See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/8685730

Modeling of the acclimation/deacclimation process of a mixed

culture degrading 4-chlorophenol

Article  in  Water Science & Technology · February 2004

DOI: 10.2166/wst.2004.0024 · Source: PubMed

CITATIONS READS

25 77

2 authors:

Germán Buitrón Jaime A Moreno

Universidad Nacional Autónoma de México Universidad Nacional Autónoma de México
200 PUBLICATIONS   2,217 CITATIONS    302 PUBLICATIONS   4,784 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Generation of gaseous biofuels from wastes View project

Continuous Higher Order Sliding Modes View project

All content following this page was uploaded by Jaime A Moreno on 06 March 2015.

The user has requested enhancement of the downloaded file.

Modeling of the acclimation/deacclimation process of a

Water Science and Technology Vol 49 No 1 pp 79–86 © IWA Publishing 2004

mixed culture degrading 4-chlorophenol
G. Buitrón* and J. Moreno**
* Environmental Bioprocesses Department, Institute of Engineering, National University of Mexico, Ap. Potal
70–472, 04510 Mexico DF, Mexico (E-mail: gbm@pumas.iingen.unam.mx)
** Automatic Control Department, Institute of Engineering, National University of Mexico, Ap. Postal 70-472,
04510 Mexico, D.F., Mexico

Abstract The variation of concentration of the toxic substrate has a negative effect on the sludge activity.
Although this loss of activity of the microorganisms under toxic starvation has been reported, this variable is
not taken into account in the operation of biological wastewater treatment plants. In order to monitor and
control the biological plant with a certain confidence it is necessary to consider the acclimation and
deacclimation processes to avoid the reactor malfunctioning, resulting in loss of efficiency or in false data
acquisition during monitoring. This paper proposes a model for the acclimation/deacclimation process. The
experiments have been done in a sequencing batch reactor during the 4-chlorophenol degradation. The
basic idea of the model is to consider the variation of the kinetic parameters of the Haldane law in terms of
the acclimation and deacclimation state. The idea is discussed, and the results obtained are presented.
Keywords Acclimation; 4-chlorophenol; deacclimation modeling; SBR process; toxic compounds

Introduction
Many of the industrial processes generating wastewater containing toxic compounds are
characterized by their variability. In the chemical, pharmaceutical, plastic and petro-
chemical industries certain production processes are in batch. The first step to biodegrade
toxic substances in a wastewater treatment plant (WWTP) is the acclimation of the
microorganisms. Nevertheless, it has been shown that this acclimation is not permanent
(Arbuckle and Kennedy, 1989; Senthilnathan and Ganczarczyk, 1989; Babcock et al.,
1992). In effect, for instance, when the containers and reactors are cleaned, a peak in the
concentration of toxic substances (shock load) is found in the wastewater treatment plant.
After this sudden increase, there is a decrease of the toxic concentration. These variations
of the toxic substrate or starvation in relation to the toxic compound have an important
effect on the sludge activity (Babcock et al., 1992).
Buitrón et al. (1994) discussed the negative effect of starvation periods on 4-chlorophe-
nol degradation for an activated sludge sequencing batch reactor (SBR) system. In that
study, the aeration was prolonged 20 to 23 hours after the toxic degradation had been com-
pleted. It was found that the activity of the acclimated population decreased drastically as a
result of starvation. Degradation time for 100 mg of 4-chlorophenol/l increased 6 times
(from 0.7 to 4.5 h). The loss in the microbial capacity was attributed to a decline in both the
enzymatic activity and the viability of the suspended cells. Although this loss of activity of
the microorganisms under toxic starvation has been reported, this variable is not taken into
account on the operation of biological wastewater treatment plants. In order to monitor and
control the biological WWTP with a certain confidence it is necessary to consider the accli-
mation and deacclimation processes to avoid the reactor malfunctioning, resulting in loss
of efficiency or in false data acquisition during monitoring. The objective of this work is to
propose a model for the acclimation/deacclimation process. The experiments have been
done in a sequencing batch reactor using 4-chlorophenol, 4CP, as model compound. This 79
 model will be used in the monitoring and control of that system in a future work. The basic
idea of the model is to consider the variation of the kinetic parameters of the Haldane law in
terms of the acclimation and deacclimation state.

Methodology
The study was divided into two sets of experiments. First, the acclimation of the biomass to
4CP was studied. In the second set, the de-acclimation of the previously acclimated bio-
G. Buitrón and J. Moreno

mass was investigated. The inoculum was obtained from a municipal WWTP. An initial
concentration of 2,000 mgVSS/L was used. For the study, 4CP was used as the sole source
of carbon and energy. The microorganisms were adapted to 4CP at three concentrations
(50, 100 and 200 mg/L). Reaction medium was complemented with nitrogen, phosphates
and oligoelements (AFNOR, 1985). Three temperatures were used (10, 20 and 25°C). For
the acclimation part, only the results obtained with 4CP at 20ºC are discussed.
The study was conducted using SBR of 7 L of useful volume. The toxic compound was
put in contact with the microorganisms and let the necessary time to obtain a previously
fixed removal efficiency (≥ 99%). Once the substrate was degraded the supernatant was
discharged, and more substrate was added. This procedure was repeated until the degrada-
tion time was around two hours. At this moment the reactor was controlled with a timer.
The degradation kinetics was followed only for the 4CP during acclimation. Acclimation
started with an initial concentration of 50 mg/L. The initial concentration was doubled
when degradation times became constant.
To study the influence of the deacclimation, the microorganisms were subjected to star-
vation periods of the substrate of 4, 8, 12, 24 and 36 hours. During starvation no source of
carbon and energy was added, this was done by extending the aeration period the necessary
time. In order to start at the same condition the previously acclimated biomass was split into
5 reactors (2 L). Before starvation, a kinetic curve was obtained. After the degradation, the
starvation period was computed and when the period established was reached, a new degra-
dation curve was followed in each reactor to determine the effect of starvation.
4CP concentration was monitored using the colorimetric technique of 4-aminoan-
tipyrine (Standard Methods, 1992). Total organic carbon was measured through a
Shimadzu 5050 device. Growth yield constant, Y, was evaluated when the biomass was
acclimated and the degradation time was constant.

Kinetic parameter determination

The evolution of the substrate in a bioreactor can be described by the standard mass balance
law
dS 1
= − µ(S) X (1)
dt Y
µ0 S
µ(S) = (2)
Ks + S + S 2 / K I
where X is the biomass concentration, S the substrate concentration, Y is the yield coeffi-
cient, and µ(S) is the biomass growth rate. Since during a cycle the change of X is very
small, it will be considered as a constant.
For inhibitory substances the biomass growth rate µ(S), as a function of the substrate
concentration S, is usually described by the Haldane model (Eq. (2)), that is illustrated in
Figure 1. In this work the effect of the acclimation/deacclimation processes will be mod-
eled through changes in the values of the parameters of the Haldane law. For this purpose it
is advantageous to introduce a reparametrization of the Haldane law (Eq. (2)). Three new
80 parameters are used, that characterize completely the Haldane law. These are (see Figure 1)
µ*, S* and Sm > S*. The maximal growth rate, µ*, is reached at the critical substrate
concentration, S*. Beyond S* the compound becomes inhibitory to the biomass. Sm is the
highest value of the substrate concentration for which the growth rate is half of the maxi-
mum value µ*, similar to Ks in the Monod model. Eq. (3) gives the expression for the
growth rate in terms of the new parameters
µ * λσ S S S*
µ(S) = , σ = * , λ = m* + − 2, λ > 0 (3)
1 + (λ − 2)σ + σ

G. Buitrón and J. Moreno

2
S S Sm
Some advantages of this new representation are: there is a clear physical meaning of the
parameters; the family of curves given by Eq. (3) is bigger than that given by Eq. (2) (when
the parameters are restricted to have only positive values); and the effects of
acclimation/deacclimation can be explained by very clear, easy, and intuitive changes in
these parameters, as will be shown in this work.
One usual method of evaluating the parameters of the Haldane law consists in solving
numerically the differential Eq. (1) for an interval of time and for different values of the
parameters, until an inflection point has been reached. This solution is very time consuming
(it needs a lot of operations), and, more importantly, the algorithm can converge to local
minima, and the global minimum will not be reached. A different strategy leads to a linear
optimization problem, for which very powerful methods can be used, the computation bur-
den is heavily reduced, and the global optimum can be found. The idea is to solve Eq. (1) by
separation of variables, leading to the expression (4):

p1 ln( S ) + p2 S + p3 S 2 + p4 = −t (4)

where,
KsY Y Y Y  1 2
p1 = , p2 = , p3 = and p4 = −  K s ln( S0 ) + S0 + S0 
µ0 X µ0 X 2 µ 0 XKi µ0 X  2KI 

Eq. (4) is linear in the parameters p1 to p4, which in turn depend (non-linearly) on the
parameters of the Haldane law and the initial conditions. Standard linear regression
algorithms can be used for the evaluation of p1 to p4, and then from these the calculation of
the kinetic parameters is straightforward.
For the objectives of this work the batch experiments conducted have a drawback: a
single degradation curve for an inhibitory substrate can be explained, within a certain small
error bound, by very different sets of kinetic parameters, even under very accurate meas-
urements. This is a problem often found in practice, but whose explanation rests in the

*/2

S
S* Sm
Figure 1 Parameters of the Haldane model determined in the study 81
 mathematical structure of the model (1). This problem has been observed by Nihtilä and
Virkkunen (1977) and Holmberg (1982). To deal with this difficulty, for each series of
experiments the parameter determination has been done introducing an additional
constraint: the parameters obtained for one experiment should not only explain correctly
the experiment itself, but also the kinetic parameters of related experiments in the series
have to correlate to each other. All these calculations have been performed with the help of
the software BioReV, developed at the Institute of Engineering, UNAM, and programmed
G. Buitrón and J. Moreno

in MATLAB.

Results and discussion

Kinetics experiments
Figure 2a represents an example of the kinetics obtained during the acclimation of the bio-
mass to the 4CP. It is an important observation that the degradation time decreased as the
number of cycles (acclimation time) increased. It was also observed that the biomass con-
centration first decreased because some organisms not adapted to the toxic died, and others
were washed out. However, after some cycles, when the degradation time became stable,
the biomass concentration increased indicating the reproduction of 4CP-degrading organ-
isms. Figure 2b presents the kinetics obtained during the deacclimation experiments. It can
also be observed here that the degradation time increases when the starvation period
increases.
These experiments show that the capacity of degrading a toxic substance is induced in
the biomass by the presence of the toxic compound (acclimation). Once the toxic is absent
for a long period of time (starvation) the degradation capacity of the biomass decreases
again. The understanding of the internal mechanisms for these phenomena and the
mathematical modeling of these processes in terms of internal states of the biomass is an
important and interesting problem. The objective of this work is just to explain phenomeno-
logically these processes as variations in the kinetic parameters of the growth law.

Effect of acclimation on the kinetic parameters

Figure 3a presents an example of the variation of the kinetic parameters as the acclimation
progress. It is clear that all kinetic parameters µ*, S*, and Sm increase as the acclimation
time increases. The same pattern occurs in all other cases. Note that the increase in the
parameters is faster at the beginning of the acclimation, and as the acclimation time grows
the kinetic parameters tend toward some constant values. This evolution can be well
approximated by a first-order dynamics on the acclimation time. So for example for µ*

90
Acclimation time, d 100 before starvation
80 a Cycle 1 0.8
4h
Cycle 3 2
70 8h
Cycle 6 5 80 b
Substrate, mg4-CP/L
Substrate, mg 4-CP/L

6
Cycle 9 12 h
60 Cycle 11
7
9 24 h
50
Cycle 15
11 60
Cycle 27 36 h
14
Cycle 41
40
40
30
20 20
10
0 0
0 4 8 12 16 20 0 1 2 3 4 5
Degradation time, h Degradation time, h

Figure 2 Kinetics of (a) acclimation; and (b) deacclimation for 100 mg4CP/L at 20°C using activated
sludge as inoculum. It can be seen that the reaction time decreases as the acclimation takes place (a). On
the contrary, when different periods of starvation of the toxic compound are introduced, there is an increase
82 in the degradation time (b)
 140
80 mu*
120
(b) S*
(a)
60 100 Sm
mu*
S* 80
40
Sm 60

20 40
20

G. Buitrón and J. Moreno

0
0
0 10 20 30 40
0 10 20 30 40
Acclimation time (d) Starvation time (h)
Figure 3 Variation of the kinetic parameters during (a) acclimation; and (b) deacclimation for 100 mg4CP/L
at 20°C using activated sludge as inoculum. (a) Corresponds to Figure 2a, and (b) corresponds to Figure
2b. In both figures the values of µ* [1/h] have been multiplied by 1,000. S*, and Sm are given in [mg/L]

(
µ * = µ *f + Γµ exp − β µ Ts ) (5)

where: µ*I is the value of µ* at the beginning of the acclimation; ∆µ is the increment in µ*
during the acclimation; Ta is the number of acclimation days; and αµ is the (inverse) time
constant with which the acclimation evolves. For S* and Sm similar equations are valid.
These coefficients are different for each kinetic parameter, and depend on factors such as
temperature, acclimation concentration, etc. They can be easily obtained by a linear regres-
sion on the data of Figure 3.
Figure 4 shows how the acclimation affects the form of the Haldane law: the more accli-
mated the biomass to the toxic substrate is, the higher and wider is the Haldane curve. This
means that the degradation velocity is increased, and that the tolerance to the toxic is
enhanced. These results are easy to interpret in terms of the kinetic parameters introduced,
and are in accordance with the experimental results. In Figure 5 it is shown how the specif-
ic and volumetric degradation rates increase as the acclimation process is in progress,
something that can be predicted by the mathematical model (5).

Effect of starvation on the kinetic parameters

Figure 3b and Figure 6 present examples of the variation of the kinetic parameters as the
starvation time increases. It is clear that all kinetic parameters µ*, S*, and Sm decrease with

0.14

0.12
Acclimation

0.1

0.08

0.06

0.04

0.02

Deacclimation
0
0 20 40 60 80 100 120 140 160

Figure 4 Variation of the form of the Haldane law during acclimation and deacclimation. The data corre-
sponds to the acclimation process of Figure 3a. 83
 32 150

28
125

qx [mg 4-CP/g SSV-h], (- - -)

24

qv [mg 4-CP/l-h], (__)

100
20

16 75
G. Buitrón and J. Moreno

12
50
8
Exp. 1 25
4 Exp. 2
Exp. 3
0 0
0 2 4 6 8 10 12 14 16 18 20 22
Operation Days
Figure 5 Increase in the specific (qx) and volumetric (qv) degradation rates as a consequence of the accli-
mation of the biomass to 4 CP. The data of Exp. 3 correspond to those of Figure 2a

0.18 250
4C 100-15
4CF-100- 15 C
4CF-100-
4C 100-20
20 C
0.16 4CF-100-
4C 100-25
25 C
4CF-200-
4C 200-15
15 C
4CF-200-
4C 200-25
25 C
200
0.14

0.12
µ** [1/h]

S* [mg/l]

150

0.1

100
0.08

0.06
50

0.04

0.02 0
0 10 20 30 40 0 10 20 30 40
Star
arvation
on T
Tim e [[h] Star
arvation
on Tim
T e [h]
[

Figure 6 Variation of µ* and S* during deacclimation for the 4CP series of experiments, for temperatures
15, 20 and 25°C, and initial substrate concentrations of 100 and 200 mg/L

an increment of the starvation time, the opposite effect of the acclimation process.
Similarly to this case, this evolution can be well approximated by a first order dynamics in
the starvation time. So for example for µ*

µ * = µ *I + ∆ µ 1 − exp( −α µ Ta ) ( ) (6)

where: µf* is the value of µ* at the end of the deacclimation process; Γµ is the decrement in
µ* during the deacclimation; Ts is the starvation time suffered by the biomass; and βµ is the
(inverse) time constant with which the deacclimation evolves. For S* and Sm similar
equations are also valid.
Again, these coefficients are different for each kinetic parameter, and depend on several
factors. Although it seems reasonable to think that these parameters should be related to
84 those in the acclimation process, there is no experimental evidence of that.
 In Figure 4 it is illustrated that the effect of the deacclimation on the form of the Haldane
law is exactly opposite to that of the acclimation: the more deacclimated the biomass is to
the toxic substrate, the lower and narrower is its Haldane curve. This means that the
degradation velocity and the tolerance to the toxic are decreased. These results are easy to
interpret in terms of the kinetic parameters introduced, and are in accordance with the
experimental results.
In Figure 6 the variation of the kinetic parameters during deacclimation is shown for

G. Buitrón and J. Moreno

different conditions. It is clear that there is an effect of the temperature on the kinetic
parameters, but no regular behavior can be deduced. The changes produced on the kinetic
parameters by the initial substrate concentration are also clear. However, no regularity is
observed in these changes.

Conclusions
The capacity of microorganisms to degrade toxic substances is an induced process, and
depends strongly on the presence or absence of the toxic compound in the medium of the
biomass. If the toxin is present continuously, then the capacity for degradation is increased,
whereas during abstinent periods the degrading capacity is lost. These phenomena have an
important influence on the operation of wastewater treatment plants. If the presence of
toxics in the reactor is only sporadic, then the efficiency of the plant will be greatly reduced,
and the danger for the biomass is high. The only solution here is to prevent these toxic
spikes from reaching the reactor. On the other side, for a plant treating toxic substances the
starvation periods have a strong negative influence on the efficiency. In these cases the
operation of the plant has to prevent the presence of such starvation periods.
In this work it has been shown that acclimation and deacclimation processes for a batch
reactor can be modeled as increments or decrements, respectively, of some parameters (µ*,
S* and Sm) of the Haldane law. The evolution of the parameters can be described by a first
order dynamics (Eqs (5) and (6)), i.e. they increase or decrease in an exponential form, and
converge to steady values. These equations are a simple mathematical model, which is able
to predict the effects of acclimation/deacclimation observed in plant operation. It is inter-
esting to note that during acclimation the Haldane law increases its size, whereas during
deacclimation it decreases. This model can be used to predict the effects of acclimation or
starvation on the operation of a plant.
Other variables as temperature or initial substrate concentration seem not to have a regu-
lar influence in these parameters of the Haldane law, and further study is necessary to
describe their effect. The study is restricted to the batch operation of the plant. It is not clear
that the model be valid for other operation forms, or under more dynamic changes in the
toxic concentration in the reactor. For this it is necessary to develop a more complete math-
ematical model of these important phenomena.

Acknowledgements
This research was supported by DGAPA-UNAM through grant PAPIIT IN112800, and
Conacyt, Mexico, through grant 34934-A.

References
AFNOR (1985) Evaluation en milieu aqueux de la biodégradabilité aérobie “ultime” des produits
organiques solubles, Normalisation Française, NFT 90-312.
Arbucke, W.B. and Kennedy, M.S. (1989). Activated sludge response to a parachlorophenol transint.
JWPCF, 61, 476 –480.
Babcock, R.W. Jr., Ro. K., Hsie, C. and Strenstom, M.K. (1992). Development of an of-line enricher reactor
process for activated sludge degradation of hazardous wastes, Wat. Environ. Res., 64, 782 –791.
85
 Buitrón, G., Capdeville, B. and Horny, P. (1994). Improvement and control of the microbial activity of a
mixed population for degradation of xenobiotic compounds. Wat. Sci. Tech., 29(7), 317 –326.
Holmberg, A. (1982). On the practical identifiability of microbial growth models incorporating Michaelis
–Menten type nonlinearities. Math. Biosci., 62, 23 –42.
Nihtilä, M. and Virkkunen, J. (1977) Practical identifiability of growth and substrate consumption models,
Biotech. Bioeng., XIX, 1831 –1850.
Senthilnathan, P.R. and Ganczarczyk, J.J. (1989). Adapatation and deadaptation kinetics of activated
G. Buitrón and J. Moreno

sludge, 43rd Purdue Industrial Waste Conf. Proc. Lewis Pub., 37, 301 –307.
Standard Methods for Examination of Water and Wastewater (1992). 18th ed. American Public Health
Association/American Water Works Association/Water Environment Federation, Washington, D.C.,
U.S.A.

86

View publication stats