CHAPTER SIX
AEROBIC DIGESTION OF WASTE ACTIVATED
AND PRIMARY SLUDGES
GA Ekama and MC Wentzel
ABSTRACT
The steady state activated sludge model
developed in Chapter 4 is applied to the special
case of aerobic digestion of waste activated sludge
(WAS), primary sludge (PS) and W AS-PS blends.
The kinetics of aerobic digestion are initially
developed for the batch aerobic digester, which
was used originally to measure the endogenous
respiration rate of the OHOs independently of their
yield coefficient. Equations for the VSS, free and
saline ammonia (FSA) and orthophosphate (OP)
concentrations versus time are also developed,
which are required to calculate the lime dose to the
digester to maintain a near neutral pH for efficient
digestion and complete nitrification.
The aerobic digester kinetics are then applied to
single and multi in-series completely mixed
digesters to develop a design chart from which the
required retention time and associated fraction
VSS removal for defined influent active fraction
and a specified effluent active fraction can be
estimated. With the fraction VSS removal known,
equations are developed to calculate the
carbonaceous, nitrification (if complete) and total
oxygen demands and outflow VSS, FSA and OP
concentrations for single and multiple reactor
aerobic digesters.
From a comparison of single and multiple reactor
digesters, it is concluded that the multiple reactor
digesters allows a significant reduction in digester
retention time, and hence volume, for the same
fraction VSS removal, i.e. same effluent active
fraction. However, this volume reduction cannot
be realised in practice due to the very high oxygen
utilization rate (OUR) in the first reactor. The
digester feed concentration therefore is limited by
the oxygen transfer rate (OTR) of the aeration
system and single, doubled, and triple reactor
digesters have to be feed decreasing influent TSS
concentrations to limit the OUR to within the OTR
capacity of the aeration system. For a specified
digester retention time an equation is developed
for the maximum digester VSS feed concentration
6.1
in terms of the maximum OUR limit in the first
reactor. Coupling this equation with a dissolved air
flotation thickening limit, allows the minimum
digester volume to be calculated for the required
design retention time. For the WAS from an 8d
sludge age activated sludge system treating the
example settled wastewater, it is concluded that a
double reactor digester gives the smallest digester
volume.
The same approach is followed for aerobic
digestion of PS and W AS-PS blends. The only
difference is that synthesis of active OHO VSS
mass from the primary sludge biodegradable
organics is included, which is assumed complete
in the first reactor of a multi-reactor digester.
Because this growth process adds a significant
oxygen demand to the first reactor, it is concluded
that for PS and WAS-PS blends, single reactor
d ig e s te rs yie ld th e hig h e s t fe e d T S S
concentrations and hence the lowest digester
volume for the specified effluent active fraction.
Again PS (gravity 6%) and WAS (flotation 5%)
thickening limits are coupled with the digester feed
TSS concentration equations in terms of a
selected maximum OUR in the first reactor and
required retention time. For the single reactor
digester, which gives the minimum digester
volume, the digester feed concentration is limited
by the PS and WAS thickening limits and yields an
OUR of 98 mgO/( R.h). The feed TSS concentration
to double and triple reactor digesters is limited by
the OUR [set at 125 mgO/( R.h)] in the first reactor
and yield larger digester volumes. Step feeding
the multiple reactor digester equalizes the OUR in
each reactor, but then digester does not achieve
the required effluent active fraction (or fraction
VSS removal), unless it is operated at the single
reactor digester retention time. Equations are also
derived to calculate the effluent nitrate (with
complete nitrification) and OP concentrations, or,
with PS feed only, the required FSA and OP dose
for uninhibited OHO sludge growth.
The activated sludge - aerobic digester model is
applied to compare a long sludge age (30d) raw
wastewater extended aeration WWTP with a short
sludge age (8d) settled wastewater treatment plant
including aerobic digestion of PS and WAS to the
same residual biodegradable organics (effluent
active fraction) as the raw wastewater system. It
is shown that the COD, N and P balance over the
two WWTPs and that the oxygen demand is the
same in both, with 54% in the activated sludge
system and 46% in the aerobic digester of the
settled wastewater system. However, the total
reactor volume of the settled wastewater system is
only 70% of the raw wastewater system.
Finally the aerobic digester model is validated by
comparing it with the USEPA aerobic digester
design guide of fraction VSS removal versus the
product of the retention time and temperature
(SRTxT). The aerobic digester model slightly
estimates the fraction VSS removal but the trend
line is similar. The model shows that retention
time and temperature do not affect the fraction
VSS removal equally so that theoretically the
SRTxT is not a parameter consistent with the
model and different fraction VSS removal versus
SRTxT lines are obtained for different influent
active fractions. It is concluded that the model
compares well with the pilot and full scale data on
which the USEPA design guide is based.
From a relationship linking the effluent active
fraction and residual specific oxygen utilization rate
(SOUR) of the digested sludge, an influent active
fraction versus digester retention time design chart
is developed for different residual SOUR (1 to 3
mgO/R.h) and corresponding fraction VSS removal.
This allows rapid selection of retention time and
fraction VSS removal from an influent active
fraction and a required sludge stability (residual
SOUR).
A comparison of the steady state model calculation
results and dynamic simulation model predictions
based on ASM #1 for the long sludge age (30d)
raw wastewater extended aeration WWTP and the
short sludge age (8d) settled wastewater treatment
plant including aerobic digestion of PS and WAS to
the same residual biodegradable organics (effluent
active fraction) as the raw wastewater system, is
given in Chapter 11. It is shown that virtually
identical results are obtained.
6.2
1 INTRODUCTION
To avoid the complexity of operating anaerobic
digesters, treating raw wastewater at a very long
sludge age (>30d), called extended aeration, is
sometimes selected. With extended aeration,
sludge stabilization takes place in the activated
sludge reactor (see Section 8 above) and the
biodegradable organic content in the waste sludge
is low enough to discharge the sludge direct to
drying beds. This approach arose out of the
general requirement for nitrification. To ensure
complete nitrification (low effluent ammonia
concentration) at the winter temperature of around
14 oC, requires sludge ages in excess of 10 days.
Once sludge ages of this magnitude were
required, it seemed practical to increase the
sludge age to 30 days because (i) the oxygen
requirement is only about 12% more and (ii) the
absence of the complexity of sludge treatment
made the larger biological reactor more attractive
both economically and operationally. However,
treating settled wastewater at a short sludge age
and treating the primary and secondary sludge in
a satellite aerobic digester in a concentrated form
leads to a significantly reduced reactor volume and
may be more attractive economically than
extended aeration.
While evaluating this alternative it became evident
that the advantages of treating settled wastewater
at short sludge ages may be combined with the
advantages of long sludge ages by including
aerobic digestion to the short sludge age system.
Furthermore, in the experimental validation of the
steady state activated sludge model developed in
Section 6 above, it became clear that the aerobic
digestion theory, which is a subset of the activated
sludge model, can make a considerable
contribution to activated sludge kinetics by
providing a means for independently determining
one of the model kinetic constants, i.e. the
endogenous respiration rate, b H (see Section
3.2.1.2 above). As a result, an investigation into
the kinetics of the aerobic digestion system was
undertaken, the results of which are described by
Marais and Ekama (1976) and Warner et al. (1984,
1986). This Section gives a brief summary of this
investigation into aerobic digestion with an
emphasis on kinetics and process design and is
followed by utilizing the activated sludge model to
the process design and operation of waste
activated and primary sludges. The procedure is
extended to anoxic -aerobic digestion in Chapter 8,
Section ??.
The questions facing the designer are inter alia, (i)
influent characterization - what are the active and
unbiodegradable concentrations in the influent to
the digester, (ii) what retention time is required in
the digester to reduce the active fraction or
remaining biodegradable organics content to a
specified value, (iii) how much oxygen must be
supplied, (iv) what is the VSS reduction in the
digester, (v) what are the effluent VSS, TKN, FSA,
NO 3 -N, TP and OP concentrations to determine
the mass flow rates of these compounds for
downstream processing and (vi) what chemical
dosages are required to maintain a near neutral
pH.
6.3
2 BASIC KINETICS
Waste activated sludge entering the aerobic
digester is characterised by dividing it into two
fractions, (i) unbiodegradable organics and (ii)
biodegradable organics. The unbiodegradable
organics comprise the unbiodegradable particulate
organics (X I) and the endogenous residue (X E )
components in the waste activated sludge from the
activated sludge system. The biodegradable
organics are part of the active OHOs (X BH ) in the
waste sludge. In aerobic digestion the active
organism biodegradable organics are reduced
through endogenous respiration or equivalently in
d yn am ic m o d e l k in e tic s th ro u g h d ea th -
regeneration. Because the unbiodegradable and
active organisms concentrations are organic, they
are best measured with the volatile settleable
solids (VSS) test. The TSS test could also be
used but because this includes the inorganic
solids, it unnecessarily complicates the aerobic
digestion model. However, equations for the
steady state model in terms of TSS are given. It
should be noted that the VSS/TSS ratio does not
remain constant during aerobic digestion -
generally the VSS concentration decreases more
than the TSS concentration resulting in a reduction
in VSS/TSS ratio during digestion. The ISS model
developed by Ekama and Wentzel (2004) will be
applied to derive the TSS based aerobic digestion
model making the effluent VSS/TSS ratio a
predictable parameter.
During digestion, OrgN and OrgP bound up in
organism mass are released which increase the
ammonia (NH 3 ) and ortho-P (OP) concentrations
in the bulk liquid. If nitrifiers are present in the
waste sludge, this ammonia may be nitrified to
nitrate, which may reduce the H 2CO 3* alk and pH.
However, with anoxic-aerobic digestion, the nitrate
generated can be denitrified to recover the
alkalinity loss and also reduce the oxygen
demand. If BEPR waste activated sludges are
a e r o b i c a l ly d i g e s t e d , h i g h b u l k l iq u i d
concentrations of OP will occur from the release of
the internally stored P in the PAO component of
the sludge mass due to endogenous mass loss of
the organism group.
In the aerobic digestion model, it is assumed that
(i) the unbiodegradable components (X E + X I) in
the influent to the digester remain unaffected with
the result that their concentration does not change
in the digester, (ii) the inorganics concentration in
the influent also does not change, i.e. (no
precipitation or dissolution of inorganics during
digestion, (iii) the unbiodegradable organics are
lumped together and denoted X Ii viz. the
unbiodegradable VSS in the influent to the
digester, (iv) the active organisms (X BH ) decrease
in the digester via endogenous respiration. This
decrease in OHO concentration gives rise to five
ancillary effects, i.e. (v) a generation of
unbiodegradable VSS in the form of “new”
endogenous residue (X E ) which is 20% (f) of the
active organisms that are “lost” in endogenous
respiration, (vi) a decrease in the ISS
concentration associated with the OHO decrease -
the lower the OHO concentration, the lower its
contribution to the measured ISS concentration via
the intra-celluar dissolved solids which precipitate
in the VSS-TSS test procedure (Ekama and
Wentzel, 2004), (vii) the utilization of oxygen for
endogenous respiration, (viii) the release of
ammonia and OP to the bulk liquid, the former of
which may be nitrified and (ix) the utilization of
oxygen for nitrification.
6.4
3 BATCH AEROBIC DIGESTION
3.1 Reduction in VSS concentration
From (i) and (iii) above, the initial VSS
concentration in the batch digester, X vi, is given by
mgVSS/R (6.1)
Generally, in the aerobic digester under batch
conditions;
mgVSS/R (6.2)
where
X = volatile solids concentration in VSS
(mgVSS/R) and subscripts v, BH, E, and I
denote volatile, active organism, endogenous
residue and unbiodegradable components of
the VSS respectively, and additional subscripts
i, e and t refer to the VSS concentrations in the
influent, effluent and at time t respectively.
Only X BHt and X Et change during digestion; X I
remains constant and equal to the influent value
X Ii, which in turn is equal to X E + X I from the
activated sludge system.
During digestion, the active organisms decrease
according to Eq 6.17, which under batch
conditions has the solution
mgVSS/R (6.3)
The endogenous residue concentration in the
digester is given by f times the change in the X BH
concentration viz.
mgVSS/R (6.4)
Substituting Eqs 6.3 and 6.4 for X BHt and X Et in Eq
6.2 and simplifying yiels,
(6.5)
Noting that the terms fX BHi and X I are constant with
time and moving these to the LHS of Eq 6.5 yields,
(6.6)
In Eq 6.6, as t 64, 60. Hence at t = 4,
mgVSS/R (6.7)
Substituting Eq 6.7 into Eq 6.6 yields,
Fig 6.1: Relationship between various components
of VSS versus digestion time in batch aerobic
digestion of activated sludge .
mgVSS/R (6.8)
The inter-relationships between the different VSS
concentrations in Eqs 6.3 and 6.8 is shown in Fig
6.1 for b H = 0.24/d at 20 o C and f = 0.20. At the
start of the batch test, the VSS (X vi) comprises only
active (X BHi) and unbiodegradable (X Ii) VSS, i.e.
X Ei=0. As the test proceeds, the active
concentration (X B H t ) decrease s and the
endogenous concentration (X Et) increases at 20%
of the X BH decrease. After a very long time
(infinity), the VSS concentration (X v4 ) comprises
only unbiodegradable VSS arising from two
sources, the original unbiodegradable VSS in the
influent (X Ii = X I) and the endogenous residue
generated during digestion, being 20% of the initial
active organism (X BHi) concentration (Eq 6.7).
Equation 6.8 indicates that if the constant VSS
concentration at time infinity, X v4 , is subtracted
from the measured X vt values, then the difference
X vt - X v4 plots as a straight line in a semi log plot of
ln(X vt - X v4 ) versus time. Practically X v4 is unknown.
To overcome this problem, different values for X v4
are chosen to plot the ln(X vt - X v4 ) versus time
graph from measured X vt data. If the chosen value
is too high, the plot will curve downwards and yield
a low correlation coefficient; if the chosen value is
too low, the plot will curve upwards and yield a low
correlation coefficient. With the correct value, the
plot will give a straight line and give the highest
correlation coefficient possible with the measured
VSS data. Once X v4 is known from the best
straight line fit to the data, the slope of the line
gives the endogenous respiration rate (b H ), and the
6.5
Fig 6.2: Theoretical batch carbonaceous,
nitrification and total oxygen utilization rates (OUR)
showing slope of the rates is dependent only on
the endogenous respiration rate b H .
intercept the initial active organism concentration
(Eq 6.8).
3.2 Oxygen utilization rate (OUR)
The carbonaceous (endogenous respiration) OUR
at time t (O ct, mgO/( R.h)) in the batch aerobic
digester is given by Eq 6.19 where X BH is the active
organism concentration at time t, viz. X BHt.
Substituting Eq 6.3 for X BHt in Eq 6.19 yields,
mgO/( R.d) (6.9)
Equation 6.9 does not take account of nitrification
of the ammonia that is released during
endogenous respiration. Nitrification during
aerobic digestion is very likely, in particular if the
waste sludge under digestion is from a nitrifying
activated sludge system. If complete nitrification
takes place, then Eq 6.9 must be modified to
include the nitrification oxygen utilization rate (O nt,
mgO/( R.h)).
The rate of loss of active organism VSS is given by
Eq 6.17. Of this concentration loss, a fraction (1-f)
is biodegradable. These biodegradable organics
are broken down catabolically and all the e - of the
organics are passed to oxygen (see Section 4.2
above). The N associated with these organics are
therefore returned to the bulk liquid as ammonia,
and if nitrification takes place, nitrified to nitrate.
Because the N content of the active organisms is
fn mgN/mgVSS, the OUR for nitrification (O nt,
mgO/( R.h)) is given by
 Fig 6.4: Measured endogenous respiration rate
versus sludge age (from Marais and Ekama, 1976)

Fig 6.3: Batch test oxygen utilization rate on

concentrated sludge from a 2.5 sludge age system
at 14 o C (from Marais and Ekama, 1976).

(6.10)

where
4.57 = mgO required per mgNH 4 + -N nitrified
to NO 3 -.

The total OUR (O tt, mgO/( R.h)) in the completely

nitrifying aerobic digester is given by the sum of
Eqs 6.9 and 6.10, viz.

(6.11)
The inter-relationships between the carbonaceous,
nitrification and total OURs in Eqs 6.9, 6.10 and
6.11 are shown in Fig 6.2 for b H = 0.24/d at 20 o C,
f = 0.20, f cv = 1.48 mgCOD/mgVSS and fn = 0.10
mgN/mgVSS. Plotting the OURs versus time in a
semi log plot, it can be seen that the OUR lines are
straight and parallel with a slope of b H , indicating
that provided nitrification is complete, the b H rate
can be determined from the slope of the
carbonaceous or total OUR results from an aerobic
digestion batch test. The intercept is a function of
the initial active VSS concentration (X BHi) and
therefore this initial concentration can be
determined from the intercept value. It can be seen
that nitrification does affect the determination of
X BHi and therefore needs to be taken into account
if the OURs include complete nitrification. In order
to ensure complete nitrification, care must be
taken that during the batch experiment, the pH is
maintained between 7 and 7.5. If the pH falls
below 7, which can readily happen due to loss of
alkalinity by nitrification, nitrification may be partial
Fig 6.5: Endogenous respiration rate versus
temperature showing the rate increases by ~3%
per oC (from Marais and Ekama, 1976).
or stop completely (see Section 15.2.3 below).
An example of experimental OURs measured in a
10 day batch aerobic digestion test on activated
sludge from a 2.5d sludge age system at 14 o C is
shown in Fig 6.3. The procedure proved to be very
reliable, accurate and convenient for determining
the b H rate. Marais and Ekama (1976) used this
method to measure b H at different sludge ages and
temperatures and concluded that b H was constant
with sludge age (see Fig 6.4 and Section 3.2.1.2
above) but increased slightly with temperature (Fig
6.5).
3.3 VSS removed and ammonia and ortho-P
concentrations
The VSS concentration removed, )X v by digestion

6.6
is given by the difference between the initial VSS
(Eq 6.1) and time t VSS (Eq 6.5) concentrations,
i.e.
which when simplified is
mgVSS/R (6.12)
Because the cumulative carbonaceous oxygen
utilized is equal to the COD of the VSS removed to
maintain a COD balance, Eq 6.12 can also be
obtained by integrating Eq 6.9 for O ct from time 0
to time t and dividing by the COD/VSS ratio fcv .
Fig 6.6: Fraction of VSS removed and bulk liquid
ammonia and ortho-P concentrations versus batch
digestion time for the same test conditions as Fig
6.2.
If the fraction of VSS removed, denoted fsr, is
defined as
mgVSS/R (6.13)
then, from Eq 6.12 and 6.13, the fraction of VSS
removed at time t, fsrt is given by
(6.14)
where
fai = influent active fraction = X BHi/X vi
The concentration of OrgN and OrgP released to
the bulk liquid as ammonia and ortho-P is equal to
the N and P content of the VSS concentration
removed. Accepting the N and P content of the
VSS to be fn and fp respectively,
6.7
where
fn = 0.10 mgN/mgVSS and
fp = 0.025 mgP/mgVSS,
then the ammonia and ortho-P concentrations at
time t in the batch test, N at and P st are given by,
mgN/R (6.15)
mgP/R (6.16)
A plot of the fraction of VSS removed and the
ammonia (N at) and ortho-P (P st) concentrations in
the bulk liquid versus time is shown in Fig 6.6 for
the same batch test conditions as in Fig 6.2. From
Fig 6.6 it can be seen that high concentrations of
ammonia (N at) and ortho-P (P st), usually greater
than 100 mgN/R and 25 mgP/R, can exist in the
aerobic digester for initial VSS concentrations
around 5000 mgVSS/R, i.e. after 10 days batch
digestion, N at = 174.6 mg/R and P st = 43.6 mgP/R; at
higher initial VSS concentrations, N at and P st will be
proportionally higher. If complete nitrification of
ammonia takes place, then the ammonia
concentration will be zero and the nitrate
concentration (N nt) equal to the original ammonia
concentration, i.e. with complete nitrification,
mgN/R (6.17)
mgN/R (6.18)
Considering that 3.57 mg/R as CaCO 3 H 2 CO 3 *
alkalinity per mgN/R is generated in the
ammonification process, which transforms free
ammonia (NH 3 ) to saline ammonia (NH 4+ ) and 7.14
mg/R as CaCO 3 H 2 CO 3 * alkalinity per mgN/R is
consumed in the nitrification process, which
transforms the saline ammonia to nitrate, then a
net 3.57 mg/R as CaCO 3 H 2CO 3 * alkalinity per
mgN/R is consumed in the aerobic digester per
mgN/R nitrate produced. For the Figs 6.25 and
6.29 batch test conditions, the alkalinity loss at 10
days digestion is 3.57x174.6 = 623 mg/R as CaCO 3
H 2 CO 3 *. This is much greater than bulk liquid
H 2 CO 3* alkalinity normally encountered in WWs
(200 - 400 mg/R as CaCO 3 ) and hence the pH will
almost certainly decline below 7.0 with the result
that nitrification will be at best only partial. To
maintain a pH above 7, lime will need to be dosed
to the digester. Because the pH begins to decline
below 7 at H 2 CO 3 * alkalinity values around 50 mg/R
as CaCO 3 (see Chapter 7, Section ??), the lime
dose required is
Lime dose = H 2 CO 3 * Alk loss - (Influent Alk - 50)
mg/R as CaCO 3 (6.19)
Hence, the lime dose for an influent H 2 CO 3*
alkalinity of 300 mg/R as CaCO 3 and 10 days
digestion (Fig 6.6) is 623 - (300-50) = 373 mg/R as
CaCO 3 . From the above, lime dosing to maintain
pH > 7 is required even at the low initial VSS
concentration of 4800 mgVSS/R of this example
batch test (i.e. XBHi = 2400 mgVSS/R and fai =
0.50). In practice, waste activated sludges would
be thickened to 4 to 6% to effect a 10 fold
reduction in digester volume, making the nitrate
and OP concentrations in the digester and effluent
an order of magnitude higher and hence the lime
dose concentration also. However, lime dosing
can be substantially reduced, if not avoided, if
anoxic-aerobic digestion is implemented to
denitrify the nitrate generated leading to a zero net
alkalinity change. Anoxic aerobic digestion is
discussed in Chapter 8, Section ??.
Fig 6.7: Diagrammatic representation of the
single reactor completely mixed flow though
aerobic digester showing symbols for the
different variables.
6.8
4 COMPLETELY MIXED AEROBIC DIGESTERS
In practice, aerobic digesters are usually of the
flow through type consisting of one or more
digesters in series.
4.1 Single complete m ixed digester.
4.1.1 VSS removal
A diagrammatic representation of the flow through
completely mixed aerobic digester is given in Fig
6.7. For the flow through system, the sludge age
(R s ) and hydraulic retention time (R h ) are equal
(see Section 5.4 above).
Applying the general mass balance equation (Eq
6.23) to the active organism, endogenous residue
and unbiodegradable VSS compounds over the
flow through completely mixed system, gives the
kinetic equations that describe this system, viz.
For the active organisms (X BH );
Dividing through by the volume of the digester, V d ,
and the mass balance time interval, dt, setting the
transient dX BH /dt to 0 and noting that (i) Q e = Q i,
(ii) R h = V d /Q i and (iii) X BH (in the digester) = X BHe
(in the effluent) from the definition of completely
mixed conditions, yields,
mgVSS/R (6.20)
By applying the same approach, yields for the
effluent endogenous residue concentration (X Ee),
mgVSS/R (6.21)
and the unbiodegradable VSS concentration (X Ie)
mgVSS/R (6.22)
The effluent VSS concentration (X ve ) is given by
the sum of the active, endogenous and
unbiodegradable VSS concentrations, viz.
mgVSS/R (6.23)
The form of Eqs 6.20 to 6.23 for the different VSS
concentrations is not very useful for design. A
more useful form is to express the equations in
terms of digester influent and effluent active
fractions of the sludge. Defining fai and fae as the
active fraction of the digester influent and effluent
sludge with respect to the VSS respectively,
 where
and (6.28b)

(6.28c)

Noting that the endogenous respiration rate, b H ,

varies with temperature in accordance with Eq
6.20, the required single completely mixed digester
retention time can be calculated from an influent
active fraction (fai) and a specified effluent active
fraction (f ae), both with respect to VSS.

To calculate the mass of sludge in the digester

Fig 6.8: Design chart for single reactor aerobic
digestion giving influent active fraction (fai) versus
retention time for different effluent active fraction
(fae ) and associated fraction of VSS removed (fsr) at
14 o C.
outflow for downstream processing requires
calculation of the effluent VSS concentration, X ve ,
which in turn requires knowledge of the VSS mass
broken down to CO 2 and water (degraded). The
fraction of VSS removed, fsr, is defined as the VSS
mass degraded, Q di (X vi - X ve), as a ratio of the
influent VSS mass, Q di X vi, like Eq 6.13, viz.
(6.29)

Noting that for the flow through system Q e = Q i and

(6.24)
The influent active fraction is usually the same as
that from the activated sludge system from which
the waste sludge originates. Consequently fai can
be directly determined from Eq 6.72 for fav and
setting fai = fav .
dividing Eq 6.29 by Q i, then substituting Eq 6.23
for X ve , Eq 6.20 for X BHe , Eq 6.21 for X Ee, noting
that X Ie + X Ii = X vi - X ai and substituting this for X Ie
into Eq 6.29 and simplifying yields,
(6.30)
Noting further from Eq 6.28 that
and substituting this into Eq 6.30 yields

The digester effluent active fraction, fae, is found (6.31)

from Eq 6.23 viz.
Hence, from Eq 6.29, the effluent VSS
concentration, X ve , is given by,

(6.25) mgVSS/R (6.32)

From Eq 6.20 and 6.21, From Eqs 6.28 and 6.31, a single completely
mixed aerobic digester design chart is constructed
(6.26) in Fig 6.8. In Fig 6.8, the intersection point of the
horizontal line representing the influent active
and from Eqs 6.24 and 6.20, fraction, fai (vertical axis) and the curved line
representing the selected effluent active fraction,
(6.27) fae, gives the required digester retention time R h .
The fraction of VSS degraded is also given at the
Substituting Eqs 6.26 and 6.27 into Eq 6.25 and same intersection point, reading from the curved
solving for the retention time yields, lines representing the f sr.

days (6.28a) 4.1.2 Oxygen demand

The carbonaceous (endogenous respiration) OUR

6.9

Fig 6.9: Carbonaceous, nitrification and total
oxygen demand in kgO/d per kg VSS load/d
versus retention time.
(O c , mgO/( R.h)) in the completely mixed aerobic
digester is given by Eq 6.19 where X BH is the active
organism concentration in the digester or effluent
flow, viz. X BH = X BHe . Substituting Eq 6.20 for X BH
in Eq 6.19 yields,
mgO/( R.h)
(6.33)
Multiplying both sides of Eq 6.33 by the volume of
the digester, V d , and noting that V d O c = FO c, i.e.
the flux carbonaceous oxygen demand (kgO/d)
and V d X BHi = Q i X BHi R h = FX BHi R h , where FX BHi is
the flux of active VSS load on the digester
(kgVSS/d), yields
(6.34)
Noting that FX BHi = fai FX vi and dividing Eq 6.34 by
FX vi yields the mass oxygen required per day per
kgVSS sludge load on the digester per day,
kgO/d per kgVSS/d (6.35)
From Eqs 6.28 and 6.31,
Hence,
kgO/kgVSS (6.36)
6.10
which is correct from a COD mass balance point of
view, in that the COD of the VSS degraded is
equal to the carbonaceous oxygen demand.
Equations 6.33 to 6.36 do not take account of
nitrification of the ammonia that is released during
the endogenous respiration (see Section 15.2.3
above). Nitrification during aerobic digestion is
very likely, in particular if the waste sludge under
digestion is from a nitrifying activated sludge
system. If complete nitrification takes place, then
Eq 6.33 to 6.36 must be modified to include the
nitrification oxygen demand (O n ), and the total
oxygen demand (O t) is the sum of the
carbonaceous (O c ) and nitrification (O n ) oxygen
demands, viz.
mgO/(R.h)
(6.37)
mgO/( R.h) (6.38)
kgO/d per kgVSS/d (6.39a)
kgO/d per kgVSS/d (6.40a)
kgO/kgVSS (6.39b)
kgO/kgVSS (6.40b)
A plot of the carbonaceous, nitrification and total
oxygen demand (kgO/d per kgVSS load on the
digester) versus retention time is shown in Fig 6.9.
Figure 6.9 shows that as retention time increases
so oxygen demand increases. The increase in
oxygen demand is very rapid over retention times
of 1 to 12 days after which the rate of increase
declines. From Eq 6.36, the carbonaceous,
nitrification and total oxygen demand is directly
proportional to the fraction of VSS removed;
because the oxygen demand does not significantly
increase at retention times longer than 15 days,
the fraction of VSS removed does not significantly
increase at retention times longer than 15 days
(Fig 6.8). This decline in digester efficiency per
unit volume at retention times longer than 15 days
is a consequence of the first order kinetic process
of endogenous respiration (or death-regeneration)
in single completely mixed reactors and can be
improved by changing the mixing regime from
completely mixed conditions towards plug flow
conditions with multiple reactors in series, which is
considered below.
4.1.3 Ammonia and ortho-P concentrations
As above for batch digestion, the concentrations of 5 IN -S E R IE S COMPLETELY MIXED
ammonia and ortho-P generated in the digester DIGESTERS
are proportional to the VSS concentration removed With N reactors in series, each with equal retention
( )X v ). Because in completely mixed digesters, the time, i.e. R hd/N, where R hd is the digester retention
reactor and effluent concentrations are equal, the time, the active, endogenous and unbiodegradable
effluent ammonia (N ae ) and ortho-P (P se ) particulate and VSS concentrations in the n th
concentrations are given by, reactor if the series, i.e. X BHn, X En , X In and X vn
respectively, are
mgN/R (6.41)
mgP/R (6.42) mgVSS/R (6.45)

From Eqs 6.41 and 6.42, it can be seen that the mgVSS/R (6.46)
higher the digester feed concentration (X vi) and
fraction of VSS removed (fsr), the higher the
effluent ammonia and ortho-P concentrations. If
complete nitrification takes place, the effluent
ammonia concentration will be zero and the mgVSS/R (6.47)
effluent nitrate concentration (N ne) will be equal to
the original ammonia concentration, as in Eqs 6.17 mgVSS/R (6.48)
and 6.18,
From Eq 6.19, the carbonaceous, nitrification (if
mgN/R (6.43) complete) and total OUR in the n th reactor of the
mgN/R (6.44) series are

If waste activated sludge were thickened to 4% mgO/( R.d) (6.49)

(40000 mgVSS/R) with dissolved air flotation
(Bratby and Marais, 1976) and fsr is say 0.35 (from mgO/R.d) (6.50)
Fig 6.8 for fai = 0.50, fae = 0.10 and R h = 28d), then
the effluent nitrate concentration with complete mgO/( R.d) (6.51)
nitrification is 0.10x40000x0.35 = 1400 mgN/R,
which would require a lime dose of 3.57x1400- Defining f sr (1 to n) as the fraction of VSS removed up
(300-50) = 4750 mg/R as CaCO 3 to keep the pH to and including the n th compartment of the series,
above 7. i.e.

then, from Eqs 6.45 to 5.157 above,

(6.52)

Subtracting the fraction of VSS removed up to and

including the (n-1) th compartment from the fraction
of VSS removed up to and including the n th
compartment, gives the fraction of VSS removed in
the n th compartment, fsr n , i.e.

(6.53)

6.11
 Fig 6.10: Design chart for double reactor aerobic
digestion giving influent active fraction (fai) versus
retention time for different effluent active fraction
(fae ) and associated fraction of VSS removed (fsr) at
14 o C.
Hence, the carbonaceous, nitrification (if complete)
and total oxygen demand in the n th compartment in
mass terms are given by
(kgO/d)/(kgVSS/d) (6.54)
(kgO/d)/(kgVSS/d) (6.55)
(kgO/d)/(kgVSS/d) (6.56)
Sim ilarly, if required, the carbonaceous,
nitrification (if complete) and total oxygen demand
up to and including the n th compartment is
(kgO/d)/(kgVSS/d)
(6.57)
(kgO/d)/(kgVSS/d)
(6.58)
(kgO/d)/(kgVSS/d) (6.59)
The ammonia and ortho-P concentrations increase
progressively through the reactor series of the
digester, increasing as N and P is released in each
6.12
Fig 6.11: Design chart for triple reactor aerobic
digestion giving influent active fraction (f ai) versus
retention time for different effluent active fraction
(fae ) and associated fraction of VSS removed (fsr) at
14 o C.
successive reactor. The ammonia and ortho-P
concentrations in the n th reactor, N an and P sn , are
therefore proportional to the fraction of VSS
removed up to and including the n th reactor, f sr 1 to n,
i.e. from Eqs 6.41, 6.42 and 6.52,
mgN/R (6.60)
mgP/R (6.61)
The effluent ammonia and ortho-P concentrations
are those in the last reactor N and are given by
Eqs 6.60 and 6.61 above when n equals the
number of reactors in series N.
The above equations for the in-series reactor
digester are useful if the digester retention time R hd
is known. However, in design, usually the
retention time is not known, and the design
objective is to calculate it to achieve a desired
degree of sludge stabilization, i.e. a desired
effluent active fraction (fae). In the same way that
Eq 6.28 relates the retention time of a single
completely mixed digester to the influent and
effluent active fractions (in terms of $ and "
respectively), an equation relating the total
hydraulic retention time R hd in N completely mixed
digesters in series to the digester influent and
effluent active fractions (also in terms of $ and "
respectively) can be derived, viz.
 days
At the limit as N 6 4, the term N(X 1/N -1) tends to
ln(X) and a plug flow or batch condition in time is
obtained, i.e.
days (6.63)
where
t= plug flow or batch reactor retention time
required to achieve a desired fae ( ") from a
known fai ( $).
Knowing R hd from the influent active fraction (fai
and hence $) and the specified effluent active
fraction (fae and hence "), the retention time of
each equal sized reactor is R hd/N, and the above
equations can be used to determine the fraction of
VSS removed, oxygen demand and ammonia,
nitrate and ortho-P concentrations. Figures 6.33
and 6.34 show the inter-relationship between the
influent active fraction (f ai) and retention time (R hd)
for different effluent active fractions (f ae) or fraction
of VSS removed (fsr) for double and triple reactor
digesters respectively at 14 oC, like Fig 6.8 for the
single reactor digester.
Fig 6.12: % reduction of active organism VSS in
single, double, triple and plugflow reactor aerobic
digesters at 20 oC (b H = 0.24/d, f=0.20).
(6.62)
Fig 6.13: Total oxygen demand in kgO/d per
kgVSS/d load on the digester for single, double,
triple and plugflow reactor digesters versus
digester retention time at 14 o C for the settled
wastewater activated sludge system waste
sludge with active fraction with respect to VSS
(fav) of 0.662.
6 COMPARISON OF SINGLE AND IN-SERIES
DIGESTERS
The % active mass remaining in single and multi
reactor digesters is compared in Fig 6.12. The
lines for the single, double and triple reactor
digesters were obtained from Eq 6.45 with n = 1, 2
and 3 respectively and for the plugflow digester
from Eq 6.63, all with X BHi = 100 mgVSS/R.
If, for example, a straight line (AB) is drawn from
the origin through the single reactor line at R hd =
7.5d (Point D), it intersects the double reactor line
at 15d (Point H). The first intersection (Point D)
gives the % active mass remaining after 7.5d
digestion in a single reactor, i.e. 35.7%, and the
second intersection (Point H) after 15d digestion in
a double reactor i.e. 12.6%. The % active mass
remaining after 15d digestion in a single digester
is given by Point F, i.e. 21.7%. Hence in a double
reactor digester (2x7.5d), the % active mass
remaining is 60% of that in a single reactor
digester at the same retention time (15d).
Similarly, in a triple reactor digester (3x5d), the %
active mass remaining (9.4%, Point K), is 40% of
that in the single reactor digester at the same
retention time (15d). To achieve 12.6% and 9.4%
active mass remaining in a single reactor digester
requires retention times of 28.6d and 40.1d
6.13
Figs 6.14 (left) and 6.15 (right): Percentage of the digester total oxygen demand in the 1 st and 2 nd
reactors of a double reactor digester and in the 1 st, 2 nd and 3 rd reactors of a triple reactor digester at
14 o C for the settled wastewater activated sludge system waste sludge with active fraction with respect
to VSS (fav) of 0.662.

respectively, which would have volumes 1.9 and
2.7 times larger than the double and triple reactor
digesters. A plugflow digester could achieve
12.6% and 9.4% active mass remaining in 8.6d
and 9.8d retention times respectively.
A comparison of the total oxygen demand in the
digester (in kgO/d per kgVSS load/d) in single,
double, triple and plugflow reactor digesters is
given in Fig 6.13. It can be seen that the digester
total oxygen demand increases in single, double,
triple and plugflow digesters at the same retention
time. This is because the fraction of VSS removed
(fsr) increases with increasing reactors in the
digester at the same retention time. Because the
total digester oxygen demand is proportional to the
fraction of VSS removed (f sr) , it can be see that
increasing the number of reactors in the digester
above 3 does not add much further advantage.
The subdivision of the total digester oxygen
demand versus retention time in double and triple
reactor digesters is shown in Figs 6.14 and 6.15.
These figures show that as the retention time
increases, leading to lower effluent active fractions
(fae ) and greater fractions of VSS removal (fsr) (Figs
6.14 and 6.15), the greater the proportion of the
digester total oxygen demand in the 1 st reactor and
the lower the proportion in the last reactor.
For the same influent (fai) and effluent (fae) active
fractions, the fraction of VSS removed (fsr) and
hence the total digester oxygen demand (FO t1 to N )
will be the same in single, double or triple reactor
digesters. The only difference is that this fsr will be
achieved in shorter digester retention times for
single, double or triple reactor digesters. For the
example settled wastewater activated sludge
system at 8 days sludge age and 14 o C (see Table
5.5), the active fraction (fav=fai) is 0.662. If the
waste sludge is digested to the same active
fraction as the 30 days sludge age raw wastewater
system of 0.235 (fav=fae), then at 14 o C, the digester
retention times are 19.1, 11.9 and 10.3 days
respectively for single, double and triple reactor
digesters. The fraction of VSS removed (fsr) will be
0.421 and the carbonaceous, nitrification (if
complete) and total oxygen demands from the
1471 kgVSS load/d, are 916, 283 and 1198 kgO/d
respectively and are the same in single, double
and triple reactor digesters (Table 6.1). From Fig
6.14, for the 11.9d double reactor digester, 69%
and 31% of the digester total oxygen demand is
required in the 1 st and 2 nd reactors respectively (i.e.
825 and 374 kgO/d), which, having the same
volume, means that the OUR in the 1 st reactor is
about twice that in the 2 nd. For the triple reactor
digester (Fig 6.15) at 10.3 days retention time, the
digester total oxygen demand is split 51%, 30%
and 19% making the total oxygen demand in each
reactor 618, 365 and 215 kgO/d. Compared with

6.14
the 1198 kgO/d in the single reactor digester, the
825 and 618 kgO/d in the 1 st reactor of the 11.9d
double and 10.3d triple reactor digesters leads to
OURs 2.21 (0.69.R hd single /(R hd double /2) and 2.84
(0.51 R hd single /(R hd triple /3) times higher than in the
single reactor digester. Because the digester feed
concentration (X vi) is limited by the maximum OUR
of the digester aeration system, the single, double
and triple reactor digesters cannot be feed the
same influent VSS concentration because this will
result in OURs that are too high in the 1 st reactor of
the multi-reactor digesters. The oxygen transfer
rate of the aeration system therefore largely
eliminates the volume reduction benefit obtained
by compartmentalising the digester. This aspect is
explored further below in Section 8.

6.15
7 AEROEROBIC DIGESTION MODEL BASED
ON TSS
Steady state and dynamic models of the activated
sludge (AS) system, e.g. WRC (1984) and
Activated Sludge Model No 1 (ASM1, Henze et al.,
1987), do not include the reactor inorganic
suspended solids (ISS) concentration directly. The
ISS is calculated from an estimated VSS/TSS ratio
(fi) for the AS. If the ISS concentration calculated
from such an estimated AS VSS/TSS ratio is
assumed to all have originated from the influent
wastewater, significant errors will be made on the
mass balance of this material around the WWTP.
To extend ASM1 to include the reactor ISS
concentration, inter alia Lesouef et al. (1992)
included entrapment of influent ISS into the AS
mixed liquor and the accumulation of ISS by active
biomass, and Gujer (1993) and Gujer et al. (1999)
additionally assigned an ISS content to non-active
sludge mass fractions such as unbiodegradable
particulate organics and endogenous residue.
Following the approach of Lesouef et al. (1992),
Ekama and Wentzel (2004) developed a model for
the ISS concentration in AS systems, which can be
readily incorporated into steady state and dynamic
simulation AS models. Based on the ISS model of
Ekama and Wentzel (2004), Ekama et al. (2006a)
developed a TSS based aerobic digestion model
and validated with the experimental data of van
Haandel et al. (1998a,b). This TSS based aerobic
digestion model is described in this Section.
7.1 The ISS model of Ekama and Wentzel
(2004)
This model is based on the accumulation and
conservation of influent ISS in the reactor (i.e.
negligible dissolution and/or precipitation of ISS in
the reactor) and an OHO ISS content (fiOHO ) of 0.15
mg ISS/mgOHOVSS. This OHO ISS (as distinct
from the influent wastewater ISS) is not “real” ISS;
it arises from intra-cellular dissolved inorganics
which precipitate in the VSS-TSS test procedure.
The ISS model was validated with data from 21
investigations conducted over the past 15 years on
30 aerobic and anoxic-aerobic nitrification
denitrification (ND) systems variously fed artificial
and real wastewater and operated from 3 to 20
days sludge age. The model predicted reactor
VSS/TSS ratio reflects the observed relative
insensitivity to sludge age. This model can be
easily integrated into steady state and dynamic
simulation AS models, and hence also into the
aerobic digestion models because the latter is a
subset of the former for the special case of zero
6.16
influent wastewater organics. The ISS model was
also validated for nitrification denitrification
biological excess P removal (NDBEPR) systems
with phosphorus accumulating organisms (PAOs),
but this aspect of the ISS model is not relevant to
aerobic digestion of WAS from aerobic and ND
systems considered in this section.
To use the ISS model, measurement of the influent
wastewater ISS concentration is required. Ekama
and Wentzel (2004) found that the conventional
test procedure for this overestimated the influent
ISS by 12 to 25% and developed a revised ISS
test method which yielded more reliable results.
The over estimation of the conventional method
could be the reason why some authors find
discrepancies between the measured influent
(fixed) ISS and the ISS mass recovered in the
reactor. For example, (1) Lesouef et al. (1992)
had to include a fixed ISS dissolution rate in the
reactor to predict the reactor ISS concentration
correctly when including an OHO ISS content
(fiOHO ) of 0.053 - clearly too much influent ISS was
entering the reactor compared with the ISS mass
measured in the reactor; (2) Gujer et al. (1999) in
ASM3 assigned ISS contents to the active OHO,
endogenous, influent particulate unbiodegradable
and biodegradable organics of 0.075, 0.225, 0.225
and 0.225 mgISS/mgVSS respectively - these lead
to significantly higher ISS concentrations in the
reactor than would be predicted by the ISS model
of Ekama and Wentzel (2004) and (3) Wentzel et
al. (2002) assigned ISS contents to the active
OHO and endogenous organics of 0.17
mgISS/mgVSS and included significant dissolution
of influent ISS in order to account for the
measured mass of ISS in the reactor from the
measured mass of influent ISS determined by the
conventional method. If the ISS model can predict
the changes in ISS concentration through a series
of aerobic digesters, then it not only provides
additional validation of the aerobic digester model
but also a framework for tracking the ISS through
WWTPs comprising ND activated sludge systems
and aerobic digestion of the WAS.
7.2 The experimental data
van Haandel et al. (1998a,b) operated a pilot scale
WWTP scheme (Fig 6.16) at 25 /C in which 500 R/d
raw municipal wastewater was fed to a 2 day
retention time aerated lagoon (R0). All the daily
waste activated sludge (WAS) from the aerated
lagoon was thickened into 30 R which served as
feed to a series of four aerobic digesters (R1 to
R4) at retention times of 1.73d, 2.14d, 3.00d and
Fig 6.16: Experimental system layout of van Haandel et al. (1998a,b). R1 to R4 are the aerobic
digesters.
5.63d respectively. From the feed to each aerobic
digester, 4, 5, 6, 7 and 8 R/d of sludge volume was
withdrawn, thickened to a volume of 0.40 R/d and
fed to five anaerobic digesters (AD0 to AD4) each
at 20d retention time. Each AD was therefore fed
a WAS with a different fraction of unbiodegradable
organics depending on the extent of aerobic
digestion before anaerobic digestion. Only the
experimental results from the aerobic digesters are
required for the aerobic digester model validation.
In this respect the TSS, VSS and ISS
concentrations and the OUR measured at the five
sampling points (Fig 6.16) are relevant and are
listed in Table 6.1.
In addition, van Haandel et al. (1998a) operated an
AS system fed the same raw wastewater at 13
different sludge ages between 3 and 10d and
temperatures between 21 and 30 /C. On sludge
harvested from this system, they conducted six
batch aerobic digestion tests at each sludge age
and temperature. From the measured oxygen
utilization rate (OUR) and volatile suspended
solids (VSS), nitrate and alkalinity concentrations
with time, they determined the endogenous
respiration rate of the OHOs to be b HT = 0.24
(1.040) (T-20) /d for the OHO yield coefficient Y H =
0.45 mgVSS/mgCOD, sludge COD/VSS ratio (f cv)
= 1.5 mgCOD/mgVSS and unbiodegradable
endogenous residue fraction (fEH ) = 0.20. This is
very close to the b HT rate determined by Marais
and Ekama (1976) of 0.24 (1.029) (T-20) between 8
and 20 /C for the same Y H and fEH and f cv = 1.48
(Fig 6.5). The b HT value of the W AS from the
6.17
aerated lagoon (R0) and in the subsequent aerobic
digesters (R1 to R4) therefore could be determined
with confidence to be 0.24(1.04) (25-20) = 0.292 /d at
25 o C.
To readers not familiar with the steady state
activated sludge model of Marais and Ekama
(1976) (or WRC, 1984 and Henze et al., 2008), the
b H20 = 0.24 /d and f EH = 0.20 may seem incorrect
compared with the ASM1 values of b’H20 = 0.62
and f’EH = 0.08. Dold et al. (1980) and Warner et
al. (1986) showed that the b H20 = 0.24 /d and fEH =
0.20 of the endogenous respiration model included
in the steady state AS model yield identical results
under aerobic conditions as the b’H20 = 0.62 /d and
f’EH = 0.08 of the death-regeneration model
included in the kinetic simulation models like
ASM1 or UCTOLD (Dold et al., 1991). The
endogenous respiration models the net effect of
the death-regeneration. The former is used in
steady state AS and aerobic digestion models
because it leads to much simpler equations with
negligible loss of accuracy. If it is accepted that
the death-regeneration approach is a more
realistic model of what happens in AS, then it
follows that the fEH = 0.20 of the endogenous
respiration model is not a good estimate of the
“real” unbiodegradable fraction of the OHO
biomass. The f’EH = 0.08 of the death-regeneration
model would be a much closer estimate of the
“real” unbiodegradable fraction of the OHO
biomass. This is important when calculating the
unbiodegradable fraction of WAS for anaerobic
digestion (Ekama et al., 2006b) but not for aerobic
digestion.
6.3 The TSS based steady state aerobic
digester model
The equations for the steady state single reactor
completely mixed aerobic digester model in terms
of TSS were derived from the assumptions
outlined in Section 6.2 above. In the interests of
brevity, their derivation is not given, but the
equations of the model in terms of both the VSS
and the TSS concentration measures are listed in
Table 6.2.
7.6.4 Aerobic digestion model validation
From the VSS based model, it can be shown that
with b HT , Y H , fEH and fcv known, the active OHO
fraction of the VSS (fav) is given by
(6.82)
where
OUR t = total endogenous oxygen utilization rate
including complete nitrification of VSS
released ammonia - mgO/( R.d) - i.e. at
zero residual influent biodegradable COD
and ammonia concentrations.
fn = TKN/VSS ratio of activated sludge = 0.10
mgN/mgVSS = 67.6 mgN/gCOD
Xv = vola tile su sp en de d s olids (V S S)
concentration.
Equation 6.82 applies provided nitrification is
complete (i.e. low effluent free and saline
ammonia, FSA concentration), which was the case
for the activated sludges from the aerated lagoon
and the four in series aerobic digesters. Thus with
known values for f cv , fn , fEH , and b HT , i.e.
1.5mgCOD/mgVSS, 0.10 mgN/mgVSS, 0.20 and
0.292/d at 25 o C, the active fraction of the VSS (fav )
can be calculated with Eq 6.82 from measured
values of OUR t and X v at the 5 sampling points (P)
in Fig 6.16. From the measured values of OUR t
and X v , the “measured” active fraction (fav ) of the
WAS in the outflow of the aerated lagoon (R0) and
the four in-series aerobic digesters (R1 to R4)
were calculated and compared with the model
predicted fav ratios (Fig 6.17).
Similarly, from the TSS based model, it can be
shown that with b HT , Y H , fEH and fcv known, the
active OHO fraction of the TSS (fat) is given by
6.18
(6.83)
where
Xt = total suspended solids (VSS)
concentration.
So from measured values of OUR t and X t at the 5
sampling points (P), the “measured” active fraction
of the TSS (fat) of the WAS in the outflow of the
aerated lagoon (R0) and the four in-series aerobic
digesters (R1 to R4) were calculated and
compared with the model predicted fat ratios (Fig
6.18).
The model predicted active fractions with respect
to VSS (fav) and TSS (fat), the OUR t and the VSS,
TSS and ISS concentrations through the aerated
lagoon and aerobic digester series were calculated
as follows. From the steady state activated sludge
model of Marais and Ekama (1976, or WRC, 1984,
Henze et al., 2008, Section 4.4), extended to
include the ISS concentration, the mass of VSS,
ISS and TSS in the reactor is related to the organic
(COD) and inorganic (ISS) loads on the reactor,
the OHO kinetic and stoichiometric constants (b HT ,
Y H , f c v , f E H and f iO H O ), the wastew ater
unbiodegradable soluble and particulate COD
fractions (f S’us, fS’up) and the sludge age (R s). With
the ISS content of the OHOs (f iOHO ) at 0.15
mgISS/mgOHOVSS as determined by Ekama and
Wentzel (2004) and assuming fS’us = 0.07, the
calculated raw wastewater characteristics fed to
the 1000 R volume 2d retention time aerated
lagoon (R0) of van Haandel et al. (1998a) at 500
R/d and 25 /C to match the measured data at P1
(Fig 6.16, i.e. VSS/TSS ratio, fi = 0.71 and OHO
active fraction, fav = 0.76) are (i) total influent COD
concentration (S ti) = 563 mgCOD/R, fS’up = 0.072
and the influent inorganic suspended solids
concentration (X Ioi) = 52.9 mg ISS/R. These raw
wastewater characteristics were used for the
theoretical calculations through the aerated lagoon
(R0) and aerobic digester sequence (R1 to R4)
w ith the theoretically calculated effluent
concentrations of the upstream aerobic digester
becoming the influent concentrations to the
downstream one.
For example, the effluent active fraction of the VSS
from the aerated lagoon (R0, fave,R0 ) is 0.76 and is
the influent active fraction to the first aerobic
digester (R1, f avi,R1 ). From Eq 6.30c, with fEH =
0.20, $ is 0.516. With b H25= 0.292/d and the
retention time (R h ) in digester R1 = 1.73d, " is
0.777 (Eq 6.66). Hence, from Eq 6.30b, the
effluent active fraction from R1 (fave,R1 ) is 0.634.
Similarly, the effluent active fraction of the TSS
from the aerated lagoon (R0, fate,R0 ) is 0.54. From
Eq 6.67, with f iOHO = 0.15, * is 0.702. With b H25=
0.292/d and the retention time (R h ) in digester R1
= 1.73d, ( is 1.158 (Eq 6.70). Hence from Eq 6.68,
the effluent active fraction from R1 (fate,R1 ) is 0.433.
The VSS/TSS ratio (fii) of the sludge from the
aerated lagoon (R0) is 0.711 (Eq 6.78) and with fave
= 0.634 and f ate = 0.433, the effluent VSS/TSS ratio
(fie ) from digester R1 is 0.683 (Eq 6.78). W ith the
$ and " values known from the influent and
effluent active fractions, the fraction of VSS
removed (f vsr) in the 1 st digester (R1) = 0.204 (Eq
6.23). Then from Eqs 6.38b, 6.41b and 6.42b, the
organic, nitrification and total oxygen utilization
rate (OUR) can be calculated from the known load
of VSS on the digester, i.e. 22.18, 6.76 and 28.94
mgO/( R.h) respectively. This calculation is then
followed through the remaining series of three
digesters (R2 to R4) with the effluent active
fractions of R1 (fave,R1 ) becoming the influent active
fractions of R2 (f avi,R2 ) and so on.
The calculated active fractions with respect to the
VSS (fav) and TSS (fat), total OUR, VSS
concentration and VSS/TSS ratio of the sludge in
the outflow from R0 to R4 at sampling points P1 to
P5 are compared with the measured values in Figs
6.17 to 6.21. It can be seen that the
correspondences are very good (correlation
coefficients R 2 >0.98). With regard to the ISS
concentration (in the 30 R sludge volume fed to the
aerobic digesters), Fig 6.22 shows the component
concentrations of the total ISS though the aerated
lagoon and four digesters in series. The “fixed” ISS
originating from the influent wastewater remains
constant at 0.882 gTSS/R as expected. The
biomass ISS concentration decreases through the
R0 to R4 series as the OHO biomass
concentration decreases which results in a
decreasing total ISS concentration through R0 to
R4. The experim entally m easured ISS
concentrations also are plotted in Fig 6.22 and it
can be seen that the theoretically predicted ISS
concentrations correspond closely with these
(R 2=0.830). The calculated and measured
VSS/TSS ratios through the R0 to R4 series also
correspond very well (R 2=0.982, Fig 6.21) as do
the active fractions with respect to VSS and TSS
(Figs 6.17 and 6.18). In Figs 6.17 to 6.22, the
close correspondence between calculated and
measured ISS concentrations is not possible
without including an OHO ISS content (fiOHO ), and
6.19
the value of 0.15 mgISS/mgOHOVSS estimated by
Ekama and Wentzel (2004) clearly also closely
applies to the van Haandel et al. (1998a) data.
The “fixed” (originating from the raw wastewater)
ISS concentration and OHO ISS content (fiOHO )
values that minimize the variance (error) between
the measured and predicted ISS concentrations at
points P1 to P5 in Fig 6.16 are 51.2 mgISS/R and
0.171 mgISS/mgOHOVSS (R 2 = 0.847). These
values are very close to the 52.9 mgISS/R and
0.150 mgISS/mgOHOVSS determined from the
Ekama and Wentzel (2004) model (R 2 = 0.830).
This not only provides additional validation for the
ISS model, but also shows that it can be used for
tracking the ISS concentration through aerobic
digestion down to very low active fractions - the fave
from R4 is 0.142, which is equivalent to an
extended aeration activated sludge system at
around 60d sludge age.
6.7.5 Simulation of aerated lagoon and aerobic
digesters series
The aerated lagoon and four in-series AerD
WWTP sequence was simulated in Aquasim
(Reichart, 1998) with ASM1, modified to include
the ISS model of Ekama and Wentzel (2004). The
theoretically predicted active fractions with respect
to the VSS (fav ) and TSS (fat), total OUR, VSS
co n ce n tra tion, VS S /TS S ra tio , an d IS S
concentration are shown plotted in Figs 6.17 to
6.22 together with the steady state model
calculated results and experimental data. It can be
seen that the ASM1 and steady state model
results match very closely and both correlate very
well with the experimental data. This shows that
the steady state aerobic digester model in terms of
VSS or TSS is sufficiently accurate to include in a
steady state plant wide WWTP model linking fully
aerobic or ND activated sludge systems to aerobic
digesters.
Table 6.1: Experimentally measured (van Haandel et al., 1998a,b) and theoretically calculated
characteristics of the activated sludge in the outflows from the 2d retention time aerated lagoon (R0) and
the four in-series aerobic digesters (R1-R4) at 1.73d, 2.14d, 3.00d, 5.63d retention time at 25 o C and fed
to anaerobic digesters AD0 to AD4 respectively (after Ekama et al., 2006).
Parameter R0 R1 R1 R2 R2 R3 R3 R4 R4
Sample Point P1 P1 P2 P2 P3 P3 P4 P4 P5 P5
Effl Infl Effl Infl Effl Infl Effl Infl Effl
Flow R/d 30 26 26 21 21 15 15 8 8
Exper Theory Exper Theory Exper Theory Exper Theory Exper Theory
TSS - X t (g/R) 4.20 4.24 3.74 3.51 2.91 2.95 2.51 2.54 2.11 2.24
VSS - X v (g/R) 3.01 3.01 2.52 2.40 1.89 1.93 1.57 1.58 1.26 1.33
VSS/TSS - fi 0.720 0.711 0.674 0.683 0.651 0.654 0.625 0.623 0.598 0.594
OUR t - mgO/(R.h) 44 43.57 29 28.94 16 17.80 8 9.49 4 3.59
fav - Eq 19 0.76 0.760 0.60 0.634 0.44 0.485 0.27 0.315 0.17 0.142
$ - Eq 1 - 0.516 - 0.777 - 1.264 - 2.371 -
" - Eq 4 - 0.777 - 1.264 - 2.371 - 6.265 -
* - Eq 9 - 0.702 - 1.158 - 2.008 - 3.941 -
( - Eq 12 - 1.158 - 2.008 - 3.941 - 10.743 -
fat - Eq 18 0.54 0.540 0.40 0.433 0.29 0.317 0.17 0.196 0.10 0.084
OHOVSS - g/R 2.29 2.29 1.51 1.52 0.83 0.93 0.42 0.50 0.21 0.19
ISSinfl - g/R - 0.88 - 0.88 - 0.88 - 0.88 - 0.88
ISSbio - g/R - 0.34 - 0.23 - 0.14 - 0.07 - 0.03
ISStot - g/R 1.19 1.23 1.02 1.11 1.02 1.02 0.94 0.96 0.85 0.91
OURc - mgO/(R.h) - 33.39 - 22.18 - 13.64 - 7.27 - 2.75
OURn - mgO/(R.h) - 10.17 - 6.76 - 4.16 - 2.22 - 0.84
OURt - mgO/(R.h) 44 43.57 29 28.94 16 17.8 8 9.49 4 3.59
VSS removed fvsr 0.204 0.163 0.195 0.250 0.181 0.169 0.157 0.197
TSS removed ftsr 0.172 0.110 0.158 0.222 0.140 0.137 0.116 0.159
Note: The effluent from R0 and the influent to R1 are the same sludge - hence the experimental
measured and theoretical calculated results are listed side by side. Likewise, the effluent from R1 and
the influent to R2 is the same sludge, and so on.

Fig 6.17: Steady state model () and
ASM1 (Ž) predicted versus measured
active fraction with respect to VSS in the
outflow sludges from the aerated lagoon
(R0) and the 4 in-series aerobic
digesters (R1 to R4).
Fig 6.18: Steady state model () and
ASM1 (Ž) predicted versus measured
active fraction with respect to TSS in the
outflow sludges from the aerated lagoon
(R0) and the 4 in-series aerobic
digesters (R1 to R4).

6.20
Tables 6.3a and b: Single, double (Table 6.3a, top) and triple (Table 6.3b, bottom) reactor aerobic
digester designs for the example settled wastewater waste activated sludge at 8 days sludge age and
14 o C; feed flow and concentration 443m 3 /d and 4000 mgTSS/R with VSS/TSS (f i), active fraction with
respect to VSS (fav ), COD/VSS ratio (f cv)and N and P content of VSS (f n and fp) of 0.83
mgVSS/mgTSS, 0.662, 1.48 mgCOD/mgVSS, 0.10 mgN/mgVSS and 0.03 mgP/mgVSS (Table 4.4
in Henze et al., 2008) with a maximum OUR limit in the first reactor of 125 mgO/( R.h) and with or
without a feed concentration thickening limit of 5%TSS.
Digester Type Single Digester Double Digester
Design ret time as.... Single Digester Double Digester Double Digester Single Digester
Apply thickening limit No Yes No Yes No Yes No Yes
Flow distribution 100 100 100/0 100/0 50/50 50/50 50/50 50/50
OUR in 1 st reactor 125 73.8 125 125 125 105 125 73.8
nd
OUR in 2 reactor - - 56.7 56.7 125 105 125 73.8
Feed Conc %TSS 8.47 5 3.84 3.84 5.94 5 8.47 5
Ret Time (d) 19.11 19.11 11.92 11.92 11.92 11.92 19.11 19.11
Effl active frac (fae) 0.235 0.235 0.235 0.235 0.31 0.31 0.235 0.235
Feed flow m 3/d 20.91 35.44 46.12 46.12 29.82 35.44 20.91 35.44
VSS removed (fsr) 0.421 0.421 0.421 0.421 0.374 0.374 0.421 0.421
Effluent TSS (X te, %) 5.514 3.254 2.501 2.501 4.096 3.446 5.514 3.254
NO 3-N in 1 st reactor 2959 1746 923 923 1846 1554 2959 1746
nd
NO 3-N in 2 reactor - - 1342 1342 1846 1554 2959 1746
st
PO 4-P in 1 reactor 888 524 277 277 554 466 888 524
PO 4-P in 2 nd reactor - - 403 403 554 466 888 524
Thickening 21.18x 12.50x 9.61x 9.61x 14.86x 12.5x 21.18x 12.50x
3
Digester volume m 400 677 550 550 356 423 400 677

Digester Type Single Digester Triple Digester

Design ret time as.... Single Digester Triple Digester Triple Digester Single Digester
Apply thick limit No Yes No Yes No Yes No Yes
Flow distribution 100 100 100/0/0 100/0/0 33/33/34 33/33/34 33/33/34 33/33/34
st
OUR in 1 reactor 125 73.8 125 125 125 116 125 73.8
nd
OUR in 2 reactor - - 73.8 73.8 125 116 125 73.8
OUR in 3 rd reactor - - 43.6 43.6 125 116 125 73.8
Feed Conc %TSS 8.47 5 2.95 2.95 5.38 5 8.47 5
Ret Time (d) 19.11 19.11 10.3 10.3 10.3 10.3 19.11 19.11
Effl active frac (fae) 0.235 0.235 0.235 0.235 0.338 0.338 0.235 0.235
Feed flow m 3/d 20.91 35.44 60.04 60.04 32.97 35.44 20.91 35.44
VSS removed (fsr) 0.421 0.421 0.421 0.421 0.358 0.358 0.421 0.421
Effluent TSS (X te, %) 5.514 3.254 1.921 1.921 3.779 3.515 5.514 3.254
NO 3-N in 1 st reactor 2959 1746 532 532 1597 1486 2959 1746
nd
NO 3-N in 2 reactor - - 846 846 1597 1486 2959 1746
rd
NO 3-N in 3 reactor - - 1031 1031 1597 1486 2959 1747
st
PO 4-P in 1 reactor 888 524 160 160 479 446 888 524
PO 4-P in 2 nd reactor - - 254 254 479 446 888 524
PO 4-P in 3 rd reactor - - 309 309 479 446 888 524
Thickening 21.18x 12.50x 7.38x 7.38x 13.44x 12.50x 21.18x 12.50x
Digester volume m 3 400 677 619 619 340 365 400 677

6.22
 Fig 6.19: Steady state model () and ASM1 (Ž)
predicted versus measured oxygen utilization
rate (OUR) in the outflow sludges from the
aerated lagoon (R0) and the 4 in-series aerobic
digesters (R1 to R4).
Fig 6.21: Steady state model () and ASM1
(Ž) predicted versus measured VSS/TSS ratio
in the outflow sludges from the aerated lagoon
(R0) and the 4 in-series aerobic digesters (R1
to R4).
6.23
Fig 6.20: Steady state model () and ASM1
(Ž) predicted versus measured VSS
concentration in the outflow sludges from the
aerated lagoon (R0) and the 4 in-series
aerobic digesters (R1 to R4).
Fig 6.22: Steady state model influent (–), OHO
biomass () and total (•), ASM model
predicted (A) and experimentally measured (Ž)
ISS concentration in the outflow sludges from
the aerated lagoon (R0) and the 4 in-series
aerobic digesters (R1 to R4).
8 THICKENING WASTE ACTIVATED SLUDGE
Once the digester retention time is known from
the degree of stabilization required (fai and fae)
and the number of reactors in series (N), the
volume of the digester (V d ) is directly proportional
to the influent flow to the digester (Q I). For the
design retention time (R hd), the lower the Q I, the
smaller the V d . Because the sludge load on the
digester is fixed by the sludge age of the
activated sludge system, the influent flow to the
digester can be reduced by waste activated
sludge thickening with dissolved air flotation (see
Section 13.3 above). Furthermore, because the
oxygen demand in the digester is fixed by the
sludge load (and active fraction) on the digester
per day, FX vi, (Eqs 6.35 to 6.40 and 6.54 to 6.59)
and the degree of stabilization required, the more
concentrated the feed sludge concentration (X ti),
the higher the oxygen utilization rate (OUR) in
mgO/( R.h) in the digester. The maximum OUR is
fixed by the oxygen transfer rate of the digester
aeration system. If the aeration system OUR
limit in the first reactor of the in-series reactor
digester, where the OUR is highest, is O t1max in
mgO/( R.h), then from Eqs 6.52 and 6.59, the
thickened sludge feed concentration X ti in %TSS
is given by
The nitrate (with complete nitrification) and
phosphate concentrations in the single or
multiple reactor digesters are found from Eqs
6.60 and 6.61 respectively, noting that the units
of X vi in these equations must be in mgVSS/R.
Equation 6.64 establishes the link between the
OUR in the 1 st reactor and the feed solids
concentration. However, WAS cannot be
thickened to more than 5 to 6% TSS. According
to Bratby and Marais (1976, 1977), the practical
upper limit for dissolved air flotation (DAF) of
activated sludge is about 5 to 6%, and to achieve
this, once the appropriate air/solids ratio has
been determined, depends almost entirely on the
way the float is removed from the flotation unit
surface. The scraper should slice off only the top
layer of the float well above the water surface
(150-200mm), the layer which has had the
longest time to drain and dewater; the scraper
should not remove the entire float above the
water level to achieve high thickened WAS
concentrations.

%TSS

(6.64)

and
(6.68)

where
X ti =digester feed TSS concentration in %
X vi =digester feed VSS concentration in %
fai =active fraction of influent VSS
O t1max = maximum OUR in first reactor in
mgO/( R.h)
n = 1 for the first reactor in the digester
N = number of in-series reactors in digester
fi = VSS/TSS ratio of waste activated sludge

Once the feed VSS or TSS concentrations (X vi or

X ti) are known from a specified O t1max from Eq
6.64, the sludge thickening required, digester
influent flow rate Q id and volume V d are given by,

(6.65)

(6.66)

(6.67)

6.24
 Figs 6.16 a to c: Digester feed concentration in
% versus digester retention time for maximum
total oxygen utilization rate in the first reactor of
100, 125 and 150 mgO/( R.h) for the waste sludge
from the example settled wastewater activated
sludge system at 8d sludge age and 14 oC with
active fraction with respect to VSS of 0.662 for
single (Fig 6.16a, top left) double (Fig 6.16b, top
right) and triple (Fig 6.16c, bottom left) reactor
aerobic digesters.

9 DESIGN EXAMPLE
The calculation results listed in Tables 6.1a and
b are for single, double and triple aerobic
digesters for the waste activated sludge (WAS)
from the 8d sludge age system treating example
settled WW at 14 o C, viz. f av = 0.662, f i = 0.83, X t
= 4000 mgTSS/R, Q W = 443 m 3/d, with fcv = 1.48
mgCOD/mgVSS, fn = 0.10 mgN/mgVSS and fp =
0.03 mgP/mgVSS. For the results in Tables 6.1a
and b, the OUR upper limit is 125 mgO/( R.h) in
the 1 st reactor, and the results are given with and
without a 5% TSS feed concentration thickening
limit. In this example, as in Figs 6.36 to 6.40, the
specified effluent active fraction (fae) is 0.235,
which is the same as that from the raw WW
activated sludge system at 30 days sludge age
0and 14 oC.
For this WAS sludge, the digester feed
concentration (X vi in %) versus digester retention
time (R hd) for O t1max values of 100,125 and 150
mgO/( R.h) are shown in Figs 6.16a to c for the
single, double and triple reactor digesters
respectively (see Table 6.1 for details). From Fig
6.16 it can be seen that the feed concentration
decreases with increasing number of reactors in
the digester. This is because the OUR in the first
reactor is the highest at O t1max and the
subsequent reactors have OURs significantly
below this (Figs 6.37 and 6.38). The feed
concentrations for the 19.1d single, 11.9d double
and 10.3d triple reactor digesters at 14 oC are
8.47% TSS (or 7.03% VSS), 3.84% TSS (or
3.19% VSS and 2.95% TSS (or 2.45% VSS) for
an O t1max of 125 mgO/( R.h) in the 1 st reactor.

6.25
Tables 6.1a and b: Single, double (Table 6.1a, top) and triple (Table 6.1b, bottom) reactor aerobic
digester designs for the example settled wastewater waste activated sludge at 8 days sludge age and
14 o C; feed flow and concentration 443m 3 /d and 4000 mgTSS/R with VSS/TSS (f i), active fraction with
respect to VSS (fav ), COD/VSS ratio (fcv )and N and P content of VSS (fn and fp ) of 0.83
mgVSS/mgTSS, 0.662, 1.48 mgCOD/mgVSS, 0.10 mgN/mgVSS and 0.03 mgP/mgVSS with a
maximum OUR limit in the first reactor of 125 mgO/( R.h) and with or without a feed concentration
thickening limit of 5%TSS.
Digester Type Single Digester Double Digester
Design ret time as.... Single Digester Double Digester Double Digester Single Digester
Apply thickening limit No Yes No Yes No Yes No Yes
Flow distribution 100 100 100/0 100/0 50/50 50/50 50/50 50/50
st
OUR in 1 reactor 125 73.8 125 125 125 105 125 73.8
OUR in 2 nd reactor - - 56.7 56.7 125 105 125 73.8
Feed Conc %TSS 8.47 5.00 3.84 3.84 5.94 5.00 8.47 5.00
Ret Time (d) 19.11 19.11 11.92 11.92 11.92 11.92 19.11 19.11
Effl active frac (fae) 0.235 0.235 0.235 0.235 0.310 0.310 0.235 0.235
Feed flow m 3/d 20.91 35.44 46.12 46.12 29.82 35.44 20.91 35.44
VSS removed (fsr) 0.421 0.421 0.421 0.421 0.374 0.374 0.421 0.421
Effluent TSS (X te, %) 5.514 3.254 2.501 2.501 4.096 3.446 5.514 3.254
NO 3-N in 1 st reactor 2959 1746 923 923 1846 1554 2959 1746
nd
NO 3-N in 2 reactor - - 1342 1342 1846 1554 2959 1746
st
PO 4-P in 1 reactor 888 524 277 277 554 466 888 524
nd
PO 4-P in 2 reactor - - 403 403 554 466 888 524
Thickening 21.18x 12.50x 9.61x 9.61x 14.86x 12.5x 21.18x 12.50x
Digester volume m 3 400 677 550 550 356 423 400 677

Digester Type Single Digester Triple Digester

Design ret time as.... Single Digester Triple Digester Triple Digester Single Digester
Apply thick limit No Yes No Yes No Yes No Yes
Flow distribution 100 100 100/0/0 100/0/0 33/33/34 33/33/34 33/33/34 33/33/34
OUR in 1 st reactor 125 73.8 125 125 125 116 125 73.8
nd
OUR in 2 reactor - - 73.8 73.8 125 116 125 73.8
rd
OUR in 3 reactor - - 43.6 43.6 125 116 125 73.8
Feed Conc %TSS 8.47 5.00 2.95 2.95 5.38 5.00 8.47 5.00
Ret Time (d) 19.11 19.11 10.30 10.30 10.30 10.30 19.11 19.11
Effl active frac (fae) 0.235 0.235 0.235 0.235 0.338 0.338 0.235 0.235
Feed flow m 3/d 20.91 35.44 60.04 60.04 32.97 35.44 20.91 35.44
VSS removed (fsr) 0.421 0.421 0.421 0.421 0.358 0.358 0.421 0.421
Effluent TSS (X te, %) 5.514 3.254 1.921 1.921 3.779 3.515 5.514 3.254
st
NO 3-N in 1 reactor 2959 1746 532 532 1597 1486 2959 1746
NO 3-N in 2 nd reactor - - 846 846 1597 1486 2959 1746
rd
NO 3-N in 3 reactor - - 1031 1031 1597 1486 2959 1747
st
PO 4-P in 1 reactor 888 524 160 160 479 446 888 524
PO 4-P in 2 nd reactor - - 254 254 479 446 888 524
PO 4-P in 3 rd reactor - - 309 309 479 446 888 524
Thickening 21.18x 12.50x 7.38x 7.38x 13.44x 12.50x 21.18x 12.50x
3
Digester volume m 400 677 619 619 340 365 400 677

6.26
 Figs 6.17 a to c: Nitrate and Ortho-P
concentration in in the 1 st (1), 2 nd (2) and 3 rd (3)
reactors for single (Fig 6.17a, top left) double (Fig
6.17b, top right) and triple (Fig 6.17c, bottom left)
reactor aerobic digesters. Versus digester
retention time for maximum total oxygen
utilization rate in the first reactor of 125 mgO/( R.h)
for the waste sludge from the example settled
wastewater activated sludge system at 8d sludge
age and 14 o C with active fraction with respect to
VSS of 0.662

Hence, the more reactors in series, the lower the
average OUR in the digester, while the average
OUR in the single reactor digester remains at
O t1max. Therefore, because the aeration system
limits the OUR to some specified maximum value
in the 1 st reactor, the benefit of digester volume
reduction with series digesters indicated by the
kinetics of endogenous respiration (Fig 6.12)
cannot be realised. The volume of the 19.1d
single, 11.9d double and 10.3d triple reactor
digesters at 14 o C with feed TSS concentrations
of 8.47%, 3.84% and 2.95% respectively are 400
m 3, 550 m 3 and 618 m 3, clearly increasing as the
number of reactors in the digester increases.
The nitrate and phosphate concentrations in
each reactor of the single, double and triple
reactor digesters versus digester retention time
are shown in Figs 6.17a to c respectively for an
O t1max of 125 mgO/( R.h) and 14 o C. From these
figures, it can be seen that in the single reactor
digester at 19.1d retention time, the nitrate
concentration is 2959 mgN/R and the ortho-P 888
mgP/R; for the 11.9d double reactor digester, the
nitrate and ortho-P concentrations in the 1 st and
2 nd reactors are 923 and 1342 mgN/R and 277
and 403 mgP/R respectively; for the 9.6d triple
reactor digester, the nitrate and ortho-P
concentrations in the 1 st, 2 nd and 3 rd reactors are
497, 799 and 981 mgN/R and 149, 240 and 294
mgP/R respectively (see Tables 6.1a and b).
For the single reactor digester the retention time
is 19.1d. For an O t1m ax of 125 mgO/( R.h), X ti is
8.47% TSS, leading to a feed flow of 20.91 m 3 /d
and hence a digester volume of 400m 3 . With the
thickening limit at 5% TSS, the feed flow is 35.4
m 3 /d giving an O t1max of 73.8 mgO/( R.h) and a
digester volume of 677 m 3. For both cases, the
fraction of VSS removed (fsr) is 0.421 and the
required f ae of 0.235 is achieved. With and
without the thickening limit, the effluent nitrate

6.27
(assuming complete nitrification) and phosphate
concentrations are 1746 and 2959 mgN/R and
524 and 888 mgP/R respectively.

For the double reactor digester, the digester can

be operated in one of three ways, i.e. (1) 100%
feed into the 1 st reactor or (2) splitting the feed
50/50 between the 1 st and 2 nd reactors in step
feed mode with the retention time set at that for
a double reactor digester, and (3) 50/50 step
feeding as in method (2) but setting the retention
time at that required for a single reactor digester.
With method 1 the OUR in the 1 st reactor is much
higher than in the 2 nd (Fig 6.14) but with step feed
methods 2 and 3 the OUR is equalized between
the 1 st and 2 nd reactors.

From Table 6.1a it can be seen that for method

1, the thickening limit does not apply, because
the feed concentration is less than 5% TSS
(3.84% TSS) to keep O t1max below 125 mgO/( R.h).
For method 1, the retention time for the double
reactor digester is 11.9d (5.96d in each reactor)
and the feed flow is 46.12 m 3 /d, giving a digester
volume of 550m 3 . Also, the required fae of 0.235
is achieved, fsr = 0.421 and the 1 st and 2 nd reactor
OUR and nitrate and phosphate concentrations
are 125 and 56.7 mgO/( R.h), and 923 and 1342
mgN/R and 277 and 403 mgP/R respectively. For
the step feed methods 2 and 3, the thickening
limit applies and when it is imposed, O t1max is
below 125 mgO/( R.h), i.e. 105 and 73.8 mgO/( R.h)
for methods 2 and 3 respectively. Note that
when 50/50 step feeding is applied to the double
reactor digester the OUR is the same in the 1 st
and 2 nd reactors. The difference between
methods 2 and 3 is essentially the nominal
retention time, 11.92d for method 2 (double
reactor) and 19.11d for method 3 (single reactor),
yielding digester volumes of 423m 3 and 356m 3
(with and without X ti limit) for method 2 and
677m 3 and 400m 3 (with and without X ti limit) for
method 3 (see Table 6.1a). The reason for the
difference in retention time is that with method 2,
which is based on a double reactor digester
retention time, the required effluent active
fraction is not achieved (f ae = 0.310). In a double
reactor digester with step feeding, the retention
time needs to be based on a single reactor
digester to achieve the required effluent active
fraction (fae) of 0.235. So step feeding to a
double reactor digester affords no advantage
over a single reactor digester. The smallest
digester conforming to the thickening and OUR
limits that achieves the required effluent active
fraction is the double digester with 100% feed
into the first reactor (method 1), viz. 550 m 3 .
Fig 6.18: Digester volume in m 3 versus digester
feed concentration in %TSS for the single (1),
double (2), and triple (3) reactor digesters for a
maximum OUR of 125 mgO/( R.h) in the first reactor
with ( V) and without ( Ž) a 5%TSS feed
concentration limit for the example settled
wastewater activated sludge system at 14 oC. Note
that the double (2) and triple (3) reactor digesters
with step feeding also plot at V and Ž when their
R hd equals that of the single digester - when their
R hd is based on multiple reactors, they do not
achieve the required effluent active fraction.
The same applies to the triple reactor digester
(Table 6.1b) as for the double reactor digester
(Table 6.1a). From Table 6.1b, it can be seen
that step feeding into a triple reactor digester
affords no advantage over the single reactor
digester, so the smallest volume conforming to
the thickening and OUR limits is 618 m 3 for the
triple reactor digester with 100% feed into the
first reactor. This 618 m 3 is greater than the 550
m 3 for the double reactor digester. This is
because moving from a double to a triple reactor
digester reduces the retention time relatively little
(from 11.9d to 10.3d) compared with moving from
a single to a double reactor digester (from 19.1d
to 11.9d), but the reduction in feed concentration
to maintain an OUR below O t1max is large,
resulting in a net increase in digester volume.
From the above, the smallest digester volume is
for the single reactor digester, but this also
requires the greatest degree of thickening (21.2x)
to achieve the feed sludge concentration of
8.48% TSS. Accepting a float solids
concentration limit of 5% requires a larger
volume for the single reactor digester, i.e.
8.48/5x400 = 677m 3 and the OUR decreases
below 125 mgO/( R.h), i.e. 125x400/677 = 73.8

6.28
 mgO/( R.h). This volume is now larger than that of
the double or triple reactor digester (Fig 6.18,
Tables 6.1a and b). The feed solids
concentrations for the double and triple reactor
digesters are well within that achievable with
DAF. Therefore, multi-reactor aerobic digestion
of WAS is advantageous not so much from a
volume saving point of view, but because it
reduces the degree of thickening required for the
feed concentration, while keeping the digester
volume low by reduction in retention time. Also,
the lower the feed sludge concentration, the
lower the nitrate and phosphate concentrations
in the digester requiring lower lime dosing
concentrations per R influent to maintain neutral
pH in the digester reactors - high lime dosages
are difficult to dissolve. The analysis shows that
the double reactor digester (with each subdivided
into eq u a l size d an o xic an d ae ro bic
compartments for denitrification) is best from
k in e tic , O U R a nd flo ta tio n th ic k e n in g
considerations.
Fig 6.20: Volatile solids reduction as a function of
digester liquor temperature (in oC) and sludge age
(SRT in d) showing experimental data from pilot
(open symbols) and fullscale plants (closed
symbols), the US design guide line from W EF
(1990) and the endogenous respiration based
aerobic digester model predicted lines at 10 and
30 o C for the example PS-WAS blend (fai = 0.619).
6.29
10 MODEL COMPARISON AND
VALIDATION
An aerobic digester model based on observed
pilot and fullscale plant performance is presented
by WEF (1990) and Metcalf & Eddy (1991). The
model is based on VSS degradation, i.e.
dM/dt = -K d M kgVSS/d (6.112)
where
M = mass of biodegradable VSS kgBVSS
Kd = VSS degradation rate /d
In theory the model recognizes the degradation
of biodegradable VSS. However, because in
practice the biodegradable VSS is not known and
cannot be measured, in application M is the
m e a s ure d V S S m a s s , w h ic h in c lu d e s
unbiodegradable VSS. This is the major
difference between the WEF (1990) aerobic
digester model and the active mass endogenous
respiration based one presented above and
results in K d values that differ significantly from
the endogenous respiration rate b HT , i.e. K d =
0.02/d at 10 o C and 0.18/d at 30 oC (see WEF,
1998, Fig 22.39, pg 22-103) compared with b H =
0.18/d at 10 o C and 0.32/d at 30 o C, the
approximate range of validity of the temperature
sensitivity coefficient of b H , 2bH = 1.029.
Although K d values are given for the WEF (1990)
aerobic digester model, in application they are
not used. The process design is based on a
design chart giving fraction VSS removal versus
the product of the digester retention time (in d)
and temperature (in o C), SRTxT (Fig 6.20). The
WEF(1990) design guide line is the best fit line
through the pilot and fullscale performance data.
The digester SRT is determined from a desired
fraction of VSS removal (usually 40% to ensure
the pilot and full scale plant performance comply
with the >38% minimum VSS removal for stable
sludge) and specified minimum digester
temperature.
The endogenous respiration based model results
calculated from Eq 6.30 for 10 and 30 oC are also
shown in Fig 6.20 for the example PS-WAS
blend which has an influent active fraction fai =
0.619. It can be seen that the lines correspond
well with most of the pilot scale data (open
symbols) and some full scale data (closed
symbols). The following conclusions can be
made from this comparison:
1. According the endogenous respiration based
model, temperature and SRT do not
compensate equally as the W EF(1990) design
chart suggests, and so different lines for
different temperatures are obtained in Fig 6.20.
2. The position of the line is strongly dependent
on the influent active fraction (see Fig 6.21).
The fixed USEPA design guideline in effect
assumes that all sludges have the same fraction
of biodegradable VSS. For PS this is not
unreasonable, but for WAS it is - the longer the
sludge age of the AS system, the lower the
biodegradable VSS (or COD) in the WAS.
3. The endogenous respiration based model
can be used with reasonable confidence for
design of aerobic digesters between 10 and
30 o C.

The advantage of the endogenous respiration

based model is that it (i) conforms to the COD
balance and structure of the activated sludge
simulation models such as ASM No 1 (Henze et
al., 1987) and UCTOLD (Dold et al., 1991), and Fig 6.22: Sludge stability in terms of SOUR in
therefore (ii) can be used to calculate digester mgO/(gVSS.h) and fraction VSS removal in the
volumes, flows and nutrient N and P doses for influent active fraction (fai) versus retention time
input to the simulation models. Simulation of the design chart.
integrated WWTP with primary sedimentation,
8d activated sludge treating settled WW and 0.127 at 20 o C, 0.150 at 14 oC or 0.120 at 22 o C.
aerobic digestion of PS and WAS with thickening Figure 6.45 shows the digester design chart (cf.
to 6 and 5% respectively at 14 oC in Aquasim 2.0 Fig 6.8) at 14 o C with the sludge effluent quality
(Reichert, 1998) yields identical results to those specified in terms of SOUR instead of f ae. For the
in Table 6.7, provided the appropriate example PS-WAS blend, with an influent active
c o n c e n tr a t i o n s o f b i o d e g r a d a b le a n d fraction fai = 0.619, to achieve an SOUR = 1.5
unbiodegradable COD, OrgN and OrgP are mgO/(gVSS.h) at 14 o C, requires a digester
removed in the PST (see Chapter 3, Section 7.3).
In the aerobic digester part of the model, care
must be taken that the calculated N (and P) dose
is added to the digester. The kinetics of the
UCTOLD and ASM No1 are set up in such a way
that if there is insufficient ammonia and/or nitrate
for growth, growth (via death-regeneration)
ceases. When this happens, high concentrations
of slowly biodegradable COD accumulate in the
digester and the OUR is zero.

The above models are based on the kinetics of

VSS reduction and generally VSS reductions in
excess of 38% are required to stabilize WW
sludges (USEPA, 1978). Also, the maximum
specific OUR (SOUR) for stable sludges
specified in the US 40 CFR Part 503 is 1.5
mgO/(gVSS.h). Not all sludges can comply with Fig 6.21: Volatile solids reduction as a function of
both requirements. Rearranging Eq 6.19 with digester liquor temperature (in o C) and sludge age
X BH = fae X v, i.e., (SRT in d) showing the USEPA design guide line
from WEF (1990) and the endogenous respiration
mgO/(gVSS.h) based aerobic digester model predicted lines at
20 o C for the example PS-WAS blend with influent
(6.113) active fractions fai = 0.50, 0.55, 0.60 and 0.65.

where SOUR = O c /X v = 1.5mgO/(gVSS.h), yields retention time of 30d and the fraction VSS
maximum effluent active fractions (fae = f av ) of removal is 0.426. From Fig 6.22, sludges with

6.30
fai< 0.56, which would be the case with PS-WAS Table 6.8: Stabilized sludge characteristics for
blends from settled WW AS systems with sludge the example WAS, PS and PS-WAS sludges
ages longer than 8d, >38% VSS removals and achieved in aerobic digestion operated as a
SOUR < 1.5 mgO/(gVSS.h) cannot be achieved process to significantly reduced pathogens
simultaneously - the SOUR < 1.5 mgO/(gVSS.h) (PSRP) conforming to the Class B biosolids
can always be achieved provided the retention (partial disinfection) conditions according to
time is long enough but the % VSS removal will USEPA Part 503 Biosolids Rule.
be lower the 38%. This demonstrates that stable Parameter WAS PS PS-WAS
sludge can be produced even though the % VSS Influent Active fraction, 0.662 0.592 0.619
removal is below 38%. In this respect, the model fai
presented in this Section is superior because an
Temperature 15oC, Retention time 60d; SRTxT = 900
effluent active fraction or SOUR representative of
a stable sludge can be selected and based on Frac VSS removed, fsr 0.491 0.438 0.458
the influent active fraction, the required digester SOUR mgO/(gVSS.h) 0.97 0.88 0.87
retention time calculated. The %VSS removal Effluent active frac, fae 0.096 0.078 0.085
then is a consequence of the other selected o
Temperature 20 C, Retention time 40d; SRTxT = 800
parameters.
Frac VSS removed, fsr 0.480 0.428 0.447
The procedures above considered the design of SOUR mgO/(gVSS.h) 1.42 1.16 1.28
aerobic digesters for the point of view of Effluent active frac, fae 0.120 0.098 0.108
producing a stable sludge. Pathogen reduction, o
which is becoming increasingly important for Temperature 30 C, Retention time 25d; SRTxT = 750
sludges that are beneficially reused, was not Frac VSS removed, fsr 0.470 0.420 0.441
considered. Retention times for pathogen SOUR mgO/(gVSS.h) 2.21 1.81 1.89
reduction are longer than for sludge stabilization.
Effluent active frac, fae 0.140 0.115 0.120
According to USEPA Part 503 Biosolids Rule
(USEPA, 1994), ambient aerobic temperature
digestion does not qualify as a process to further Table 6.8 shows that the partial disinfection
reduced pathogens (PFRP) and so the effluent (Class B) and stabilisation requirements conform
sludge does not qualify as a Class A (disinfected) quite closely at the lower temperatures of 15 and
sludge, which is required to contain <1000 most 20 o C. For the example the WAS, PS and PS-
probable number (MPN) faecal coliforms per g WAS sludges, the fraction of VSS removed (fsr) is
dry mass solids. To qualify as a PFRP and > 40% and the SOUR is < 1.5 mgO/(gVSS.h). At
produce a Class A sludge, thermophilic the high temperatures (>20 o C) partial disinfection
temperatures between 55 and 60 oC at retention is more easily achieved than stabilisation. At the
times >10d are required. However, ambient ~25d required for partial disinfection, the SOUR
aerobic digestion does qualify as a process to exceeds 1.5 mgO/(gVSS.h). To achieve SOUR
significantly reduce pathogens (PSRP) to < 1.5 gO/(gVSS.h), the retention time should be
produce a Class B (partially disinfected) sludge, somewhat longer than 25d, i.e. ~30d to meet the
which is defined as < 2x10 6 MPN faecal coliforms stabilisation requirements. While sludges with
per g dry mass solids. To qualify as a PSRP, different influent active fractions (fai) would give
digestion retention times must exceed 40d at somewhat different values, the results for the
20 o C and 60d at 15 o C. For mesophilic anaerobic example sludges are given to demonstrate the
digestion (~37 o C), retention times >15d are general trends.
required for PSRP. This gives SRTxT products
of 600 to 900 with the lower values at the high Because ambient temperature aerobic digestion
temperatures. Accepting SRTxT products for and mesophilic anaerobic digestion do not qualify
partial disinfection (PSRP, Class B) for the as PFRPs and therefore do not produce a Class
aerobic digestion of 900, 800 and 750 for 15, 20 A (disinfected) sludge, thermophilic aerobic
and 30 o C, yields retention times of 60, 40 and digestion and mesophilic anaerobic digestion
25d respectively. The model predicted systems have been combined in the dual
stabilisation achieved under these conditions are digestion system. In this system, disinfection
listed in Table 6.8 for the example W AS, PS and (>65 o C for >2h) is achieved in the 1 st stage
PS-WAS blend. autothermal thermophilic aerobic digester with
the oxygen and/or air and final stabilisation is
achieved in the 2 nd stage mesophilic (or
thermophilic) anaerobic digester. Successful full-
scale dual digester system trials are reported by

6.31
Messenger et al. (1993a, b), Messenger and
Ekama (1993a, b) and Pitt and Ekama (1994,
1995).
11 REFERENCES
Bratby J and Marais GvR (1977). Flotation -
Chapter 5 in Solid-Liquid Separation
Equipment Scale-up, Ed D Purchase, Uplands
Press, England.
Bratby J (1978) Aspects of sludge thickening by
dissolved-air flotation. Wat. Pollut. Contr.,
77(3) 421-432.
Dold PL, Wentzel MC, Billing AE, Ekama GA and
Marais GvR (1991). Activated sludge
sim ulation program s: Nitrification and
nitrification/denitrification systems (version
1.0). Water Research Commission, P O Box
824, Pretoria 0001, South Africa. TT 52/91,
ISBN 0 947447 19 9.
Henze M, Grady C P L, Gujer W, Marais GvR
and Matsuo T (1987). Activated sludge model
No.1. IAW PRC Scientific and Technical
Reports No 1, International Association on
Water Pollution Reseach and Control,
IAW PRC, London, ISSN 1010-707X. 33pp.
Marais GvR and Ekama GA (1976) The activated
sludge process part 1 - Steady state
behaviour. Water SA 2(4) 163-200.
Metcalf & Eddy (1991) Wastewater engineering:
Treatment, disposal, reuse. McGrw-Hill Intnl
Series, New York, ISBN 0-07-100824-1.
Messenger JR, de Villiers HA, Laubscher SJA,
Kenmuir K and Ekama GA (1993) Evaluation
of the dual digestion system: Part 1 -
Overview of the Milnerton experience. Water
SA 19(3) 185-191.
Messenger JR, de Villiers HA and Ekama GA
(1993) Evaluation of the dual digestion
system: Part 2 - Operation and performance
of the pure oxygen aerobic reactor. Water SA
19(3) 193-200.
Messenger JR and Ekama GA (1993)
Evaluation of the dual digestion system: Part
3 - Considerations in the process design of
the aerobic reactor. Water SA 19(3) 201-208.
Messenger JR and Ekama GA (1993)
Evaluation of the dual digestion system: Part
4 - Simulation of the temperature profile in the
batch fed aerobic reactor. Water SA 19(3)
209-215.
Pitt AJ and Ekama GA (1994) Dual digestion of
sewage sludge with air. Procs of IAW Q
Specialized Conference on Anaerobic
digestion AD 94, Cape Town, Vol 2, 140-143.
Pitt AJ and Ekama GA (1996) Dual digestion of
sewage sludge with air and pure oxygen.
Procs 69th W ater Environment Federation
6.32
Annual conference and exhibition, Dallas TX,
2, 69-82.
Reichert P (1998) Concepts underlying a
computer programme for the identification and
simulation of aquatic systems. Swiss Federal
Institute for Environmental Science and
Technology (EAWAG), CH-8600, Dubendorf,
Switzerland.
Samson KA and Ekama GA (2000) An
assessment of sewage sludge stability with a
specific oxygen utilization rate (SOUR) test
method. Water Sci. Technol. 42(9) 37-40.
Warner APC, Ekama GA and Marais GvR (1985)
Comparison of aerobic and anoxic-aerobic
digestion of waste activated sludge. Water
Sci. Technol. 17(8) 1475-1478.
Warner APC, Ekama GA and Marais GvR (1986)
The activated sludge process Part 4 -
Application of the general kinetic model to
anoxic-aerobic digestion of waste activated
sludge. Water Research 20(8) 943-958.
WEF (1990) Operation of municipal wastewater
treatment plants, Manual of practice MOP11,
Vol III, Water Environment Federation,
Alexandria, VA, USA.
WEF (1998) Design of Municipal WW TP, 4 th
Edn., Water Environment Federation (WEF)
Manual of Practice MOP 8, WEF, Alexandria,
VA, American Society of Civil Engineers
(ASCE) Manual and Report on Engineering
Practice No 76, ASCE Vol 3, ASCE, Reston,
VA, USA.
USEPA (1978) Sludge treatment and disposal -
Vol 1. US Environmental Protection Agency,
625/4-78-012, Tech. Trans, Washington, DC,
USA.
USEPA (1994) A plain English guide to the EPA
Part 503 biosolids rule. US Environmental
Protection Agency, Office of wastewater
management (4204), EPA/832/R-93/003,
EPA, Washington, DC, USA.