Bioresource Technology 102 (2011) 5504–5513

Contents lists available at ScienceDirect

Bioresource Technology
journal homepage: www.elsevier.com/locate/biortech

Development and calibration of bio-kinetic model for surfactant biodegradation

with combined respirometric and titrimetric measurements
V. Aravinthan ⇑, M.A. Hoque
Australian Centre for Sustainable Catchments, University of Southern Queensland (USQ), Qld 4350, Australia

a r t i c l e i n f o a b s t r a c t

Article history: Substrate removal mechanism in aerobic activated sludge processes was lately modeled using the simul-
Received 25 June 2010 taneous storage and growth (SSAG) phenomenon. The SSAG model was further refined with titrimetric
Received in revised form 25 August 2010 components and successfully calibrated using both respirometric and titrimetric measurements for com-
Accepted 26 August 2010
mon substrate acetate. However, the improved SSAG model calibration was not verified with other
Available online 31 August 2010
organic substrates. Furthermore, very few studies are available in the literature on surfactant bio-kinetics,
which generally use off-line experimental measurements with limited model-based interpretation.
Keywords:
Therefore, the aim of this paper is to demonstrate its applicability for surfactant biodegradation using
Sodium dodecyl sulfate (SDS)
Model calibration
on-line measurements. Batch experiments were conducted using sodium dodecyl sulfate (SDS) as a test
Parameter estimation surfactant. Model calibration was done successfully for three different SDS concentrations using respiro-
Simultaneous storage and growth (SSAG) metric, titrimetric and combined respirometric–titrimetric measurement approaches. The parameter
Respirometric and titrimetric estimation results from all three stated combinations were statistically evaluated and found to be very
measurements close validating the model.
Ó 2010 Elsevier Ltd. All rights reserved.

1. Introduction encounter stringent effluent discharge conditions imposed by

Surfactants are widely used in the manufacture of detergents
and personal care products (Hotantai and Nardello-Rataj, 2001).
Consequently, the presence of such organic contaminants in raw
wastewater has remarkably increased. Substantial research has
been conducted to investigate the biodegradation of surfactants
under aerobic and/or anaerobic conditions over the last decade
(Huber et al., 2000; Qin et al., 2005; Sharvelle et al., 2007). It is in-
deed a complicated process since surfactants have heavy molecular
weights and complex chemical structures. In addition, surfactants
exhibit characteristics that make them prone to escape with the
effluent as well as to adsorb onto the sludge during primary tank
settling (Jahan, 2005). Though significant research has been per-
formed that focused on the extent and pathways of surfactant deg-
radation, the resulting recalcitrant products, their biological
activities and applications (Ahmed et al., 2010; Hrenovic and Iva-
nkovic, 2007; Lara-Martin et al., 2006; Qin et al., 2005; Thanomsub
et al., 2006), there is little evidence reported regarding the deter-
mination of biodegradation kinetics (Chen et al., 2005; Mohan
et al., 2006) using activated sludge models.
In-depth understanding of substrate removal mechanisms via
improved models is essential for process optimization and control
in full-scale wastewater treatment plants (WWTPs). These plants
⇑ Corresponding author. Tel.: +61 7 4631 2299; fax: +61 7 4631 2526.
E-mail address: aravintv@usq.edu.au (V. Aravinthan).
Environmental Protection Agencies because of the prevailing envi-
ronmental concerns regarding the presence of organic wastewater
contaminants in water resources, albeit at very low concentrations
(Kolpin et al., 2002). Activated sludge models have been evolving
from a simple growth-based concept (Gernaey et al., 2002b; Guisa-
sola et al., 2005; Vanrolleghem et al., 2004) to more complicated
models using the simultaneous storage and growth (SSAG) phe-
nomenon (Beccari et al., 2002; Pratt et al., 2004; Sin et al., 2005)
to interpret the organic carbon removal mechanisms in a mixed
culture. Researchers commonly employ on-line monitoring tools
such as respirometry and titrimetry for experimental investigation
on aerobic biodegradation, precise model calibration and parame-
ter estimation purposes (Gernaey et al., 2002b; Hoque et al., 2010;
Petersen et al., 2001; Sin and Vanrolleghem, 2007). These tools
investigate the biodegradation rate of organics employing high fre-
quency data collection that preserves all the bio-kinetic informa-
tion during the oxidation period. The SSAG model proposed by
Sin et al. (2005) was calibrated using the experimental oxygen up-
take rate (OUR) of carbon-based compound such as acetate biodeg-
radation. This model was further extended by Hoque et al. (2010),
introducing the titrimetric components in each step of the growth
and storage phases in the SSAG process, along with the consider-
ation of the dynamic carbon dioxide transfer rate (CTR) in the li-
quid phase. However, calibration of the recently developed SSAG
has been done using an easily biodegradable compound such as
acetate. Therefore, the model needs to be verified using different

0960-8524/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.biortech.2010.08.101
 V. Aravinthan, M.A. Hoque / Bioresource Technology 102 (2011) 5504–5513 5505

Nomenclature

ASM3 activated sludge model no. 3 pK1 negative logarithm of the first acidity constant in the
bH endogenous decay coefficient of biomass (day1) CO2equilibrium
bSTO endogenous decay of storage products (day1) pKNH4 negative logarithm of the equilibrium constant for NHþ4
CHaObNc elemental composition of biomass (C-mol) dissociation
CHpOq elemental composition of storage products (C-mol) qMAX maximum substrate uptake rate (day1)
CHyOz elemental composition of substrate (C-mol) SCO2 CO2 concentration in liquid phase (mmol/L)
CT,init total inorganic carbon in the aqueous medium (mmol SCO2 CO2 saturation concentration at 1 atm (mmol/L)
CO2/L) SDS sodium dodecyl sulfate
CTR CO2 transfer rate (mmol CO2/L day) SHCO3 bicarbonate concentration in liquid phase (mmol/L)
DO dissolved oxygen SNH ammonium concentration (mg N/L)
fSTO fraction of substrate used for storage (mg COD XSTO/mg SO dissolved oxygen concentration in liquid phase (mg/L)
COD SS) SS readily biodegradable substrate concentration (mg
fXI inert fraction of biomass (mg COD/mg COD) COD/L)
Hp proton concentration in liquid phase (meq/L) SSAG simultaneous storage and growth
iNXI nitrogen content of the inert fraction of biomass (g N/g XH biomass concentration (mg COD/L)
COD XI) XH(0) initial biomass concentration (mg COD/L)
iNBM nitrogen content of biomass (g N/g COD XH) XI inert particulate COD (mg COD/L)
K1 forward reaction rate for aqueous CO2 equilibrium XNHacc nitrogen accumulation (mg N/L)
(day1) XS slowly degradable particulate COD (mg COD/L)
KS substrate affinity constant (mg COD/L) XSTO storage products concentration (mg COD/L)
KLa oxygen mass transfer coefficient (day1) YH,S yield coefficient for growth on substrate (mg COD XH/
K L aCO2 CO2 mass transfer coefficient (day1) mg COD SS)
kh hydrolysis rate (day1) YH,STO yield coefficient for growth on storage products (mg
kNHacc nitrogen accumulation rate of biomass (day1) COD XH/mg COD XSTO)
kSTO maximum storage rate of biomass (day1) YSTO yield coefficient for storage on substrate (mg COD XSTO/
KX hydrolysis saturation constant (mg COD/mg COD) mg COD SS)
K1 regulation constant of biomass controlling degradation s first order time constant (day)
rate of XSTO (mg COD XSTO/mg COD XH) lMAX,S maximum growth rate of biomass on substrate (day1)
K2 a lumped parameter related to the affinity of biomass to lMAX,STO maximum growth rate of biomass on storage products
storage fraction of biomass (mg COD XSTO/mg COD XH) (day1)
MSE mean squared error cS degree of reduction of substrate (mol electron/C-mol)
OUR oxygen uptake rate (mg O2/L day) cSTO degree of reduction of storage products (mol electron/
OURend endogenous oxygen uptake rate (mg O2/L day) C-mol)
PHB polyhydroxybutyrate cX degree of reduction of biomass (mol electron/C-mol)

substrates to interpret these biodegradation mechanisms to be
confident in its use for wider application. Moreover, the existing
models describing the surfactant biodegradation were based on
first-order or a simple Monod function without considering its bio-
degradation pathway. These models were calibrated using off-line
substrate depletion measurements that result in an inaccurate
parameter estimation process due to the non-availability of high-
frequent measurements.
Hence, the objective of this paper is to demonstrate the applica-
bility of the improved SSAG model (Hoque et al., 2010) for surfac-
tant biodegradation using on-line measurements considering the
biodegradation pathway of surfactants. Sodium dodecyl sulfate
(SDS), a common anionic surfactant used in household products,
was used as a calibration substrate for the batch experiments.
Moreover, three different calibration approaches: using the respi-
rometric measurements alone, the titrimetric measurements alone
and combined respirometric-titrimetric measurements were per-
formed to estimate the parameters more precisely and to validate
the proposed model.
Hoque et al. (2010) for acetate biodegradation was further ex-
tended here by newly introducing hydrolysis component.
The proposed SSAG model for SDS biodegradation includes the
stoichiometric parameters involved in titrimetry in each step of
the growth and storage phases along with consideration of the
non-linear carbon dioxide transfer rate in the liquid phase. The ma-
jor steps, other than hydrolysis, during the aerobic biodegradation
of SDS are the formation of storage products, aerobic growth on the
substrate, aerobic growth on the storage, endogenous respiration,
respiration on storage products, aqueous CO2 equilibrium and
stripping of CO2 (see Table 1 for the process matrix).
The conversion of sodium dodecly sulfate to alcohol (1-dedeca-
nol) occurs through a hydrolysis process that releases H+ in the li-
quid medium (Eq. 1). The proton production during hydrolysis can
be estimated by using the matrix shown in Table 1, where the
parameter ‘‘C” represents the molecular weight of the substrate,
SDS (576 g COD/mol)
C12 H25 O4 SNa þ H2 O ! C12 H26 O þ NaSO4 þ Hþ ð1Þ

The alcohol undergoes a multi-step oxidation process producing

2. Model development
A bio-kinetic model was proposed to describe both the respiro-
metric and the titrimetric behavior resulting from the aerobic bio-
degradation of the surfactant, SDS, in an activated sludge system.
Fig. 1 illustrates the proposed model diagram with processes in-
volved during SDS biodegradation. The SSAG model developed by
lauric acid in a liquid medium (Yao, 2006). During the modeling,
the course of oxidation was consolidated into one step to keep
the proposed model simple. While the SDS biodegradation pathway
shows proton production during the lauric acid formation, a frac-
tion of the proton is consumed for lauroyl-CoA synthesis. The net
H+ production that occurs in the liquid medium is shown in Table 1
considering lauric acid as a readily biodegradable compound (SS) to
be used for biomass growth. The stoichiometry related to the
5506 V. Aravinthan, M.A. Hoque / Bioresource Technology 102 (2011) 5504–5513

SO
SO

XH XI
Endogenous
H2O
Growth on Respiration
substrate Growth on
XS SS Storage Endogenous
SO
Respiration
Hydrolysis SO on Storage
Storage
XSTO
SO

Fig. 1. Model diagram for aerobic biodegradation of SDS.

processes such as aerobic growth on substrate, formation of storage 1ðmg CODÞCHa Ob Nc þ 1ðmg CODÞO2
products from substrate and aerobic growth on storage indicates 1
the CO2 production rate that can be determined simply from Eqs. ! ðmmolÞCO2 þ ð. . .ÞH2 O þ iNBM ðmg NÞNH3 ð5Þ
8cX
(2)–(4) respectively (readers are referred to Hoque et al. (2010)
1ðmg CODÞCHp Oq þ 1ðmg CODÞO2
for the derivation)
1
  ! ðmmolÞCO2 þ ð. . .ÞH2 O ð6Þ
1 1 8cSTO
ðmg CODÞCHy Oz þ 1  ðmg CODÞO2 þ iNBM ðmg NÞNH3
Y H;S Y H;S
  The kinetics and stoichiometry corresponding to the processes of
1 1 c
! 1 ðmg CODÞCHa Ob Nc þ  S ðmmolÞCO2 þ ð. . .ÞH2 O ð2Þ aqueous CO2 equilibrium and CO2 stripping are kept the same as
8cS Y H;S cX
  those used for acetate biodegradation modeling (Hoque et al.,
1 1
ðmg CODÞCHy Oz þ 1  ðmg CODÞO2 2010).
Y STO Y STO
 
1 1 c
! 1 ðmg CODÞCHp Oq þ  S ðmmolÞCO2 þ ð. . .ÞH2 O ð3Þ
8c Y STO cSTO 3. Methods
 S 
1 1
ðmg CODÞCHp Oq þ 1  ðmg CODÞO2 þ iNBM ðmg NÞNH3
Y H;STO Y H;STO 3.1. Batch study
 
1 1 c
! 1 ðmg CODÞCHa Ob Nc þ  STO ðmmolÞCO2 þ ð. . .ÞH2 O
8cSTO Y H;STO cX Batch experiments were conducted using sodium dodecyl sul-
ð4Þ fate (SDS) as a test surfactant. A titrimetric respirometer was in-
stalled in the laboratory that was equipped with dissolved
where CHaObNc, CHyOz and CHpOq represent the elemental compo- oxygen (DO) and pH sensors along with a reactor having a capacity
sition of biomass, substrate (lauric acid) and storage products of 3.5 L (Hoque et al., 2008). Both pH and DO were monitored every
respectively. The yield coefficients for growth on substrate, storage 5 s interval and pH was controlled at a set point of 7.8 ± 0.03. Data
on substrate and growth on storage products are expressed as YH,S, acquisition of the analogue signals from the sensors was processed
YSTO and YH,STO respectively. The degree of reduction of the substrate by a personal computer equipped with the Labview software pack-
(cS), the storage products (cSTO) and the biomass (cX) can be calcu- age. Compressed air was supplied for the proper aeration in the
lated as 4 + y  2z, 4 + p  2q and 4 + a  2b  3c respectively. The bioreactor. The sludge was collected from Wetalla Water Reclama-
coefficient related to ammonia uptake (iNBM) is expressed as g N tion Plant (operated by Toowoomba City Council), Australia. The
per g COD biomass unit basis that can be determined from the rela- sludge was acclimatized with surfactant (SDS) for 15 days prior
tion 14c/8cX (Sin, 2004). to the commencement of the main experiments to allow the
According to the principles of SSAG model, part of the readily microorganisms to perform at their maximum capacity. Basic
biodegradable compound, lauric acid, is considered to be uptaken nutrient solution consisting of phosphate buffer, minerals and
for heterotrophic biomass growth while the rest of it is consumed NH4Cl was fed at the beginning of experimental run along with
for simultaneous storage formation. Proton production takes place allylthiourea to inhibit nitrification (Beccari et al., 2002; Guisasola
as lauric acid leaves H+ in the liquid medium before it is consumed et al., 2005). SDS with varying concentration of 50, 75 and 100 mg
by biomass cell that has been supplemented due to ammonia COD/L was used as test substrate to investigate the biodegradation
assimilation for biomass growth. In Table 1, the parameter ‘‘p” rep- mechanism under aerobic condition. All raw data related to DO and
resents the fraction of NHþ 4 in the liquid phase which is derived as
pH was processed using a spreadsheet program as prescribed by
1=ð1 þ 10pHpKNH4 Þ by Gernaey et al. (2002a). During the aerobic Gernaey et al. (1998, 2001). OUR was calculated based on the pro-
growth on storage process, it is assumed that the biomass accumu- cedure explained in Gernaey et al. (2001) using the experimentally
lates nitrogen too within the cell along with carbon source for their determined value for the oxygen transfer coefficient (KLa). Re-
subsequent growth purpose (see the ammonium balance in aeration procedure was followed to calculate the parameter KLa
Table 1), which is also assumed for acetate biodegradation model- (ASCE, 1996).
ing Hoque et al., 2010.
Moreover, the biological reactions during endogenous respira- 3.2. Model parameter estimation
tion as well as the respiration on storage lead to CO2 production
in the system thus influence the process titrimetry that can be esti- The proposed model was calibrated with both the respirometric
mated using the stoichiometric expression as mentioned in Eqs. 5 and titrimetric measurements where assays were conducted for
and 6 accordingly. Table 1 presents the production of CO2 for the SDS pulses of 50, 75 and 100 mg COD/L to the activated sludge.
respective oxygen uptake of (1  fXI) g COD (as derived by Sin Three different calibration approaches were applied: using the res-
and Vanrolleghem (2007)). The parameter cSTO in Eq. 6 refers to pirometric measurements alone, the titrimetric measurements
the degree of reduction of the storage products alone and combined respirometric–titrimetric measurements
 V. Aravinthan, M.A. Hoque / Bioresource Technology 102 (2011) 5504–5513 5507

followed by model parameter estimation. The results were com-

 XH
pared for validation of the proposed SSAG model. Non-linear tech-
nique was employed for the parameter estimation process using


S
: SSKþK
S
the MATLAB optimisation toolbox (R2007a). Minimization of the

ð1  et=s Þ  kNHacc  M s  X H

1  e s  kMAX;S  MS  X H

ð1  e s Þ  lMAX;S  M S  X H

k1 SCO2  k1 10pk1 pH SHCO3

mean squared error (MSE) between the model and the experimen-

1 ðX STO =X H Þ
tal output was calculated as the main criterion for curve fitting.

ðX STO =X H Þ2

K L aCO2 ðSCO2  SCO2 Þ

The model parameters KS, qMAX, YH,S, YSTO, YH,STO, K1, K2, kh, KX and
kh  M X SDS=XH  X H

s were estimated along with calculation of 95% confidence inter-

lMAX;STO K 2 þK
vals. A description of the model parameters is presented in Nomen-

bSTO  XSTO


Kinetics

t
t

b H  XH
clature. Default values assigned in the ASM3 model for the
parameter fXI (0.2) were assumed for the current analysis. The val-


ues for the parameters bH and bSTO were fixed at 0.00036 per min
(1  fXI)
(0.52 day1) for better curve fitting, though it is higher than the
S0(g O2)

1Y H;STO
STO
1Y H;S

Y H;STO
Y STO

ASM3 default one (0.2 day1). For successful model calibration,

Y H;S

 1Y

1

Carvalho et al. (2001) and Beccari et al. (2002) assumed the param-

–
–

–
–
eter bH in their study as 0.72 and 0.041 day1 respectively. In the
iNBM  iNXIfXI

current study, a value of 0.2 day1was considered for the parame-

SNH (g N)

Assuming that ammonia required for the biomass growth during storage to growth process is taken from the internal source (cell) instead of the external environment.
ter kNHacc at the beginning of the parameter estimation process,
iNBM
iNBM

and was revised later for better curve fitting. The ASM3 prescribed
–

–
–
–

values for the parameters iNXI (0.02 g N/g COD XI) and iNBM (0.07 g
N/g COD XH) were fixed during the proposed model calibration.
 ðiNBM f14X1 iNXI ÞP

The parameter fSTO was fixed at 0.65 for successful model calibra-
SHp (mol)

þ C3

tion. The maximum storage rate (kSTO) and the maximum growth
iNBM P

iNBM P

rate of the biomass (lMAX,S) were calculated from the estimates

14

14

–
1
–
1
C

of the parameters qMAX and fSTO based on the procedure explained



in Sin et al. (2005) where the parameter lMAX,STO is assumed to be

 ccSTO
X


the same order of magnitude as lMAX,S. The relationship



STO
 ccs
X
 ccS
SCO2 (mol)

Y H;STO

OURend(0) = (1  fXI).bH.XH(0) was employed to calculate the initial

1
Y STO
Y H;S

1
1

concentration of the biomass, XH(0).




1fX1
8cSTO

8cSTO
8cx

1
1

1
8cs

8cs

The parameter k1(1.0762 per min) was kept at the same value it
1

1
–
–

was used for the acetate biodegradation study (Hoque et al., 2010),
SHCO3 (mol)

and total inorganic carbon in the aqueous medium, CT,init was ad-
justed for different assays to fit the experimental profile with the
model one. The initial concentrations of CO2 and HCO3 in the reac-
–
–
–

–
–
1
–

tor were calculated using their relationship with CT,init (Sin, 2004).
During the model calibration, the value for K L aCO2 was calculated as
XSDS (g COD)

0.055 min1 from the oxygen transfer coefficient (KLa) using the
relationship between their diffusivity coefficients (Sperandio and
1

Paul, 1997; Sin and Vanrolleghem, 2007), whereas the parameter

–
–

–
–
–
–

pK1 was taken as 6.39 (Sperandio and Paul, 1997). The default val-
SS (g COD)

ues suggested by Stumm and Morgan (1996) for the parameters

Process matrix involved in the proposed model for aerobic biodegradation of surfactant.

pK NH4 (9.25) and SCO2 ð0:017 mmol=LÞ were assumed during the
 Y STO
 Y 1H;S
1

parameter estimation process. Besides, the degree of reduction of

1
–

–
–
–
–

the substrate (cS) and biomass (cX) were calculated as 5.67 and
XSTO (g COD)

4.2 using the elemental composition of lauric acid (C12H23O2) and

biomass (CH1.8O0.5N0.2) respectively. The degree of reduction coef-
 Y H;STO

ficient cSTO was kept at 4.5 assuming polyhydroxybutyrate (PHB)

1

1

types storage compounds formed in biomass cell to consume it la-

–
–
–

–
–

ter for their growth.

XNHacc (g N)

iNBM
iNBM

4. Results and discussion

–

–
–
–
–

4.1. Experimental observations on surfactant biodegradation

XH (g COD)

Activated sludge was acclimatized with the anionic surfactant,

1
–
–
1

–
–
–

SDS, for about 15 days before starting the main experimental work
in order to allow the microorganisms to perform at their maximum
Aqueous CO2 equilibrium
Aerobic growth on XSTO

capacity. Batch experiments with varying initial SDS concentra-

Endogenous respiraton
Aerobic growth on SS

tions were conducted where pH was maintained at 7.8 ± 0.03.

SNH accumulation

Formation of XSTO

The OUR profiles corresponding to the SDS concentrations of 50,

XSTO respiraton

CO2 stripping

75 and 100 mg COD/L at time t = 0 are presented in Fig. 2 with the

Hydrolysis

titrimetric measurements. The OUR increases to a maximum level

Process

due to the consumption of SDS under the feast period, then drops
Table 1

to a level higher than the endogenous OUR level followed by a

gradual declination until it reaches the endogenous level. A similar
5508 V. Aravinthan, M.A. Hoque / Bioresource Technology 102 (2011) 5504–5513

0.7 0.8

SDS = 100 mg COD/L

0.6 OUR 0.7

0.6

Acid/Base added (meq/L)

0.5

OUR (mg O2 /L.min)

0.5
0.4

Base added 0.4

0.3 (proton produced)
Acid added
(proton consumed) 0.3

0.2
0.2

0.1 0.1

0.0 0.0
0 100 200 300 400 500 600 700 800
Time (min)

0.6 0.6

SDS = 75 mg COD/L

0.5 OUR 0.5

Acid/Base added (meq/L)

0.4 0.4
OUR (mg O2/L.min)

0.3 Base added 0.3

(proton produced)
Acid added
(proton consumed)
0.2 0.2

0.1 0.1

0.0 0.0
0 100 200 300 400 500 600 700 800
Time (min)

0.5 0.4
SDS = 50 mg COD/L
OUR
0.4
0.3
Acid/Base added (meq/L)
OUR (mg O2 /L.min)

0.3
Acid added
(proton consumed) 0.2
Base added
0.2
(proton produced)

0.1
0.1

0.0 0.0
0 100 200 300 400 500 600
Time (min)

Fig. 2. OUR with titrimetric profiles for three different SDS concentrations in an activated sludge system (at pH 7.8).

pattern was also observed in the acetate biodegradation study, in et al., 2005; Van Loosdrecht et al., 1997; Van Loosdrecht and
which the consumption of previously stored products occurred, Heijnen, 2002). In case of SDS, the tail part of the OUR is found
resulting in a tail in the OUR profile (Guisasola et al., 2005; Sin to be much more prominent compared to that for the acetate
 V. Aravinthan, M.A. Hoque / Bioresource Technology 102 (2011) 5504–5513 5509

a 0.6

0.5

OURexp

OUR (mg O2/L.min)

0.4
OURmod

0.3

0.2

0.1

0.0
0 100 200 300 400 500 600 700 800
Time (min)

b 0.5

0.4

Hp exp
Hp mod
0.3
Hp (meq/L)

0.2

0.1

0.0
0 100 200 300 400 500 600 700 800
Time (min)

c 0.6 0.6

0.5 0.5

OURexp
OUR (mg O2/L.min)

0.4 OURmod 0.4

Hp exp
Hp (meq/L)

Hp mod
0.3 0.3

0.2 0.2

0.1 0.1

0.0 0.0
0 100 200 300 400 500 600 700 800
Time (min)

Fig. 3. Model calibration using (a) respirometric data alone (b) titrimetric data alone and (c) combined respirometric–titrimetric data (SDS = 75 mg COD/L).

biodegradation study, indicating a significant influence of storage
products in the overall biodegradation process.
Concurrently, SDS biodegradation results in base addition to the
reactor under exogenous state where the cumulative pulses are
proportionally increased with the initial concentration (Fig. 2).
Base addition in the reactor represents the proton production in
the system which is the net result of SDS and ammonia uptake,
endogenous respiration and the CO2 production as described in
Section 2. Fig. 2 depicts that the base addition profile has a steeper
slope until the OUR reaches its maximum level. This is followed by
a mild base pulse rate that represents the oxidization of respective
intermediate products. The CO2 stripping leads to a drop in the
5510 V. Aravinthan, M.A. Hoque / Bioresource Technology 102 (2011) 5504–5513

Table 2
Model calibration results using respirometric data alone for three different concentration studies (confidence intervals are shown in brackets as percentages).

Parameters SDS 100 mg COD/L (Confidence interval, %) SDS 75 mg COD/L (Confidence interval, %) SDS 50 mg COD/L (Confidence interval, %)
Parameters estimated
qMAX (1/min) 0.0131 ± 6.1  104 (4.65) 0.0125 ± 9.2  104 (7.36) 0.0123 ± 2.6  103 (21.13)
kh (1/min) 0.0216 ± 1.1  103 (5.09) 0.0191 ± 9.6  104 (5.03) 0.0182 ± 2.0  103 (10.98)
Ks (mg COD/L) 0.63 ± 0.087 (13.8) 0.62 ± 0.061 (9.84) 0.65 ± 0.115 (17.69)
KX (mg COD Xs/mg COD XH) 0.44 ± 0.049 (11.14) 0.42 ± 0.045 (10.71) 0.39 ± 0.084 (21.54)
YH,S (mg COD XH/mg COD SS) 0.64 ± 0.062 (9.69) 0.64 ± 0.087 (13.59) 0.64 ± 0.203 (31.71)
YSTO (mg COD XSTO/mg COD SS) 0.84 ± 0.33 (39.28) 0.84 ± 0.17 (20.2) 0.84 ± 0.218 (25.95)
YH,STO (mg COD XH/mg COD XSTO) 0.73 ± 0.128 (17.53) 0.73 ± 0.144 (19.73) 0.73 ± 0.075 (10.27)
K1 (mg COD XSTO/mg COD XH) 0.83 ± 0.315 (37.95) 0.7 ± 0.106 (15.14) 0.63 ± 0.151 (23.97)
K2 (mg COD XSTO/mg COD XH) 8.2  105 ± 1.5  104 (182.9) 8.1  106 ± 2.8  105 (345.7) 9.8  107 ± 2.08  106 (212.2)
s (min) 6.51 ± 0.46 (7.07) 11.72 ± 1.05 (8.96) 9.29 ± 1.28 (13.78)
Parameters assumed
bH (1/min) 0.00036 0.00036 0.00036
bSTO (1/min) 0.00036 0.00036 0.00036
kNHacc (1/min) 0.000056 0.000056 0.000056
fSTOb ðmg COD X STO =mg COD Ss Þ 0.65 0.65 0.65
fXI (mg COD/mg COD) 0.2 0.2 0.2
Parameters calculated
kSTO (1/min) 0.007161 0.006857 0.006721
lMAX,S (1/min) 0.002911 0.002812 0.002755
lMAX,STO (1/min) 0.002911 0.002812 0.002755
XH (mg COD/L) 200 200 200
MSEa 2.05  104 7.58  105 1.15  104
a
MSE refers to the mean squared error which is calculated from sum of squared errors divided by number of observations.
b
Parameters were fixed by trials for the better fit of experimental profile with the model.

titrimetric profile to the background proton consumption (acid
addition) rate which was also observed before adding SDS to the
reactor when pH was maintained at 7.8.
4.2. Results and discussion of model calibration
The proposed model was found to be satisfactory in explaining
the experimental OUR and Hp measurements corresponding to
SDS biodegradation in the activated sludge process. The proposed
SSAG model was calibrated with varying initial SDS concentra-
tion. Fig. 3 represents the model calibration results for the SDS
concentration of 75 mg COD/L. The model calibration graphs for
the SDS concentrations of 100 and 50 mg COD/L are not shown
in this paper due to page constraints. Table 2 shows the parame-
ter estimation results and their confidence intervals using on-line
respirometric measurements for three different SDS concentration
studies. Observation shows that the hydrolysis related kinetic
parameters such as kh and KX increase with initial SDS concentra-
tions. The parameter kh is estimated as 27.5 and 31.1 day1 for
the SDS concentrations of 75 and 100 mg COD/L respectively. Be-
sides, the parameter KX shows a relatively higher value (0.44) for
the higher SDS concentration (100 mg COD/L). There is little evi-
dence reported in the literature about SDS biodegradation kinet-
ics. Most of it refers to a first order or simple Monod model
(Zhang et al., 1999; Chen et al., 2001) ignoring the hydrolysis
phase though SDS was found to be hydrolyzed before it under-
went the oxidation process (Yao, 2006). Moreover, they calibrated
the model using off-line substrate depletion measurements that
resulted in an inaccurate parameter estimation process due to
the constraints involved in the collection of frequent bio-kinetic
information from the system. In this study, model calibration
was performed with on-line measurements where the parameter
estimation process gives the hydrolysis rate relatively higher than
ASM3 default values (3.4 day1). Information regarding the kinet-
ics of SDS hydrolysis is very limited in the literature; however
the hydrolysis rate was noticed to be significant (18.5 day1) by
Lopez Zavala et al. (2004) when they investigated the biodegrada-
tion of faeces under aerobic conditions. In addition, Karahan et al.
(2006) estimated the hydrolysis rate as high as 30 day1 for sol-
uble starch biodegradation using sequencing batch reactor in
their experimental study. While Carucci et al. (2001) estimated
the parameters kh (22.9 h1) and KX (12.3) very high for filtered
wastewater biodegradation, contrasting results were found by
Beccari et al. (2002) who observed the respective parameters as
low as 0.0082 h1 and 0.0001 using simultaneous storage and
growth for model calibration.
Parameter estimation shows that the substrate affinity con-
stant, KS lies between 0.62 and 0.65 mg COD/L for all three initial
SDS concentration studies indicating the affinity to be as strong
as observed in the acetate biodegradation study (Hoque et al.,
2010). The calculated maximum biomass growth rate lMAX,S ranges
from 3.97 to 4.19 day1, while a faster storage formation rate (9.7–
10.3 day1) is observed from the model calibration and parameter
estimation process. It is noteworthy that similar sludge behavior
was observed during the acetate biodegradation study where the
calculated parameter kSTO was found to be higher than that for
the parameter lMAX,S (Hoque et al., 2010). However, the parameter
lMAX,S was found to be less (0.67–2.01 day1) for the acetate bio-
degradation. Using SDS as a test substrate, Chen et al. (2001) iden-
tified a high biomass growth rate (8.88 day1), whereas Zhang
et al. (1999) estimated the parameter lMAX,S as 2.76 day1 applying
Monod kinetics in their model. While Anderson et al. (1990)
showed the estimated parameter for lMAX,S to range between
0.67 and 2.01 day1, the growth rate was observed by Marchesi
et al. (1997) to be significantly higher (28.32 day1) when the
exponential growth based model was calibrated with residual
SDS measurements.
From the parameter estimation process, the yield coefficients
YH,S , YSTO and YH,STO are found to be 0.64, 0.84 and 0.73 for all
three SDS concentrations. The literature reports combined yield
coefficient (Y) ranges from 0.34 (Chen et al., 2001) to 0.915 (Zhang
et al., 1999) when a Monod kinetic based model was used for
calibration. The current study also reveals the biomass yield for
storage YSTO higher when compared to other yield coefficients in
the process. A similar observation was noted for the acetate bio-
degradation study where the yield coefficient YSTO (0.88) was
 V. Aravinthan, M.A. Hoque / Bioresource Technology 102 (2011) 5504–5513 5511

Table 3
Model calibration results using titrimetric data alone for three different concentration studies (confidence intervals are shown in brackets as percentages).

Parameters SDS 100 mg COD/L (confidence interval, %) SDS 75 mg COD/L (confidence interval, %) SDS 50 mg COD/L (confidence interval, %)
Parameters estimated
qMAX (1/min) 0.0129 ± 8.3  104 (6.43) 0.0125 ± 1.6  103 (12.8) 0.0123 ± 1.08  103 (8.78)
kh (1/min) 0.0212 ± 3.0  103 (14.15) 0.016 ± 8.6  104 (5.38) 0.0181 ± 3.06  103 (16.9)
Ks (mg COD/L) 0.64 ± 0.037 (5.78) 0.59 ± 0.085 (14.4) 0.63 ± 0.117 (18.57)
KX (mg COD Xs/mg COD XH) 0.44 ± 0.096 (21.82) 0.42 ± 0.074 (17.62) 0.41 ± 0.055 (13.41)
YH,S (mg COD XH/mg COD Ss) 0.64 ± 0.168 (26.25) 0.64 ± 0.114 (17.8) 0.64 ± 0.079 (12.34)
YSTO (mg COD XSTO/mg COD Ss) 0.84 ± 0.089 (10.6) 0.84 ± 0.085 (10.12) 0.84 ± 0.021 (2.5)
YH,STO (mg COD XH/mg COD XSTO) 0.73 ± 0.194 (26.58) 0.73 ± 0.19 (26.02) 0.73 ± 0.046 (6.3)
K1 (mg COD XSTO/mg COD XH) 0.83 ± 0.212 (25.54) 0.7 ± 0.209 (29.86) 0.65 ± 0.059 (9.08)
K2 (mg COD XSTO/mg COD XH) 8.14  105 ± 1.73  104 (208.4) 8.2  106 ± 1.1  105 (134.2) 9.85  107 ± 1.79  106 (181.7)
s (min) 5.01 ± 0.55 (10.98) 8.18 ± 0.68 (8.31) 6.54 ± 0.56 (8.56)
Parameters assumed
bH (1/min) 0.00036 0.00036 0.00036
bSTO (1/min) 0.00036 0.00036 0.00036
kNHacc (1/min) 0.000056 0.000056 0.000056
fSTObðmg CODX STO =mg CODSs Þ 0.65 0.65 0.65
fXI (mg COD/mg COD) 0.2 0.2 0.2
k1 (1/min) 1.0762 1.0762 1.0762
K L aCO2 (1/min) 0.055 0.055 0.055
CT,init (mmol/L) 1.3 1.2 1.3
Parameters calculated
HCO3 (mmol/L) 1.2511 1.1549 1.2511
CO2 (mmol/L) 0.0489 0.0451 0.0489
kSTO (1/min) 0.007064 0.006849 0.006729
lMAX,S (1/min) 0.002867 0.002809 0.002764
lMAX,STO (1/min) 0.002867 0.002809 0.002764
XH (mg COD/L) 200 200 200
MSEa 5.84  105 3.98  105 1.64  105
a
MSE refers to the mean squared error which is calculated from sum of squared errors divided by number of observations.
b
Parameters were fixed by trials for the better fit of experimental profile with the model.

Table 4
Model calibration results using combined respirometric–titrimetric data for three different concentration studies (confidence intervals are shown in brackets as percentages).

Parameters SDS 100 mg COD/L (confidence interval, %) SDS 75 mg COD/L (confidence interval, %) SDS 50 mg COD/L (confidence interval, %)
Parameters estimated
qMAX (1/min) 0.0131 ± 2.86  104 (2.18) 0.0125 ± 4.73  104 (3.78) 0.0123 ± 8.78  104 (7.14)
kh (1/min) 0.0216 ± 9.8  104 (4.54) 0.0189 ± 1.7  103 (8.99) 0.0182 ± 2.35  103 (12.91)
Ks (mg COD/L) 0.63 ± 0.078 (12.38) 0.6 ± 0.087 (14.5) 0.63 ± 0.113 (17.94)
KX (mg COD Xs/mg COD XH) 0.44 ± 0.03 (6.82) 0.42 ± 0.046 (10.95) 0.39 ± 0.075 (19.23)
YH,S (mg COD XH/mg COD Ss) 0.64 ± 0.053 (8.28) 0.64 ± 0.075 (11.72) 0.64 ± 0.057 (8.9)
YSTO (mg COD XSTO/mg COD Ss) 0.84 ± 0.018 (2.14) 0.84 ± 0.024 (2.86) 0.84 ± 0.019 (2.26)
YH,STO (mg COD XH/mg COD XSTO) 0.73 ± 0.039 (5.34) 0.73 ± 0.052 (7.12) 0.73 ± 0.043 (5.89)
K1 (mg COD XSTO/mg COD XH) 0.83 ± 0.034 (4.09) 0.7 ± 0.047 (6.71) 0.63 ± 0.067 (10.64)
K2 (mg COD XSTO/mg COD XH) 8.1  105 ± 3.2  104 8.2  106 ± 2.1  105 9.85  107 ± 1.73  106
s (min) 6.35 ± 0.17 (2.68) 11.7 ± 0.5 (4.27) 9.01 ± 0.66 (7.33)
Parameters assumed
bH (1/min) 0.00036 0.00036 0.00036
bSTO (1/min) 0.00036 0.00036 0.00036
kNHacc (1/min) 0.000056 0.000056 0.000056
fSTObðmg CODX STO =mg CODSs Þ 0.65 0.65 0.65
fXI(mg COD /mg COD) 0.2 0.2 0.2
k1 (1/min) 1.0762 1.0762 1.0762
K L aCO2 (1/min) 0.055 0.055 0.055
CT,init (mmol/L) 1.3 1.2 1.3
Parameters calculated
HCO3 (mmol/L) 1.2511 1.1549 1.2511
CO2 (mmol/L) 0.0489 0.0451 0.0489
kSTO (1/min) 0.007153 0.006849 0.00674
lMAX,S (1/min) 0.002911 0.002809 0.002768
lMAX,STO (1/min) 0.002911 0.002809 0.002768
XH (mg COD/L) 200 200 200
MSEa 8.68  105 7.45  105 8.32  105
a
MSE refers to the mean squared error which is calculated from sum of squared errors divided by number of observations.
b
Parameters were fixed by trials for the better fit of experimental profile with the model.

found to be significant when compared to the yield coefficient YH,S the sludge was collected from the same treatment plant which
(0.71) (Hoque et al., 2010). It confirms the common sludge behav- is subjected to alternating aerobic and anoxic environments lead-
ior showing remarkable storage formation approach during both ing the microorganisms to store the substrate to consume for
the SDS and acetate biodegradation process. The reason is that, growth in the absence of an external source.
5512 V. Aravinthan, M.A. Hoque / Bioresource Technology 102 (2011) 5504–5513
100
80
Surfactant, X S (mg COD/L)
60
40
20
0 0
0 50
Fig. 4. Model validation using off-line measurements for surfactant biodegradation (SDS = 75 mg COD/L).
The proposed SSAG model was successfully calibrated with the
titrimetric measurements (Fig. 3b). The estimated model parame-
ters are presented in tabular form along with their confidence
intervals for checking the estimation accuracy (Table 3). The
estimated model parameters kH (23–30.5 day1), KX (0.41 and
0.44 mg/mg), KS (0.59–0.64 mg COD/L), kSTO (9.69 and 10.17 day1),
lMAX,S (3.98 and 4.13 day1), YH,S (0.64), YSTO (0.84) and YH,STO (0.73)
were found to be very close to the results obtained using respiro-
metric measurements alone (Table 2).
Fig. 3c shows the model calibration outcome based on com-
bined respirometric–titrimetic measurements for the initial SDS
concentration of 75 mg COD/L. Observation shows that the model
profiles are well fitted with both the experimental OUR and Hp
data. The parameter estimation result is presented in Table 4.
The estimated model parameters are consistent with the respec-
tive parameters that were estimated using either respirometric
data alone or titrimetric data alone (see Tables 2–4). The confi-
dence intervals for all the estimated parameters are reasonable ex-
cept that for K2. A similar problem was noticed during the acetate
biodegradation modeling where the SSAG model parameters K1
and K2 were identified as interdependent under the feast phase
of the biodegradation process (Sin et al., 2005).
4.3. Proposed model evaluation
The proposed model has been evaluated using three different
initial surfactant (SDS) concentrations (50, 75 and 100 mg COD/L)
in an activated sludge system. The model has also been calibrated
by using on-line respirometric measurements alone, titrimetric
measurements alone and validated using combined respirometric
titrimetric measurements, illustrating satisfactory calibration re-
sults (Fig. 3). The estimated parameters are shown in tabular form
(Tables 2–4) along with parameter estimation errors calculated for
95% confidence intervals and mean squared error (MSE). The esti-
mated parameters using respirometry alone gives almost the same
values as estimated from the titrimetric data alone, as well as from
the combined approach, thereby confirming the accuracy of the
proposed model. The calculated confidence intervals and mean
squared error are acceptable from a statistical perspective. Along
with on-line measurements, the proposed model was validated fol-
lowing off-line methods where the soluble COD and ammonium in
the liquid medium were monitored during the oxidation period.
Fig. 4 shows the time profile study with model simulation that also
validates the proposed SSAG model.
The model consists of the kinetic parameter kNHacc that was
fixed to 0.08 day1 for better model calibration. In the proposed
10
8
Ammonium, S NH (mg N/L)
6
Xs model
4
Xs measured
SNH model
SNH measured 2
100 150 200 250 300
Time (min)
SSAG model, polyhydroxybutyrate (PHB) was assumed to be stored
in the biomass cell during the process. Contrary to several reports
in the literature that indicate the formation of PHB during acetate
biodegradation, there was no strong result to show what kind of
storage products are formed during SDS biodegradation. However
the lauric acid synthesis generated acetyl group compound that
was assumed to form PHB for storage in the biomass cell as ob-
served in the acetate biodegradation process. The degree of reduc-
tion of storage products, cSTO was calculated as 4.5 by considering
the formula as CH1.5O0.5. The iNBM content corresponding to the
biomass composition CH1.8O0.5N0.2 was calculated as 0.083 g N/g
COD XH which lies within the range of 7–8.6% reported by Henze
et al. (2000) as typical value for the parameter iNBM.
5. Conclusions
The SSAG model was improved to interpret the SDS biodegrada-
tion process in aerobic activated sludge systems by introducing the
hydrolysis process and non-linear carbon dioxide transfer rate in
the liquid phase and considering the relevant stoichiometric
parameters in each step of the model. The proposed SSAG model
was successfully calibrated for three different SDS concentrations
using respirometric, titrimetric and combined respirometric–titri-
metric measurement approaches. The parameter estimation re-
sults from all three stated combinations were statistically
evaluated and found to be very close validating the model. Further-
more, off-line measurements of COD and ammonium concentra-
tions confirm the accuracy and validity of the model.
Acknowledgements
The authors would like to thank the Faculty of Engineering and
Surveying (FoES), University of Southern Queensland (USQ) and
Australian Centre for Sustainable Catchments of University of
Southern Queensland for the laboratory and financial supports.
References
Ahmed, E.H., Raghavendra, T., Madamwar, D., 2010. An alkaline lipase from organic
solvent tolerant Acinetobacter sp. EH28: application for ethyl caprylate
synthesis. Bioresource Technology 101 (10), 3628–3634.
Anderson, D.J., Day, M.J., Russell, N.J., White, G.F., 1990. Die-away kinetic analysis of
the capacity of epilithic and planktonic bacteria from clean and polluted river
water to biodegrade sodium dodecyl sulfate. Applied and Environmental
Microbiology 56 (3), 758–763.
ASCE, 1996. Standard Guidelines for In-Process Oxygen Transfer Testing. American
Society of Civil Engineers, ASCE-18-96, p. 48.
 V. Aravinthan, M.A. Hoque / Bioresource Technology 102 (2011) 5504–5513
Beccari, M., Dionisi, D., Giuliani, A., Majone, M., Ramadori, R., 2002. Effect of
different carbon sources on aerobic storage by activated sludge. Water Science
and Technology 45 (6), 157–168.
Carucci, A., Dionisi, D., Majone, M., Rolle, E., Smurra, P., 2001. Aerobic storage by
activated sludge on real wastewater. Water Research 35 (16), 3833–3844.
Carvalho, G., Nopens, I., Novais, J.M., Vanrolleghem, P.A., Pinheiro, H.M., 2001.
Modelling of activated sludge acclimisation to a non-ionic surfactant. Water
Science and Technology 43 (7), 9–17.
Chen, G., Strevett, K.A., Vanegas, B.A., 2001. Naphthalene, phenanthrene and
surfactant biodegradation. Biodegradation 12, 433–442.
Chen, H.J., Tseng, D.H., Huang, S.L., 2005. Biodegradation of octylphenol
polyethoxylate surfactant Triton X-100 by selected microorganisms.
Bioresource Technology 96 (13), 1483–1491.
Gernaey, K., Vanrolleghem, P., Verstraete, W., 1998. On-line estimation of
Nitrosomonas kinetic parameters in activated sludge samples using titration
in-sensor-experiments. Water Research 32 (1), 71–80.
Gernaey, K., Petersen, B., Nopens, I., Comeau, Y., Vanrolleghem, P.A., 2002a.
Modeling aerobic carbon source degradation processes using titrimetric data
and combined respirometric–titrimetric data: experimental data and model
structure. Biotechnology and Bioengineering 79 (7), 741–753.
Gernaey, K., Petersen, B., Dochain, D., Vanrolleghem, P.A., 2002b. Modeling aerobic
carbon source degradation processes using titrimetric data and combined
respirometric–titrimetric data: structural and practical identifiability.
Biotechnology and Bioengineering 79 (7), 754–767.
Gernaey, K., Petersen, B., Ottoy, J.-P., Vanrolleghem, P.A., 2001. Activated sludge
monitoring with combined respirometric–titrimetric measurements. Water
Research 35 (5), 1280–1294.
Guisasola, A., Sin, G., Baeza, J.A., Carrera, J., Vanrolleghem, P.A., 2005. Limitations of
ASM1 and ASM3: a comparison based on batch oxygen uptake rate profiles from
different full-scale wastewater treatment plants. Water Science and Technology
52 (10), 69–77.
Henze, M., Grady Jr., C.P.L., Gujer, W., Marais, G., Matsuo, T., 2000. Activated Sludge
Models: ASM1, ASM2, ASM2d and ASM3. IAW Scientific and Technical Report
No. 9, London, UK.
Hoque, M.A., Aravinthan, V., Porter, M., 2008. Respirometric and titrimetric
techniques for monitoring aerobic biodegradation of surfactant. Research
Journal of Biotechnology, 399–405 (special issue 2008).
Hoque, M.A., Aravinthan, V., Pradhan, N.M., 2010. Calibration of biokinetic model for
acetate biodegradation using combined respirometric and titrimetric
measurements. Bioresource Technology 101 (5), 1426–1434.
Hotantai, Louis, Nardello-Rataj, Veronique, 2001. The main surfactants used in
detergents and personal care products. OCL 8 (2), 141–144.
Hrenovic, J., Ivankovic, T., 2007. Toxicity of anionic and cationic surfactant to
Acinetobacter junii in pure culture. Central European Journal of Biology 2 (3),
405–414.
Huber, M., Meyer, U., Rys, P., 2000. Biodegradation mechanisms of linear alcohol
ethoxylates under anaerobic conditions. Environmental Science and Technology
34 (9), 1737–1741.
Jahan, K., 2005. Detergents, Water Encyclopedia, Part. Waste Water Treatment. John
Wiley & Sons Inc., New Jersey.
Karahan, O., Van Loosdrecht, M.C.M., Orhon, D., 2006. Modeling the utilization of
starch by activated sludge for simultaneous substrate storage and microbial
growth. Biotechnology and Bioengineering 94 (1), 43–53.
Kolpin, D.W., Furlong, E.T., Meyer, M.T., Thurman, E.M., Zaugg, S.D., Barber, L.B.,
et al., 2002. Pharmaceuticals, hormones, and other organic wastewater
contaminants in US streams, 1999–2000: a national reconnaissance.
Environmental Science & Technology 36 (6), 1202–1211.
5513
Lara-Martin, P.A., Petrovic, M., Gomez-Parra, A., Barcelo, D., Gonzalez-Mazo, E., 2006.
Presence of surfactants and their degradation intermediates in sediment cores
and grabs from the Cadiz Bay area. Environmental Pollution 144 (2), 483–491.
Lopez Zavala, M.A., Funamizu, N., Takakuwa, T., 2004. Modeling of aerobic
biodegradation of feces using sawdust as a matrix. Water Research 38 (5),
1327–1339.
Marchesi, J.R., White, G.F., Russell, N.J., House, W.A., 1997. Effect of river sediment
on the biodegradation kinetics of surfactant and non-surfactant compounds.
FEMS Microbiology Ecology 23, 55–63.
Mohan, P.K., Nakhla, G., Yanful, E.K., 2006. Biokinetics of biodegradation of
surfactants under aerobic, anoxic and anaerobic conditions. Water Research
40 (3), 533–540.
Petersen, B., Gernaey, K., Vanrolleghem, P.A., 2001. Practical identifiability of model
parameters by combined respirometric–titrimetric measurements. Water
Science and Technology 43 (7), 347–356.
Pratt, S., Yuan, Z., Keller, J., 2004. Modelling aerobic carbon oxidation and storage by
integrating respirometric, titrimetric, and off-gas CO2 measurements.
Biotechnology and Bioengineering 88 (2), 135–147.
Qin, Y., Zhang, G.Y., Kang, B.A., Zhao, Y.M., 2005. Primary aerobic biodegradation of
cationic and amphoteric surfactants. Journal of Surfactants and Detergents 8
(1), 55–58.
Sharvelle, S., Lattyak, R., Banks, M.K., 2007. Evaluation of biodegradability and
biodegradation kinetics for anionic, nonionic, and amphoteric surfactants.
Water Air and Soil Pollution 183 (1–4), 177–186.
Sin, G., 2004. Systematic Calibration of Activated Sludge Models. Ph.D. Thesis.
Faculty of Agricultural and Applied Biological Sciences, Ghent University,
372pp.
Sin, G., Guisasola, A., DePauw, D.J.W., Juan, A.B., Carrera, J., Vanrolleghem, P.A., 2005.
A new approach for modelling simultaneous storage and growth processes for
activated sludge systems under aerobic conditions. Biotechnology and
Bioengineering 92 (5), 600–613.
Sin, G., Vanrolleghem, P.A., 2007. Extensions to modeling aerobic carbon
degradation using combined respirometric–titrimetric measurements in
view of activated sludge model calibration. Water Research 41, 3345–
3358.
Sperandio, M., Paul, E., 1997. Determination of carbon dioxide evolution rate using
on-line gas analysis during dynamic biodegradation experiments.
Biotechnology and Bioengineering 53, 243–252.
Stumm, W., Morgan, J.J., 1996. Aquatic Chemistry: Chemical Equilibria and Rates in
Natural Waters, third ed. Wiley, New York.
Thanomsub, B., Pumeechockchai, W., Limtrakul, A., Arunrattiyakorn, P., Petchleelaha,
W., Nitoda, T., Kanzaki, H., 2006. Chemical structures and biological activities of
rhamnolipids produced by Pseudomonas aeruginosa B189 isolated from milk
factory waste. Bioresource Technology 97 (18), 2457–2461.
Van Loosdrecht, M.C.M., Heijnen, J.J., 2002. Modelling of activated sludge
processes with structured biomass. Water Science and Technology 45 (6),
12–23.
Van Loosdrecht, M.C.M., Pot, M.A., Heijnen, J.J., 1997. Importance of bacterial storage
polymers in bioprocesses. Water Science and Technology 35 (1), 41–47.
Vanrolleghem, P.A., Sin, G., Geraney, K.V., 2004. Transient response of aerobic and
anoxic activated sludge activities to sudden substrate concentration changes.
Biotechnology and Bioengineering 86 (3), 277–290.
Yao, G., 2006. Dodecyl Sulfate Pathway Map. University of Minnesota, viewed 28
November 2007, <http://umbbd.msi.umn.edu/dds/dds_map.html>.
Zhang, C.L., Valsaraj, K.T., Constant, W.D., Roy, D., 1999. Aerobic biodegradation
kinetics of four anionic and nonionic surfactants at sub- and supra-critical
micelle concentrations (CMCs). Water Research 33 (1), 115–124.