Water Environment J - 2007 - Morgenroth - Sequencing Batch Reactor Technology Concepts Design and Experiences Abridged
Water Environment J - 2007 - Morgenroth - Sequencing Batch Reactor Technology Concepts Design and Experiences Abridged
E x pe r ie n c es
(A bridged)
E. Morgenroth, Dipl-lng and P. A. Wilderer, Dr-lng, Dr.h.c*
Fiu
static, mixed
React
aerobic, anoxic
Settle Sludge
wastage
J."w Draw & (Idle)
or aerated or anaerobic
Cycle time >
t phases of SBR cycle (phases are repeated and order may be
upon process objective)
Inthent Effluent
holding tank buffer tank
treatment (optional) (optional)
\ /
COD-removal mix
may consist of a short fill phase followed by an aerated (short SRT) aerate
phase, settle and draw (a). The same cycle can be used to settle
achieve nitrification; then, however, only the sludge age
has to be increased (b). To achieve nitrogen removal, an
unaerated phase has to be introduced in the SBR cycle 6) fill
(c). Oles and Wilderer@)proposed a SBR cycle with three COD-removal mix
nitrification
fill phases, where the second and the third fills are used (long SRT)
to supply the denitrifiers with the required source of
electron donors. To achieve enhanced biological phos-
phorus removal, an anaerobic phase must be included in
(c)
the program (d). The first fill phase is used to denitrify COD-removal
any nitrate remaining in the reactor after draw. The nitrification
readily biodegradable COD of the second fill phase is denitrification
then made available for the release of phokphorus and the
accumulation of poly-P-hydroxybutyrate. The COD of
the third fill phase is used as a carbon source for (4
denitrification. COD-removal
nitrification
denitrifhtion
Draw enhanced biol.
Many different decanters are being used to remove P-removal
the treated supernatant from the reactor (Fig. 4), and
three categories can be differentiated. The decanters Fig. 3. Examples of SBR cycles for removal of COD,
shown in (a) and (b) remove the supernatant from the Nand t? (Cycle design depends upon composition
surface of the reactor. In (a) a motor is u6ed to rotate the
trough with the overflow weir; therefote the speed of of sewage. Note that given cycles are just examples
rotation to lower the weir controls the rate of draw. In (b) of many possible design options.)
Fig. 4. Vpical decanter mechanisms: (a) motorized unit rotates header pipe,
(b) floating header type, and (c) fixed-depth decantefi7)
Fig. 5, Unaerated mixing can be provided using (a) submergedjets, (b) floating
mixing device (low speed), (c) submerged horizontal (or vertical) propeller
Fig. 6. Aeration of SBR using (a) jet aeration, (b) floating surface aerator (high
speed), and (c) compressedair
a decanter is shown which floats on the surface of the by Ardern and Lockettcl), a number of other variable-
water, and the draw rate is determined by the hydrostatic volume processes have evolved. In the 1950s, Pasveer(*)
pressure at the inlet of the decanter. In (c) a decanter is proposed fill-and-draw operation of aeration ditches,
shown which is installed at a fixed depth in the SBR. therefore making final clarifiers unnecessary. In contrast
to the SBR concept, the Pasveer ditches are continuously
Mix fed, also during the settle and decant phase. T h e CAST
Different mixing devices are illustrated in Fig. 5. In (cyclic activated sludge treatment) process can be
the jet mixer (a), pumps force the activated-sludge sus- considered as a further development of the Pasveer
pension through jet nozzles. A rotor floating on the system. It is similar to the SBR in that different pro-
surface (b) can be used for mixing, and mixing can also be cesses take place at different times in the cycle. However,
provided using a submerged propeller (c). the CAST reactor is divided into three sections to
provide a spatial separation of a separate ‘fill’ zone and a
Aerate ‘draw’ zone. Sludge is pumped from the final section
A number of devices are available to aerate the of the tank into a small influent chamber to achieve
reactor (Fig. 6), some of which can also be used for a selector effectc9).T h e sequential operation of the SBR
unaerated mixing (a, b). If air is introduced into the can also be applied in biofilm systemd’O). In the
suction point of the jet (a), it can be used for aeration. If ‘sequencing batch biofilm reactor’ (SBBR), the same
the floating rotor is operated at a higher speed, it acts as phases as described in Fig. 1 are applied in a reactor
a surface aerator (b). Air or pure oxygen can be packed with a biofilm support medium. However,
introduced using bubble aeration (c). because the biomass is fixed on the support, no sedimen-
tation phase is required. Therefore, the volumetric
Other Sequentially Operated Sewage-Treatment exchange ratio is not limited by the sludge volume, and
Systems up to 100% of the bulk liquid in the reactor can be
In addition to the classical SBR systems as invented exchanged during each cycle.
Corn p a r ison Of C 0 n t i n u o us- FI o w PI ants a n d function of the comparable recirculation rate ( a Qin the
SBRs continuous-flow systeml:
1
Depending upon the mode of operation, SBR systems fexr = - (1)
can be compared with ‘plug-flow reactors’ (PFRs) or l+a
‘continuous-flow stirred tank reactor’ (CFSTR) From Eq. (1) and Fig. 8 it can be seen that normal
systemscll). In Table 1, mass-balance equations for SBR recycle ratios (a)of 1-4 (including return sludge and
and continuous-flow systems are compared. It can be internal recycles) in the continuous-flow system are
seen that the differential equation for the SBR with comparable to f,, in the range 0.2-0.5.
dump fill is the same as the differential equation of the In the foregoing discussion it was pointed out that
PFR, where the residence time in the PFR (t)is similar the micro-organisms, in both the SBR and the
to the reaction time in the SBR. T h e mass-balance continuous-flow activated-sludge system, are periodi-
equation for the SBR with slow fill over an extended cally exposed to a series of process conditions. Therefore
period of the cycle time and for small volumetric the activated sludge which develops should possess
exchange ratios (fexr = AV/(V,+AV)) resembles the similar physical and metabolic properties. Theoretically,
mass-balance equation of the CFSTR. In Fig. 7, two both systems should be applicable for identical purposes;
SBR cycles are shown with their corresponding however, in practice both systems have their specific
continuous-flow system. T h e recirculatidn of sludge and niches. T h e SBR has the advantage of a more flexible
treated water in the PFR system is comparable to the
water remaining in the SBR after draw Po). From that,
the volumetric exchange ratio f,,, can be expressed as a $A list of abbreviations can be found at the end of the paper
I-
I
-
A
T I
’
cve-
> aQ Qw
settle
draw
aQ Qw’
For the design of SBR systems, similar principles can be From the above analysis it can be seen that SBRs and
applied as for the design of continuous-flow activated- continuous-flow systems are comparable; on the other
hand, both systems have their specific niches of appli-
W C T = University of Capetown
cation. Comparing the required volume of a SBR and a
continuous-flow system, the volume of the SBR will be qualifies for decentralization of sewage-treatment
larger than the corresponding aeratiorl tanks in the systems on the basis of savings in construction and
continuous-flow system. However, if the volume of the maintenance of long-distance sewers.
SBR is compared to the total volume of the continuous-
flow plant (aeration tanks plus secondary clarifier), the
SBR usually requires less volume. The reduction in size Conclusions
is not the main advantage of the SBR mare important is
the inherent flexibility of operation offered by the SBR. 1. SBRs were the first applications of activated-sludge
The combination of metabolic reactions and sedimen- technology and were converted into continuous-flow
tation in one tank is advantageous for obvious reasons, systems because of various problems at the time. With
and the duration of the process phases can be adjusted modern technology these problems can be overcome.
easily to any actual conditions without the need to 2. The SBR offers considerable flexibility in the oper-
structurally retrofit existing tanks. Sedimentation pro- ation of sewage-treatment plants where different
ceeds in the SBR unaffected by currents - as in phases (fill, react, settle, draw, idle) can be combined
continuous-flow clarifiers. A SBR plant consisting of a in many different combinations.
number of reactors and holding tanks can be designed to 3. Various mechanical devices are availablefor draw, mix
be compact. Modular design allows adjustment of the and aerate. Most of these devices are proprietory.
plant's capacity by adding an additional reactor, or taking 4. The design of SBR systems can be derived from
reactors out of service, respectively, as the influent empirical procedures used for the design of
conditions change. This aspect is of specific interest not continuous-flow activated-sludge plants. However,
only for industry but also for municipalities (e.g. tourist for better reliability of the calculations, SBRs should
resorts, industrial dischargers with varying operation, be designed on the basis of mass balances. Cycle times
and expanding villages). Finally it should be realized that can be optimized using numerical simulations.
a SBR plant needs to be automatically controlled, which 5. A comparison of the mass balances for SBRs and
may be considered as a positive or a negative feature. continuous-flow plants reveals that a SBR with a
With modern data-communication systems, SBR plants short fill phase is comparable to a plug-flow reactor,
can easily be remote controlled, and SBR technology whereas a SBR with a slow fill and a small volumetric
exchange ratio can be compared to a continuous-flow (7) KETCHUM,L. H. J. Design and physical features of sequencing batch
stirred-tank reactor. reactors. Wat. Sci. Technol.,1997,35, (I),11-18.
6. Even though SBR and continuous-flow systems are (8) PASVEER,A. Contribution to the development in activated sludge
comparable, both technologies offer their specific treatment. J. Proc. Inst. Sewage Purif., 1959, (4),436.
advantages. The SBR allows for a more flexible (9) DEMOULIN,G., GORONSZY,M. C., WUTSCHER,K. AND FORSTHUBER, E. Co-
operation. A continuous-flow system may be advan- current nitrificationldenitrificationand biological P-removal in cyclic
tageous for complex sludge and wastewater recycles activated sludge plants by redox controlled cycle operation. Wat. Sci.
within the system, and for handling highly variable Techno/.,1 9 9 7 , (I),
~ 215-224.
influx from combined sewers. (10) WILOERER,P. A. Sequencing batch biofilm reactor technology. In:
Harnessing Biotechnology for the21st Century (M. R. Ladisch and A.
Bose. Eds.) Am. Chem. Sot., 1992,475-479.
(11) IRVINE,R. L. AND KETCHUM,L. H. Sequencing batch reactors for bio-
L i s t of A b b r e v i a t i o n s
logical wastewater treatment. CRC Critical Reviews in Environmental
Control, 1989,18, (4),255-294.
A = cross-sectional area of plug-flow reactor (12) GONZALEZ-MARTINEZ, S. AND WILOERER, P. A. Phosphate removal in a
Cs, CS,in,ACs = substrate concentration, influent sub- biofilm reactor. Wat. Sci. Technol., 1991,23,(7-9), 1405-1415.
strate, substrate removed
(13) ORHON,D. AND ARTAN,N. Modelling of Activated-Sludge Systems.
cx = sludge concentration Technomic, Lancaster, 1994.
CFSTR = continuous-flow stirred-tank reactor (14) HENZE,M., GRAOY,C. P. L., GUJER,W., MARAIS,G.v.R. AND MATSUO,T.
fDN = fraction of total electron acceptor Activated Sludge Model No. 1. lAWPRC Scientific and Technical
demand associated with anoxic reactions
Reports, No. 1, IAWQ London, 1987.
far = AV/(Vo+AV), volumetric exchange
ratio in the SBR
MX = mass of sludge Discussion
n = number of SBRs
D r J. Barnard (Reid Crowther International), opening the
PFR = plug-flow reactor discussion, said that he was pleased to see that the ‘professors’
Q = flow-rate and the ‘practitioners’ had finally got together. In laboratory
rs,v = volumetric conversion rate work the rapid-fill technique was always used, but in practice
SVI = sludge-volume index the slow-fill method had been employed. He said that he had yet
SFV,SFx = safety factors to find a SBR with a SVI of less than 180 ml/g. He considered
t,, tf, t,, tD, tN,=phase duration with c: cycle, f: fill, that with rapid fill this could be achieved, but for the larger
t,, t,, td, ti r: react, D: denitrification, N: nitrifi- plant an extra holding tank and associated pumping would be
cation, s: settle, w: wastage, d: draw, i: idle required.
V = volume With respect to the simultaneous nitrification/denitrifi-
Vo, AV,VR = minimal volume, exchanged volume, cation which was essential, he referred to the work which had
been carried out at the University of Capetown on this phenom-
total volume of the SBR
enon in terms of sludge bulking, which was observed to result
VQ N = reactor volume in a continuous-flow from nitrous oxide, etc. He asked the authors to comment upon
plant dedicated to denitrification and this aspect.
nitrification M r A. Boon (Hyder Consulting) considered that it was
X = distance in the PFR important to provide conditions in the aeration tank which
YH,net = net. yield including decay and produc- would encourage the activity and growth of protozoa in order
tion of particulate products to achieve effluents of high clarity (low SS and BOD); the role
a.Q = recycle in a continuous-flow system of such protozoa had been conclusively demonstrated earlier.
= correction factor anoxic respiration He asked if the authors would agree that correct design and
r operation of a SBR, or a plug-flow aeration system, should
t = xA/Q, residence time in the PFR
ensure that the D O concentration in the mixed liquor at the end
of the aeration period would be about 4-6 mg/l in order to
provide the conditions necessary for the selection and growth of
References appropriate protozoa.
M r S. Williams (Thames Water) commented that he had
been operating three nutrient-removal plants in the Thames
(1) AROERN, E. AND L O C K EW. ~ ,T. Experiments on the oxidation of sewage
Water area, all of which had SSVI values of 60-70 ml/g.
without the aid of filters. J. SOC.Chem. lnd., 1914, 33, 523. He referred to the authors’ remarks with regard to the
(2) DUCKWORTH, W. H. Aeration experiments with activated sludge. Proc. possibility of sludge production from a SBR being lower, and
Ann. Meeting. Manchester District Branch of the Assoc. Managers of wondered if this was a feature of a ‘feasdfamine’ situation and
Sewage Disposal Works, 1914,50. occurred in the plug-flow type reactor in which it was the plug-
(3) MELLING,S. E. Purification of Salford sewage along the line of the flow nature which resulted in a lower sludge yield and not the
Manchester experiments. 1 Soc. Chem. lnd., 1914,33,1124. SBR feature.
(4) IRVINE,R. L. AND BUSH,A. W. Sequencing batch biological reactors - an M r F. Xiong (Air Products PIC)said that, because it was a
overview. J. Wat. Pollut. ControlFed., 1979,51,235. plug-flow system, the oxygen demand or uptake rate would
(5) IRVINE,R. L., WILOERER, P. A. AND FLEMMING, H.-C. Controlled unsteady reduce during the reaction phase, and he asked the authors what
they advised as the best strategy for oxygenation control.
state processes and technologies - an overview. Wat. Sci. Technol.,
M r A. Ware (Cegelec Projects Ltd) asked for the authors’
1997,35,(l), 1-10. opinions of baffled pre-reactor zones and on the value of
(6) OLES, J. AND WILOERER, P. A. Computer-aided design of sequencing internal sludge recycle into pre-reactor zones. He said that
batch reactors based on the IAWPRC activated sludge model. Wat. Sci. there had been other reference to this aspect - resulting in
Technol., 1991,23,(4-6), 1087-1095. improved sludge settlement characteristics.
M r T. Pitman (Great Johannesburg Metropolitan limited to SBR systems. The extent of the decreasing sludge
Council) expressed concern about the comparison and consis- yield to be expected in a SBR was still under discussion and, in
tency of effluent quality. He said that, in his experience, a the literature, a large range had been reported for the sludge
continuous-flow process would consistently achieve (a) effluent production in SBRs.
amm. N concentrations of less than 0.5 mg/l and (b) consistent In answer to Mr Xiong, the apthors said that in a plug-flow
nitrification. He said that from figures he had seen in the or SBR system the respiration rates would vary with space or
literature, the SBR gave a sine-wave effect of effluent amm. N time, respectively. In order to save energy, adjustment of the
and nitrate concentrations which were out of phase. He asked if aeration intensity according to the oxygen requirements, was
this was normal for the process or whether it was possible to necessary. Therefore, at the end of a plug-flow reactor or at the
avoid this apparent effect. end of a SBR cycle, the aeration could be reduced. Mr
M r E.De Jong (Degremont UK Ltd) said that, although Morgenroth said that the yet unanswered question was whether
problems with sludge settlement within the sequencing batch those initial peaks of the oxygen requirement must be fully
biofilm reactor process had been overcome, there was still the satisfied at the inlet of the plug-flow reactor or at the beginning
need to remove the sludge by some kind of backwash, resulting of the SBR. From present results, he concluded that the
in a sludge of very low DS concentration (2-3 mg/l) which aeration device did not have to fully meet the initial high oxygen
required subsequent settlement and treatment. He thought requirements. No adverse effects on sludge settling or effluent
that, with submersed media, the success of this was always quality were observed, and more economic design would result
dependent upon a proper backwash. from this procedure.
Responding to Mr Ware, the authors said that if the
treatment plant had to treat a continuous waste stream and
consisted of only a single SBR, a baffle was necessary to avoid
Authors’ R e p l y short-circuiting during the draw period. The initial idea of the
SBR was to have a completely-mixed system where the con-
In response to Dr Barnard, the authors referred to their ditions varied only over a period of time, compared with the
presentation where they had shown that the flow regime in a continuous-flow system where conditions varied over space. By
SBR was, in many ways, comparable to continuous-flow also introducing a baffle and an internal sludge recycle, a spatial
systems. In continuous-flow systems, nitrate was recycled to the gradient was achieved in the SBR. This spatial gradient could
anoxic zone, comparable to the SBR where nitrate (remaining increase the initial substrate to organism ratio (So/&), leading
in the reactor after draw) was available for denitrification in the to improved sludge settlement characteristics. However, an
next cycle. Denitrification could be further enhanced by initial high substrate loading could also be achieved by
operating the SBR with multiple-fill phases which could be operating the SBR with a rapid fill, and by introducing a baffle
considered to be a combination of pre- and post-denitrification. and pumping of sludge, the simplicity and ease of operation of
Therefore nitrogen removal in the SBR did not have to rely the SBR was lost. Mr Morgenroth said that he had mixed
upon simultaneous nitrification/denitrification,and significant feelings about the positive effect of combining spatial and time-
problems with nitrous oxide had not been observed. dependent gradients in one system.
They agreed that, if a SBR was operated incorrectly, In reply to Mr Pitmann, the authors said that they were not
sludge bulking problems would result - just as in any aware of such a sine-wave effect of effluent concentrations
conventional continuous-flow process. However, SBRs which unless it was derived from inhomogenous loading of different
were operated with rapid fill and a large volumetric exchange SBRs within a system. However, in such a case it would be
ratio had been shown to produce sludges with excellent settling expected that a continuous-flow system would exhibit the same
properties. effects.
Responding to Mr Boon, the authors agreed with his In conclusion, the authors said that they agreed with Mr
comments about the importance of protozoa to the success of De Jong’s comments confirming that, in a sequencing batch
the process. biofilm reactor, proper backwashing was essential. Biomass had
Replying to Mr Williams, they agreed that for the variation to be removed from the reactor to prevent (for example)
between very high and low organic concentrations, in either clogging. Biomass which was subsequently removed must be
plug-flow or SBR regimes, lower sludge yields had been separated from the backwash water. However, the volume
reported. During the high initial substrate concentrations, required for the solid-liquid separation in a biofilm system was
bacteria made inefficient use of the substrate. Those ‘feast and smaller than in a sedimentation tank in a comparable activated-
famine’ conditions leading to a smaller sludge yield were not sludge system.