• Pergamon Wal. Sci T~ch. Vol. 38, No.6, pp. 303-310, 1998.

IAWQ
o 1998 Published by ElaevieT Science lid.
PrImed In 0Jeat Brilaln. All tllhta reaervcd
0273-1223198519'00 + 0'00
PIT: 80273-1223(98)00590-3

A SIMPLIFIED METHOD FOR ASSESSING

THE SATURATION EFFICIENCY AT
FULL-SCALE DISSOLVED AIR
FLOTATION PLANTS

S. Steinbach and J. Haarhoff

Department ofCivil Engineering, Rand Afrikaans University, Box 524, Auckland
Park, 2006 South Africa

ABSTRACT
This peJlef' proposes • simplified yet comprehensive method for measuring the efficiency of air saturation
systems used in dis50lved air flotation. It is based on • review of methods suggested by other researchers In
this field, but &eVmI1 impovc:ments to c:xistin& methods are: made. A saturation air composition difl'c:rc:nt
than that of atmospheric air and incomplete Iftcipilation in the measuring unit after pressure release are
taken into consideration. 1be devdopcd measurement and calculation poccdure provide en&incc:n aPd
operators alike with an easy tool to assess the saturation efficiency of their sysIem at full-scale OAF plants u
weD as at pilot and 1aboralcry SQle plants. C 1998 Published by Elsevier Science Ltd. All rights reserved

KEYWORDS

pissolved air flotation~ Henry's law~ measurement method~ saturation efficiency

INTROOUcnON

Air saturation systems are used in dissolved air flotation (OAF) to dissolve air into water under pressure. The
eflieiency ofthese systems determines a major part of the operating cost of the OAF process. Measuring the
sanaration efficiency is firstly required for practical verification upon commissioning of a new system. and
secondly for periodical checking ofthe saturation system against reduced flow rate or media blockage.

'fhis paper presents a simplified yet comprehensive method for determining the saturation efficiency at full·
scale DAF plants as well as at pilot and laboratory scale plants. A number of methods for assessing the
saturation efficiency have been suggested in the past, which all suffer from one or more shortcomings. To
overcome these shortcomings, all the complicating factors are discussed and considered in the development
oftbc improved measurement procedure.

DEFINITION OF SATURATION EFFICIENCY

There are different ways to define saturation efficiency which implies that a given saturation efficiency is
rather meaningless without a complete and explicit description of definitions and assumptions. In this case,
.aturation efficiency is derIDed in terms of absolute air concentrations and described as the ratio between the
actUal air mass transferred to the theoretical air mass transferable:
_ actual air mass trans!e"ed c,-e/l
116 - theoretical air mass transferable =-C';'·--..;.C=.. (1)
• /I

303
304 S. STEINBACH and J. HAARHOFF

with 1'/, • saturation efficiency; C, • actual mass concentration of air in water after passing through the
saturation system; C• • mass concentratioo of air in incoming water; C*, = theoretical mass concentration of
air in water that would have attained equilibrium with saturation air at saturation pressure.

PREVIOUSLY REPORTED METHODS FOR MEASURING THE SATURATION

EFFICIENCY

DitTerent methods for measuring the saturation efficiency have been described in the literature and the
relevant references are listed in Table I. The methods can be classified into two broad categories, depending
on whether they measure the amount of air precipitated upon pressure release or the total amount of gas
dissolved under pressure. The amount of air precipitated from a depressurized solution can be detennined in
two ways; continuous measurement with an air flow meter, or batch measurement by liquid displacement.
Because ofspace limitations, none ofthe previously reported methods will be discussed in further detail here
but they can be found in the according literature or in summarized form in HaarhotT and Steinbach (1997).

Table I. Reported techniques for measuring the saturation efficiency

Method Precipitated air measured Precipitated air measured Direct measurement of

with air meter by liquid displacement dissolved air
Rererence Bratby &. Marais (1975) Conway el aJ (1981) Leininger &. Wall (1974)
(date) Roberts eloJ (1978) Henry &. Gehr (1981) Packham &. Richards (1975)
Rees el aJ (1980) Lovett el oJ (1984) Henry &. Gehr (1981)
Shannon &. Buisson (1980) Vosloo et aJ (1986)
Casey &. Naoum (1986)
Smits (1990)
Rykaart (1991)

All these methods, however, sutTer from one or more ofthe following shortcomings:
• An inadequate or unclear definition ofsaturation efficiency is used.
• The water entering the saturation system is always assumed to be at equilibrium with the atmosphere.
• The composition of the saturation air may ditTer from that of atmospheric air. This fact is either
disregarded or a fixed theoretical air composition is assumed. .
• Where the precipitated air volume is measured, complete precipitation during the measurement IS
assumed.

SUGGESTED MEASURING TECHNIQUE FOR ASSESSING SATURATION

EFFICIENCY

The simplest method is based on the measurement of the precipitated air volume after depressurization, as
evidenced by the fact that 11 out of 14 references (see Table 1) described variations of this method. Batch
tests, in addition, are simpler than continuous tests, as they require a simple calibrated air container rather
than a more sophisticated air-flow meter. For these reasons, the method proposed here will be based on
measuring the precipitated air volume with a batch system.

Equipment

The equipment used to determine the saturation efficiency is shown in Fig. 1, together with its relevant
dimensions. The measuring bottle( I) represents the main part of the measuring unit and consists of a metal
base(2), a perspox cylinder(3), a plastic cone(4) and a graduated measuring cylinder(S). The perspex
cylinder(3) is flanged on a metal disk(2), which is made of stainless steel. This disk also contains the
inlet(7) and outlet(S) of the measuring unit and a partition(9), which divides the apparatus into an inlet and
 Saturation efficiency at full-scale dissolved air flotation plants 30S

an outlet section. The partition prevents short-circuiting of the water and. therefore. air bubbles from being
drawn into the outlet. The plastic cone(4) is held on top of the perspex cylinder(J) by clamps and a 500mL.
graduated measuring cylinder(5). provided with a valve(6). is glued on top of the cone. The pressure is
released across a nozzle(10) BCr~wed .on~o the inlet(7) of the measuring bottle. The nozzle(10) is made of
sta.inless steel and has a lrnm onfice m ItS center. The length of the straight nozzle channel is Smm and a
small plastic plate is positioned Srnm downstream of the nozzle exit in order to cause more turbulence and
enhaDce precipitation. To control the water level in the measuring bottle a glass funnel(ll) with an
overl1ow(12) is connected to the outlet(8) ofthe bottle by a plastic tube. The valves used at the inlet(7) and
outlet(8) ofthe measuring bottle as well as at the connecting point with the saturation system are ~ inch ball
valves.

1: measuring bottle

. . ;¥
2: metal base Nozzle(IO)
3: perspex cylinder 6
... / 5 11 impinging
4: plastic cone surface
5: graduated measuring E .,..,~I 50 _ ..levd 1..
12 t.~~
cylinder E \0 ..
.,.., N .l~.v.f:l2 :::!.. .
6: valve ~ 4 ..:.' .,,' .... f
7: inlet valve .,..,~
8: outlet valve ~
00
9: partition 00

10: nozzle
II: funnel
12: ovep;rt1_ow
_ _-,

Saturation 2 1J1l!n;()t
system
.,, .,,
.
: 24mm ;
"
to flotation
unit measurin~ unit

Figure I. Equipment for measuring saturation efficiency

!"fea.uring procedure
Th measuring unit is connected in parallel to the flotation unit. with part of the outlet of the saturation
~em directed to the measuring bottle(I). Both the inlet(7) and the outlet(8) valves of the measuring bottle
sy open to allow a continuous flow through the bottle and the valve(6) on top of the graduated cylinder is
at'~ to atmosphere. Before the actual measurement two arbitrary water levels in the graduated cylinder(5)
~;detennined by adjusting the overflow funnel(II.12).

The measurement starts at the upper. level ~ with closing the valve(6) on t~p of the graduated cylinder. The
erflow(IZ) is diverted to a measunng cybnder and the funnel(ll) then adjusted to level 2. When the water
Ov 1 in the measuring bottle reaches level Z. the measuring cylinder is removed from the overt1ow(l2) and
e
Jev volume of the water is measured. The volume of the air precipitated corresponds to the difference
~een level I and level 2 in the graduated cylinder(5). The water temperature is measured in the outlet of
tb measuring unit and recorded. After the measurement. the valve(6) on top of the graduated cylinder is
o;e,.ed again. and the measurement can be repeated.
306 S. STEINBACH and J. HAARHOFF

CALCULATION PROCEDURE FOR DETERMINING THE SATURATION

EFFICIENCY

The determination of the saturation efficiency from the measured data is based on Henry's law and a mass
balance across the measuring bottle. The shortcomings of previously reported methods will also be
discussed and corrected in this section.

lIeory'. law

Atmospheric air consists, for practical purposes, only of two major gases; nitrogen and oxygen. The
volumetric or molar fraction of nitrogen in atmospheric) air is considered to be 0.791 and the molar fraction
ofoxygen 0.209.

BO'I. nitrogen and oxygen obey Henry'. law over the temperature and pressure ranges used in dissolved air
flotation (0-30°C, 200-800kPa). Henry'. law predicts a linear relationship between the partial pressure of
the gas in the atmosphere above the water to the quantity ofgas dissolved in the water:

C, =11 ·44.6· *".(27~1 5).(f;~) (2)

with 1m' of~ containing 44.6molat O"C and a pressure of IOI.3kPa; C,· mass concentration ofgu i in
water [g·m'); " • molar fraction of gas i in gas phase [0); M, .. molecular weight of
gas i [g·morl ); H, • Henry'. constant of gas i [0); T .. absolute temperature [K); PlOIaJ .. absolute, applied
pressure [kPa). Value. for Henry's constants for nitrogen and oxygen are given in Table 2.

Table 2. Henry'. constants for nitrogen and oxygen. HN" Henry's constant for nitroge~ Ho· Henry's
constant for oxygen

49.8 54.0 58.2 62.3 66.5 70.7

23.9 26.4 28.8 31.3 33.8 36.3

Air cODcetltntl.D 01 the water eDterille tbe .atuntioD system

Different water sources can be used as feed for the saturation system. In the case where natural surface
water is used it is assumed that the water is saturated with nitrogen, as there are only a few nitrogen•
consuming compounds in the water. Tho oxygen concentration, however. can be considerably below
saturation. It i. therefore recommended to use part of the treated water (after flotation) as feed for the
saturation .ystem, rather than natural surface water, in order to ensure saturation concentration of both,
nitrogen and oxygen, in the water. The air mass concentration in the incoming water can then be calculated
by Henry', law (sec equation 2).

Compolitio.. 01 tbe latur.tlo.. air

From Henry'. law it is obvious that the solubility of air is not only detennined by pressure and temperature
but also by the actual air composition. Depending on the saturation .ystem used, the composition of the
saturation air may differ from that of atmospheric air due to the different solubilities of nitrogen and oxygen.
Thi. phenomenon is found in continuously operating pressure saturators. which are used in most DAF
applications. Since oxygen is two times more soluble than nitrogen the oxygen will disappear faster at first
until equilibrium is roached and the same quantities of oxygen and nitrogen leave the saturator in the
pressurized water as enter it in the compressed air and the incoming water. By then the nitrogen content of
the gas in the saturator will be higher than in atmospheric air, leaving a nitrogen-enriched atmosphere. The
air composition inside the ..turator at equilibrium can be derived from a mole balance across the satur~or
(Haarhoff and Steinbach, 1996). Values for the volumetric or molar fraction of oxygen in the saturator I1t at
 Saturation efficiency at full-scale dissolved air flotation plants 307

equilibrium over a temperature range of 5 to 30°C and a pressure range of 300 to 700kPa gauge are given in
Fig. 2.
It is important to note, that it can take hours from saturator start-up to reach equilibrium composition
(lIaarhoff and Stein~ach, 1996). To ensure that the saturation efficiency is measured at equilibrium, it is
recommended to WllIt at least 12 hours after start-up before the tirst measurement is taken.

0.144
--T- 5C
0.142 ,,
:i:
,,
, --T - 10C
1:1 0.140 ~. " -T-ISC
~ "
--T-20e
~ 0.138
o
~" ~
'- .................
'0 0.136 • T-2Se
~ -.- T- 30e
'-
1:1 ~~~ " '-
-
.!as 0.134 --, --., -
---- r--::::t---;-.r--::- r:---;
~ '--
~-.., ~..........., '-
J:.. 0.132
-----
~ 0.130
~ ...... "-
a
~ 0.128
Molar fraction of nitrogen: I
------. ----.
........... ~-- ~--

---
ySN = l-yso I ...... '-----.
;... 0.126
Atmospheric pressure = 101.3kPal
0.124
300 350 400 450 500 SSO 600 6S0 700
Saturator gauge pressure [kPaj

Figure 2. Equilibrium composition of saturator air

Air precipitation after pressure release

All the techniques, which have been proposed so far to measure the precipitated air volume after
depressurization, made the implicit assumption that all the excess air would completely precipitate during
he measurement. This amount of precipitated air was considered to be equal to the amount of air, which
~ been dissolved under pressure in the saturation system. It was proven (Steinbach and Haarho~ 1997),
hat not all the. ex':Css air comes out ~f ~Iutio~ during the ~easurement and ~t the water leaving t;he
:neasuring unit IS ~11l supc:rsaturated wit? atr. This oversaturation has to be taken Into account, as Otherwise
the saturation effiCiency Will be underestimated.
Effective air precipitation is mainly a function ofthe injection nozzle. The nozzle which was selected for the
turation efficiency measurement fulfills two important requirements. Firstly, it provides a high release
S3f}iciency to minimize the effect of incomplete precipitation on the measured saturation efficiency.
eS~ndlY the release efficiency is reasonably constant for different saturation pressures. During the
ft' ."e~ent about 92"10 of all the excess air is released.
JJ1ea.....
Calculation procedure
The following steps have to be carried out to calculate the saturation efficiency from the measured data:
• Allow enough time for the satur~on. syste~ to r~h equilibrium ~fore the flfSt measurement.is taken.
• calculate the air mass concentratlon In the Incommg water by applymg Henry's law (see equation 2):

C
•.0
=0209.44.6.(E..).(2711S)
Ho T
.(..l!L)
101.3
(3)
308 S. STEINBACH and J. HAARHOFP

N
-r- .
C. P =0.791·44.6· ( H28) . (273.15) ( 101.3
p" ) (4)

~=~P+~~ ~
with C.o - mass concentration ofoxygen in the water entering the saturation system [g·m..)]; Ce,N - mass
concentration of nitrogen in the water entering the saturation system [g·m"]; T'" measured water
temperature (K];P. - atmospheric pressure [kPa).
• Calculate the theoretical air mass concentration in the water that would be at equilibrium with the
saturation air at saturation pressure by applying Henry'. law:

C;p =Y•.o '44.6.(~~).e7~15) ·C~i.3) (6)

C;,N =Y•.N '44.6.(~:).(27~IS),C~i.3) (7)

C; =C;.o + C;~ (8)

with C·"o • theoretical mass concentration of oxygen in water that would have attained equilibrium with
saturation air at saturation pressure [g.m"]; C.I,J{ - theoretical mass concentration of nitrogen in water
that would have attained equilibrium with saturation air at saturation pressure [g.m..)]; y..o - molar
fraction of oxygen in saturation air [-] (for saturators see Fig. 2); yl,J{ - molar fraction of nitrogen in
saturation air;p, - total applied saturation pressure [kPa].
• Convert the measured volume of precipitated air into mass concentration:

a, =V, ·1.2929 ·e7~1S) .L~~.3) (9)

with the density of air at O°C and 101.3kPa given by 1.2929 kg.m·' ; a,· precipitated air mass [g·m..)];
V,I - precipitated air volume [l·m.]).
• Determine the saturation efficiency:

~ O~
'1, = O.92.(C; -C.>
where the incomplete air release during the measurement is taken into account by using the factor 0.92.

If the suggested measuring and calculation procedure, as described above, are followed, the saturation
efficiency can be determined with an accuracy of± 6%. To achieve higher accuracy, a 00 meter is required
for the measurements and the calculation procedure has to be carried out by computer because of its
increased complexity. A detailed description of this more exact method can be found in Haarhoff and
Steinbach 0997).

Eumple

The following example illustrates how to determine the saturation efficiency from the measured data using
tho calculation procedure outlined above. The input parameters for the example are given in Tablo 3.

• Air mas. concentration in the incoming water:

C. P .. 0.209.44 6.
.
(E..)
31.3
.(273.1S)
293.15 .(101.3)
101.3
=8.9 g.m"
C,.N" 0.791'44.6·C~~3).(~~~:~~)·G~~:D = 14.8 s·m"
C. =8.9+14.8=23.7 8· m"'
 Saturation efficiency II full-scale dissolved air flotation plants 309

• Theoretical air mass concentration in water at equilibrium with saturation air at saturation pressure:

C;,o = 0.133.44.6'(33 2 ) .(273.15) .(601.3) = 33 S g.m.)

13 293.15 101.3 .

C;,N = 0.867.44.6'(6228) . (273.lS) .(601.3) =961 g.m·J

.3 293.15 101.3 .
C; = 33.5 + 96.1 = 129.6 8·m· J

Conversion ofthe measured volume ofprecipitated air into mass concentration:

•
a = 70. 1.2929. (273.lS) .
P 293.15
(Ion)
Ion = 84.3 g m ..J

Determination of the saturation efficiency:

•
84.3
'7~ = 0.92. (129.6 _ 23.7) = 87010

Table 3. Input parameters for calculation example

-Water temperature T= 20°C" 293.15K

lIenrY's constant (from Table 2) Bo" 31.3
HN=62.3
Atmospheric. pressure p... 101.3kPa
Total satur~on pres~~ P," 500 + 101.3 =60 1. 3kPa
S turation lUI' composition (e.g. for a packed ' Yl,o=0.133
~ator) (from Fig. 2) y,,N=0.867
M~red precipitated air volume Vp ·701.m.)

CONCLUSIONS

A review of measurement methods for saturation efficiency showed that the methods are either not
• explicitly formulated in terms of definitions, or make some crude assumptions which cannot always be
justified.
This paper proposes a simple measurement procedure, based on a clear definition of saturator efficiency,
• which takes variables into account, which had hitherto been ignored.
The proposed method does not require any sophisticated equipment and the saturation efficiency can be
• determined easily from the measured data by hand and on site.

REFERENCES

sratby J. and Marais G.v.R. (1975) Saturator performance in dissolved-air (pressure) flotation. Wat. ReJ. 9,
929-936.
esseY T.J. and Naoum I.E. (1986) Air saturators for use in dissolved air flotation processes. Wat. Supply 4,
69-82.
conway R.A, Nelson R.F. and Young BA (1981) High-solubility gas flotation. Journal WPCF 53, 1198-
1205.
hoff 1. and Rykaart E.M (1995) Rational design ofpacked saturators. Wat. Sci. Tech. 31(3-4), 179-190.
~off J. and Steinbach S. (1996) A model for the prediction ofthe air composition in pressure saturators.
Wal. Res. 30, 3074-3082.
J-Ja8rl1off ,. and Steinbach S. (1997) A comprehensive method for measuring the air transfer efficiency of
pressure saturators. Wat. Res. 31,981-990.
310 S. STEINBACH and J. HAARHOFP

Henry J.O. and Gehr R. (1981) Dissolved air flotation for primary and secondary clarification. Sewage
Collection and Treatment Report SCAT-9, Canada.
Leininger K.V. and Wall OJ. (1974) Available air measurements applied to flotation thickener evaluations.
Deeds & Data, WPCF II, 3-7.
Lovett DA, Travers S.M. and Maas R.L. (1984) Treatment of abattoir wastewater by dissolved air flotation.
Part I: Wastewater not pretreated. Meat Research Report No.9, CSIRO Meat Research Laboratory,
CaMon Hil~ Australia.
Packham R.F. and Richards W.N. (1975) The determination of dissolved air in water. Technical
Memorandum 1M /06, Water Research Centre, Medmenham, England.
Rees A.1., Rodman OJ. and Zabel T.F. (1980) Evaluation of dissolved-air flotation saturator performance.
Water Research Centre TR 143. Water Research Centre, Medmenham, England.
Roberts K.L., Weeter D.W. and Ball R.O. (1978) Dissolved air flotation performance. Proceedings 33rd
Industrial Waste Conference. Purdue University, 194-199.
Rykaart E.M. (1991) Die modellering van lugversadigers vir opgeloste-lug flottasie (The modelling of air
saturators for dissolved air flotation). BEng Thesis, Department of Civil Engineering, Rand Afrikaans
University. Johannesburg.
Shannon W.T. and Buisson D.H. (1980) Dissolved air flotation in hot water. Wat. Res. 14,759-765.
Smits J. (1990) Bepaling van de verzadigingsgraad van gesatureerd water bij flotatie (Determination of the
degree ofsaturation of saturated water during flotation). Thesis submitted to the HTS Den Haag, based
on an investigation at Drinkwaterleiding Rotterdam.
Steinbach S. and Haarhoff J. (1997) Air precipitation efficiency and its effect on the measurement of
saturator tfficiency. Proceedings International Conference on Dissolved Air Flotation, April 1997,
London.
Vosloo P.B.B., Williams P,G. and Rademan R.O. (1986) Pilot and full-scale investigations on the use of
combined dissolved-air flotation and filtration (DAFF) for water treatment. Wat. Pollut. Control 85.
114-121.