Article pubs.acs.org/est
Formation of Distinct Soluble Microbial Products by Activated Sludge: Kinetic Analysis and Quantitative Determination Bing-Jie Ni,† Fang Fang,† Wen-Ming Xie,† Juan Xu,† and Han-Qing Yu*,† †
Department of Chemistry, University of Science & Technology of China, Hefei 230026, China S Supporting Information *
ABSTRACT: Soluble microbial products (SMP) released by microorganisms in bioreactors are classified into two distinct groups according to their different chemical and degradation kinetics: utilization-associated products (UAP) and biomassassociated products (BAP). SMP are responsible for effluent chemical oxygen demand or for membrane fouling of membrane bioreactor. Here an effective and convenient approach, other than the complicated chemical methods or complex models, is developed to quantify the formation of UAP and BAP together with their kinetics in activated sludge process. In this approach, an integrated substrate utilization equation is developed and used to determine UAP and their production kinetics. On the basis of total SMP measurements, BAP formation is determined with an integrated BAP formation equation. The fraction of substrate electrons diverted to UAP, and the content of BAP derived from biomass can then be calculated. Dynamic quantification data are obtained for UAP and BAP separately and conveniently. The obtained kinetic parameters are found to be reasonable as they are generally bounded and comparable to the literature values. The validity of this approach is confirmed by independent SMP production tests in six different activated sludge systems, which demonstrates its applicability in a wide range of engineered system regarding SMP production. This work provides a widely applied approach to determine the formation of UAP and BAP conveniently, which may offer engineers with basis to optimize bioreactor operation to avoid a high effluent soluble organics from SMP or SMP-based membrane fouling in membrane bioreactors.
■
INTRODUCTION Soluble microbial products (SMP) are soluble organic compounds released by microorganisms during normal metabolism in bioreactors.1,2 SMP are important because they are ubiquitously present and often form the majority of the effluent chemical oxygen demand (COD) from biological wastewater treatment systems.3 Therefore, their presence is of a particular interest in terms of achieving discharge consent levels for wastewater treatment plants.4,5 Also of special contemporary importance is the role of SMP in membrane fouling of membrane bioreactor (MBR), which are becoming more prevalent worldwide.3,6 SMP are classified into two groups based on the type of bacterial metabolism from which they are derived (different production mechanisms): utilization-associated products (UAP) that are produced directly during substrate utilization and biomassassociated products (BAP) that are produced indirectly via the hydrolysis of a biomass component.2,7 UAP and BAP have distinct chemical and degradation kinetics, i.e., UAP are more readily biodegraded and BAP have a much larger molecular weight (MW).8 Thus, BAP might be the most likely cause for SMP-based fouling,6,9 because they accumulate more in MBRs due to the normally long solids retention time (SRT) in MBRs,10 the relatively refractory characteristic of BAP,11 and their larger size.11,12 © 2011 American Chemical Society
If UAP and BAP could be accurately and quantitatively determined, it would become possible to determine which type of SMP is responsible for COD or for fouling membranes, and effective control methods may become evident.8,13,14 The effective and convenient determinations of the UAP and BAP concentrations can then offer engineers with basis to optimize reactor operation to avoid a high effluent soluble COD from SMP or SMP-based membrane fouling. In particular, if a sufficiently high SMP concentration is detected, post-treatment can then be applied to reduce SMP concentration. Therefore, the quantitative determination of the formation of UAP and BAP by activated sludge and their kinetics is highly desired. However, differentiating UAP from BAP has proven to be a challenge, since both components are heterogeneous materials, not specific compounds.15−17 The combination of experimental and modeling approaches has been proven to be an applicable way to quantitatively determine UAP and BAP.8,11,18 Several approaches have been used to model the UAP and BAP production kinetic. Rittmann and co-workers have developed a Received: Revised: Accepted: Published: 1667
August 8, 2011 December 7, 2011 December 20, 2011 December 20, 2011 dx.doi.org/10.1021/es202756d | Environ. Sci. Technol. 2012, 46, 1667−1674
Environmental Science & Technology
Article
although it was fed with the same synthetic wastewater as SBR1. For SBR3, the SRT was set to 20 d, but it was fed with a fatty-acids-rich wastewater at approximately 800 mg L−1 of COD in order to test the approach with different types of wastewater composition. This fatty-acids-rich wastewater was the effluent of a laboratory-scale anaerobic acidogenic reactor fed with a sucrose-rich wastewater.20 Butyrate, acetate, and propionate were its main constituents. Other operating parameters for SBR2 and SBR3 were the same as those for SBR1. Correspondingly, two additional experiments with the sludge taken from SBR2 and SBR3 were conducted to produce SMP using the similar approach for Experiments I−III. The two sets of experiments were conducted with acetate wastewater (1,000 mg COD L−1) and fatty-acids-rich wastewater (500 mg COD L−1) to evaluate the SMP production by activated sludge from SBR2 and SBR3, respectively. Mixed liquor samples were centrifuged for 15 min at 12,000 rpm at 4 °C and then were prefiltered through 0.45-μm acetate cellulose membranes to represent the SMP, which were used for SCOD analysis. The measurement of MLVSS and COD was performed according to the Standard Methods.21 The concentrations of butyrate, acetate, and propionate were measured using a gas chromatograph (6890NT, Agilent Inc., USA) equipped with a flame ionization detector and a 30 cm × 0.25 mm × 0.25 mm fused-silica capillary column (DB-FFAP). The net SMP concentration (as COD) was calculated by subtracting the butyrate, acetate, and propionate concentrations (as COD) from the SCOD concentration. For SBR1 and SBR2, their total SMP was determined from the difference as (SCOD − 1.07×Acetate). For SBR3, the total SMP was determined from the difference as (SCOD − 1.82×Butyrate − 1.51×Propionate − 1.07×Acetate). The constants, 1.07, 1.82, and 1.51 g COD/g substrate, are the conversion factors from acetate butyrate and propionate to COD, respectively. Integrated UAP and BAP Production Equation. The total SMP are divided into UAP and BAP (eq 1)
series of SMP models, which are summarized as a unified SMP and EPS (extracellular polymeric substances) theory.2,19 In this unified theory, eight SMP-associated model parameters are introduced. Later, a very complex SMP model has been incorporated into ASM1 and ASM3.18 These existing SMP models have heterogeneous model structures. A common problem of them is that the models are too complex and overparameterized with strong parameter correlations. Indeed, the available measurements for model calibration are usually limited. Thus, the validity of the SMP model structure and the obtained parameter values are questionable. Moreover, in most cases, few SMP data, and mostly only steady state Soluble COD (SCOD) data, are available for parameter estimation. Thus, model parameters are often estimated using trial and error methods, and no parameter confidence interval is given and no independent model validation is provided. Therefore, an effective and convenient approach, other than the complicated chemical methods or complex models, to quantitatively and qualitatively determine UAP and BAP and their production kinetics, is highly desirable. This paper describes an effective and convenient approach to determine the UAP and BAP by activated sludge. The integrated form of the Monod-based equations for substrate utilization production and BAP formation are formulated and used to construct the objective function for determining UAP and its production kinetics. The independent experimental data from six different activated sludge systems are used to validate this approach and demonstrate its applicability in a wide range of engineered system regarding SMP production. Moreover, a comparison between this approach and others is performed to obtain better understanding of UAP and BAP formation.
■
MATERIALS AND METHODS Experimental Setup and Analysis. Activated sludge used to produce SMP was taken from a bench-scale sequencing batch reactor (SBR) with a working volume of 2 L (SBR1). The SBR1 was fed with a synthetic wastewater (see details in the Supporting Information, SI) at COD of 800 mg L−1 with acetate as carbon source. The SRT was set at 20 days by controlling the amount of sludge wasted from the reactor in each cycle. The other reactor operation is the same as the previous work.11 Three different sets of experiments (I−III) were conducted at three different substrate levels (400, 800, 1,600 mg COD L−1, respectively) to evaluate the SMP production by activated sludge. Each experiment was carried out in triplicates. The seeding sludge was sampled from the SBR when no substrate was present in the medium and was washed twice with distilled water to remove external soluble organic material. A reactor was inoculated with 1 L of diluted activated sludge at 1500 mg L−1 MLVSS (mixed liquor volatile suspended solids) and aerated it continuously to keep DO over 2 mg L−1. Then, the wastewater with predetermined concentrations was added. The composition of the feeding wastewater was the same as that of the synthetic wastewater for SBR1. Samples were taken at given intervals for the analysis of COD, SMP, and MLVSS. The experiments were conducted at 20 °C and pH of 7.0 ± 0.05. To further validate the proposed approach, two other activated sludge SBRs with different operating conditions were also used to generate SMP production data. These two SBRs, i.e., SBR2 and SBR3, had the same configuration as SBR1. Nevertheless, for SBR2, the SRT was set to 30 d instead of 20 d,
dSSMP dS dS = UAP + BAP (1) dt dt dt A Monod-based kinetics for UAP formation is used in this work. For the flow of electrons from the external substrate, external substrate is partially used for biomass synthesis, and part of external electrons is diverted to the formation of UAP. UAP are released to the aqueous solution. Substrate oxidation and respiration of the electrons to reduce oxygen and generate the energy needed to fuel for formation of active biomass and UAP.2 The rate of UAP formation is proportional to the substrate utilization rate,2 and the kinetics is expressed as eq 2. Eq 3 is the kinetic equation for active biomass, while the kinetics for external substrate is described with eq 4 μ dSUAP SS = kUAP H XH dt YH KS + SS (2)
1668
SS dXH = (1 − kUAP)μH XH dt KS + SS
(3)
μ dSS SS =− H XH dt YH KS + SS
(4)
dx.doi.org/10.1021/es202756d | Environ. Sci. Technol. 2012, 46, 1667−1674
Environmental Science & Technology
Article
consumption and SMP formation in the batch tests of SBR1 are shown in Figure 1. The acetate concentration decreased rapidly
The biomass in turn undergoes endogenous decay and releases cell lysis products, which are the source of BAP. The BAP formation rate is proportional to the biomass concentration, with kBAP as the first-order rate coefficient for BAP production (eq 5)
dSBAP = kBAPXH (5) dt For the UAP production, the weighted nonlinear least-squares analysis (WNLSA) is used to construct the objective function and to determine the kinetic parameters by comparing the model predictions with the observed substrate concentrations and t by using the known values. Thus, the estimation of the kinetic parameters is successively performed by minimizing the sum of the squared weighted errors (SSWEUAP) based on n observations using the spreadsheet method, as shown in the following equation t=
+
−
KS YH (1 μH XHo + YH − kUAP)SS 0 ⎛ (X + Y (1 − kUAP)SS 0 − Y (1 − kUAP)SS) ⎞ × ln ⎜ Ho ⎟ KSSS ⎝ ⎠
1 ln ((XHo + YH(1 − kUAP)SS 0 μH (1 − kUAP) − YH(1 − kUAP)SS)) KS X YH ln Ho KSSS 0 μH XHo + YH(1 − kUAP)SS 0 1 ln XHo − YH(1 − kUAP)
(6)
The difference between the predicted and the observed SS values can be estimated by multiplying tiobs − tipred by the local slope of the substrate utilization curve, ΔSS/Δt. Therefore, the logical weighting factor is the local slope of the substrate utilization curve n
SSWEUAP =
⎡ ΔSS
∑ ⎢⎣
Δt
i=1 n
≈
pred
⎤ ⎥⎦
2
)
2
∑(
SSiobs
i=1
(
tiobs − ti
−
pred SSi
)
Figure 1. Model fitting results of the integrated SMP production equations to the substrate utilization and SMP production data of SBR1. The model (solid line) is fitted to one data set of substrate and SMP concentrations (scatter, 800 mg COD L−1 initial acetate concentration), which resulted in the parameter values. Model curves obtained with the same parameter values are shown for other two data sets (400 and 1600 mg COD L−1 initial acetate concentration) for comparison: (A) substrate utilization; (B) total SMP formation; (C) UAP production; and (D) BAP formation.
(7)
whereSSiobs and SSipre are the ith measured and predicted substrate concentrations, respectively. Similarly, the kinetic BAP parameters can be estimated from n
SSWEBAP =
i=1
obs SBAPi
2
pred obs − SBAPi) ∑ (SBAPi
(8)
and continuously within the initial 3 h, as the external substrate was consumed. Because of the SMP production, a complete depletion of the SCOD pool in the mixed liquid would not occur despite of the complete consumption of acetate after 3 h. External substrate consumption resulted in an increase in the amount of total SMP. The SMP concentrations increased to 20, 45, and 75 mg COD L−1 at ∼3 h gradually (Figure 1) in the three tests (I−III), respectively. In the rapid growth period, the acetate consumption resulted in an increase in total SMP. In addition, with the evolution of microbial decay, SMP might also be produced gradually from the biomass decay.18,22 These two processes might lead to an increase in total SMP. These observations are in agreement with those reported by Jarusutthirak
pre SBAPi
where and are the ith observed and predicted BAP concentrations, respectively. More details (mathematical development) about the integration of eq 6 leading to eq 8 and parameters definitions could be found in the SI. All of the calculations were done with a personal computer by using a spreadsheet program (Microsoft Excel 2003).
■
RESULTS
SMP Production in Activated Sludge. The profiles of measured substrate consumption and SMP formation in all the sets of batch tests were similar. The time dependence of acetate 1669
dx.doi.org/10.1021/es202756d | Environ. Sci. Technol. 2012, 46, 1667−1674
Environmental Science & Technology
Article
Figure 2. Contour plots of the objective functional used for parameter estimation (SSWEUAP) as a function of different parameter combinations: μH vs KS; μH vs YH; μH vs kUAP; and YH vs kUAP. The plots were drawn using the final optimal parameters (Table 1) as midpoint of the intervals with q1 order of magnitude change (except YH, which is always lower than 1) on both sides of the intervals.
and Amy.5 They are also in accordance with the unified theory for SMP production proposed by Laspidou and Rittmann,2,19 which couple the production and degradation of SMP with the formation of EPS for aerobic systems. Determination the Formation of UAP and BAP. The experimental data of the batch test II of SBR1 for substrate utilization were analyzed by using this new approach. The UAP formation was determined based on the estimation of rate coefficients (μH and KS) and the two yield coefficients (YH and kUAP), by fitting eq 6 to the substrate utilization data using the spreadsheet method with WNLSA (eqs 7 and 8). The input value for XH0 was estimated based on the results of several experiments using the baseline endogenous OUR level prior to substrate addition.22 In the present work, the initial concentration of active heterotrophic biomass XH0 in this batch test was estimated to be 480 mg COD L−1. With the calculated UAP data with eq 7, the BAP production as well as the rate coefficients kBAP could then be determined by fitting eq 17 (SI) to the BAP production data by using the spreadsheet method with the WNLSA (eq 8). The spreadsheet for fitting the integrated substrate utilization equation to the data is shown in Figure S1 (SI). The values of the rate coefficients and yield constants, which were used as fitting parameters, are given in Rows 1−4. Similarly, the spreadsheet for fitting the integrated BAP production equation to the BAP data is shown in Figure S2 (SI). The determination results of the UAP and BAP production in batch test II of SBR1 are shown in Figure 1. More details of the spreadsheet fitting method employed here can be found in the SI. UAP and BAP Formation Kinetic Analysis. Uncertainty analysis of a model structure is important as it tells which parameter combinations can be estimated under given measurement accuracy and quantity. Uncertainty estimates are also crucial
to compare rates for different microorganisms or under environmental conditions. In the uncertainty analysis, contour plots of the objective function for the degrees of correlation between parameters were evaluated. The contour plots in Figure 2 were calculated around the optimum for different combinations of parameters. The plots were drawn using the final optimal parameters for the data fitting as midpoint of the intervals with 1 order of magnitude change (except YH, which is always lower than 1) on both sides of the intervals. The contours of the objective function for half-saturation constant (KS) vs maximum growth rate (μH) showed a well-defined valley, where the optimum values of μH and KS resides. This indicates a good identifiability of these parameters. Generally, there is a greater sensitivity of the model to μH compared to KS. There is a greater change in SSWE on the μH compared with the KS axis. The more detailed information for the UAP and BAP formation kinetic parameters and the related sensitivity analysis has been presented in the SI. Verification of the Approach. Verification was performed through comparing more measured and predicted results in different activated sludge systems with different sets of data (Table S2). In this work, one data set of SBR1 (set II) was first used to determine UAP and BAP, and the obtained values of the coefficients were used to generate the two sets of additional curves including acetate, SMP, UAP, and BAP for comparison with the other two data sets of SBR1 (sets I and III with initial 400 and 1600 mg COD L−1 in Figure 1). The maximum difference between the measured and calculated values was 15%, and about 65% of the results had a difference of less than 5%. Furthermore, the simulation shows no systematic deviations, suggesting the validity of this approach for UAP and BAP determination. 1670
dx.doi.org/10.1021/es202756d | Environ. Sci. Technol. 2012, 46, 1667−1674
Environmental Science & Technology
Article
wastewater. The kBAP was estimated to be 0.0049 h−1, which was comparable to those of SBR1 and SBR2. The experimental trends and absolute concentrations for all components are described well by this approach. These trends include the reduction of substrate, the increase of BAP, and the leveling off of UAP, as the original substrate becomes depleted (Figure 3). Model and experimental data agree well for both SBR2 and SBR3 independently, strongly supporting the validity of the proposed approach and its flexibility of applying to different sets of data in various systems regarding SMP production.
The proposed approach was further validated using the experimental results from SBR2 and SBR3 to evaluate its flexibility of applying to different sets of data in various systems. The experimental conditions for these two experiments were substantially different from those in Figure 1. The SRT was 30 d for SBR2 other than 20 d, whereas the added wastewater was the fatty-acids-rich wastewater for SBR3. Also, the initial substrate concentrations were different. Figure 3A-D illustrates
■
DISCUSSION In this work a simple kinetic model including only five SMPrelated parameters was introduced and integrated to determine the SMP production. Dynamic batch data were determined for BAP and UAP separately. Such an approach was validated using independent SMP production tests. Moreover, with the integrated Monod-based equation in this work UAP and BAP could be determined through progress curves from a few batch experiments or even one batch experiment. The motivation for developing this approach is the complexity and the timeconsuming feature of the existing SMP production models. This work just uses a computer spreadsheet program to determine UAP and BAP accurately from experimental data. It provides a flexible simulation tool capable of accounting in a straightforward manner. The obtained parameter coefficients represent the overall values for the different microorganisms present and are comparable to other values previously published. Table 1 lists the values of the kinetic parameters for SMP production reported in the literature for comparison. The value of kUAP (0.042 mg CODUAP mg−1 CODS) estimated from this approach is generally in accord with those reported in other studies. However, it is significantly lower than the values reported by Lu et al.24 and Aquino and Stuckey,25 due to their different UAP formation kinetics. In addition, the usually long SRT in MBRs of Lu et al.24 and Aquino and Stuckey25 might result in a higher UAP production coefficient. The YH estimated from this approach, 0.531 g CODX g−1 CODS, falls in the range of the reported values (Table 1) but is relatively lower when compared with the values reported in some literature,18,24,26 because a significant amount of substrate electrons is diverted to the UAP, rather than the active biomass. The value of the estimated KS, 30.59 g CODS m−3, is between 0.7 and 500 g CODS m−3, as shown in Table 1. The heterotrophs in activated sludge have a wide range of KS value, depending on the cultivation conditions such as substrate type, nutrient concentration, and many other factors. The μH value reported in literature varies largely between 0.05 and 1.67 h−1 (Table 1), whereas the μH value estimated in this work, 0.268 h−1, is comparable with these values. The difference is partially attributed to the different heterotrophic microbial communities. The value of kBAP (0.0041 h−1) estimated from this approach is also in the range of the reported values (Table 1) but is significantly lower than the value (0.0215 h−1) reported by Jiang et al.18 because the large-size BAP could accumulate more in MBRs by membrane rejection in their experiments. Different model structures in their studies might be another reason for such a difference. In summary, it should be noted that substrate type, reactor operating conditions, and microbial communities are important factors affecting the SMP production. Since the data are obtained from mixed-culture experiments (activated sludge), the parameters reported here are highly dependent on these factors.
Figure 3. Comparison between the model simulations and the experimental data for the verification of the approach: (A) substrate profiles and SMP data of SBR2; (B) UAP and BAP concentrations of SBR2; (C) substrate profiles and SMP data of SBR3; and (D) UAP and BAP concentrations of SBR3.
the simulated and measured results for substrate, SMP, UAP, and BAP from the experiments of SBR2 and SBR3. For SBR2, the estimated kUAP and kBAP values were 0.047 g CODUAP g−1 CODS and 0.0065 h−1, respectively, which were higher than the values obtained from experiments of SBR1, suggesting that a higher SRT might favor BAP production. For SBR3, the estimated kUAP value was 0.052 g CODUAP g−1 CODS, indicating the higher UAP production in SBR3 by feeding the fatty-acids-rich 1671
dx.doi.org/10.1021/es202756d | Environ. Sci. Technol. 2012, 46, 1667−1674
Environmental Science & Technology
Article
Table 1. Comparison among the Values of the Kinetic Parameters for SMP Formation Reported in the Literature sources this study Lu et al.24 Laspidou and Rittmann19 Laspidou and Rittmann22 Ahn et al.26 Aquino and Stuckey25 Jiang et al.18 a
kUAP (g CODUAP g−1 CODS)
YH (g CODX g−1 CODS)
μH (h−1)
KS (g CODS m−3)
kBAP (h−1)
0.042 ± 0.006 (28.6%)a 0.38 0.05 0.05 0.05 0.2 0.0963
0.531 ± 0.03 (11.3%) 0.67 0.34 0.34 0.43−0.56 0.2 0.57
0.268 ± 0.018 (13.4%) 0.25 0.403 0.403 0.05−0.14 1.67 0.25
30.59 ± 3.5 (22.9%) 20.0 0.7 20.0 20.0 500 4.0
0.0041 ± 0.0002 (9.8%) 0.0167 0.00014 0.0215
95% confidence intervals are presented in bracket as absolute percentage of the parameter estimates, i.e., (confidence interval/parameter) × 100.
that in activated sludge systems.27 Thus, the evaluation results of the experimental data from Hsieh et al.27 using this approach indicated that this convenient method to determine UAP and BAP could be applicable to qualitatively describe a wide range of systems regarding SMP production. Furthermore, the proposed approach is further evaluated by using the experimental results of Jiang et al.18 As shown in Figure 4C and D, the model predictions matched the experimental data, suggesting that this approach was appropriate to determine the UAP and BAP formation. However, the estimated kBAP value, 0.021 h−1, was much higher than that calculated using this approach (0.0041 h−1). In their work the total SCOD was the estimate of the BAP concentration under starvation conditions, which might be an overestimation of the real BAP concentration, resulting in the higher kBAP value. On the contrary, their kUAP value of 0.035 CODUAP g−1 CODS was slightly lower than that of this work (0.042 g CODUAP g−1 CODS). Jiang et al. also obtained the SMP production kinetic parameters using a complex mathematical model.18 Their estimated kUAP and kBAP values, 0.0963 g CODUAP g−1 CODS and 0.0215 h−1, were comparable to those estimated using this approach. However, their model structure included many complex mass balance equations and had to be calibrated using extensive experimental data and complex model solution method. Again, only one data set was needed to obtain UAP and BAP concentrations in this approach. As for the UAP production, even though the model was able to reasonably predict the UAP changing trends, the difference between the simulated and measured UAP concentrations appeared to be high, especially after 3 h. This was attributed to the fact that UAP were produced directly in substrate utilization (0−3 h). When the external substrate became depleted, the UAP underwent net consumption by active heterotrophic biomass. Thus, UAP showed a net accumulation between 0 and 3 h (Phase 1), but most of the UAP was degraded between 3 and 8 h (Phase 2). However, the relatively long duration of the UAP tests and the high MW for the UAP in phase-2 suggest that some or most of the “high-MW UAP” may have been BAP, which might not be retracted from the SCOD to represent the true UAP concentration.18 Finally, SMP data reported in the previous work are also used to validate the proposed approach.11 In Ni et al., a complicated method combining chemical analysis and modeling approach is applied to determine UAP and BAP quantitatively.11 As shown in Figure 4E and F, the model predictions matched the experimental data, suggesting that this approach was appropriate to determine the UAP and BAP formation again. Ni et al. also obtained the SMP production kinetic parameters and the UAP and BAP formation data.11 However, their model included 12 nonsteady-state kinetic equations and 10 SMP-related parameters
To further validate this approach, the predicted SMP and external substrate profiles in Figure 4A were compared with the
Figure 4. Comparisons between the model simulations and the experimental data for the verification of the approach: (A) SMP data set of Hsieh et al.;27 (B) UAP and BAP determinations corresponding to A; (C) BAP data set of Jiang et al.;18 (D) UAP data set of Jiang et al.;18 (E) SMP data set of Ni et al.;11 and (F) UAP and BAP data set of Ni et al.11.
corresponding experimental data from Hsieh et al. for a transient batch experiment.27 Overall, the model outputs captured all the experimental trends for this experiment in terms of SMP production. The external substrate declined at an increasing rate over the initial 10 h. During the period of rapid microbial growth, the SMP increased gradually. Also, the UAP and BAP concentrations were determined as shown in Figure 4B. The two most appropriate estimates of the kUAP and kBAP were 0.03 g CODUAP g−1 CODS and 0.002 h−1, respectively, which were comparable to the values obtained in the present work. In this approach, only one data set was needed to obtain UAP and BAP concentrations by using a convenient kinetic approach, as demonstrated in this work. In addition, the production of SMP in the pure culture system of Hsieh et al. is much higher than 1672
dx.doi.org/10.1021/es202756d | Environ. Sci. Technol. 2012, 46, 1667−1674
Environmental Science & Technology
Article
complex mathematical models. A spreadsheet program (e.g., Microsoft Excel 2003) can also allow convenient application of the approach. Dynamic quantification data for UAP and BAP and reasonable kinetic parameter values can be obtained separately and conveniently. The successful applications in six different SMP production systems demonstrate that this approach is highly applicable in a wide range of engineered system regarding SMP production, which may offer engineers with the basis to optimize reactor operation to avoid a high effluent soluble COD from SMP or SMP-based membrane fouling in MBRs.
and has to be combined with an extensive chemical analysis method, such as MW determination and dissolved organic carbon measurements, to determine the UAP and BAP formation. In contrast, in the present work, a simple kinetic model including only 4 nonsteady-state kinetic equations and 5 SMPrelated parameters could be used to determine the UAP and BAP formation successfully, indicating its effectiveness and convenience again. Smith et al. also developed a convenient method for evaluation of biochemical reaction rate coefficients to avoid the complexity of existing statistical methods for analysis of biochemical rate equations.28 However, their method could not be applied to determine the SMP productions in activated sludge. They proposed a computer spreadsheet program to estimate the kinetic parameters for biochemical reactions, based on the conventional Monod equation for utilization of a substrate in a batch reactor.28 In such a Monod equation, the SMP production is not considered and has been lumped into the yields for cell synthesis. Thus, their methods cannot be used for the determination of UAP and BAP. In addition, the substrate electrons for the synthesis of SMP, which are accounted for the biomass yield in Smith et al.,28 would result in a higher biomass yield value. Actually, SMP are important sinks for electrons and carbon derived from original substrate. The diversion of electrons and carbon for SMP production could otherwise be invested in the cell yield and the growth rate. An ignoring of SMP formation by their method could lead to an overestimation of true cellular growth rates. In the present approach, the SMP formation by activated sludge has been incorporated into the integrated Monod equation (not the conventional Monod equation) for UAP and BAP determination. For better comparison, the experimental data in this work are also used to estimate the bioreaction kinetics using Smith et al. method. The best-fit biomass yield on substrate for Smith et al. method is 0.60 g CODX g−1 CODS, which is higher than the corresponding value of biomass yield on substrate consumed by using this approach (i.e., 0.531 g CODX g−1 CODS). Obviously, Smith et al. method results in an overestimation of true cellular growth rates. The electron distribution for active biomass and UAP production should be 4.2% and 53.1%, respectively. Thus, a significant amount of substrate electrons is distributed to produce SMP in addition to the active cells. In this work, a synthetic influent is used to cultivate activated sludge. Thus, the initial SMP concentrations (BAP0) in the experiments were zero. However, changes in initial SMP concentration originating from the influent wastewater or accumulated from inoculated sludge do not limit the applicability of the proposed approach, because they can be included in the integrated form of the equation. In the real wastewaters, the biodegradable soluble COD (SS) can be estimated via oxygen uptake rate profiles.29 The inert fraction soluble COD (SI) is the source of SMP originating from the influent wastewater, which can then be determined independently by subtracting SS from the soluble COD to give the fraction SI. Thus, in the integrated SMP production equations, the obtained SI values can be adopted as the initial SMP concentration (BAP0). In summary, the integrated method presented in this work is an effective and convenient approach to determine the UAP and BAP production of activated sludge. This approach is attractive for the complex SMP production processes because a simple kinetic model including only 4 nonsteady-state kinetic equations and 5 SMP-related parameters is developed and applied successfully, other than the complicated chemical methods or
■
ASSOCIATED CONTENT
S Supporting Information *
Additional methods and results. This material is available free of charge via the Internet at http://pubs.acs.org.
■ ■
AUTHOR INFORMATION
Corresponding Author
*Fax: +86-551-3601592. E-mail: hqyu@ustc.edu.cn.
ACKNOWLEDGMENTS The authors wish to thank the Natural Science Foundation of China (50878203 and 50738006) and the Fundamental Research Funds for the Central Universities (WK2060190007) for the partial support of this study.
■
REFERENCES
(1) Barker, D. J.; Stuckey, D. C. A review of soluble microbial products (SMP) in wastewater treatment systems. Water Res. 1999, 33, 3063−3082. (2) Laspidou, C. S.; Rittmann, B. E. A unified theory for extracellular polymeric substances, soluble microbial products, and active and inert biomass. Water Res. 2002, 36, 2711−2721. (3) Rosenberger, S.; Laabs, C.; Lesjean, B.; Gnirss, R.; Amy, G.; Jekel, M.; Schrotter, J. C. Impact of colloidal and soluble organic material on membrane performance in membrane bioreactors for municipal wastewater treatment. Water Res. 2006, 40, 710−719. (4) Rittmann, B. E.; McCarty, P. L. Environmental biotechnology: principles and applications. Mc-Graw Hill: New York, NY, 2001. (5) Jarusutthirak, C.; Amy, G. Understanding soluble microbial products (SMP) as a component of effluent organic matter (EfOM). Water Res. 2007, 41, 2787−279. (6) Jarusutthirak, C.; Amy, G. Role of soluble microbial products (SMP) in membrane fouling and flux decline. Environ. Sci. Technol. 2006, 40, 969−974. (7) Grady, C. P. L., Jr.; Daigger, G. T.; Lim, H. C. Biological Wastewater Treatment, 2nd ed.; Marcel Dekker: New York, 1999; pp 282−284. (8) Ni, B. J.; Rittmann, B. E.; Yu, H. Q. Soluble microbial products and their implications in mixed culture biotechnology. Trends Biotechnol. 2011, 29, 454−463. (9) Jeong, T.-Y.; Cha, G.-C.; Yoo, I.-K.; Kim, D.-J. Characteristics of bio-fouling in a submerged MBR. Desalination 2007, 207, 107−113. (10) Meng, F.; Chae, S. R.; Drews, A.; Kraume, M.; Shin, H. S.; Yang, F. Recent advances in membrane bioreactors (MBRs): Membrane fouling and membrane material. Water Res. 2009, 43, 1489−1512. (11) Ni, B. J.; Zeng, R. J.; Fang, F.; Xie, W. M.; Sheng, G. P.; Yu, H. Q. Fractionating soluble microbial products in the activated sludge process. Water Res. 2010, 44, 2292−2302. (12) Jiang, T.; Kennedy, M. D.; De Schepper, V.; Nam, S. N.; Nopens, I.; Vanrolleghem, P. A.; Amy, G. Characterization of soluble microbial products and their fouling impacts in membrane bioreactors. Environ. Sci. Technol. 2010, 44, 6642−6648. (13) Smith, P. J.; Vigneswaran, S.; Ngo, H. H.; Ben-Aim, R.; Nguyen, H. A new approach to backwash initiation in membrane systems. J. Membr. Sci. 2006, 278, 381−389. 1673
dx.doi.org/10.1021/es202756d | Environ. Sci. Technol. 2012, 46, 1667−1674
Environmental Science & Technology
Article
(14) Ciston, S.; Lueptow, R. M.; Gray, K. A. Controlling biofilm growth using reactive ceramic ultrafiltration membranes. J. Membr. Sci. 2009, 342, 263−268. (15) Al-Halbouni, D.; Dott, W.; Hollender, J. Occurrence and composition of extracellular lipids and polysaccharides in a full-scale membrane bioreactor. Water Res. 2009, 43, 97−106. (16) Kimura, K.; Naruse, T.; Watanabe, Y. Changes in characteristics of soluble microbial products in membrane bioreactors associated with different solid retention times: Relation to membrane fouling. Water Res. 2009, 43, 1033−1039. (17) Okamura, D.; Mori, Y.; Hashimoto, T.; Hori, K. Identification of biofoulant of membrane bioreactors in soluble microbial products. Water Res. 2009, 43, 4356−4362. (18) Jiang, T.; Myngheer, S.; De Pauw, D. J. W.; Spanjers, H.; Nopens, I.; Kennedy, M. D.; Amy, G.; Vanrolleghem, P. A. Modelling the production and degradation of soluble microbial products (SMP) in membrane bioreactors (MBR). Water Res. 2008, 42, 4955−4964. (19) Laspidou, C. S.; Rittmann, B. E. Non-steady state modeling of extracellular polymeric substances, soluble microbial products, and active and inert biomass. Water Res. 2002, 36, 1983−1992. (20) Mu, Y.; Yu, H. Q. Biological hydrogen production in a UASB reactor with granules I: Physicochemical characteristics of hydrogenproducing granules. Biotechnol. Bioeng. 2006, 94, 980−987. (21) APHA, Standard Methods for the Examination of Water and Wastewater, 19th ed.; American Public Health Association: New York, 1995. (22) Laspidou, C. S.; Rittmann, B. E. Evaluating trends in biofilm density using the UMCCA model. Water Res. 2004, 38, 3362−3372. (23) Ni, B. J.; Fang, F.; Rittmann, B. E.; Yu, H. Q. Modeling microbial products in activated sludge under feast-famine conditions. Environ. Sci. Technol. 2009, 43, 2489−2497. (24) Lu, S. G.; Imai, T.; Ukita, M.; Sekine, M.; Higuchi, T.; Fukagawa, M. A model for membrane bioreactor process based on the concept of formation and degradation of soluble microbial products. Water Res. 2001, 35, 2038−2048. (25) Aquino, S. F.; Stuckey, D. C. Integrated model of the production of soluble microbial products (SMP) and extracellular polymeric substances (EPS) in anaerobic chemostats during transient conditions. Biochem. Eng. J. 2008, 38, 138−146. (26) Ahn, Y. T.; Choi, Y. K.; Jeong, H. S.; Chae, S. R.; Shin, H. S. Modeling of extracellular polymeric substances and soluble microbial products production in a submerged membrane bioreactor at various SRTs. Water Sci. Technol. 2006, 53, 209−216. (27) Hsieh, K. M.; Murgel, G. A.; Lion, L. W.; Schuller, M. L. Interactions of microbial biofilms with toxic trace metals 1. Observation and modeling of cell growth, attachment, and production of extracellular polymer. Biotechnol. Bioeng. 1994, 44, 219−231. (28) Smith, L. H.; McCarty, P. L.; Kitanidis, P. K. Spreadsheet method for evaluation of biochemical reaction rate coefficients and their uncertainties by weighted nonlinear least-squares analysis of the integrated Monod equation. Appl. Environ. Microbiol. 1998, 64, 2044−2050. (29) Sin, G.; Guisasola, A.; de Pauw, D. J. W.; Baeza, J. A.; Carrera, J.; Vanrolleghem, P. A. A new approach for modelling simultaneous storage and growth processes for activated sludge systems under aerobic conditions. Biotechnol. Bioeng. 2005, 92, 600−613.
1674
dx.doi.org/10.1021/es202756d | Environ. Sci. Technol. 2012, 46, 1667−1674