Facultad de Ingeniería Universidad de Concepción Barrio Universitario, Concepción, Chile Tel: 56 41 2204197 Fax: 56 41 2243750 E-mail: chuilinir@udec.cl
Facultad de Ingeniería Universidad de Concepción Casilla 160-C Correo-3 Concepción, Chile Tel: 56 41 2204534 Fax: 56 41 2243750 E-mail: easpe@udec.cl
Facultad de Ingeniería Universidad de Concepción Casilla 160-C Correo-3 Concepción, Chile Tel: 56 41 2203663 Fax: 56 41 2243750 E-mail: mroeckel@diq.udec.cl
Several food processing industries discharge their liquid effluents with high organic load to the environment causing ecological and public health problems. Salmon-plant effluents have high
protein (4.08- 22.5 g COD L Several authors have suggested that
HNO Nitrite is an intermediate in the denitrification process and depending on the pH and temperature of the medium could prevail as the ionized or the non-ionized form. The pH can affect directly the
bacterial growth and its enzymatic activities (Campos and Flotats, 2003),
including denitrifying enzymes, and indirectly affect the denitrification rate
through changes in the concentration of HNO Almeida
et al. (1995) and Wild et
al. (1995) have developed kinetic models to predict denitrification
rates. However, these models have considered the apparent net pH
effect, Anoxic activated sludges usually
include The rate of nitrate reduction has
been represented as a function of the nitrate concentration by the Monod
equation (Almeida et al. 1995) and the rate of nitrite reduction has been
represented by a Haldane-type kinetics, since high HNO Although it has been reported that the denitrification rate varies with a change in the pH (Almeida et al. 1995; Glass and Silverstein, 1998), the quantification of the direct pH effect on this rate or its inclusion in the kinetics of denitrification have not been reported in the literature. There are different expressions for the pH-inhibition function for biological processes reported in the literature. Ramsay and Pullammanapallil (2005) modelled the effect of the pH, up to pH 7.0, on the acidogenic bacteria by an empiric relationship of the exponential type. Siegrist et al. (2002) included a non-competitive quadratic inhibition factor due to pH inhibition up to pH 7.0. Since the former relationships are empiric ones and for pH values below 7.0, the Michaelis function (Segel, 1975), initially proposed to quantify the dependence of the enzymatic activity on the pH, seems the more appropriate expression to model the effect of the pH on the denitrification rate. This function has been modified to quantify the inhibitory effect of a wide range of pH values on the acetogenic step of the anaerobic process (Angelidaki et al. 1993; Batstone et al. 2002). The goal of this work was to include the effect of the pH on the kinetics of nitrate reduction and nitrite reduction in the modelling of these rates, and evaluate its predictive value using experimental data obtained with a mixed denitrifying sludge for treatment at 37ºC of a salmon plant effluent. Validation of the inclusion of the pH effect in the kinetics of denitrification will be assessed by calculating the fitting deviations of the reported models with and without the inclusion of pH function. An improvement of the predictive values of these models will broaden the use of the kinetics to different operational conditions. An adapted, stable biomass, able to carry
out denitrification and methanogenesis was obtained from 3 L anaerobic and anoxic reactors that were at steady state for at least a year. These reactors were
fed with 50% of a salmon-plant effluent and 50% of a synthetic substrate (v/v);
the latter to adapt the sludge to nitrate. The
average industrial effluent composition was (g L Kinetic assays at constant pH were
carried out in 1-L batch reactors with a useful volume of 800 mL of anoxic medium.
They were seeded with 20% (v/v) of the adapted inoculum (50% anaerobic and 50%
anoxic); the purpose of seeding the reactors with a large microbial
concentration was to ensure a constant biomass throughout the assays. The reactors were fed with a mixture (v/v) of 50% salmon
plant effluent and 50% of a synthetic substrate of
the following composition (g L Anoxic conditions in the reactors
were obtained by gassing for approximately 1 min with N Samples were withdrawn every 40 min
for 12 hrs and every 2 hrs for 16 hrs in kinetic assays performed at an initial
nitrate concentration of 10 and 80 mg NO In each sample, nitrite, nitrate, total ammonia nitrogen (TAN), volatile suspended solids (VSS) and total organic carbon (TOC) were measured as described elsewhere (Aspé et al. 2001; Sánchez et al. 2005). VSS were measured at the beginning and at the end of each assay. Total ammonia nitrogen was measured according to Standard Methods (APHA, 1992). Analyses were performed in triplicate. ## Maximal rates of nitrate and nitrite reduction.The kinetic analysis used a simplified denitrification model:
Although other reduced products of denitrification have been reported, pH was thought to affect specially the nitrate and nitrite reduction rates. The nitrate consumption rate can be described by:
Where k is
a rate constant and _{1}_{} is
any nitrate concentration function to be used in the kinetics of
nitrate reduction. The rate of nitrite consumption is:
Where Nitrite is simultaneously formed and consumed, thus:
The initial maximal
rates of nitrate consumption (r _{ nitrite}, was obtained from a nitrite concentration vs. time plot
at different pHs and measuring the slope after nitrite has reached its maximal
accumulation.## Modelling of the pH effectSince the sludge used to denitrify in the present work is an anaerobic sludge, it is feasible to quantify the inhibitory effect of the pH on the denitrifying bacteria through mathematical functions representing the effect of the pH on anaerobic (acidogenic) bacteria. Several studies have included the inhibitory effect of the pH on the acidogenic bacteria as a factor that multiplies the substrate consumption rate:
Where
i.
i degradation as
a function of the substrate concentration used.
The Michaelis function used to model the effect of the pH on enzymatic reactions was used to model the effect of the pH on nitrate and nitrite consumption rate (Segel, 1975):
Or:
where: ^{+}], is the proton concentration; Ks is the
lowest proton concentration where _{1}r is equal to ½r; _{max}Ks is the highest proton concentration where _{2}r is equal to ½r; _{max}pKs, is the logarithm of the lowest pH at which _{1}r is
equal to ½r; _{max}pKs is the logarithm of the highest pH at which _{2}r is equal to ½r._{max}If the pK values at which r _{max}vs pH will occur at a value significantly lower than the theoretical
maximum and, consequently, the pH values at the half-maximum points will not
correspond to the pK values (Segel, 1975) and, thus, the Michaelis pH function
must be modified by an empiric factor (A) so that it reaches r as a central value (Glass et al. 1997). Therefore, equation
(7) assumes the
following form:_{max}
The concentration pKs values were obtained by fitting of the _{2}r/r _{max}vs pH plot by equation (8).The deviation between the experimental data and the values given by the model using kinetic parameters reported in the literature, were calculated considering the experimental data as the true values.
Assays at different pH values and
different nitrate concentrations were carried out to determine the inhibitory
effect of the pH on denitrification. Figure 1a and Figure 1b shows the nitrate and nitrite concentration variation in assays performed at pH
8.0 and an initial nitrate concentration of 10 mg NO Figure 2 shows the _{3}^{-}-N L^{-1}.
Maximal nitrate consumption rates were obtained from the initial slope of the
nitrate consumption vs. time plot at different pH values. As shown, the r/r ratio and, hence, the apparent
maximum rate of nitrate consumption, reached a peak between pH 7.5 and 8.0 at
initial nitrate concentrations of 10 mg NO_{max}_{3}^{-}-N L^{-1} and 80 mg NO_{3}^{-}-N L^{-1}.At an initial nitrate concentration
of 80 mg NO The optimum rate was
attained at pH 7.5 to 8.0, and no differences in rates were observed between
both initial nitrate concentrations. This pH range is in agreement with those
reported by Glass et al. (1997), who reported an optimum pH for denitrification
close to 8.0 in an active sludge reactor with a nitrate and nitrite
concentrations of 1350 mg NO
The effect of the pH on the apparent
maximum rate of nitrate reduction at an initial
nitrate concentration of 10 mg NO pKs of the factor accounting for pH
inhibition were calculated by fitting of the experimental data by equation
(8).
Fitting was carried out by the TableCurve 2D, program using the
Levenberg-Marquardt method. The fitting of the experimental data gave the
following parameter values: _{2}pKs = 6.27 ± 0.204; _{1}pKs = 9.04 ± 0.175; maximum specific
rate at the optimum pH = 1.32 ± 0.24 [mg NO_{2}_{3}^{-}-N
(g VSS h)^{-1}] and A = 1.01 ± 0.085. As shown in Figure 2, the
function gives a good prediction of the r/r ratio at pH < 8.0, as the predicted values lie within
the experimental error range. At pH values over 8 (8.5 and 9), the function was
not able to predict values within the experimental error range. _{max}
Figure 3 shows the apparent maximum specific
rates of nitrite reduction at initial nitrate concentrations of 80 mg NO
The
separate effect of the pH on the maximum rate of nitrite reduction at an
initial nitrate concentration of 10 mg NO pKs of the factor accounting for pH
inhibition were calculated by fitting of the experimental data by equation (8).
Fitting was carried out by the TableCurve 2D, program using the Levenberg-Marquardt
method. Figure 3 shows the fitting of the experimental data by this
function for an initial nitrate concentration of 10 mg NO_{2}_{3}^{-}-N
L^{-1}; as shown, this function fitted the experimental data in the
whole pH range studied. The fitting of the experimental data gave the following
parameter values: pKs = 6.41 ± 0.072, _{1}pKs = 8.93 ± 0.065, r_{2}_{max} = 1.07 ± 0.08 [mg NO_{2}^{-}-N
(g VSS h)^{-1}] and A = 1.09 ± 0.024.## Quantification of the inhibitory effect of the pH in reported kinetic modelsSeveral authors have modelled the
nitrate consumption rate by the Monod equation. However, the kinetic parameters
they reported were obtained with different carbon sources, at a different
temperature and/or different pH than the conditions used in this work (mainly
protein as carbon source, 37ºC and pH range 6.5-9.0). Wild et al. (1995) reported
kinetic parameters at 20ºC with acetate as carbon source; moreover, Wild et al.
(1995) performed their experiments at pH 7.0. On the other hand, Almeida et al.
(1995) reported kinetic parameters for denitrification obtained at 28ºC, pH
7.0, with acetate as carbon source and using a pure culture of The inclusion of the Michaelis function, which accounts for the effect of the pH, in the Monod and in the Haldane model gives the following equations: For nitrate reduction:
Similarly, for nitrite reduction:
Figure 4a and Figure 4b show the fittings by the Monod model and the
pH-modified Monod model with the kinetic parameters reported by Soto et al.
(2007) of the experimental data for nitrate consumption at pH 6.5 and 9.0,
respectively, and at an initial nitrate concentration of 80 mg NO Modelling by the pH-modified Monod
of the experimental data reported by Almeida et al. (1995) for nitrate and
nitrite consumption at pH 7.0 validated the use of the Michaelis function to
quantify the effect of the pH in the denitrification kinetic. The Michaelis
parameters calculated from the data reported by
Almeida et al. (1995) were pK pKs values lie out
of this pH range._{2}This function allows modelling of the kinetics of denitrification by one equation that integrates kinetic variables and the pH effect. These results shows that the effect of the pH should be taken into account in the modelling, operation and design of bacterial processes.
The results of this work indicate that the inclusion of the pH effect through the Michaelis function in the kinetics of nitrate reduction and nitrite reduction clearly improves its predictive potential as it reduces the deviation between the predicted and experimental values from an average of 33.8% and 53.5% to 10.5% and 10.7% for nitrate reduction (Monod) and nitrite reduction (Haldane), respectively. Thus, not only the
pH-dependent HNO ALMEIDA, J.S.; JULIO, S.M.; REIS,
M.A.M. and LARRONDO, M.J.T. Nitrite inhibition of denitrification by ANGELIDAKI, I.; ELLEGAARD, L. and
AHRING, B. A mathematical model for dynamic simulation of anaerobic digestion
of complex substrates: focusing on ammonia inhibition. APHA, AWWA, WPCF, Standard Methods
for the Examination of Water and Wastewater. 20 ASPÉ,
E.; MARTÍ, M.C.; JARA, A. and
ROECKEL, M. Ammonia inhibition in the anaerobic treatment of fishery effluents. BATSTONE, D.; KELLER, J.;
ANGELIDAKI, I.; KALYUZHNYI, S.; PAVLOSTATHIS, S.; ROZZI, A.; SANDERS, W.;
SIEGRIST, H. and VAVILIN, V. CAMPOS, E. and FLOTATS, X. Dynamic
simulation of pH in anaerobic processes. GLASS, C. and
SILVERSTEIN, J. Denitrification kinetics of high nitrate concentration water:
pH effect on inhibition and nitrite accumulation. GLASS, C.; SILVERSTEIN, J. and OH,
J. Inhibition of denitrification in activated sludge by nitrite. KUNAK, M.; KUCERA, I. and VAN
SPANNING, R. Nitric oxide oscillations in MOSQUERA-CORRAL, A.; SANCHEZ, M.;
CAMPOS, J.L.; MÉNDEZ, R. and LEMA, J.M. Simultaneous methanogenesis and
denitrification of pretreated effluents from a fish canning industry. RAMSAY, I.R. and PULLAMMANAPPALLIL,
P.C. Full-scale application of a dynamic model for high-rate anaerobic
wastewater treatment systems. RIVER, L.; ASPÉ, E.; ROECKEL, M. and
MARTÍ, M.C. Evaluation of clean technology processes in the marine products processing
industry. SÁNCHEZ, O.; ASPÉ, E.; MARTI, M.C. and
ROECKEL, M. Rate of ammonia oxidation in a synthetic saline wastewater by a
nitrifying mixed-culture. SEGEL, Irwin. Effects of pH and
temperature. In: SEGEL, Irwined. SIEGRIST, H.; VOGT, D.;
GARCIA-HERAS, J.L. and GUJER, W. Mathematical model for meso- and thermophilic
anaerobic sewage sludge digestion. SOTO, O.; ASPÉ, E. and ROECKEL, M.
Kinetics of cross-inhibited denitrification of a high load wastewater. WILD, D.; VON SCHULTHESS, R. and
GUJER, W. Structured modelling of denitrification intermediates. Note: Electronic Journal of Biotechnology is not responsible if on-line references cited on manuscripts are not available any more after the date of publication. |