Evaluation of wastewater characterization methods

Water Science & Technology Vol 52 No 10– 11 pp 61–68 Q IWA Publishing 2006
E.-H. Choi*, B. Klapwijk**, A. Mels*** and H. Brouwer****
*Department of Environmental Engineering, Pusan National University, South-Korea
(E-mail: choiieh@pusan.ac.kr)
**Sub-department of Environmental Technology, Wageningen, the Netherlands
(E-mail: bram.klapwijk@wur.nl)
***LeAF (Lettinga Associates Foundation), Wageningen, the Netherlands
(E-mail: Adriaan.Mels@wur.nl)
****TNO-MEP, Apeldoorn, the Netherlands
(E-mail: H.H.Brouwer@mep.tno.nl)

Abstract Wastewater contains various organic components with different physical and biochemical
characteristics. ASM No. 1 distinguishes two categories of biodegradable organic matter in wastewater,
rapidly and slowly biodegradable. In general there are two methods for wastewater characterization: based
on filtration in combination with a long-term BOD test or based on a respirogram. By comparing both
approaches, we showed that in wastewater three categories of organic compounds with different
biodegradation rates can be distinguished. These categories are referred to as readily biodegradable, rapidly
hydrolysable and slowly hydrolysable organic matter. The total biodegradable COD can be found from a
long-term BOD-test combined with a curve-fit and the readily biodegradable and rapidly hydrolysable from a
respirogram. The slowly hydrolysable is the difference between total biodegradable COD and the sum of
readily biodegradable and rapidly hydrolysable COD. Simulation with characterization based on filtration for
a pre-anoxic reactor with a certain N-removal compared with the N-removal of the same plant with
wastewater according to the modified characterization shows different results of each wastewater,
especially with regard to the effluent nitrate concentration.
Keywords activated sludge model; wastewater characterization; biodegradability; respirogram

Introduction
Wastewater contains various organic components with different physical and biochemical
characteristics. In ASM No 1 (Henze et al., 1987), the organic matter is subdivided in
four categories: 1) readily biodegradable (soluble), 2) soluble inert, 3) slowly biodegrad-
able (particulate) and 4) particulate inert. In the Netherlands a method has been devel-
oped by STOWA (Dutch foundation for applied water research) to standardize the
characterisation of wastewater (STOWA, 1996; Roeleveld and Loosdrecht, 2002). The
COD is divided into a soluble and particulate fraction by filtration over a membrane filter
with a pore size of 0.1mm. The soluble influent COD (CODinf,sol) is subdivided into bio-
degradable and inert by assuming that inert COD can be found from the effluent COD of
low loaded activated sludge plant (CODeff,sol). The total biodegradable COD (BCOD) is
based on a long-term BOD-test combined with a curve-fit. In Table 1 the calculations
necessary to find the concentrations of the four organics components are presented.
Another approach to characterize a wastewater with reference to the biodegradable
organic components is the Respirogram-method (Ekama, 1986; Brouwer et al., 1998). In
this method a respirogram (Figure 1) is used to calculate the concentration readily biode-
gradable COD (SS) and slowly biodegradable COD (XS). A peak in such a respirogram
presents the readily biodegradable COD (Ss) while the slow decrease of the respiration
rate represents the slowly biodegradable COD (XS). 61
 Table 1 Translation of COD and BOD to components in ASM No.1 components (STOWA)

Components Name Calculation Remark

SI Soluble inert organic material 0.9*CODeff,sol Low loaded plant,

high efficiency
0.9*CODeff,sol – 1.5* BODeff High loaded plant,
low efficiency
SS Readily biodegradable substrate CODinf,sol-Sl
E.-H. Choi et al.

XS Slowly biodegradable substrate BCOD-SS

XI Particulate inert organic material CODinf,tot-Sl-Ss-Xs

The aim of this article is to prove that the biodegradable organic components in munici-
pal wastewater should be subdivided into three categories (readily biodegradable, rapidly
hydrolysable and slowly hydrolysable) and to propose a Modified Respirogram-method for
estimating the concentrations of these three categories of organic components in waste-
water. By simulating the performance of a pre-anoxic wastewater treatment plant we com-
pared the effect of the STOWA- method of wastewater (2 categories of biodegradable
organic matter) with the Modified Respirogram-method (3 categories of biodegradable
organic matter).

Materials and methods

BCOD measurement
An OXITOPw pressure system was used to measure the cumulative biochemical oxygen
consumption of wastewater over a time period of 25 days. The total BOD (BODtotal) and
the rate constant kBOD were estimated by using the following equation:

BODt ¼ ð1 2 e2kBOD
t ÞBODtotal ð1Þ

During BOD measurements growth and decay of biomass takes place. During the test a
part of the BCOD is converted into inert bacterial residuals, resulting in an underestima-
tion of the actual BCOD. This can be corrected for by using equation 2. The factor fBOD
has a value of 0.15 according to Roeleveld & Van Loosdrecht (2002)
1
BCOD ¼ BODtotal ð2Þ
1 2 fBOD

STOWA methods for estimating the organic components

CODtotal and CODsoluble (after filtration with 0.1 mm NC 10 membrane filter, Schleicher
& Schuell) were determined using a spectrophotometer (Lange Group, Düsseldorf,

62 Figure 1 Respirogram of wastewater

Germany). The calculation of the four organic fractions is explained in Table 1. Waste-
water samples of Emmen and Venray were used for the characterisation.

Respirogram method for estimating the organic components

The experimental set-up consisted of a batch-vessel (3 l), which was attached to a con-
tinuous respirometer (RA-1000; Manotherm) (Spanjers, and Klapwijk, 1990). A known
volume of pre-aerated wastewater is added to endogenous activated sludge in the batch-

E.-H. Choi et al.

vessel (2 l). Allythiourea (ATU) (20 mg per l activated sludge) is added to suppress nitrifi-
cation. Subsequent to the addition of wastewater the respiration rate increases due to the
oxidation of organic compounds. When the respiration rate has returned to the initial rate,
i.e. the rate before wastewater addition (taking dilution into account), the respirogram test
is complete. In the tests reported in this article the pH, temperature and the S/X-ratio
were 7.41, 20.08C, and 0.07, respectively. The cumulative oxygen consumption under the
peak was used to calculate the concentration of readily biodegradable COD by dividing
by (1-Y). For Y we assumed the default value 0.67 of ASM No.1. The consumption of
oxygen during depletion of slowly biodegradable COD is found as the total cumulative
oxygen consumption minus the oxygen consumption in the peak minus the cumulative
endogenous oxygen consumption. The concentration of slowly biodegradable COD is
found by dividing that value by 1-Y. Influent wastewater of the municipal wastewater
treatment plant (WTP) Emmen and Venray were used for the characterisation.

Validation with GPS-X

The validation has been done for STOWA- and Modified Respirogram-method results by
GPS-X simulator (Hydromantis, 1999), using ASM No. 1. Kinetic and stoichiometric par-
ameters for simulation were based on Henze et al. (1987).

Simulation of Venray wastewater treatment plant

Influent of WTP Venray was characterized by each method (STOWA- and Modified
Respirogram-method) resulting in two different wastewater characterisations. For these
characterisations a simulation was performed to see the effect of those two types of
‘wastewater’ on the performances of a treatment plant. The Venray WTP (Netherlands)
was selected as model plant. This system operates as an Anoxic-Aerobic (A/O) process.
The influent flow (Q) is 18,950 m3/d and the return sludge flow rate and the interior
return flow rate from the aerobic to anoxic tanks are operated at 2Q and 1Q, respectively.
Sludge is wasted on a secondary clarifier of 463.5 m3. The anoxic and the aeration tank
have the same volume, i.e. 9,475 m3. GPS-X was used as simulator with ASM No.1 and
Two-step mantis model.

Results and discussions

Emmen wastewater
Figure 2 shows the concentrations of biodegradable organic matter in wastewater of
Emmen characterized by the STOWA- and the Respirogram-method. The STOWA-
method assesses a much higher concentration of SS compared to the other characteriz-
ation (Figure 2a). The estimation of SS by the Respirogram-method is quite accurate as it
is based on the real oxygen consumption. The STOWA-method therefore overestimates
the concentration of SS. The concentrations of XS are also different for both methods
(Figure 2b). Two times the STOWA-method gives a higher value, one time the Respiro-
gram-method gave a higher value and one time the results were almost the same.
Figure 2c shows that the sum of SS and XS (total BCOD based on the BOD-test) in the
STOWA-method is much higher than the sum of SS and XS in the Respirogram-method. 63
E.-H. Choi et al.

Figure 2 Concentrations of biodegradable organic components in a wastewater sample of WTP Emmen;

comparison of the STOWA- with the Respirogram-method

We concluded that the difference is the result of a biodegradable component that is not
visible in a respirogram from activated sludge with a low S/X ratio. Only a component
with a low hydrolysis rate constant is not visible in such a short-term respirogram. Based
on these observations we concluded that there are three categories of biodegradable
organic matter in wastewater and that could be defined as: readily biodegradable, rapidly
hydrolysable and slowly hydrolysable organic matter (Sollfrank & Gujer, 1991). Because
of the slow hydrolysis rate, the slowly hydrolysable fraction can only be detected in a
long-term respirogram test like a BOD-test.

Venray wastewater
In the foregoing experiments the BCOD of the Emmen wastewater was based on a
BOD5-test and by using a default value for kBOD the BCOD was estimated. To get more
accurate data on the BCOD of wastewater compared with the biodegradable matter esti-
mated by the Respirogram-method we performed a BOD-test over 25 days with
OXITOPw with the Venray wastewater. Figure 3 shows a fitted BOD-curve for the deter-
mination of kBOD, BODtot and BCOD. A simulation with a kBOD of 0.44 d21 matched the
measured data. This value is in the range reported by Roeleveld and Loosdrecht (2002).
BCOD can be calculated by using 0.15 for fBOD as by Roeleveld and Loosdrecht (2002).
The concentration of BODtot and BCOD was 339 mg/l and 398 mg/l, respectively. The
result of the wastewater characterisation with the STOWA-method is shown in Table 2.
The concentrations of readily biodegradable and slowly biodegradable were also
measured with the Respirogram-method. The concentration of SS was 72 mg/l according
to the Respirogram-method and the concentration of Xs 49 mg/l. Figure 4 shows the dia-
gram of each categories concentration comparing the STOWA- with the Respirogram-
method. The BCOD was 75% of total COD and 25% was estimated as inert. With the
STOWA-method, 21% of BCOD was estimated to be SS but the Respirogram-method
64 resulted in 13%. For the Venray wastewater a big discrepancy between both methods was
 E.-H. Choi et al.
M S

Figure 3 Fitted BOD curve for the determination of kBOD, BODtot and BCOD

Table 2 Components in wastewater according to the STOWA-method

Component CODtot CODdis CODeff.sol BOD5 BCOD SI Ss Xs XI

[mg/l] [mg/l] [mg/l] [mg/l] [mg/l] [mg/l] [mg/l] [mg/l] [mg/l]

Concentration 534 139 312 293 398 28 111 288 107

found for XS. In the Respirogram-method 53% of the BCOD is an unknown biodegrad-
able component.
It is suggested to combine the STOWA- and the Respirogram-method in the so-called
Modified Respirogram-method. The concentration of the total biodegradable COD is
found from the long-term BOD-test combined with a curve-fit. The total biodegradable
COD is subdivided into three categories: readily biodegradable, rapidly hydrolysable and
slowly hydrolysable. The readily biodegradable COD and the rapidly hydrolysable COD
is measured with a respirogram, while the slowly biodegradable is the difference
between total biodegradable COD and the sum of readily biodegradable and rapidly
hydrolysable COD.

Validation with GPS-X

The data that were collected with both characterisation methods were used to simulate a
respirogram (Figure 5). The simulation with the data of the Respirogram-method is, as
expected, almost similar as the measured respirogram, whereas the simulation with the

Figure 4 Diagram of each category’s concentration comparison of STOWA- with Respirogram-method 65

E.-H. Choi et al.

Figure 5 Validation of Xs of WTP Venray using the GPS-X simulator; dotted for the STOWA -method and
lined for the Respirogram-method

STOWA-method is rather different. The oxidation of SS takes somewhat more time and
the oxygen uptake rate subsequent to the peak is higher than measured as a result of the
higher XS concentration. This figure again illustrates that the STOWA-method results in
higher values for the sum of SS and XS compared to the Respirogram-method.
So, what was called slowly biodegradable (XS) in the Respirogram-method is in the
Modified Respirogram-method defined as rapidly hydrolysable organic matter (SSH)
while the unknown fraction of biodegradable organic matter is defined as slowly
hydrolysable organic matter (XS). It is assumed that XS is hydrolyzed into SSH and after
that into Ss.
IWA ASM No.1 gives a default value for the maximum specific hydrolysis rate of
slowly biodegradable organic matter of 3 d21. The Two-step mantis model (Hydromantis,
1999) recommends a default value of 25 d21 for the maximum specific hydrolysis rate of
SSH and 2.5 d21 for Xs. We estimated khr (hydrolysis rate for rapidly hydrolysable) and
kh (for slowly hydrolysable) by calibration (Figure 6) as 3.7 d21 and 0.6 d21, respect-
ively. The estimated rates are not accurate, because they are based on one example. How-
ever they provide some indication with respect to these constants.
Respiration rate[mg O2/l.h]

M S S

Figure 6 Calibrated hydrolysis rates for rapidly and slowly hydrolysable substrate of the wastewater sample
66 of WTP Venray
Table 3 Wastewater composition (WW1: wastewater 1; WW2: wastewater 2)

SI SS SSH XS XI

WW1 (STOWA-method) 28 111 2 288 107

WW1 (Modified Respirogram-method) 28 72 49 278 107
WW2 (STOWA-method) 38 178 2 32.5 89
WW2 (Modified Respirogram-method) 38 27 67 117 89

E.-H. Choi et al.

Simulation of Venray WTP
We simulated a wastewater treatment plant (pre-anoxic) for two samples of wastewater
characterised by the two methods (Table 3).
In wastewater 1 the concentration of Ss is a little bit lower in the Modified Respiro-
gram-method. In wastewater 2 the concentration of Ss is much lower than in the
STOWA-method.
ASM No. 1 was used in the simulation for the wastewater characterised by the
STOWA-method while the two-step Mantis model of GPS-X was used for the other
characterisation.
Figure 7 shows some results for both simulations. In the case of wastewater 1 the
calculated NO3 – N concentration in the anoxic tank and the effluent is almost the
same in both of STOWA- and Modified Respirogram-method. In the case of
wastewater 2, the concentration in the anoxic tank is significantly lower with the
STOWA-method, because more readily biodegradable substrate is available for the
denitrification process. The nitrate concentration is in the STOWA-method too low
because Ss is overestimated as was already shown earlier. Figure 8 presents the results
of the COD in the effluent. As can be seen there is almost no difference between all
types of characterisation.
As is shown already the Modified Respirogram-method gives a better characteris-
ation of the wastewater. In this simulation we showed that there is a clear difference
with respect to N-removal using the STOWA- or the Modified Respirogram-method,
whereas the COD removal looks the same for both characterisations. Therefore we
suggest using the Modified Respirogram-Method especially for cases where N-removal
is simulated.

Figure 7 Profile of NO3-N concentrations comparing characterization with two and three categories
biodegradable organic matter (Sno_ano ¼ NO3-N in anoxic tank; Sno_eff ¼ NO3-N in effluent; DNR ¼
Denitrification rate in anoxic tank) 67
E.-H. Choi et al.

Figure 8 Concentrations of COD and organic compounds in effluent

Conclusions
In this study, we showed that in wastewater three categories of organic matter with differ-
ent biodegradation rates can be distinguished which are referred to as readily biodegrad-
able, rapidly hydrolysable and slowly hydrolysable organic matter. The total
biodegradable COD can be found from a continuous BOD-test over 20 days combined
with a curve fit. It is suggested to combine the STOWA- and Respirogram-method in the
so-called Modified Respirogram-method. The concentrations of readily biodegradable and
rapidly hydrolysable are found with the Respirogram-method, while the slowly hydroly-
sable is the difference between total biodegradable COD and the sum of readily biode-
gradable and rapidly hydrolysable COD. Simulation results show that the prediction of
effluent NO3-N depends on the characterization method used.

References
Brouwer, H., Klapwijk, A. and Keesman, K.J. (1998). Identification of activated sludge and wastewater
characteristics using respirometric batch-experiments. Wat. Res., 32(4), 1240– 1254.
Ekama, G.A., Dold, P.L. and Marais, G.v.r. (1986). Procedure for determining influent COD fractions and the
maximum specific growth rate of heterotrophs in activated sludge process. Wat. Sci. Tech., 18(6),
91 – 114.
Henze, M., Grady, C.P.L., Jr., Gujer, W., Marais, G.v.R. and Matsuo, T. (1987). Activated sludge model
No. 1. In IAWPRC, Sci. and Technol. Report No. 1, IAWPRC, London.
Hydromantis (1999). Technical Reference, Version 4.0, Hydromantis, Hamilton.
Roeleveld, P.J. and Loosdrecht, M.C.M.van (2002). Experience with guidelines for wastewater
characterization in the Netherlands. Wat. Sci. Tech., 45(6), 77 – 87.
Sollfrank, U. and Gujer, W. (1991). Characterization of domestic wastewater for mathematical modelling of
the activated sludge process. Wat. Sci. Tech., 23(4 –6), 1057– 1066.
Spanjers, H. and Klapwijk, A. (1990). On-line meter for respiration rate and short-term biochemical oxygen
demand in the control of the activated sludge process. Advances in Water Pollution Control, Proc. 5th
IAWPRC Workshop, 26 July 3 August, Yokohama and Kyoto, Japan.
STOWA (1996). Methoden van influentkarakterisering. Inventarisatie en richtlijnen(Dutch). STOWA report
number 1996-08, Utrecht. The Netherlands.

68