26 Water 44.

2_3 F2083 28/7/01 10:10 am Page 203

Applying fine bubble aeration to small aeration tanks

Water Science and Technology Vol 44 No 2–3 pp 203–210 © 2001 IWA Publishing and the authors
Ph. Duchène, E. Cotteux and S. Capela
Cemagref, Unité de recherche qualité et fonctionnement hydrologique des systèmes aquatiques, Parc de
Tourvoie, B.P. 44, 92163 Antony Cedex, France.

Abstract Because the aeration system in an activated sludge plant typically represents a large part of the
total energy requirements, designers and operators need accurate oxygen transfer information to make the
aeration system as energy efficient as possible. This paper presents clean water tests performed at 38
wastewater treatment plants. The Specific Aeration Efficiency results (SAE, kgO2/kWh) are reported for:
(1) large open channels (volume higher than 1000 m3), (2) small open channels, (3) total floor coverage
cylindrical tanks, and (4) cylindrical tanks with a grid arrangement. Some practical guidelines can be drawn,
some of them being: (1) high SAE can be achieved at small aeration tanks (<1000m3), applying cylindrical
tanks with a total floor coverage arrangement of diffusers, volumetric blowers, and moderate air flow rates
per diffuser area; (2) the high investment cost of this configuration can be justified with respect to a grid
layout characterized by spiral liquid circulation which affects the oxygen transfer; (3) small open channels
can meet sufficient SAE values but fail to meet in this range of tank volumes those of total floor coverage
cylindrical tanks.
Keywords Fine bubble; aeration; small wastewater treatment plants; spiral flows; SAE; general layout

Introduction
The activated sludge process is likely to remain the main method of wastewater treatment;
even if its intensive operation tasks may result in the choice of other biological processes in
certain situations, such as in small towns and lesser developed areas.
In France, operating costs are a particularly important factor for the owners of waste-
water treatment plants (WWTP), especially given public subsidies for investment. Energy
requirements often account for one third of the total operating costs and aeration can
represent often 60% to 65% of these power costs.
Optimization of gas-liquid oxygen transfer is therefore recognized as one of the major
economic factors. Reliable methods for measuring oxygen transfer under process
conditions have only been available over the last few years (Redmon et al., 1983; ASCE,
1996; Capela et al., 1999). Their implementation is still delicate and the results are impact-
ed not only by the configuration of the aeration system and the tank but also by the local
wastewater characteristics and the loading rate. Whilst research into diffused aeration has
brought some interesting initial results (Gillot et al., 1997; Wagner and Pöpel, 1998;
Capela, 1999; Gillot et al., 2000), it has so far failed to provide a global, modeled under-
standing of the oxygen transfer efficiency.
Comparisons between aeration devices in identical or similar arrangements, and certain
comparisons in different diffusers layouts are currently more accessible through classic
oxygen transfer clean water tests (Héduit and Racault, 1983a; ASCE, 1992; ATV, 1996).
For nearly 30 years, through its measurements performed at WWTP, Cemagref has been
contributing to the selection of optimal aeration devices and energy efficient configura-
tions (Héduit and Racault, 1983b).
Whilst ceramic fine bubble diffusers offered the best performance (Héduit and Racault,
1983b), they have since been replaced in new plants and, except for a few very large
WWTP, by surface aeration systems. In the 1980s, the launch of elastomer membrane
diffusers that could be shut down without causing any major damage, resulted in a renewal 203

Downloaded from https://iwaponline.com/wst/article-pdf/44/2-3/203/32329/203.pdf

by Tufts University user
on 31 July 2018
26 Water 44.2_3 F2083 28/7/01 10:10 am Page 204

of the use of fine bubble aeration concomitant in practically all cases with the objectives of
nitrification and denitrification.
In the 1990s the large recently built WWTPs (> 5,000 p.e.) were almost exclusively
equipped with this type of diffusers. The advantages brought in terms of noise reduction,
expected performance and, more commercially speaking, the reduction of aerosol, resulted
in a massive increase in the use of fine bubble diffusers with flexible membranes, even in
very small aeration tanks and in non optimized configurations.
Ph Duchène et al.

Frequently in the case of small tanks and for economic and technical reasons, the design
differed from the predominant model found in the large WWTPs in France, namely in the
open channel with a dissociation of the aeration and mixing functions, where large blade
mixers create a controlled horizontal velocity over the diffusers (Deronzier et al., 1998).
The various adaptations to small aeration tanks cover the following major types:
• the miniature open channels
• the cylindrical tanks with a total floor coverage arrangement
• the cylindrical tanks with diffusers located in one or several separate grids.
The purpose of this article is to analyze the aeration performance results obtained at 25
WWTPs (taking as a point of reference a set of results obtained at large WWTPs) and to
draw practical conclusions in the form of guidelines for the design of small tanks with fine
bubble diffusers.

Material and methods

Oxygen transfer measurement
The oxygen transfer measurements were carried out in clean water according to the non-
steady state method and the protocol described by Héduit and Racault (1983a), defined as a
variation for experienced operators in the European standard CEN (1999).
The wire power was measured at the terminals of each motor using an energy analyzer
Prowatt 3 Chauvin-Arnoux. The air flow rates were determined using a Pitot tube coupled
to a flow meter Zephyr Solomat, or in a few cases calculated from data supplied by the
constructor.
The horizontal water velocities in the tanks were measured at 20 points over 30 seconds
using an OTT C2 hydrometric propeller (Deronzier and Duchène, 1997). The mean hori-
zontal velocity of the liquid was calculated as follows: (i) if the velocity field is transversal-
ly homogenous then the horizontal velocity is the arithmetic mean of the velocities
measured locally; (ii) otherwise, the horizontal velocity corresponds to the mean of the
local velocities weighted by the number of diffusers concerned.
The parameters acquired from the oxygenation measurements in clean water are pre-
sented in Table 1.
The saturation concentration (Cs) used in the calculation of SOTR is measured in situ by
the Winkler chemical titration. It is a specific particularity of the method employed. The Cs
value corresponds to a mean hydrostatic pressure of 36% of the diffuser submergence

Table 1 Used performance parameters

Notation Unit Expression

Standard Oxygen Transfer Rate SOTR kgO2.h–1 SOTR = KLa20Cs,20.V.10–3

Standard Oxygen Transfer Efficiency SOTE %/m SOTE = OC20

W0.h

Standard Aeration Efficiency SAE kgO2.h–1.kWh–1 SAE = OC20

PW

KLa20: oxygen transfer coefficient at 20°C in clean water (h–1); Cs,20: DO saturation concentration
(mg.L–1);V: aeration tank water volume (m3); WO2: mass flow of oxygen in air stream (kgO2.h–1); h: diffuser
204 depth (m); PW: wire power (mixer + blower) (kW)

Downloaded from https://iwaponline.com/wst/article-pdf/44/2-3/203/32329/203.pdf

by Tufts University user
on 31 July 2018
26 Water 44.2_3 F2083 28/7/01 10:10 am Page 205

(mean of 90% of the data). This value is significantly different from the saturation at
mid-depth used in certain European countries, and results in SOTR and SAE values are
anything from 3% lower for a depth of 2.5m to 8% lower for a depth of 7.5m, in relation to
those measured elsewhere (i.e. in Germany).

Measurement sites
The measurement sites are presented in Table 2. They consist of large open channels (LOC)

Ph Duchène et al.
(volumes over 1000 m3 and classic geometry), of small open channels (SOC), of cylindrical
tanks with a total floor coverage arrangement (FC), and of cylindrical tanks with diffusers
located in separate grids not covering the entire surface of the tank (GC).
Such a set of full-scale data merits the most complete statistical analysis possible, which
is done elsewhere (Capela, 1999). We shall focus here on those aspects leading to the most
evident practical conclusions concerning the diffuser layout and the aeration tank in rela-
tion to the specific aeration efficiency (SAE). Furthermore, several measurements were
performed at each site: with and without the simultaneous operation of mixers, at different
water depths or with different air flow rates.

Results and discussion

The results (Table 3) correspond to the best SAE value obtained at each site. The SAE val-
ues result from measurements made with the simultaneous operation of the mixers in the
open channels and from measurements made with only aeration in operation in the case of
the cylindrical tanks. UL is the mean horizontal water velocity (clean water,without
aeration).
The SAE value integrates a large number of parameters:
• the diffuser layout and the configuration of the aeration tank
• the efficiency of the blower
• the air flow rate per perforated surface area or per diffuser pores
• the efficiency of the diffusers
• the efficiency of the mixer.
The mean SAE values therefore only provide an initial evaluation criterion. They
evidence a highly significant difference between the large open channels (LOC) and the

Table 2 Site characteristics

Type Site number Volume (m3) Diffuser Air flow Air flow rate(2)
depth (m) rate(1) (Nm3.h–1.m–3) (Nm3.h–1.m–2)

LOC 13 1034–15830 3.8–7.6 0.4–1.6 43.0–148.0

SOC 7 117–820 2.5–5.4 0.5–2.7 28.7–95.2
FC 7 78–1434 3.1–6.1 0.9–2.7 67.5–113.0
GC 11 62–1320 3.8–5.2 0.8–2.5 63.3–152.0
(1) air flow rate per m3 of aeration tank volume
(2) air flow rate per m2 of perforated area of diffusers

Table 3 Oxygen transfer efficiencies as a function of the aeration tank type

Average SAE Range of SAE Increase in Change in P(mixer) UL (cm/s)

(kgO2.kWh–1) (kgO2.kWh–1) SOTR by SAE by P(blower)
mixing (%) mixing (%) (%)

LOC 3.41 2.5–4.3 + 43 + 27 12.6 36.2

SOC 1.95 1.3–2.6 + 36 + 10 23.6 35.3
FC 3.11 2.3–3.9 + 10 –12 24.8 16.0
GC 2.12 1.1–3.1
205

Downloaded from https://iwaponline.com/wst/article-pdf/44/2-3/203/32329/203.pdf

by Tufts University user
on 31 July 2018
26 Water 44.2_3 F2083 28/7/01 10:10 am Page 206

cylindrical tanks with a total floor coverage arrangement (FC), on the one hand, and
the miniaturized open channels (SOC) and cylindrical tanks with grids arrangement (GC)
on the other hand.
In the case of the large open channels, that are taken here as a point of reference, the rota-
tion of the liquid, above a mean velocity of 30 cm/s, allows us to minimize the spiral flow
effect and particularly the spiral flows which occur between the grids of diffusers or outside
the grids (Deronzier et al., 1998). It should be noted that the mean SAE value of these aera-
Ph Duchène et al.

tion tanks is quite clearly penalized by two results obtained with a type of diffusers proba-
bly less efficient (mean SAE = 3.55 kg O2/kWh for the best 11 results).

Cylindrical aeration tanks

From comparable data in terms of volume, water depth and specific air flow rate, the
mean SAE measured in tanks with a total floor coverage arrangement is 47% higher
than that measured in the case of grids arrangement. This result must, however, be
tempered by the results obtained in tanks with a grid arrangement where the air is supplied
by centrifugal blowers, the yield of which is 20% to 30% less than that of volumetric blow-
ers. Setting these measurements aside, the difference in SAE between the two arrange-
ments is close to 40%, which matches previous results (Deronzier et al., 1998). The grid
arrangement in fact induces vertical liquid circulation movements, which accelerate the
rise of the air bubbles, reduces their contact time and the oxygen transfer efficiency (Fujie
et al., 1997).
Interesting conclusions can be drawn from the comparison of the measurements per-
formed with a horizontal water velocity induced by the mixers.
• In all cases, despite the relatively high stirring powers (+25% on average), the mean hor-
izontal velocities is low (16 cm/s on average). This can be explained: i) by the use of
small blade mixers (just a few dozen centimetres in diameter) with high speed (a few
hundred revolutions per minute) and not, as in the large open channels, large blade mix-
ers (diameter from 1.8 to 3 m) with slow speed (a few dozen revolutions per minute)
which are more energy efficient for the direct thrust and the total induced flow rate and,
ii) by the fact that in these tanks the air bubbles reduce the efficiency of the thrust of the
horizontal axis mixers.
• The mean increase in the SAE due to the mixing is slightly higher in the tanks with a total
floor coverage arrangement (15% on average)1 than in the tanks with a grid arrangement
(on average 8%)1. We can suppose that in the case of the grid arrangement, when the aer-
ation is in operation, the bubble clouds constitute a headloss and the horizontal water
velocity tends to flow around the grids of diffusers, resulting in a lower effective
horizontal velocity on the diffusers than that calculated on the basis of the velocities
measured without aeration.
The influence of the horizontal velocity therefore clearly corresponds to that obtained in
large open channels for the tanks with a total floor coverage arrangement and would appear
to be only halved in the tanks with grids (Deronzier et al., 1998).
As a result of these slight increases in SAE, the simultaneous operation of mixers lowers
significantly SAE values, in the cylindrical aeration tanks. It can be concluded from this
that in these tanks the alternate operation of aeration and mixing must be ensured (which
improves the denitrification rate) and sufficient mixing by aeration alone (“self cleaning”
of the floor).
All these findings lead us to recommend, for small aeration tanks, a configuration in
cylindrical tanks with a total floor coverage arrangement. The periods of operation of aera-

206 1 Means established on reduced numbers in comparison with Table 3.

Downloaded from https://iwaponline.com/wst/article-pdf/44/2-3/203/32329/203.pdf

by Tufts University user
on 31 July 2018
26 Water 44.2_3 F2083 28/7/01 10:10 am Page 207

tion must be dissociated from those of mixing. Furthermore, in the tanks with a total floor
coverage arrangement, it must be possible to come back up the diffusers sector by sector,
given the frequency of problems and the relatively short life time of the elastomer diffusers
(a few years). The additional cost in comparison with the grid arrangement can be offset
extremely rapidly by owners, especially in the French context.
The other parameters for optimization of the aeration system in clean water are common,
and concern the number of diffusers, their density at the bottom of the tank, the air flow rate

Ph Duchène et al.
per perforated area of the diffusers, the water depth, and the type of blower. The first four
factors are all related : at a fixed air flow rate, the greater the number of diffusers, the higher
the SOTE. The water depth has practically no effect on the SAE. Various tanks studied with
two water depths failed to evidence any significant difference in the SAE values (Table 4).
This corroborates the results of Wagner and Pöpel (1998).
In practice, it is only under a depth of 3 m that the water depth has a negative effect on the
SAE. On the contrary, very high water depths do not improve the SAE in clean water, but
are likely to impact the alpha factor due to the progressive adsorption of the surfactants at
the interface of the bubbles
The variability in the results of the grid arrangement can be explained to a large extent by
the surface percentage of the tank occupied by the grids of diffusers. The best results,
exceeding 2.5 kgO2/kwh in SAE, were obtained in tanks where grids were added, often
after initially disappointing clean water tests and where the diffuser layout comes close to a
total floor coverage arrangement. In such circumstances, they no longer offer any econom-
ic advantages from the investment view sufficient to justify this non total floor coverage
arrangement.
Naturally, the case of the cylindrical tanks can be extrapolated to that of the parallelepi-
pedic tanks. Nevertheless, interesting performance results have been recorded in a tank
with the diffusers located in separate grids and equipped with a vertical axis large blade
mixer in a central position with a thrust from the surface towards the bottom
(OCNBenfeld=3.54 kgO2/kWh), with the mixer operating simultaneously. This result
remains to be confirmed, as the preliminary studies did not enable any modeling to be
made.

Miniaturization of the open channels

Economic constraints have a particular impact on the design of these small open channels.
In the case of small tanks, it is not easy to place the aeration tank, which is always ring-
shaped, around a central focal zone such as an anaerobic tank, an anoxic tank or even occa-
sionally a selector.
Two major types can be distinguished (Table 5): Type I, with simple miniaturization of
large open channels with an outside radius/inside radius ratio less than 1.5 (i.e. with large
central focal zone) ; and Type II, the inside radius of curvature of which is very small, with
the outside radius/inside radius ratio between 2 and 5 (i.e. with small central focal zone).
The example of the aeration tank no 3 allows us to illustrate the SAE range in the case of
Type I tanks. The aeration tank is equipped with two, diametrically opposite grids and a

Table 4 Influence of water depth on the oxygen transfer efficiency at the aeration tank of Graye/Mer (FC)

Water Volume Air flow rate SOTE Ps (W.m–3) SOTR SAE

depth (m) (m3) (Nm3.h–1.m–2) (%/m) (kgO2.h–1) (kgO2/kWh)

6.1 475 99.5 5.2 34.7 53.9 3.2

3.1 243 112.6 5.5 40.5 31.5 3.2
Ps: specific wire power(kW), Ps = Pw/V 207

Downloaded from https://iwaponline.com/wst/article-pdf/44/2-3/203/32329/203.pdf

by Tufts University user
on 31 July 2018
26 Water 44.2_3 F2083 28/7/01 10:10 am Page 208

Table 5 Characteristics of miniaturized aeration open channels

(SB: small blades; LB: large blades)

Site Type V H DD Mixer Pmixer Pblower Ps SOTE SAE (kgO2 Handicaps

no (m3) (m) (%) type (kW) (kW) (W.m–3) (%/m) .kWh–1)

1 I 216 3.1 6.7 SB 2.64 9.49 56 7.2 1.30 (1), (3), (5)
2 I 135 3.3 7.0 SB 2.33 5.02 47 4.2 1.48 (1), (3)
3 I 117 2.5 2.5 SB 1.58 6.52 69 4.8 1.27 (1), (2), (3), (4),
Ph Duchène et al.

(5), (6)
4 I 320 3.0 4.8 LB 1.06 7.32 26 5.6 2.81 (3)
5 II 823 5.4 7.5 LB 2.31 20.59 28 3.8 1.90 (2), (8)
6 II 820 5.4 5.0 LB 2.43 10.13 16 5.5 2.49 (2), (7), (8)
7 II 865 5.5 4.2 LB 4.41 17.60 26 5.2 2.26 (7), (8)

small blade mixer located immediately downstream of one of the grids. The air supply is
provided by a centrifugal blower and the diffuser depth is 2.25 m. The negative impacts on
oxygen transfer can be approached as follows.
1. The use of a small blade mixer (consuming an additional 1 to 1.2 kW, and lowering the
SAE by approximately 16%).
2. The mixer positioned in the bubbles clouds. In this case, aeration slows down the mean
velocity by 15 cm/s, instead of 1 to 2cm/s when the mixers are not directly affected by
aeration, which lowers the oxygen transfer by something in the order of 25% (Deronzier
and Duchène, 1997).
3. The small size of the tank implies a low volume/wetted surface ratio in relation to large
open channels, which overall doubles the mixing power required, with a subsequent
impact on the SAE of approximately 7%.
4. The mean horizontal velocity without aeration is only 21.5 cm/s, suggesting an oxygen
transfer reduced by something of the order of 10% in relation to the optimum obtained
above velocities of 30 cm/s.
5. The air supply from a centrifugal blower: loss of yield of at least 25% in comparison to a
volumetric blower (i.e. ROOTS type).
6. The low diffuser depth (<3 m): presumed effect on the SAE of 5 to 10%.
In the case of the site no 3, if the SAE values measured are corrected by the various coeffi-
cients above, the result would come close to 3 kgO2/kWh. The fact that this value is still way
below the performance measured on the large open channels comes from the fact that: (i) the
various coefficients 1. to 6. have been relatively reduced above, so as not to exaggerate the
importance of each factor analyzed separately, and that (ii) in this particular case, the longi-
tudinal arrangement of tubular diffusers with very low horizontal velocities is probably an
additional negative factor for the oxygen transfer efficiency.
In the case of the Type II tanks, with wide tanks in relation to their overall length and
with a small inside radius of curvature, the specific phenomena impacting oxygen transfer
are of two kinds.
7. The absence of homogeneity across the width of the field of horizontal velocities under
the combined effect of the relatively low power required, of the considerable width of
the tank and of the short distance between the mixer and the first downstream grid,
which all help to create counter currents.
8. The lesser efficiency of the large blade mixers due to the fact that the surface area con-
cerned by the mixers is relatively small in relation to the cross section of the tank and to
the fact that interior counter currents are created more easily as a result of the small
inside radius of the tank.
The various problems affecting the aeration performance of each tank are listed in the
208 last column of Table 5, with the codes used (n) in the comments above. The tank no 4 pres-

Downloaded from https://iwaponline.com/wst/article-pdf/44/2-3/203/32329/203.pdf

by Tufts University user
on 31 July 2018
26 Water 44.2_3 F2083 28/7/01 10:10 am Page 209

ents more or less the best example of what can be achieved in small tanks, evidencing fur-
thermore performance levels below those of the majority of cylindrical tanks with a total
floor coverage arrangement.

Conclusions
Fine bubble aeration by elastomer membrane for small aeration tanks, defined here as
having volumes of less than 1,000 m3 and especially for aeration tanks of less than

Ph Duchène et al.
500–600 m3, is often employed at the expense of the energy efficiency of oxygen transfer.
However, within this size range, the use of cylindrical tanks with a total floor coverage
makes it possible to reach easily SAE values that are worthwhile for local authorities, above
3 kgO2/kWh in clean water, provided that volumetric blowers are employed and moderate
specific air flow rates implemented.
The extra investment cost of this type of arrangement, maintaining diffusers that can be
raised sector by sector, does not warrant the continuing use of diffuser grids which induce
spiral flows uncontrollable by a simultaneous mechanical mixing.
The majority of tests performed in miniaturized open channels have so far resulted in
relative failures as far as SAE is concerned. Subject to certain precautions – efficiency of
the mixer and air supply equipment, minimum distance both upstream and downstream
between the mixer and a single array of diffusers, avoiding excessively small inside
radii – it is perfectly possible to obtain sufficient SAE values, even though they do not,
within this size range, exceed those of cylindrical or parallelepiped tanks with a total floor
coverage.

Acknowledgements
The authors thank M. Pierre Mauricrace for technical support, M. Alain Héduit for discus-
sion and Ile-de-France Regional Council for financial support.

References
ASCE (1992). ASCE Standard measurement of oxygen transfer in clean water. American Society of Civil
Engineers.
ASCE (1996). Standard Guidelines for In-Process Oxygen Transfer Testing. American Society of Civil
Engineers.
ATV (1996). Messung des Sauerstoffzufuhr von Belüftungseinrichtungen in Belebungsanlagen in
Reinwasser und in belebtem Schlamm.
Capela, S. (1999). Influence des facteurs de conception et des conditions de fonctionnement des stations
d’épuration à boues activées sur le transfert d’oxygène. Thèse de doctorat, Université Paris XII – Val de
Marne, 146 p. + annexes.
Capela, S., Gillot, S., Héduit, A. (1999). Oxygen transfer under process conditions: comparison of measure-
ment methods. 72nd Annual Conference and Exposition Water Environment Federation, New Orleans,
Louisiana, USA.
CEN (1999). PrEN 12255-15: Measurement of the oxygen transfer in clean water in activated sludge aera-
tion tanks. 20 p.
Deronzier, G., Duchène, Ph. (1997). Vérification de la vitesse horizontale dans les chenaux d’épuration
pourvus d’un système d’aération par insufflation d’air. TSM, 3, pp. 35–41.
Deronzier, G., Duchène, Ph. and Héduit, A. (1998). Optimization of oxygen transfer in clean water by fine
bubble diffused air system and separate mixing in aeration ditches. Wat. Sci. Tech., 38(3), 35–42.
Fujie, K., Kasakura, T., Tsuchiya, K. and Tomita, Y. (1997). Effects of air diffuser arrangement on hydrody-
namics and oxygen transfer in activated sludge aeration tanks. Env. Tech., 18, 663–668.
Gillot, S., Deronzier, G. and Héduit, A. (1997). Oxygen transfer under process conditions in an oxidation
ditch equipped with fine bubble diffusers and slow speed mixers. 70th Annual Conference and
Exposition Water Environment Federation , Chicago, Illinois, USA.
Gillot, S., Capela, S. and Héduit, A. (2000). Effect of horizontal flow on oxygen transfer in clean water and
in clean water with surfactants. Wat. Res., 34, 678–683. 209

Downloaded from https://iwaponline.com/wst/article-pdf/44/2-3/203/32329/203.pdf

by Tufts University user
on 31 July 2018
26 Water 44.2_3 F2083 28/7/01 10:10 am Page 210

Heduit, A. and Racault, Y. (1983a). Essais d’aérateurs. Enseignements tirés de 500 essais en eau claire
effectués dans 200 stations d’épuration différentes – I) Méthodologie,. Wat. Res., 17(1), 97–103.
Heduit, A. and Racault, Y. (1983b). Essais d’aérateurs. Enseignements tirés de 500 essais en eau claire
effectués dans 200 stations d’épuration différentes – II) Résultats,. Wat. Res., 17(1), 97–103.
Redmon, D.T., Boyle, W.C. and Ewing, L. (1983). Oxygen transfer efficiency measurements in mixed
liquor using off-gas techniques. JWPCF, 55(11), 1338–1347.
Wagner, M.R. and Pöpel, H.J. (1998). Oxygen transfer and aeration efficiency – Influence of diffuser sub-
Ph Duchène et al.

mergence, diffuser density, and blower type. Wat. Sci. Tech., 38(3), 1–6.

210

Downloaded from https://iwaponline.com/wst/article-pdf/44/2-3/203/32329/203.pdf

by Tufts University user
on 31 July 2018