/2002 The hose unscrewed from the fixed end of the #1 Cannon. The operator called
me when he got the high flow alarm. We put the hose back on and checked the
other 3.
3/22/2002 The sensor program stopped again Friday night about 7:15pm. I checked the 3/22/2002 3/23/2002
sensor panels and they were working. I restarted the control room PC and 7:15:00 PM 7:00:00 AM
everthing started working ok.
3/29/2002 B Heat Flux panel quit 3 times today
3/30/2002 B heat flux panel out, changed to waterlances
4/1/2002 Cannons placed in program mode B sensor panel 4/1/2002 4/2/2002
9:00:00 AM 2:00:00 PM
4/2/2002 Placed in program mode B sensor panel quit after interface replaced Replaced 4/2/2002 4/5/2002
all 12 contact modules on B panel 3/4/02. Placed back in Auto Mode Friday 3:15:00 PM 9:00:00 AM
4/5/02 after panel ran all night
4/11/2002 Sensors updating error- Rebooted PC 4/11/2002 4/11/2002
11:37:00 AM 12:15:00 PM
4/23/2002 3:00:00 #2 Cannon Vertical Axis Alarm - Vertical Axis sticking- Replaced vertical axis 4/23/2002 4/23/2002
AM 3:00:00 AM 7:00:00 PM
5/20/2002 Pump giving fail to start. Checked strainer. Removed wires from pressure 5/20/2002 5/20/2002
switch, jumped switch and pump ran. Put switch back in service and pump ran. 8:30:00 AM 10:30:00 AM
6/4/2002 #1 valve failed to open. Turned manual valve off and cycled COAX valve with 6/4/2002 6/4/2002
overide. Alarm cleared
6/4/2002 #3 Lance backing out of blow through tube. Tightened back up 6/4/2002 6/4/2002
8:15:00 AM 8:15:00 AM
6/4/2002 Bushing came out of #2 cannon. Put back in 6/4/2002 6/4/2002
6/7/2002 #1 Cannon Isolation Valve failed to open. Cycled valve with overide button. 6/7/2002 6/7/2002
Valve seems to be working
7/6/2002 Channel 17 Module B05 drawing straight line from 2:50pm to 8:43 pm. Disabled
ch 17 from 7/6/02 8:00 pm to 7/8/02 7:00 am. Module was replaced 7/9/02.
7/16/2002 #2 Y axis resolver fault. Y axis position failure. Got cannon to rest position and
disabled cannon. Ran until 7/17/02 on other 3 cannons. Replaced resolvers
7/17/02
7/28/2002 #3 Y axis resolver fault has been coming and going for about 2 weeks. Would
not clear 7/28/02. Replaced set of resolvers
7/29/2002 #4 vertical position failure, #4 horizontal position failure. Swapped to 7/29/2002 7/30/2002
waterlances. Calibrated #4 cannon 7/30/02 7:15 am and returned to service 10:25:00 PM 8:10:00 AM
11/11/2002 Lance backed out of ball. Put lance back in and tack welded 11/11/2002 11/12/2002
3:46:00 PM 8:15:00 AM
12/5/2002 Ch 8 cannot clean zones disabled. Enabled ch 8. No problems 12/6/02. 12/5/2002 12/5/2002
12/5/2002 Ch 19 cannot clean zones disabled. Enabled ch 19. Same problem 12/06/02. 12/5/2002
Blew WL 71 Heat flux went from 5.3 to 29.8 Enabled Sensor again 12/6/02
1/31/2003 Found water in oil when sampling. Changed oil. 1/31/2003 1/31/2003
2/7/2003 Changed oil in pump 2/7/2003 2/7/2003
2/19/2003 Changed oil in pump, 3 gallons oil and 1-1.5 quarts water 2/19/2003 2/19/2003
A-1
12762100
Water Cannon Reliability
Problems
DateProblem Description StartDate EndDate
2/24/2003 Replace seals. Oil seal had blisters (pitting) where faces tend to stick when 2/24/2003 2/24/2003
pump is off. New seal made of different material. The water seal had lost its 6:45:00 AM 2:35:00 PM
axial movement, probably due to girt accumulating behind the seal.
2/27/2003 Vertical position failure, horizontal position failure on Cannon 3. Operator 2/27/2003 2/27/2003
switched to water lances. Cannon was all the way right and all the way to 1:38:00 AM 10:19:00 AM
bottom. Found CB114 breaker tripped. Recalibrated cannon.
2/28/2003 Horizontal position failure Cannon 4. Recalibrated cannon.
3/23/2003 Upgraded to new balls on all 4 cannons 3/23/2003 3/23/2003
3/22/2003 Cannon 3 horizontal position failure, reset Circuit Breaker and calibrated 3/22/2003 3/22/2003
cannons 3 and 4 8:00:00 AM
3/31/2003 Hose backed off of piping end of cannon 1. Reinstalled and tightened
4/1/2003 Channel 8 reading 100.0 heat flux. Changed from pair 2 to pair 4.
4/4/2003 Y Axis resolver fault #1 cannon coming and going. Replaced both resolvers.
4/13/2003 Lance backed out of ball on #3 cannon
5/16/2003 Got sample from #1 Cannon pump. Small amount of water
5/23/2003 6:22:00 Channel 7 cannot clean zone disabled
AM
PM
PM
6/13/2003 Got sample from #1 Cannon pump. Small amount of water, less than last
sample
7/29/2003 Changed oil in #1 Pump
7/30/2003 Replaced #4 Resolvers, Getting Y axis resolver fault
9/17/2003 Replaced pump seals 9/17/2003 9/18/2003
6:30:00 AM 2:45:00 PM
9/24/2003 Y axis resolver fault Cannon 4. Replaced resolvers. Did not correct problem. 9/24/2003 9/25/2003
Replaced Resolver Interface. 4:30:00 PM 6:00:00 PM
12/6/2003 Cannon 2 Isolation Valve Fail To Open Alarm would not reset. Valve operated 12/6/2003 12/6/2003
from the PC in the Maintenance mode and showed open and closed. Reset the
fault manually in the PLC. He will add a reset in the ladder logic next week.
2/3/2004 Y axis would not move. It was all the way to the bottom left. Overrode the limit 2/3/2004 2/4/2004
on the xaxis and moved the cannon to the center of the x, but could not get the 12:00:00 PM 2:00:00 PM
y axis to move. Replaced both axes. Used Unit 2 Cannon 4 parts.
A-2
12762100
B
CLEANING EVENT CHARACTERIZATION
B-1
12762100
Cleaning Event Characterization
Table B-1
Zone Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Total
101 27 40 43 27 59 70 56 73 74 87 142 85 783
102 29 54 57 20 27 39 36 95 149 89 131 117 843
103 29 36 45 19 36 37 37 75 87 116 204 130 851
104 38 44 55 25 67 85 71 98 105 59 212 194 1053
105 45 55 84 44 70 62 74 116 168 73 156 186 1133
106 52 40 41 47 82 86 79 93 87 59 147 123 936
201 26 25 30 17 27 42 45 104 101 50 150 281 898
202 99 145 42 40 30 45 36 50 52 45 165 302 1051
203 20 24 29 14 26 32 31 407 567 273 431 425 2279
204 20 24 29 14 25 31 23 403 559 270 425 423 2246
205 88 104 155 107 123 173 166 180 201 74 202 250 1823
206 103 109 308 113 121 345 338 205 45 35 110 33 1865
207 24 34 35 16 64 48 71 256 104 34 117 102 905
208 25 33 35 14 65 47 62 252 97 32 111 103 876
301 127 86 438 128 135 162 185 248 202 112 161 266 2250
302 127 86 438 128 135 162 186 248 202 111 162 265 2250
303 22 28 44 12 46 88 60 75 98 38 84 82 677
304 27 69 74 46 50 61 64 64 57 32 72 50 666
305 27 69 74 50 49 60 55 60 50 29 67 50 640
306 21 26 32 104 29 116 64 113 179 93 446 597 1820
307 21 26 32 103 30 117 64 113 179 93 446 597 1821
308 76 57 110 128 61 139 156 79 46 44 108 565 1569
309 201 269 110 43 150 191 165 130 67 54 231 702 2313
310 201 269 110 44 150 190 165 130 67 54 231 702 2313
311 201 269 110 43 153 192 166 130 69 54 233 705 2325
312 27 67 74 46 48 60 55 60 50 29 67 49 632
313 103 109 308 113 120 344 336 202 45 27 69 33 1809
314 103 109 308 113 120 344 335 202 45 27 67 32 1805
315 25 33 35 15 64 47 63 255 97 33 111 100 878
316 45 55 83 45 65 61 64 113 161 62 109 186 1049
317 45 55 83 46 67 61 73 117 168 65 114 184 1078
318 48 55 83 44 67 61 73 118 167 64 115 183 1078
319 43 55 83 45 67 61 73 116 166 64 116 182 1071
401 26 25 30 17 25 42 41 105 101 37 100 265 814
402 26 25 30 17 25 42 41 105 101 37 98 264 811
403 26 25 30 17 24 41 32 101 94 34 92 264 780
404 25 36 45 19 33 37 28 71 79 106 152 130 761
405 25 36 45 19 34 37 28 71 79 105 154 130 763
406 27 41 43 27 59 70 57 73 73 75 95 79 719
407 27 41 42 27 59 70 56 73 73 74 97 80 719
408 43 33 39 13 40 37 35 87 314 103 198 180 1122
409 43 35 39 13 40 42 33 87 314 103 195 181 1125
410 24 33 34 12 46 44 31 55 62 40 79 92 552
411 30 30 34 12 45 56 31 136 146 75 189 416 1200
412 30 30 34 12 44 57 31 136 146 75 189 416 1200
413 28 40 44 14 28 41 29 44 44 21 62 57 452
414 28 40 44 14 29 41 29 44 44 21 62 57 453
415 131 164 216 92 113 230 212 179 166 56 123 175 1857
416 58 75 234 60 111 104 109 87 78 73 83 99 1171
417 61 75 234 60 112 105 116 93 89 74 87 99 1205
418 58 76 234 60 112 105 118 91 85 74 88 99 1200
419 30 30 34 12 44 56 32 136 146 75 188 416 1199
Total 2861 3449 5102 2330 3451 4916 4616 6754 6745 3639 8043 11783 63689
B-2
12762100
1000 1000 1000 1000

500 500 500 500
0 0 0 0
1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112
1000 1000 1000 1000
500 500 500 500
0 0 0 0
1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112
1000 1000 1000 1000
500 500 500 500
0 0 0 0
1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112
1000 1000 1000 1000
500 500 500 500
0 0 0 0
1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112
1000 1000 1000 1000
500 500 500 500
0 0 0 0
1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112
1000 1000 1000 1000
500 500 500 500
0 0 0 0
1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112
1000 1000 1000 1000
500 500 500 500
0 0 0 0
1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112
1000 1000 1000 1000
500 500 500 500
0 0 0 0
1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112
1000 1000 1000 1000
500 500 500 500
0 0 0 0
1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112
1000 1000 1000 1000
500 500 500 500
0 0 0 0
1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112
1000 1000 1000 1000
500 500 500 500
0 0 0 0
1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112
1000 1000 1000 1000
500 500 500 500
0 0 0 0
1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112
1000 1000 1000 1000
500 500 500 500
0 0 0 0
1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112
Figure B-1
B-3
12762100
250
200
150
Count
100
50
0
20030101 20030215 20030327 20030529 20030706 20030813 20030920
FastFile
Figure B-2
B-4
12762100
250
200
150
Count
100
50
0
20030101 20030215 20030327 20030529 20030706 20030813 20030920
FastFile
Figure B-3
B-5
12762100
250
200
150
Count
100
50
0
20030101 20030215 20030327 20030529 20030706 20030813 20030920
FastFile
Figure B-4
B-6
12762100
250
200
150
Count
100
50
0
20030101 20030215 20030329 20030531 20030708 20030815 20030922
FastFile
Figure B-5
B-7
12762100
250
200
150
Count
100
50
0
20030101 20030213 20030329 20030530 20030705 20030815 20030922
FastFile
Figure B-6
B-8
12762100
250
200
150
Count
100
50
0
20030101 20030207 20030314 20030528 20030706 20030818 20030929
FastFile
Figure B-7
B-9
12762100
Cleaning Events with Change < 50 DegF
250
200
150
Count
100
50
0
20030101 20030215 20030327 20030529 20030706 20030813 20030920
FastFile
Figure B-8
Cleaning Events with Change Less Than 50°F for 2003
B-10
12762100
Cleaning Events with Droop > 0 DegF
250
200
150
Count
100
50
0
20030101 20030215 20030327 20030529 20030706 20030813 20030920
FastFile
Figure B-9
Cleaning Events with Droop Greater Than 0°F for 2003
B-11
12762100
250
200
150
Count
100
50
0
20030101 20030216 20030328 20030530 20030707 20030814 20030921
FastFile
Figure B-10
B-12
12762100
250
200
150
Count
100
50
0
20030101 20030215 20030331 20030601 20030708 20030814 20030920
FastFile
Figure B-11
B-13
12762100
250
200
150
Count
100
50
0
20030101 20030218 20030406 20030606 20030706 20030812 20030924
FastFile
Figure B-12
B-14
12762100
250
200
150
Count
100
50
0
20030101 20030311 20030412 20030610 20030720 20030827
FastFile
Figure B-13
B-15
12762100
250
200
150
Count
100
50
0
20030101 20030128 20030413 20030530 20030806 20030810 20030923
FastFile
Figure B-14
B-16
12762100
Cleaning Events with Droop < 50 DegF
250
200
150
Count
100
50
0
20030101 20030215 20030327 20030529 20030706 20030813 20030920
FastFile
Figure B-15
Cleaning Events with Droop Less Than 50°F for 2003
B-17
12762100
Temperature Droop
4000
3000
Frequency
2000
1000
0
0 100 200 300 400 500
Temperature (Deg F)
Figure B-16
Cleaning Events Distribution of Temperature Droop for 2003
B-18
12762100
Temperature Gain
3000
Frequency
2000
1000
0
0 100 200 300 400 500
Temperature (Deg F)
Figure B-17
Cleaning Events Distribution of Temperature Gain for 2003
B-19
12762100
Temperature Change
1500
1000
Frequency
500
0
0 100 200 300 400 500
Temperature (Deg F)
Figure B-18
Cleaning Events Distribution of Temperature Change for 2003
B-20
12762100
Initial Temperature
2500
2000
Frequency
1500
1000
500
0
400 500 600 700 800 900 1000 1100
Temperature (Deg F)
Figure B-19
Cleaning Events Distribution of Initial Temperature for 2003
B-21
12762100
Table B-2
Cleaning Event Means by Sensor for 2003
TStart TMean TDroop TGain TChange
S01 791.0 810.0 22.167 38.058 60.22
S02 789.5 796.3 25.207 25.898 51.11
S03 799.7 836.3 6.076 54.124 60.20
S04 798.0 828.1 9.223 44.110 53.33
S05 785.9 809.0 9.523 38.235 47.76
S06 772.0 787.6 13.956 29.951 43.91
S07 866.1 887.3 63.392 49.979 113.37
S08 855.9 858.8 28.631 16.649 45.28
S09 861.9 885.7 119.647 72.790 192.44
S10 857.8 895.0 85.882 79.422 165.30
S11 778.4 779.1 64.764 21.442 86.21
S12 801.9 815.3 30.305 36.798 67.10
S13 800.0 815.4 11.365 30.645 42.01
S14 790.9 809.1 6.398 27.618 34.02
S15 802.8 809.4 35.138 26.423 61.56
S16 792.5 802.5 44.650 31.454 76.10
S17 797.5 798.1 31.060 15.253 46.31
S18 790.5 788.6 12.828 4.996 17.82
S19 805.6 805.9 5.840 4.279 10.12
S20 807.7 817.4 38.927 33.910 72.84
S21 787.2 805.0 35.171 34.870 70.04
S22 803.4 820.4 32.696 36.020 68.72
S23 752.1 751.6 24.196 10.904 35.10
S24 777.3 795.7 53.807 41.171 94.98
B-22
12762100
Table B-3
Cleaning Event Sensor vs. Temperature Droop for 2003
Temperature Droop (DegF)
Sensor (0,100] (100,200] (200,300] (300,400] (400,500]

1 448 5 1 0 0
2 350 7 0 0 0
3 260 0 0 1 0
4 269 3 0 0 0
5 238 0 0 0 0
6 456 1 0 1 0
7 570 193 5 1 0
8 562 13 0 0 0
9 498 767 84 0 0
10 199 59 23 5 0
11 200 33 5 0 0
12 405 0 1 0 0
13 975 3 0 0 0
14 313 0 0 0 0
15 1001 34 0 0 0
16 308 28 0 0 0
17 495 8 0 0 0
18 935 0 0 0 0
19 617 3 0 0 0
20 1304 129 0 0 0
21 651 62 0 0 0
22 341 34 2 0 0
23 542 13 0 0 0
24 310 87 2 0 0
B-23
12762100
Table B-4
Cleaning Event Sensor vs. Temperature Gain for 2003
Temperature Gain (DegF)
Sensor (0,100] (100,200] (200,300] (300,400] (400,500]
1 463 0 0 1 0
2 358 0 0 0 0
3 588 5 1 0 0
4 428 9 0 0 0
5 286 0 0 0 0
6 468 0 0 0 0
7 675 83 0 0 0
8 556 8 0 0 0
9 1208 147 0 0 0
10 198 89 0 0 0
11 237 0 0 0 0
12 401 2 0 0 0
13 1202 5 0 0 0
14 410 9 0 0 0
15 1035 0 0 0 0
16 343 2 0 0 0
17 495 1 0 0 0
18 846 0 0 0 0
19 593 2 0 0 0
20 1451 2 0 0 0
21 735 0 0 0 0
22 356 7 0 0 0
23 538 0 0 0 0
24 397 1 0 0 0
B-24
12762100
Table B-5
Cleaning Event Sensor vs. Temperature Change for 2003
Temperature Change (DegF)
Sensor (0,100] (100,200] (200,300] (300,400] (400,500]
1 437 25 1 1 0
2 344 13 1 0 0
3 579 14 1 1 0
4 419 18 0 0 0
5 284 2 0 0 0
6 465 2 0 1 0
7 350 296 122 5 0
8 538 35 5 0 0
9 56 655 613 33 0
10 32 189 48 14 5
11 168 63 6 1 0
12 392 14 1 0 0
13 1188 49 0 0 0
14 415 12 0 0 0
15 960 74 1 0 0
16 281 64 0 0 0
17 474 29 0 0 0
18 942 1 0 0 0
19 618 9 2 0 0
20 1052 388 26 0 0
21 509 226 4 0 0
22 313 40 24 2 0
23 549 20 0 0 0
24 230 141 34 0 0
B-25
12762100

1400
1200
1000
Count
800
600
400
200
0
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Sensor
1200
1000
Count
800
600
400
200
0
1
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Sensor
30
25
Count
20
15
10
5
0
1
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Sensor

5
4
Count
3
2
1
0
1
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Sensor
Figure B-20
Cleaning Events Temperature Change vs. Sensor for 2003
B-26
12762100
Temperature Droop by Sensor
300
200
DegF
100
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Sensor
Figure B-21
Cleaning Events Surface Temperature Droop vs. Sensor for 2003
B-27
12762100
Temperature Gain by Sensor
300
250
200
DegF
150
100
50
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Sensor
Figure B-22
Cleaning Events Surface Temperature Gain vs. Sensor for 2003
B-28
12762100
Temperature Change by Sensor
400
300
DegF
200
100
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Sensor
Figure B-23
Cleaning Events Surface Temperature Change vs. Sensor for 2003
B-29
12762100
Sensor 1 Sensor 2 Sensor 3 Sensor 4 Sensor 5 Sensor 6
200
150
200 400
250
100
100
Frequency
Frequency
Frequency
Frequency
Frequency
Frequency
100
0 100
0 40
0 50
0 40
0
0 300 0 300 0 300 0 300 0 300 0 300
DegF DegF DegF DegF DegF DegF
150
80
Frequency
Frequency
Frequency
Frequency
Frequency
Frequency
30
0 40 100
40 80
20
40
15
0 50
0
0
0 300 0 300 0 300 0 300 0 300 0 300

700
100 200
150 300
300 600
40
Frequency
Frequency
Frequency
Frequency
Frequency
Frequency
100
300
20
0
0
0 300 0 300 0 300 0 300 0 300 0 300

150 300
250
250
100 200
150
80
Frequency
Frequency
Frequency
Frequency
Frequency
Frequency
100
0 100
40
0 50
0
0
0 300 0 300 0 300 0 300 0 300 0 300
Figure B-24
Cleaning Events Temperature Droop Distribution vs. Sensor for 2003
B-30
12762100
100 200
60 120
Frequency
Frequency
Frequency
Frequency
Frequency
Frequency
40 80
40 80
100
100
0
0
0 300 0 300 0 300 0 300 0 300 0 300
40
40
60
Frequency
Frequency
Frequency
Frequency
Frequency
Frequency
80
100
40 80
20
20
30
40
0 40
0
0
0 300 0 300 0 300 0 300 0 300 0 300
700
150 300
350
100
Frequency
Frequency
Frequency
Frequency
Frequency
Frequency
100
40
300
150
20
0 40
0 40
0
0
0 300 0 300 0 300 0 300 0 300 0 300

500
80
250
40
150
40
Frequency
Frequency
Frequency
Frequency
Frequency
Frequency
40
20
20
200
0 100
0 50
0
0
0 300 0 300 0 300 0 300 0 300 0 300
Figure B-25
Cleaning Events Temperature Gain Distribution vs. Sensor for 2003
B-31
12762100
100 200
60 120
Frequency
Frequency
Frequency
Frequency
Frequency
Frequency
40 80
40 80
100
100
0
0
0 300 0 300 0 300 0 300 0 300 0 300
40
40
60
Frequency
Frequency
Frequency
Frequency
Frequency
Frequency
80
100
40 80
20
20
30
40
0 40
0
0
0 300 0 300 0 300 0 300 0 300 0 300
700
150 300
350
100
Frequency
Frequency
Frequency
Frequency
Frequency
Frequency
100
40
300
150
20
0 40
0 40
0
0
0 300 0 300 0 300 0 300 0 300 0 300

500
80
250
40
150
40
Frequency
Frequency
Frequency
Frequency
Frequency
Frequency
40
20
20
200
0 100
0 50
0
0
0 300 0 300 0 300 0 300 0 300 0 300
Figure B-26
Cleaning Events Temperature Change Distribution vs. Sensor for 2003
B-32
12762100
12762100
Export Control Restrictions
Access to and use of EPRI Intellectual Property is granted
with the specific understanding and requirement that
responsibility for ensuring full compliance with all applicable
U.S. and foreign export laws and regulations is being
undertaken by you and your company. This includes an
obligation to ensure that any individual receiving access
hereunder who is not a U.S. citizen or permanent U.S.
resident is permitted access under applicable U.S. and foreign
export laws and
regulations. In the event you are uncertain whether you or
your company may lawfully obtain access to this EPRI
Intellectual Property, you acknowledge that it is your
obligation to consult with your company’s legal counsel to
determine whether this access is lawful. Although EPRI may
make available on a case by case basis an informal assessment
of the applicable U.S. export classification for specific EPRI
Intellectual Property, you and your company acknowledge
that this assessment is solely for informational purposes and
not for reliance purposes. You and your company
acknowledge that it is still the obligation of you and your
company to make your own assessment of the applicable
U.S. export classification and ensure compliance accordingly.
You and your company understand and acknowledge your
obligations to make a prompt report to EPRI and the
appropriate authorities regarding any access to or use of
EPRI Intellectual Property hereunder that may be in violation
of applicable U.S. or foreign export laws or regulations.
About EPRI
EPRI creates science and technology solutions for
the global energy and energy services industry.
U.S. electric utilities established the Electric Power
Research Institute in 1973 as a nonprofit research
consortium for the benefit of utility members, their
customers, and society. Now known simply as EPRI,
the company provides a wide range of innovative Program: 1009572
products and services to more than 1000 energy- Nuclear Power

related organizations in 40 countries. EPRI’s
multidisciplinary team of scientists and engineers
draws on a worldwide network of technical and
© 2004 Electric Power Research Institute (EPRI), Inc. All rights reserved. Electric Power Research
business expertise to help solve today’s toughest
Institute and EPRI are registered service marks of the Electric Power Research Institute, Inc.
energy and environmental problems. EPRI. ELECTRIFY THE WORLD is a service mark of the Electric Power Research Institute, Inc.
EPRI. Electrify the World Printed on recycled paper in the United States of America
EPRI • 3412 Hillview Avenue, Palo Alto, California 94304 • PO Box 10412, Palo Alto, California 94303 • USA
800.313.3774 • 650.855.2121 • [email protected] • www.epri.com
12762100

Water Cannons

Uploaded by

Copyright:

Available Formats

Water Cannons

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Water Cannons

Uploaded by

Copyright:

Available Formats

Demonstration of Clyde Bergemann

Water Cannons at Alabama Power

Final Report, November 2004

EPRI Project Manager

(A) MAKES ANY WARRANTY OR REPRESENTATION WHATSOEVER, EXPRESS OR IMPLIED, (I)

(B) ASSUMES RESPONSIBILITY FOR ANY DAMAGES OR OTHER LIABILITY WHATSOEVER

ORGANIZATION(S) THAT PREPARED THIS DOCUMENT

Southern Company Services, Inc.

Alabama Power Company, Inc.

This report was prepared by

Southern Company Services, Inc.

Alabama Power Company, Inc.

2 TECHNOLOGY DESCRIPTION................................................................................ 2-1

3 TEST PROGRAM ..................................................................................................... 3-1

4 SUMMARY AND CONCLUSIONS............................................................................ 4-1

A WATER CANNON RELIABILITY............................................................................ A-1

B CLEANING EVENT CHARACTERIZATION ........................................................... B-1

Figure 2-1 SmartCannon Cleaning Areas................................................................................ 2-2

Table 2-1 CBI External Inputs ................................................................................................. 2-7

Results & Findings

Key findings of the program are as follows:

Water Cannon Water Lances

Technology Installation and Integration at Miller 1

Feedwater Flow Superheat Spray Flow

Reheat Outlet Temperature – Left Reheat Outlet Temperature – Right

A AH Gas Out Temperature B AH Gas Out Temperature

Unit 1 Generated Megawatts

Characterization of Water Cleaning Events

Heat Flux Characteristics

1000 1000 1000 1000

Notes: 1. All temperatures in °F.

Cleaning Events with Change > 0 DegF

20030101 20030215 20030327 20030529 20030706 20030813 20030920

Cleaning Events with Change > 100 DegF

20030101 20030215 20030327 20030529 20030706 20030813 20030920

Cleaning Events with Change < 50 DegF

20030101 20030215 20030327 20030529 20030706 20030813 20030920

0 100 200 300 400 500

Notes: 1. All temperatures in °F.

Temperature Droop by Sensor

Temperature Gain by Sensor

Temperature Change by Sensor

Counts of Events with TChange > 0 DegF

Counts of Events with TChange > 100 DegF

Counts of Events with TChange > 200 DegF

Counts of Events with TChange > 400 DegF

Heat Flux vs. Load

(100,200] (300,400] (500,600] (700,800]

Surface Temperature vs. Load

(100,200] (300,400] (500,600] (700,800]

Waterwall Tube Life

Field Testing and Data Collection

Model Test Case Application

Model Field Data Application

Damage Index from Long-Term Data

σ = ε ⋅ E ⋅ ∆T Temperature induced stress

Summary of Waterwall Tube Damage Analysis

Case 1: Dni=23.9 Case 2: Dni=23.9

Case 3: Dni=23.9 Case 4: Dni=23.9

Case 5: Dni=0.0* Case 6: Dni=14.8