lable at ScienceDirect

Journal of Environmental Management 146 (2014) 59e68

Contents lists avai

Journal of Environmental Management

journal homepage: www.elsevier .com/locate/ jenvman

Microbial kinetic for In-Storage-Psychrophilic Anaerobic Digestion(ISPAD)

Mahsa Madani-Hosseini, Catherine N. Mulligan, Suzelle Barrington*

Department of Building, Civil and Environmental Engineering, Concordia University, 1455 de Maisonneuve, Montr�eal H3G 1M8, Canada

a r t i c l e i n f o

Article history:Received 9 May 2014Received in revised form21 July 2014Accepted 22 July 2014Available online

Keywords:Psycrophilic anaerobic digestionSwine manureKinetic coefficients

* Corresponding author. Tel.: þ1 1 450 773 6155x6E-mail address: suzellebarrington@sympatico.ca (S

http://dx.doi.org/10.1016/j.jenvman.2014.07.0280301-4797/© 2014 Elsevier Ltd. All rights reserved.

a b s t r a c t

In-Storage-Psychrophilic-Anaerobic-Digestion (ISPAD) is a wastewater storage tank converted into ananaerobic digestion (AD) system by means of an airtight floating geo-membrane. For process optimi-zation, ISPAD requires modelling with well-established microbial kinetics coefficients. The present ob-jectives were to: obtain kinetics coefficients for the modelling of ISPAD; compare the prediction of theconventional and decomposition fitting approach, an innovative fitting technique used in other fields ofscience, and; obtain equations to predict the maximum growth rate (mmax) of microbial communities as afunction of temperature. The method consisted in conducting specific Substrate Activity Tests (SAT) usingISPAD inoculum to monitor the rate of degradation of specific substrates at 8, 18 and 35 �C. Microbialkinetics coefficients were obtained by fitting the Monod equations to SAT. The statistical procedure ofLeast Square Error analysis was used to minimize the Sum of Squared Errors (SSE) between the measuredISPAD experimental data and the Monod equation values. Comparing both fitting methods, thedecomposition approach gave higher correlation coefficient (R) for most kinetics values, as compared tothe conventional approach. Tested to predict mmax with temperature, the Square Root equation betterpredicted temperature dependency of both acidogens and propionate degrading acetogens, while theArrhenius equation better predicted that of methanogens and butyrate degrading acetogens. Increasingtemperature from 18 to 35 �C did not affect butyrate degrading acetogens, likely because of theirdominance, as demonstrated by microbial population estimation. The estimated ISPAD kinetics co-efficients suggest a robust psychrophilic and mesophilic coexisting microbial community demonstratingacclimation to ambient temperature.

© 2014 Elsevier Ltd. All rights reserved.

1. Introduction

In-Storage-Psychrophilic-Anaerobic-Digestion (ISPAD) is theconversion of a standard exterior storage tank into an anaerobicdigester by means of an airtight geo-membrane cover collectingbiogas (King et al., 2011b). With ISPAD, the tank is gradually filledover an extended period of at least 100 days, thus compensating forits psychrophilic temperature fluctuating with ambient climaticconditions. For ISPAD, microbial acclimation, protein degradation(King et al., 2011a, 2011b) and biogas generation were investigated(Giard et al., 2013; Nohra et al., 2003). Nevertheless, process opti-mization is still required through modelling with well-establishedmicrobial kinetic coefficient values, such as maximum microbialgrowth rate (mmax) and fluctuation with temperature, microbialyield (Y), and substrate half saturation constant (Ks). In AD systems,microbial kinetics coefficients are useful modelling parameters to

03; fax: þ1 450 773 3133.. Barrington).

evaluate the behaviour of microbial population, the rate of sub-strate degradation and the biogas production (Jim�enez et al., 2006).Based on the determination of kinetics coefficients, Nwabanne et al.(2009) concluded that the rate of digestion of a municipal solidwaste digester could be corrected through inoculation.

A variety of methods have been used in kinetic parameteridentification, because the AD process is characterized by its highcomplexity and non-linearity. This causes variability in reportedkinetic parameter values even when the same operational andenvironmental conditions have been evaluated (Donoso-Bravoet al., 2011). The mode of operation (e.g., batch vs. continuous)and the environmental and operational conditions (e.g., pH, tem-perature, organic load) are other factors resulting in kineticparameter variability (Pavlostathis and Giraldo-Gomez, 1991). TheISPAD system operates at a temperature varying with that ofambient and under a steadily decreasing loading rate betweenbatch-feeding, as compared to constant parameters in conventionalAD systems. Therefore, ISPAD systems operate under very specificmicrobial kinetic values which need to be determined.

Nomenclature, subscript and expressions

SubscriptM Acetoclastic methanogensAB Butyrate degrading acetogenicAP Propionate degrading acetogenicA Acidogensglu Glucosepr Propionatebut Butyrateac Acetate

SymbolAD Anaerobic DigestionISPAD In-Storage-Psychrophilic Anaerobic DigestionVFA Volatile Fatty AcidGRG General Reduced Gradientmmax Maximum microbial growth rateKs Half saturation constantY Yield coefficientKd Decay rate constantm Microbial growth rateX Microbial population densityBMP Biochemical Methane Production

SAT specific Substrate Activity TestSSE Sum of Square ErrorTObsi Observed dataTPredi Model predicted outputn Number of dataK Reaction rateR* Universal gas constantEa Activation energyb Regression coefficientT TemperatureTmin Apparent minimum temperatureMAE Mean Absolute ErrorMSE Mean Squared ErrorFB Fractional BiasRSME Root Mean ErrorNMSE Normalized Mean Squared ErrorMAPE Mean Absolute Percentage ErrorR2 Coefficient of determinationR Correlation coefficientCOD Chemical Oxygen DemandTS Total SolidVS Volatile SolidVSS Volatile Suspended SolidVDS Volatile Dissolved SolidFS Fixed Solid

M. Madani-Hosseini et al. / Journal of Environmental Management 146 (2014) 59e6860

The objectives of this study were to: 1) statistically fit laboratoryspecific substrate activity test (SAT) results to that of the Monodequation applied in series to the various AD microbial groups togenerate ISPAD kinetics coefficients and microbial densities; 2)compare the prediction accuracy of the conventional fittingmethod to that of the decomposition approach, and; 3) comparethe prediction accuracy of both the Arrhenius and Square Rootequations to describe the relationship between temperature andmmax. The laboratory data was produced using SAT conducted withISPAD inoculum acclimated to swine manure and individual sub-strates namely, glucose, propionate, butyrate and acetate, at 8, 18and 35 �C. The curve fitting process was limited to known kineticparameter ranges and yielded microbial population densities, X, forall 4 main degradation groups.

2. Kinetic coefficient determination

A variety of laboratory AD methods are used to obtain data forthe computation of microbial kinetics coefficients through statis-tical fitting. Three different feeding methods are used, namelybatch (Donoso-Bravo et al., 2009; Flotats et al., 2003), continuous(Batstone et al., 2009; Bernard et al., 2001) and fed-batch (Redzwanand Banks, 2004; Rodrigues et al., 2003). Two batch tests arecommonly used in the determination of AD kinetics coefficientsbecause of their simplicity and short experimental duration:biochemical methane production (BMP) and specific substrate ac-tivity test (SAT). The BMP assay is a long-term batch incubation ofcultures with periodic biogas and reactor content sampling toprovide a temporal profile of substrate consumption and methaneproduction from a known initial concentration of active biomassand substrate (Chynoweth et al., 1993; Shelton and Tiedje, 1984).However, the SAT is a shorter assay measuring the rate of con-sumption of individual substrates used by one of the main ADmicrobial groups. Such an assay provides more specific data for thefitting of individual AD process. Donoso-Bravo et al. (2009) usedstarch, glucose and acetic acid as the main substrates in SAT to

obtain kinetics coefficients for hydrolysis, acidogenesis and meth-anogenesis, respectively. Using SAT, Flotats et al. (2003) also esti-mated kinetics coefficients for the AD of valerate.

The experimental data is used to estimate kinetics coefficientsby changing their value till the model predicts a response corre-sponding to that obtained experimentally. The selection of anappropriate set of modelling equations is important. In kineticscoefficient estimation, the simple Monod equation is accurate andsimple when a limited number of parameters are unknown. For AD,the kinetic values of mmax, Ks, Y, and X are interdependent, whichmakes their estimation difficult, challenging and tedious. Used inother scientific fields, the decomposition approach is applied in thiswork as an innovative method of estimating interdependent pa-rameters through data fitting (Bahn et al., 1996; Harjunkoski andGrossmann, 2001). This approach requires the decomposition ofthe designed algorithms into sub-problems and the solving of in-dividual parameters by data fitting, in order of importance. Theadvantages of this approach are to: 1) reduce the possibilities ofdropping into local minima in the process of optimization; 2)reduce the computation time; and 3) increase the fitting accuracycompared to other approaches (Jiang and Cheng, 2005).

Using the decomposition approach, each sub-problem can besolved through an objective or cost function, recognized as themost common fitting tool. The most useful cost function is the Sumof Square Error (SSE) (Batstone et al., 2009; Donoso-Bravo et al.,2010; L�opez and Borzacconi, 2010; Noykova and Gyllenberg, 2000)which assumes that the standard deviation of the measurementerrors is constant:

SSE ¼Xn

i¼1

�tobsi � tpredi

�2(1)

where the observed data, the model-predicted outputs, and thenumber of data are shown as tobsi , tpredi , and n, respectively.

To optimize the cost function, several algorithms were devel-oped, namely grouped under the Exact and Heuristic methods. An

M. Madani-Hosseini et al. / Journal of Environmental Management 146 (2014) 59e68 61

Exact method is a mathematical procedure generating a sequenceof solutions improving the order of a class of problems. Thismethoduses convexity assumptions for the cost function to obtain theoptimized values. The main drawback is the risk of getting trappedin local minima, specifically when the cost function is non-linear.The common approach to solve this trap is to start the search us-ing several randomly selected initial parameters, which is called amulti-start strategy. Solving the problem more quickly, the Heu-ristic methods are multi-start by nature, but are highly sensitive tothe initial parameters. Also, the Heuristic methods are better atfinding an approximate solution when classic methods fail. This isachieved by trading optimality, completeness, accuracy, and/orprecision for speed.

Both the Exact and Heuristic methods were used in AD kineticdetermination. Examples for the Exact methods are: LevenbergeMarquardt (García-Ochoa et al., 1999; Lokshina et al., 2001),Sequential Quadratic Programming (Aceves-Lara et al., 2005; Sales-Cruz and Gani, 2004); Multiple Shooting (L�opez and Borzacconi,2010; Müller et al., 2002), and Direct Search also called SimplexAlgorithm (Haag et al., 2003; M€osche and J€ordening, 1999;Simeonov, 1999). Applications of the Heuristic methods are:Simulated Annealing (Haag et al., 2003), Genetic Algorithms(Wichern et al., 2009), and Particle Swarm Optimization (Wolfet al., 2008).

The Heuristic methods require a certain level of knowledge andexperience, making them difficult and expensive to use, besidestheir sensitivity to initial values and stopping criteria. Conse-quently, Heuristic methods may be more inefficient than the Exactmethods with local optima issues.

Because of limitations offered by Heuristic methods, this studyused an Exact method, namely the multi-start strategy, to obtainISPAD kinetics coefficients. Since the risk of getting trapped in localminima is sensitive to the initialization parameters, the kineticscoefficient obtained for the Keshtkar et al. (2001) model were used.The Keshtkar et al. (2001) model was designed for manure sub-strates, includes inhibition factors for volatile fatty acid (VFA), pHand free NH3, and describes a cyclic batch reactor, similar to ISPAD.This model considers five steps: 1) the hydrolysis of particulatesubstrate by extracellular enzymes; 2) the consumption of solublesubstrates by acid forming bacteria; 3) the consumption of VFA; 4)the formation of acetate by propionate and butyrate degradingacetogens, and finally; 5) the consumption of acetate to generatemethane by methanogens.

3. Temperature functions

In AD systems, two equations are generally used to predicttemperature dependence, namely the Arrhenius and the SquareRoot equations. According to the Arrhenius equation (Equation (2)),the reaction rate roughly doubles for a temperature increase of10 �C:

K ¼ Ae�EaR*T (2)

where K is the reaction rate, A is a constant, R* is the universal gasconstant (0.008314 kJ/mol K), T is temperature (K) and Ea is theactivation energy (kJ/mol).

The Square Root equation is simpler and purely descriptive ofthe evolution of microbial growth rate with temperature(Ratkowsky et al., 1982). The Square Root equation describes a lessthan optimal temperature adaptation of microbial growth in pureculture:

ffiffiffiffiK

p¼ b

�T � Tmin

�(3)

where K is the reaction rate (or growth rate in the case of microbes)at temperature T (K), Tmin is the apparent minimum temperaturefor growth (K), and b is the regression coefficient.

In most AD models, the Arrhenius equation was applied to thegrowth rate, m (Siegrist et al., 2002; Sinechal et al., 1979; Srisertpolet al., 2010), the maximum growth rate, mmax (Angelidaki andAhring, 1993; Hashimoto, 1983), the saturation constant,Ks (Dague et al., 1998; Siegrist et al., 2002), the hydrolysis rate, K, thedeath rate, Kd (Donoso-Bravo et al., 2009; McKinney, 1963; Siegristet al., 2002; Veeken and Hamelers, 1999), the inhibition constants,Ki (Siegrist et al., 2002), and the yield coefficient from substrate tobiomass, Y (McKinney, 1963). However in some cases, the SquareRoot equation provided a better temperature prediction than theArrhenius.With granular sludge adapted to 10 �C over 235 days, thetemperature dependence for the methanogenic conversion of ac-etate was well described by the Arrhenius equation, but the SquareRoot equation better predicted the propionate, butyrate and mixedVFA activities (Rebac et al., 1995). In another study, the Arrheniusequation was a poor predictor of temperature dependence for themethanogenic activity of a biomass adapted to lower temperaturesof 5e29 �C (Kettunen and Rintala, 1997). Accordingly, both theArrhenius and Square Root equations will be tested to predicttemperature dependence of the maximum growth rate for the ADmicrobial communities.

4. Material and methods

4.1. Inoculum characterization and analytical procedure

In 2004, a full-scale swine manure ISPAD facility was estab-lished at Saint-Francois-Xavier in the central region of theProvince of Quebec, Canada. This facility consisted of a circularconcrete tank measuring 30 m in diameter by 3.66 m in depth,and covered with an airtight membrane (GTI, Fredericton, NB,Canada). The tank received manure from the swine facility on aweekly basis and was emptied twice yearly. Manure samplesfrom this ISPAD installation were brought to the laboratory foranalysis using standard methods (Eaton and Franson, 2005) andSATs.

Samples were analysed for solids (total solids, TS; volatile solids,VS; total suspended solids, TSS, and volatile suspended solids, VSS),COD (total and soluble chemical oxygen demand), VFA, anions andcations. Total solids were determined by drying at 103 �C overnight(VWR, Sheldon Manufacturing, model 1327F, OR, USA). Volatilesolids were determined by incineration at 500 �C for two hours(Barnstead Thermodyne model 48000, IA, USA). Suspended solidswere separated from the supernatant by centrifuging at 1000 rpmfor 10 min at 4 �C. Chemical oxygen demand was measured usingthe potassium perchromate method and a spectrophotometer(Hach model DR 2800, CO, USA). The pH of all samples wasmeasured using a pH meter (Corning model 450, NY, USA).

Volatile fatty acids (VFAs) (acetic, propionic and butyric acids)were analysed on a gas chromatograph equipped with a flameionization detector. Anions (NO2

- , NO3- , PO4

3�, Cl�) were analysedusing a polymer-based chromatography column, 250 mm � 41 mmOD (model PRP-X100, Hamilton, NV, USA), on a high-performanceliquid chromatograph (model P4000 & AS3000, TSP). Conductiv-ity data were obtained by using a Waters Millipore detector model432. The parameters were:mobile phase 4.0mMp-hydroxybenzoicacid, pH 8.5 with 2.5% methanol, 100 mL injection, 1.8 cm3/min flowrate at 40 �C. Cations (Naþ, NH4

þ, Kþ) were similarly analysed on acation resin-based chromatography column, 250 mm � 41 mmOD (model PRP-X200, Hamilton) with: mobile phase 4.0 mM nitricacid with 30% methanol, 20 mL injection, and 1.8 cm3/min flow rateat 40 �C. Table 1 shows the measured manure characteristics.

Table 2Kinetic coefficient used as initial values (Keshtkar et al., 2001).

Process Parameter Unit Value

1. Acidogenesis mmaxA 1/d 5.0Ksglu mg/L 500YA mg/mg 0.077

2. Propionate degradingacetogenic

mmaxAP 1/d 0.54Kspr mg/L 259YAP mg/mg 0.094

3. Butyrate degradingacetogenic

mmaxAB 1/d 0.68

Ksbut mg/L 176YAB mg/mg 0.083

4. Methanogenesis mmaxM 1/d 0.60Ksac mg/L 120YM mg/mg 0.04

M. Madani-Hosseini et al. / Journal of Environmental Management 146 (2014) 59e6862

4.2. Computer simulation and statistical fitting method

The experimental data to compute the kinetic parameters wasobtained from SATs conducted by King (2011). The SATs wereconducted at 8, 18 and 35 �C using ISPADmanure as inoculum. Fourindividual liquid substrate assays were conducted, where,excluding hydrolysis, each assay applied the substrate used by oneof the main AD microbial group: glucose, acetate, propionate andbutyrate. Describing the substrate uptake behaviour of the ISPADmicrobial communities, this data provided parameters for curvefitting and kinetics coefficient determination (Donoso-Bravo et al.,2009).

Each experimental uptake data set can be fitted mathematicallyto an AD equation simulating the specific substrate uptake ratesuch as the Monod equation (Donoso-Bravo et al., 2009; Goudaret al., 1999; Robinson and Tiedje, 1983). For this, the Monodequation (Equations (4)e(6)) was programmed using Excel(Microsoft 2010), where mi, dXi/dt and dSj/dt were computed insteps of short time increment:

mi ¼ mmaxiSj

Sj þ Ksii ¼ A; AP; AB; M (4)

dXi

dt¼ ðmi � KdiÞXi (5)

dSjdt

¼ �1Yi

dXi

dtj ¼ glu;pr;but; and ac (6)

where S is the concentration of substrate in mg/L, X is the con-centration of active microbial biomass in mg/L, Y is the yield ofmicrobial biomass from the substrate in mg biomass/mg substrate,Ks is the half-saturation constant in mg substrate/L, Kd is the bac-terial decay constant which is considered equal to 5% of mmax, jrepresents the substrates of glucose (glu), propionate (pr), butyrate

Table 1Characteristics of the experimental manure.1

Characteristic Unit Freshmanure

ISPADmanure

SolidsTS g/L 48.01 38.71VS g/L 34.34 25.40FS g/L 13.63 13.31VSS g/L 27.38 24.01VDS g/L 6.96 1.38pH e 6.90 7.46CODTotal g/gVS 2.43 1.99Soluble g/gVS 0.88 0.08VFAAcetic mg/gVS 142.04 0.33Propionic mg/gVS 60.31 0.00Butyric mg/gVS 40.69 0.00AnionsCl� mg/gVS 33.81 21.33NO�

2 mg/gVS 2.93 0.04NO�

3 mg/gVS 0.00 0.00PO3�

4 mg/gVS 15.61 6.94SO2�

4 mg/gVS 0.00 17.38CationsNaþ mg/gVS 19.19 13.43NHþ

4 mg/gVS 108.88 79.16Kþ mg/gVS 70.23 34.51

ATPATP mg/gVS 12 16.7Active %VSS 0.4e1.2 0.5e1.6

1 King (2011)

(but), and acetate (ac), and A, AP, AB and M represent acidogens,propionate degrading acetogens, butyrate degrading acetogens,and acetoclastic methanogens, respectively.

As the kinetic values of mmax, Ks, Y, and X are interdependent,the decomposition approach was used and kinetics coefficientswere determined in order of importance. Since the Exact methodcan produce local minima traps, the fitting process was initiatedby computing X values using the kinetics coefficients of Keshtkaret al. (2001) at 35 �C (Table 2). Microbial population densities, X,are critical in determining ISPAD kinetics coefficients (Goudaret al., 1999; Rebac et al., 1999), especially considering their lackof reference in the literature. The fitted X value pertained toacidogens, propionate degrading acetogens, butyrate degradingacetogens, and acetoclastic methanogens. Once the X values wereestablished, the kinetics coefficients were fitted for 35 �C, usingthe Monod equation results and boundary values for mmax, Ks, andY based on ADM1 (Batstone et al., 2002) (Table 3). Also, the 35 �Cvalues for Y were presumed to apply to 18 and 8 �C conditionsbecause of weak sensitivity to temperature. The 35 �C values for Xand Y were used to fit the 18 and 8 �C data. The Microsoft solver(Microsoft 2010) was used to minimizing the SSE between calcu-lated and experimental values. To assess the effectiveness of thedecomposition approach, the fitting process was also performedusing the conventional approach of fitting all kinetic parameters ofmmax, Ks, Y, and X simultaneously at 35 �C, and then using theoptimized X and Y values for the fitting of mmax and Ks at 8 and18 �C.

Since the difference between the experimental substrate uptakedata and the modelled values produced smooth nonlinear graphs,the Generalized Reduced Gradient algorithms (GRG) of the ExcelMicrosoft solver were used to reduce the likelihood of local mini-mum traps. An example of such a trap is the convergence definingthe algorithm stopping criterion which is the amount of relativechange specified in the last five iterations. To further reduce thelikelihood of local minimum traps, the following features in GRGwere enabled: 1) “multistart” for repeated run startingwith specificdecision variable values; 2) “random seeds” where GRG generatescandidate starting points, and; 3) forward and central “derivatives”to find the optimal trajectory for further iterations.

The trade-off between the accuracy of the solution, the optimalvalues of kinetic coefficients, and the time and difficulty level of thealgorithm is an important issue. Specifically, central derivativeswere used because their approach is more accurate when changingrapidly at the current point, but this operation requires morerecalculations. The algorithm was allowed to use the multistartmethod set at 10�3 for the convergence feature. Finally, populationsize was set at 20 and 3 random seeds.

To compare the temperature prediction accuracy of the Arrhe-nius and the Square-Root equations for mmax, theses models were

Table 3ISPAD kinetic coefficients at 8, 18 and 35 �C.

Process Parameter Unit Range of reported data(Batstone et al., 2002)

Value

8 �C 18 �C 35 �C

1. Acidogenesis mmaxA 1/d 0.4e21.12 0.64 2.90 6.40Ksglu mg/L 22e1280 219.20 167.64 140.20YA mg/mg 0.01e0.17 0.123 0.123 0.123XA mg/L ND1 7.54 7.54 7.54

2. Propionate degradingacetogenesis

mmaxAP 1/d 0e1.64 0.011 0.063 0.12Kspr mg/L 20e1146 392.00 163.70 100.50YAP mg/mg 0.019e0.089 0.053 0.053 0.053XAP mg/L ND 18.32 18.32 18.32

3. Butyrate degradingacetogenesis

mmaxAB 1/d 0.021e2.64 0.023 0.22 0.23Ksbut mg/L 12e450 411.38 450 450YAB mg/mg 0.026e0.079 0.034 0.034 0.034XAB mg/L ND 85.96 85.96 85.96

4. Methanogenesis mmaxM 1/d 0.0192e1.2 0.045 0.20 0.40Ksac mg/L 11e930 533.77 213.30 193.33YM mg/mg 0.014e0.076 0.019 0.019 0.019XM mg/L ND 23.59 23.59 23.59

1 No Data.

M. Madani-Hosseini et al. / Journal of Environmental Management 146 (2014) 59e68 63

fitted with the obtained mmax values (Table 3). The initially usedcoefficients of Ea and Tmin were based on the literature (El-Mashadet al., 2005; Finster, 2008). Parameters of both models were thenoptimized using GRG algorithms, minimizing SSE between thecalculated and the experimental values. The prediction accuracy oftwo equations was then compared by the Mean Absolute Per-centage Error (MAPE), the Fractional Bias (FB), the Root MeanSquare Error (RMSE), the Normalized Mean Square Error (NMSE),and the Coefficient of Determination (R2) (Table 5) (Kusiak andWei,2012).

5. Results and discussion

5.1. Characteristics of ISPAD manure

The analytical results of analyses performed on the ISPADmanure are presented in Table 1. Differences between the freshmanure characteristics reported in the literature and ISPADmanurecharacteristics may therefore be considered to represent the resultof processes occurring in the ISPAD tank. The low COD and VSvalues reflect the consumption of organic matter by the AD process(King, 2011). The reduced concentrations of VFAs in the ISPADmanure, as well as the resulting increase in pH, confirm thatthe process was well acclimated to operating conditions(Kotsyurbenko, 2005).

5.2. The experimental substrate consumption rate

For temperatures of 8, 18, and 35 �C, the substrate consumptionrate data obtained from SAT assays can be classified into threetypes: exponential, linear and linear-exponential curves. Theglucose activity curves at temperatures 8, 18 and 35 �C, showed asmooth exponential trend with an early low uptake rate followedby a higher rate associated with microbial growth. The curve shapeis stretched out in time as temperature decreases, showing thatacidogens consumed glucose faster at higher temperatures.

The propionate consumption rate at 8 �Cwas linear compared toexponential at 18 and 35 �C (Fig. 1). This linear 8 �C curve resultedfrom either a slow growth at low temperatures (Arbeli et al., 2006;McHugh et al., 2004; €Oztürk, 1993) or inhibition by acetate andbutyrate (Vavilin and Lokshina, 1996). Methane production frompropionate is known to be slower than that from butyrate and ac-etate, because of its thermodynamically unfavourable AD process

(Gijzen et al., 1988). Propionate consumption doubled by increasingthe temperature from 18 to 35 �C.

As opposed to other substrates in Fig. 1, butyrate consumingacetogens demonstrated a linear trend for 8, 18 and 35 �C, with thecurve slope increasing especially between 8 and 18 �C. It wasdemonstrated that low temperatures of 3e9 �C favour the degra-dation of butyrate over propionate (Nozhevnikova et al., 2000).There was no important increase in butyrate consumption ratebetween 18 and 35 �C, indicating a lack of microbial temperaturesensitiveness. Finally, the strong slope of the 8 �C linear curve, asopposed to propionate, suggests a robust butyrate-consumingacetogen population with good growth. Accordingly, butyrate didnot accumulate at low temperatures as opposed to propionate, asalso found by Langenhoff and Stuckey (2000).

Acetate consumption curves were linear-exponential at 8, 18and 35 �C but of a slope similar to that of butyrate. Furthermore, thecurve slope increased with temperature confirming the change inbiomass activity, despite the fact that their linear-exponential trendreflected limited growth.

5.3. The decomposition versus the conventional fitting approach

Using the decomposition approach, the curve fitting methodproduced reliable kinetic coefficients, with a goodness of fit rep-resented by a Correlation Coefficients (R) for non-linear curves(Ting and Shiqiang, 2011), ranging from 0.87 to 0.99 (Fig. 1).However, the conventional approach produced R values rangingfrom 0.58 to 0.99 (Fig. 2). Both approaches produced similar Rvalues for the uptake rate curves such as for propionate, butyrate,and acetate curves at 35 �C, but, for other curves, the R value for theconventional approach was lower. For example, the R value forglucose, butyrate, and acetate curves at 8 �C are respectively 52, 67,and 26% higher if the fitting process is performed using thedecomposition approach. The goodness of fit was also comparedusing the Mean Squared Error (MSE), and the Mean Absolute Error(MAE) besides the R value for all steps at all 3 temperatures(Table 4). For the decomposition approach, both MSE and MAEwere lower as compared to the conventional approach, especially at8 �C, compared to the conventional approach. The conventionalapproach was slightly more accurate at predicting propionatedegrading acetogenesis at both 18 and 35 �C, and butyratedegrading acetogenesis at 18 �C.

In general, the conventional approachmay result in a drop of thelocal minima in finding different kinetic parameters. Rather, the

Fig. 1. Glucose, propionate, butyrate and acetate uptake rate by ISPAD biomass using decomposition approach: experimental data (D), and model prediction (�).

M. Madani-Hosseini et al. / Journal of Environmental Management 146 (2014) 59e6864

decomposition approach offers advantages in terms of improvingdata fitting.

5.4. Maximum growth rate coefficient and temperature effect

Acidogens, propionate degrading acetogens, butyrate degradingacetogens and methanogens demonstrated a mmax decreasing by afactor of 10, 10.9, 10 and 8.8 from 35 to 8 �C, respectively (Table 3).For all microbial groups, temperature and mmax were positivelycorrelated due to activation of enzymatic reactions (Kayranli andUgurlu, 2011).

Growing faster than the other AD microbial groups (Kashyapet al., 2003), the acidogens (glucose degrading bacteria) exhibitedthe fastest mmax at all temperatures, followed by the methanogensand the butyrate degrading acetogens, and then, with the slowestgrowth rate, the propionate degrading acetogens. At 35 �C, theacidogens demonstrated a mmax of 6.40 d�1 as compared to thepropionate and butyrate acetogens with a mmax of 0.12 and 0.23 d�1,respectively. Methanogens demonstrated a mmax of 0.40 d�1 at35 �C. When temperature increased from 18 to 35 �C, there was noincrease in mmax for butyrate consuming bacteria, but by increasingthe temperature from 8 to 18

�C, mmax increased by a factor of 10.

To predict the impact of temperature on mmax for each stage ofAD, both the Arrhenius and the Square Root equations were tested(Table 5) for accuracy through their R2 values and the statisticalparameters MAPE, RMSE, NMSE and FB. The FB is a nonlinear

operator, which is used to represent the relative difference betweenmodel and experimental data. It varies between þ2 for extremeunder prediction, to �2 for extreme over prediction. The otherstatistical estimators of MAPE, RMSE, and NMSE represent theoverall deviation between experimental data and model results.

For the Arrhenius equation, the optimized Ea values for each stepof the AD process were in the range of 38.07e43.42 kJ/mol, as re-ported in the literature (Ngozi-Olehi et al., 2010). For the SquareRoot equation, the optimized Tmin was in the range of 250e262 K(Table 5) for psychrophilic organisms (Bowman, 2001).

For acidogens, despite close R2 values of 0.95 and 0.97, theSquare-Root equation provided a better prediction, as compared tothat of Arrhenius, because of lower FB, MAPE and NMSE. King(2011) showed that ISPAD acidogen growth rate versus tempera-ture did not obey the Arrhenius equation, because of coexistingpsychrophilic and acclimated mesophilic population.

Similarly for acetogens consuming propionate (Table 5) andalthough both the Square Root and Arrhenius equations gave aclose R2 of 0.94 and 0.92 respectively, the former shows a betterprediction, because of a lower FB. The butyrate degrading acetogensshowed a different behaviour, with a mmax better estimated by theArrhenius equation, because of its higher R2 and lower FB values(Table 5). Finally for the methanogens consuming acetate, R2 forboth the Arrhenius and Square-Root equations were close, but theArrhenius equation provided a better prediction with a lower FBvalue.

Fig. 2. Glucose, propionate, butyrate and acetate uptake rate by ISPAD biomass using non-decomposition approach: experimental data (D), and model prediction (�).

M. Madani-Hosseini et al. / Journal of Environmental Management 146 (2014) 59e68 65

5.5. Apparent half-saturation coefficient and temperature effect

The apparent half-saturation constant, Ks, indicates the con-centration at which the microbial group is able to process thesubstrate at half of its maximum growth rate. It also represents theaffinity of the microbial population for the substrate. A relationshipwas observed between temperature and Ks for all substrate con-sumption, except for butyrate degrading acetogenesis (Table 3),confirming that Ks increases at low temperatures (Lawrence andMcCarty, 1969; Lin et al., 1987; Nedwell, 1999). Taken up by aform of temperature sensitive active transport mechanism, the

Table 4Comparison of conventional and decomposition approaches.

Process Approach R

8 �C 18 �C 35 �C

1. Acidogenesis Decomposition 0.99 0.99 0.98Conventional 0.65 0.98 0.96

2. Propionate degradingacetogenesis

Decomposition 0.87 0.91 0.99Conventional 0.87 0.84 0.99

3. Butyrate degradingacetogenesis

Decomposition 0.97 0.98 0.93Conventional 0.58 0.99 0.93

4. Methanogenesis Decomposition 0.99 0.99 0.99Conventional 0.78 0.99 0.99

substrate is likely to become increasingly less available at lowertemperatures. Also the ability of microbes to sequester the sub-strate declines as temperature dropped (Nedwell, 1999). Also astemperature dropped, viscosity within the cell membrane increasesreducing the effectiveness of substrate transport during meta-bolism and a minimum is reached at temperatures solidifying themembrane lipids.

The Ks for the acidogens remained within a range of140e219 mg/L, but increased slightly with lower temperatures.Temperature had a variable impact on Ks for propionate andbutyrate degrading acetogenesis. Specifically, propionate

MAE MSE

8 �C 18 �C 35 �C 8 �C 18 �C 35 �C

38 13 84 2488 496 14301448 88 125 542,071 24,072 37,07614 67 58 297 7493 644314 63 26 284 12,659 109741 62 188 4192 6886 63,914

409 30 348 233,745 1839 178,18662 47 28 6958 4553 1608

795 55 46 974,922 5572 3829

Table 5Prediction accuracy of Arrhenius and Square Root equations for mmax.

Process Model R2 MAPE RMSE NMSE FB Ea (kJ/mol) Tmin (K)

1. Acidogenesis Arrhenius 0.95 0.39 0.019 0.014 0.041 43.42 e

Square root 0.97 0.26 0.014 0.007 0.021 e 262.942. Propionate degrading acetogenesis Arrhenius 0.92 0.59 0.0005 0.026 0.061 39.62 e

Square root 0.94 0.44 0.0004 0.016 0.032 e 260.543. Butyrate degrading acetogenesis Arrhenius 0.97 0.34 0.0005 0.009 0.13 42.38 e

Square root 0.91 0.59 0.001 0.031 0.24 e 251.904. Methanogenesis Arrhenius 0.95 0.46 0.001 0.016 0.11 38.07 e

Square root 0.92 0.54 0.001 0.024 0.18 e 250.67

M. Madani-Hosseini et al. / Journal of Environmental Management 146 (2014) 59e6866

degrading acetogens showed a Ks substantially larger of 392 mg/Lat 8 �C as compared to that of 100 mg/L at 35 �C. However, Ks

increased slightly from 8 to 18 �C and did not change from 18 to35 �C for butyrate degrading acetogens. For butyrate consumingacetogens, Ks increased with temperature from 8 to 18 �C, asopposed to acidogens, propionate degrading acetogens, andmethanogens. For methanogens, Ks decreased by a factor of 2.7when the temperature dropped from 35 to 8 �C. Westermann et al.(1989) also observed a drop of substrate affinity for methanogenswhen temperature decreased. Methanogens are clearly less tem-perature dependent when substrate concentration is reduced tosub-saturating levels.

The propionate degrading acetogens showed the lowest Ks valueof 100.2 mg/L at 35 �C indicating their highest affinity to substrate(Table 3). However, methanogens showed the lowest desire toconsume substrate with Ks of 533.8 mg/L at 8 �C. A high Ks formethanogens indicated that methanogenic growth might not besensitive to low concentrations of acetate (Chen, 2010).

5.6. Biomass density and temperature effect on yield coefficient

The anaerobic process generally exhibits a yield coefficient Yvarying from 0.01 to 0.17 mg of biomass produced per mg of sub-strate consumed (Table 3), compared to the values determined inthis work of 0.019e0.123. As for Jia et al. (1996), acidogens had thehighest Y of 0.123 mg of biomass produced per mg of substrateconsumed, indicating their higher microbial biomass productionfor each gram of substrate consumed and their lower sensitivity tothe effect of pH and substrate concentration (Beccari et al., 1996; Linand Chen, 1999; Shin et al., 1995). The propionate and butyratedegrading acetogens exhibited a respective Y of 0.053 and 0.034while the methanogens exhibited Y of 0.019 mg of biomass pro-duced per mg of substrate consumed.

The biomass density of each AD groups lacks documentationbecause of difficulties in measuring individual populations. In thisproject, the decomposition effect was used to determine the con-centration of each microbial population X by minimizing SSE be-tween the experimental data and the selected model (Haag et al.,2003). The relative density among AD microbial populations asfound in the literature (Kalyuzhnyi, 1997; Torre andStephanppoulos, 1986) indicates that 50% of the total biomass isgenerally associated with the acidogens (Torre andStephanppoulos, 1986). However for the present ISPAD inoculum,acidogens and propionate degrading bacteria represented only 5.4and 13.5% of total population while the methanogens and thebutyrate degrading bacteria made up 17.4 and 63.5% of the ADbiomass. The different ISPAD biomass distribution, as compared tothe literature, resulted from the fact that the ISPAD inoculum hadnot been fed for over 60 days, when exposed to the SAT. Accord-ingly, the AD groups demonstrating the highest population werethose associated with the substrate remaining to be degraded, suchas butyrate.

6. Conclusions

The concept of In-Storage-Psychrophilic-Anaerobic-Digestion(ISPAD) is a sequentially-fed batch anaerobic system with unde-fined kinetic coefficients. For temperatures of 8, 18 and 35 �Ccovering the range of ISPAD operating conditions, the objective ofthis study was therefore to obtain kinetic coefficients namely mmax,Ks and Y, to test the innovative decomposition approach to obtain amore accurate fit, and to establish a temperature function for themaximum growth coefficient (mmax) corresponding to each ADmicrobial consortium. Microbial population densities were alsocalculated to determine methane production potentials undervarious ambient temperatures. The kinetics coefficients were ob-tained by fitting laboratory Substrate Activity Test (SAT) results tothe Monod model.

The kinetic coefficients obtained indicated that, at a low tem-perature of 8

�C:

1) except for the acidogens demonstrating an initial lag phase, theorganisms consuming VFAs exhibited limited growth comparedto temperatures of 18 and 35 �C;

2) the propionate-consuming organisms at 8 �C had not acclimatedas much as the other organisms within the AD community andshowed almost no degradation resulting in the accumulation ofpropionate;

3) the butyrate degrading acidogens did demonstrate acclimation,with a linear consumption rate even at 8 �C.

Also, among organisms and temperatures, the kinetic co-efficients indicated that:

1) acidogens were acclimated even at 8 �C, degrading glucose at anexponential rate increasing with temperature;

2) propionate degrading acetogens exhibited activity at 18 and35 �C, but very little at 8 �C;

3) butyrate degrading acetogens exhibited activity at 8 �C,increasing with temperatures;

4) methanogens exhibited a linear consumption rate for acetate at8 �C, and an exponential rate at 35 �C, suggesting the existenceof two groups of methanogens, one acclimated to cold condi-tions and another remaining mesophilic.

To predict the impact of temperature on mmax for each ISPADmicroorganisms group, the Square Root equation performed betterfor both acidogens and propionate degrading acetogens, while theArrhenius equation performed better for methanogens and buty-rate degrading acetogens.

Acknowledgement

The authors acknowledge the financial contribution of Geo-membrane Technology Inc. (Fredericton, New Brunswick, Canada)

M. Madani-Hosseini et al. / Journal of Environmental Management 146 (2014) 59e68 67

and the Natural Science and Engineering Research Council ofCanada (NSERC).

References

Aceves-Lara, C., Aguilar-Garnica, E., Alcaraz-Gonzlez, V., Gonzlez-Reynoso, O.,Steyer, J., Dominguez-Beltran, J., Gonzlez-lvarez, V., 2005. Kinetic parametersestimation in an anaerobic digestion process using successive quadratic pro-gramming. Water Sci. Technol. 52, 419e426.

Angelidaki, I., Ahring, B.K., 1993. Thermophilic anaerobic digestion of livestockwaste: the effect of ammonia. Appl. Microbiol. Biotechnol. 38, 560e564.

Arbeli, Z., Brenner, A., Abeliovich, A., 2006. Treatment of high-strength dairywastewater in an anaerobic deep reservoir: analysis of the methanogenicfermentation pathway and the rate-limiting step. Water Res. 40, 3653e3659.

Bahn, O., Haurie, A., Kypreos, S., Vial, J.-P., 1996. A Decomposition Approach toMultiregional Environmental Planning: a Numerical Study. Springer.

Batstone, D.J., Keller, J., Angelidaki, I., Kalyuzhny, S.V., Pavlostathis, S.G., Rozzi, A.,Sanders, W.T.M., Siegrist, H., Vavilin, V.A., 2002. Anaerobic Digestion Model No.1 (ADM1). IWA Publishing.

Batstone, D.J., Tait, S., Starrenburg, D., 2009. Estimation of hydrolysis parameters infull-scale anerobic digesters. Biotechnol. Bioeng. 102, 1513e1520.

Beccari, M., Bonemazzi, F., Majone, M., Riccardi, C., 1996. Interaction betweenacidogenesis and methanogenesis in the anaerobic treatment of olive oil milleffluents. Water Res. 30, 183e189.

Bernard, O., Hadj-Sadok, Z., Dochain, D., Genovesi, A., Steyer, J.P., 2001. Dynamicalmodel development and parameter identification for an anaerobic wastewatertreatment process. Biotechnol. Bioeng. 75, 424e438.

Bowman, J., 2001. Methods for psychrophilic bacteria. Methods Microbiol. 30,591e614.

Chen, Q., 2010. Kinetics of Anaerobic Digestion of Selected C1 to C4 Organic Acids.University of Missouri, Columbia.

Chynoweth, D.P., Turick, C.E., Owens, J.M., Jerger, D.E., Peck, M.W., 1993. Biochemicalmethane potential of biomass and waste feedstocks. Biomass Bioenergy 5,95e111.

Dague, R.R., Banik, G.C., Ellis, T.G., 1998. Anaerobic sequencing batch reactor treat-ment of dilute wastewater at psychrophilic temperatures. Water Environ. Res.70, 155e160.

Donoso-Bravo, A., Mailier, J., Martin, C., Rodríguez, J., Aceves-Lara, C.A.,Wouwer, A.V., 2011. Model selection, identification and validation in anaerobicdigestion: a review. Water Res. 45, 5347e5364.

Donoso-Bravo, A., P�erez-Elvira, S.I., Fdz-Polanco, F., 2010. Application of simplifiedmodels for anaerobic biodegradability tests. Evaluation of pre-treatment pro-cesses. Chem. Eng. J. 160, 607e614.

Donoso-Bravo, A., Retamal, C., Carballa, M., Ruiz-Filippi, G., Chamy, R., 2009. Influ-ence of temperature on the hydrolysis, acidogenesis and methanogenesis inmesophilic anaerobic digestion: parameter identification and modeling appli-cation. Water Sci. Technol. 60, 9e17.

Eaton, A.D., Franson, M.A.H., 2005. Standard Methods for the Examination of Water& Wastewater. American Public Health Association, Washington, D.C.

El-Mashad, H.M., Van Loon, W.K.P., Zeeman, G., Bot, G., 2005. Rheological propertiesof dairy cattle manure. Bioresour. Technol. 96, 531e535.

Finster, K., 2008. Anaerobic Bacteria and Archaea in Cold Ecosystems, Psychro-philes: from Biodiversity to Biotechnology. Springer, pp. 103e119.

Flotats, X., Ahring, B.K., Angelidaki, I., 2003. Parameter identification of thermo-philic anaerobic degradation of valerate. Appl. Biochem. Biotechnol. 109, 47e62.

García-Ochoa, F., Santos, V.E., Naval, L., Guardiola, E., Lopez, B., 1999. Kinetic modelfor anaerobic digestion of livestock manure. Enzyme Microb. Technol. 25,55e60.

Giard, D., Choiniere, D., Cordeau, S., Barrington, S., 2013. Biogas generation from in-storage psychrophilic anaerobic digestion. Environ. Technol. 34, 1007e1015.

Gijzen, H.J., Zwart, K.B., Verhagen, F.J.M., Vogels, G.P., 1988. High-rate two-phaseprocess for the anaerobic degradation of cellulose, employing rumen microor-ganisms for an efficient acidogenesis. Biotechnol. Bioeng. 31, 418e425.

Goudar, C.T., Sonnad, J.R., Duggleby, R.G., 1999. Parameter estimation using a directsolution of the integrated Michaelis-Menten equation. BBA-Protein Struct. M1429, 377e383.

Haag, J.E., Vande Wouwer, A., Queinnec, I., 2003. Macroscopic modelling andidentification of an anaerobic waste treatment process. Chem. Eng. Sci. 58,4307e4316.

Harjunkoski, I., Grossmann, I.E., 2001. A decomposition approach for the schedulingof a steel plant production. Comput. Chem. Eng. 25, 1647e1660.

Hashimoto, A.G., 1983. Thermophilic and mesophilic anaerobic fermentation ofswine manure. Agr. Wastes 6, 175e191.

Jia, X.S., Furumai, H., Fang, H.H.P., 1996. Yields of biomass and extracellular poly-mers in four anaerobic sludges. Environ. Technol. 17, 283e291.

Jiang, X., Cheng, D.-C., 2005. A novel parameter decomposition approach to faithfulfitting of quadric surfaces. Pattern Recogn. 3663, 168e175.

Jim�enez, A.M., Borja, R., Martín, A., Raposo, F., 2006. Kinetic analysis of the anaer-obic digestion of untreated vinasses and vinasses previously treated withPenicillium decumbens. Environ. Manag. 80, 303e310.

Kalyuzhnyi, S.V., 1997. Batch anaerobic digestion of glucose and its mathematicalmodeling. II. Description, verification and application of model. Bioresour.Technol. 59, 249e258.

Kashyap, D., Dadhich, K., Sharma, S., 2003. Biomethanation under psychrophilicconditions: a review. Bioresour. Technol. 87, 147e153.

Kayranli, B., Ugurlu, A., 2011. Effects of temperature and biomass concentration onthe performance of anaerobic sequencing batch reactor treating low-strengthwastewater. Desalination 278, 77e83.

Keshtkar, A., Ghaforian, H., Abolhamd, G., Meyssami, B., 2001. Dynamic simulationof cyclic batch anaerobic digestion of cattle manure. Bioresour. Technol. 80,9e17.

Kettunen, R., Rintala, J., 1997. The effect of low temperature (5e29 C) and adaptationon the methanogenic activity of biomass. Appl. Microbiol. Biotechnol. 48,570e576.

King, S., Schwalb, M., Giard, D., Whalen, J., Barrington, S., 2011a. Effect of ISPADanaerobic digestion on ammonia volatilization from soil applied swine manure.Appl. Environ. Soil Sci. 2012, 1e8.

King, S.M., 2011. In-Storage Psychrophilic Anaerobic Digestion of Swine Manure.McGill University.

King, S.M., Barrington, S., Guiot, S.R., 2011b. In-storage psychrophilic anaerobicdigestion of swine manure: acclimation of the microbial community. BiomassBioenergy 35, 3719e3726.

Kotsyurbenko, O.R., 2005. Trophic interactions in the methanogenic microbialcommunity of low-temperature terrestrial ecosystems. FEMS Microbiol. Ecol.53, 3e13.

Kusiak, A., Wei, X., 2012. A data-driven model for maximization of methane pro-duction in a wastewater treatment plant. Water Sci. Technol. 65, 1116e1122.

Langenhoff, A.A.M., Stuckey, D.C., 2000. Treatment of dilute wastewater using ananaerobic baffled reactor: effect of low temperature. Water Res. 34, 3867e3875.

Lawrence, A.W., McCarty, P.L., 1969. Kinetics of methane fermentation in anaerobictreatment. J. (Water Pollut. Control Fed.), R1eR17.

Lin, C.-Y., Chen, C.-C., 1999. Effect of heavy metals on the methanogenic UASBGranule. Water Res. 33, 409e416.

Lin, C.Y., Noike, T., Sato, K., Matsumoto, J., 1987. Temperature characteristics of themethanogenesis process in anaerobic digestion.Water Sci. Technol.19, 299e300.

Lokshina, L.Y., Vavilin, V.A., Kettunen, R.H., Rintala, J.A., Holliger, C.,Nozhevnikova, A.N., 2001. Evaluation of kinetic coefficients using integratedMonod and Haldane models for low-temperature acetoclastic methanogenesis.Water Res. 35, 2913e2922.

L�opez, I., Borzacconi, L., 2010. Modelling of slaughterhouse solid waste anaerobicdigestion: determination of parameters and continuous reactor simulation.Waste Manag. 30, 1813e1821.

McHugh, S., Carton, M., Collins, G., O'Flaherty, V., 2004. Reactor performance andmicrobial community dynamics during anaerobic biological treatment ofwastewaters at 16e37 C. FEMS Microbiol. Ecol. 48, 369e378.

McKinney, R.E., 1963. Mathematics of complete-mixing activated sludge. Trans. Am.Soc. Civ. Eng. 128, 497e523.

M€osche, M., J€ordening, H.-J., 1999. Comparison of different models of substrate andproduct inhibition in anaerobic digestion. Water Res. 33, 2545e2554.

Müller, T.G., Noykova, N., Gyllenberg, M., Timmer, J., 2002. Parameter identificationin dynamical models of anaerobic wastewater treatment. Math. Biosci. 177,147e160.

Nedwell, D.B., 1999. Effect of low temperature on microbial growth: lowered affinityfor substrates limits growth at low temperature. FEMS Microbiol. Ecol. 30,101e111.

Ngozi-Olehi, L.C., Ayuk, A.A., Oguzie, E.E., Ejike, E.N., Ijioma, B.C., 2010. Kinetics ofbiogas potential from animal and domestic waste. Rep. Opin. 2, 17e21.

Nohra, J.A., Barrington, S., Frigon, J.C., Guiot, S.R., 2003. Storage psychrophilicanaerobic digestion of swine slurry. Resour. Conserv. Recycl. 38, 23e37.

Noykova, N.A., Gyllenberg, M., 2000. Sensitivity analysis and parameter estimationin a model of anaerobic wastewater treatment processes with substrate inhi-bition. Bioprocess Eng. 23, 343e349.

Nozhevnikova, A., Rebak, S., Kotsyurbenko, O., Parshina, S., Holliger, C., Lettinga, G.,2000. Anaerobic production and degradation of volatile fatty acids in lowtemperature environments. Water Sci. Technol. 41, 39e46.

Nwabanne, J., Onukwuli, O., Ifeakandu, C., 2009. Biokinetics of anaerobic digestionof municipal waste. Int. J. Environ. Res. 3, 511e516.

€Oztürk, M., 1993. Degradation of acetate, propionate, and butyrate under shocktemperature. J. Environ. Eng. 119, 321e331.

Pavlostathis, S.G., Giraldo-Gomez, E., 1991. Kinetics of anaerobic treatment. WaterSci. Technol. 24, 35e59.

Ratkowsky, D.A., Olley, J., McMeekin, T.A., Ball, A., 1982. Relationship betweentemperature and growth rate of bacterial cultures. J. Bacteriol. 149, 1e5.

Rebac, S., Gerbens, S., Lens, P., van Lier, J.B., Stams, A.J.M., Keesman, K.J., Lettinga, G.,1999. Kinetics of fatty acid degradation by psychrophilically grown anaerobicgranular sludge. Bioresour. Technol. 69, 241e248.

Rebac, S., Ruskova, J., Gerbens, S., Van Lier, J.B., Stams, A.J.M., Lettinga, G., 1995.High-rate anaerobic treatment of wastewater under psychrophilic conditions.J. Ferment. Bioeng. 80, 499e506.

Redzwan, G., Banks, C., 2004. The use of a specific function to estimate maximummethane production in a batch-fed anaerobic reactor. J. Chem. Technol. Bio-technol. 79, 1174e1178.

Robinson, J.A., Tiedje, J.M., 1983. Nonlinear estimation of monod growth kineticparameters from a single substrate depletion curve. Appl. Environ. Microbiol.45, 1453e1458.

Rodrigues, J.A.D., Ratusznei, S.M., Zaiat, M., 2003. Fed-batch and batch operatingmode analysis of a stirred anaerobic sequencing reactor with self-immobilizedbiomass treating low-strength wastewater. J. Environ. Manag. 69, 193e200.

M. Madani-Hosseini et al. / Journal of Environmental Management 146 (2014) 59e6868

Sales-Cruz, M., Gani, R., 2004. Aspects of modelling and model identification forbioprocesses through a computer-aided modelling system. Comput. AidedChem. Eng. 8, 1123e1128.

Shelton, D.R., Tiedje, J.M., 1984. General method for determining anaerobicbiodegradation potential. Appl. Environ. Microbiol. 47, 850e857.

Shin, H.S., Jung, J.Y., Bae, B.U., Paik, B.C., 1995. Phase-separated anaerobic toxicityassays for sulfate and sulfide. Water Environ. Res. 67, 802e806.

Siegrist, H., Vogt, D., Garcia-Heras, J.L., Gujer, W., 2002. Mathematical model formeso- and thermophilic anaerobic sewage sludge digestion. Environ. Sci.Technol. 36, 1113e1123.

Simeonov, I.S., 1999. Mathematical modeling and parameters estimation of anaer-obic fermentation processes. Bioprocess Eng. 21, 377e381.

Sinechal, X.J., Installe, M.J., Nyns, E.-J., 1979. Differentiation between acetate andhigher volatile acids in the modeling of the anaerobic biomethanation process.Biotechnol. Lett. 1, 309e314.

Srisertpol, J., Srinakorn, P., Kheawnak, A., Chamniprasart, K., 2010. Mathematicalmodeling and parameters, estimation of an anaerobic digestion of shrimp ofculture pond sediment in a biogas process. Int. J. Energ. Environ. 4, 213e220.

Ting, W., Shiqiang, Z., 2011. Study on linear correlation coefficient and nonlinearcorrelation coefficient in mathematical statistic. Stud. Math. Sci. 3, 58e63.

Torre, A.D., Stephanppoulos, G., 1986. Mixed culture model of anaerobic digestion:application to theevaluationof startupprocedures.Biotechnol. Bioeng.28,1106e1118.

Vavilin, V.A., Lokshina, L.Y., 1996. Modeling of volatile fatty acids degradation ki-netics and evaluation of microorganism activity. Bioresour. Technol. 57, 69e80.

Veeken, A., Hamelers, B., 1999. Effect of temperature on hydrolysis rates of selectedbiowaste components. Bioresour. Technol. 69, 249e254.

Westermann, P., Ahring, B.K., Mah, R.A., 1989. Temperature compensation inMethanosarcina Barkeri by modulation of hydrogen and acetate affinity. Appl.Environ. Microbiol. 55, 1262e1266.

Wichern, M., Gehring, T., Fischer, K., Andrade, D., Lübken, M., Koch, K., Gronauer, A.,Horn, H., 2009. Monofermentation of grass silage under mesophilic conditions:measurements and mathematical modeling with ADM1. Bioresour. Technol.100, 1675e1681.

Wolf, C., McLoone, S., Bongards, M., 2008. Biogas plant optimization using geneticalgorithms and particle swarm optimization. In: Signals and Systems Confer-ence, Galway, Ireland, pp. 244e249.