Kinetic parameters estimation in an anaerobic digestion process using successive quadratic programming.

In this work, an optimization method is implemented in an anaerobic digestion model to estimate its kinetic parameters and yield coefficients. This method combines the use of advanced state estimation schemes and powerful nonlinear programming techniques to yield fast and accurate estimates of the aforementioned parameters. In this method, we first implement an asymptotic observer to provide estimates of the non-measured variables (such as biomass concentration) and good guesses for the initial conditions of the parameter estimation algorithm. These results are then used by the successive quadratic programming (SQP) technique to calculate the kinetic parameters and yield coefficients of the anaerobic digestion process. The model, provided with the estimated parameters, is tested with experimental data from a pilot-scale fixed bed reactor treating raw industrial wine distillery wastewater. It is shown that SQP reaches a fast and accurate estimation of the kinetic parameters despite highly noise corrupted experimental data and time varying inputs variables. A statistical analysis is also performed to validate the combined estimation method. Finally, a comparison between the proposed method and the traditional Marquardt technique shows that both yield similar results; however, the calculation time of the traditional technique is considerable higher than that of the proposed method.     

Estimation of kinetic parameters of a model for deammonification in biofilms and evaluation of the model.

A systematic approach to estimate and evaluate parameters for deammonification in biofilms from available experimental data was evaluated. Parameter estimation was based on a regional steady state sensitivity analysis to select relevant parameters and design of experiments based on a local identifiability analysis. The calibrated model was evaluated under different experimental conditions. Nine of the 16 kinetic and stoichiometric parameters had a significant influence on model predictions. Of these nine parameters only six kinetic parameters were identifiable from batch experiments regardless of the experimental design. More parameters were not identifiable due to high correlations between growth rates and the corresponding affinity constant for oxygen. Data from a batch experiment at 2 mg/L dissolved oxygen (DO) were used to estimate inactivation rates and affinity constants for oxygen for ammonium oxidisers (AO), nitrite oxidisers (NO) and anaerobic ammonium oxidisers (AN). In addition, it was found that not only kinetic and stoichiometric parameters but also the external mass transfer resistance significantly affected model predictions. The resulting model was able to reproduce batch test and continuous reactor operation where DO concentrations were similar to those in the batch experiment used for parameter estimation. However, the model overestimated deammonification for a batch experiment at a much higher DO concentration (5 mg/L). Thus, parameter values that are identifiable and are estimated for given environmental conditions may not necessarily be valid for significantly different experimental conditions.     

Improved nitrogen removal in upflow anaerobic sludge blanket (UASB) reactors by incorporation of Anammox bacteria into the granular sludge.

Upflow anaerobic sludge blanket reactors may offer a number of advantages over conventional mixed-tank, SBR, and biofilm reactors, including high space-loading, low footprint, and resistance to shocks and toxins. In this study, we assessed the use of upflow anaerobic sludge blanket (UASB) reactor technology as applied to anaerobic ammonia removal, or Anammox. Four 200 ml UASB reactors were inoculated with 50% (by volume) anaerobic granular sludge and 50% flocular sludge from different sources (all with the potential for containing Anammox organisms). Tools used to assess the reactors included basic analyses, fluorescent in-situ hybridisation, and mathematical modelling, with statistical non-linear parameter estimation. Two of the reactors showed statistically identical Anammox activity (i.e., identical kinetic parameters), with good ammonia and nitrite removal (0.14 kgNHx m(-3) reactor day(-1), with 99% ammonia removal). The third reactor also demonstrated significant Anammox activity, but with poor identifiability of parameters. The fourth reactor had no statistical Anammox activity. Modelling indicated that poor identifiability and performance in the third and fourth reactors were related to an excess of reduced carbon, probably originating in the inoculum. Accumulation of Anammox organisms was confirmed both by a volume loading much lower than the growth rate, and response to a probe specific for organisms previously reported to mediate Anammox processes. Overall, the UASB reactors were effective as Anammox systems, and identifiability of the systems was good, and repeatable (even compared to a previous study in a rotating biological contactor). This indicates that operation, design, and analysis of Anammox UASB reactors specifically, and Anammox systems in general, are reliable and portable, and that UASB systems are an appropriate technology for this process.     

Multi-wavelength fluorometry for anaerobic digestion process monitoring.

Applicability of multi-wavelength fluorometry for anaerobic digestion process monitoring was investigated in a 3.5 L upflow anaerobic sludge bed (UASB) lab-scale reactor. Both off-line and on-line monitoring of key process parameters was tested. Off-line emission spectra were measured at an angle of 90 degrees to the excitation beam using a cuvette. On-line measurements were carried out using a fiber optic probe in the external recirculation line of the digester. Fluorescence spectra were correlated to available analytical measurements to obtain partial least square regression models. An independent set of measurements was used to validate the regression models. Model estimations showed reasonable agreement with analytical measurements with multiple determination coefficients (R2) between 0.6 and 0.95. Results showed that offline fluorescence measurements can be used for fast estimation of anaerobic digestor effluent quality. At the same time, the on-line implementation of multi-wavelength fluorescence measurements can be used for realtime process monitoring and, potentially, for on-line process control.     

Modelling of an EGSB treating sugarcane vinasse using first-order variable kinetics

An expanded granular sludge bed (EGSB) anaerobic reactor treating sugar cane vinasse was modelled using a simple model with two steps (acidogenesis and methanogenesis), two populations, two substrates and completely mixed conditions. A first-order kinetic equation for both steps with time-variant kinetic coefficients was used. An observer system was used to estimate the evolution of kinetic constants over time. The model was validated by comparing methane flow predictions with experimental values. An estimation of evolution of populations of microorganisms was also performed. This approach allows calculation of specific kinetic constants that reflect biological activity of microorganisms. Variation of specific kinetic constants reflects the influence of the fraction of raw vinasse in the feed. High salt concentrations in the reactor may have inhibited the process.     

Effects of feeding strategies on the performance of an anaerobic discontinuous reactor containing immobilized biomass with circulation system for liquid-phase mixing.

Data on the influence of feeding strategy on the performance of a fed-batch anaerobic sequencing reactor containing biomass immobilized on polyurethane foam and subjected to liquid phase circulation are presented and discussed. Six-hour cycles, temperature of 30 degrees C and circulation flow rate of 6 L/h were used. During each cycle 890 mL of synthetic domestic wastewater, with organic matter concentration of 500 mgCOD/L were fed to the reactor. The feeding strategies were implemented using fill times of 6 min (batch mode), 60, 120, 240 (fed-batch/batch mode) and 360 min (fed-batch mode). The system attained high efficiency and stability for all the operating conditions, and the substrate removal efficiency based on filtered samples presented a slight decrease from 85% to 81% when fill time was increased from 6 min to 360 min. A model considering a first-order kinetic equation was fitted to the experimental data. The apparent kinetic parameters for both batch and fed-batch phases were estimated, thus permitting evaluation of the influence of the feeding strategy on the reactor performance. The current system may be considered flexible in terms of the operating conditions it is subjected to.     

Treatment of textile dyes in two-phase and single-phase anaerobic bio-treatment systems.

This research integrates two different concepts of anaerobic biotechnology- two-phase anaerobic treatment and anaerobic granular sludge bed technology, in treatment of colored wastewaters from textile industries. Four anaerobic reactors based on upflow anaerobic sludge blanket (UASB) technology were used as acid reactors and an expanded granular sludge bed (EGSB) reactor was used as a methane reactor. A conventional single-phase anaerobic reactor, working on EGSB technology was run in parallel to compare the performances of the two systems. Reactors were operated at different hydraulic retention times. The results from the study, which span over a period of 400 days, indicated that the two-phase system produces a higher quality of effluent in terms of color, COD and suspended solids than single-phase anaerobic treatment when operated under similar conditions. Alkalinity requirement of two-phase system was also observed to be lower than that of single-phase system which is important regarding design consideration.     

ADM1 application for tuning and performance analysis of a multi-model observer-based estimator.

Anaerobic digestion model no.1 (ADM1) was used for tuning and performance analysis of the multi-model observer based estimator (mmOBE). The mmOBE was based on the variable structure model (VSM) of the anaerobic digestion model, which consists of several local submodels, each of which describes a typical process state. Depending on the hydraulic retention time, ADM1 simulated the methanogenic, organic overload, and acidogenic states of the process. These simulations allowed for optimising tunable parameters of the mmOBE. Owing to relatively slow process dynamics, a data acquisition interval as large as one day was sufficient to obtain acceptable accuracy. The simulations of mmOBE performance showed excellent rate of mmOBE convergence to ADM1 outputs. Moreover, mmOBE successfully estimated key kinetic parameters, such as maximal transformation rates of CODs, VFAs, and methane. These estimations can be used in the development of the advanced knowledge-based process system, which uses both available measurements and estimations of key kinetic parameters for extended diagnosis of failures and process trend analysis.     

Comparison of hydrolytic enzyme systems in pure culture and activated sludge under different electron acceptor conditions

Enzymatic hydrolysis under different electron acceptor conditions in nutrient removal activated sludge treatment processes is a weak link in the Activated Sludge Model no. 2 (Henze et al., 1995). An experimental study was undertaken to gain insight into the hydrolysis process with specific focus on hydrolysis kinetics and rates under different electron acceptor conditions. Two pure cultures, Bacillus amyloliquefaciens (Gram positive) and Pseudomonas saccharophila (Gram negative) were chosen for the study. In addition, activated sludge grown in an anaerobic-aerobic system was tested for enzymatic activity using starch as the model substrate. The hydrolytic enzymes were found to be released into the bulk in pure cultures whereas the enzyme activity was found to be mainly associated with the cell surfaces in activated sludge. Further, it was observed that the development of the hydrolytic enzyme system in Bacillus amyloliquefaciens and P. saccharophila is strongly suppressed under anoxic and anaerobic conditions. However, the effect of anaerobic and aerobic incubation on hydrolytic enzyme activity in activated sludge was found to be small. Starch hydrolysis kinetic data from batch experiments with activated sludge followed substrate saturation kinetics that were linear with biomass concentration. Finally, the similar hydrolytic enzyme activities observed under anaerobic and aerobic phases of a sequencing batch reactor are explained by considering the aspects of enzyme location and enzyme system development under aerobic and anaerobic phases. It is proposed that the floc bound enzymes are recycled in a single sludge system so that an equilibrium exists between enzyme loss and synthesis at steady state.     

Application of an anaerobic hybrid reactor for petrochemical effluent treatment

An anaerobic hybrid reactor (UASB/Filter) was used for petrochemical wastewater treatment in mesophilic conditions. The seeded flocculent sludge from a UASB plant treating dairy wastewater, acclimatized to the petrochemical wastes in a two-stage operation. After start up, under steady-state conditions, experiments were conducted at OLRs of between 0.5 and 24 kg TCOD m(-3) d(-1), hydraulic retention times (HRT) of 4-48 h and up-flow velocities 0.021-0.25 mh(-1). Removal efficiencies in the range of 42-86% were achieved at feed TCOD concentrations of 1,000-4,000 mg L(-1). The results of reactor performance at different operational conditions and its relations are presented and discussed in this paper. Then, the obtained data are used for determination of kinetic models. The results showed that a second-order model and a modified Stover-Kincannon model were the most appropriate models for this reactor. Finally, the biogas production data were used for the determination of biogas production kinetics.     

Treatment of organic synthesis wastewater using anaerobic packed bed and aerobic suspended growth bioreactors.

The performance of an anaerobic mesophilic packed bed reactor, with a mixture of GAC and tezontle, followed by an aerobic suspended growth system was studied for the treatment of organic chemical wastewater with a high COD concentration (22-29 g/L). The testing of the anaerobic-aerobic system was conducted in an experimental set-up for almost 2.5 years. Different operational conditions were evaluated. The anaerobic reactor showed performance stability and COD removals higher than 80% were obtained with loads up to 16.6 kg x m(-3) x d(-1). The acclimation of the aerobic biomass to the substrate in the anaerobic effluent was very quick and COD removals higher than 94% were obtained even at high organic loads. The combined anaerobic-aerobic system allowed total COD removals higher than 99.5% and the accomplishment of the discharge requirements of 200 mgCOD/L when the anaerobic reactor was operated with loads of 8-11 kg x m(-3)x d(-1) and the aerobic reactor with 0.33 kg x kg(-1) x d(-1), being the total HRT of 4.4. The average TKN removal in the anaerobic-aerobic system was 97%, the average for the anaerobic reactor being 52% and that one for the aerobic system being 94%.     

Simulation of DEHP biodegradation and sorption during the anaerobic digestion of secondary sludge.

Di-ethylhexyl phthalate (DEHP) has commonly been found in the sludge of municipal wastewater treatment plants especially during anaerobic processing. It is slowly biodegradable under anaerobic conditions. Due to its high hydrophobicity, sorption-desorption processes can be rate-limiting for the compound biodegradation. In this study, the anaerobic biodegradation of DEHP was investigated through batch kinetic experiments and dynamic transitions of a continuous stirred tank reactor (CSTR) fed with secondary sludge contaminated with DEHP. A widely accepted model (ADM1) was used to fit the anaerobic digestion of secondary sludge and was properly extended to account for DEHP removal, in which mass transfer processes are also involved. It was shown that DEHP removal was limited by the transfer of DEHP within the solid fraction. The criterion selected for the distinction of the two sites was whether the compound sorbed in those sites was bioavailable for biodegradation or not. Thus, the aqueous phase and the surface of the biosolids were considered as suitable sites for the compound to be bioavailable and the main bulk of the solid matrix was regarded as sites, where the compound remains "protected" against biodegradation. The model, fitted to the batch experimental data, was able to predict DEHP removal in the CSTR operated at various HRTs.     

High nitrification rate at low pH in a fluidized bed reactor with chalk as the biofilm carrier.

A typical steady state bulk pH of about 5 was established in a nitrifying fluidized bed with chalk as the only buffer agent. In spite of the low pH, high rate nitrification was observed with the nitrification kinetic parameters in the chalk reactor similar to those of biological reactors operating at pH>7. Various methods were used to determine the reasons for high rate nitrification at such low pH including (i) determination of bacterial species, (ii) microsensor measurements in the biofilm, and (iii) comparison of nitrification performance at low pH with a non-chalk fluidized bed reactor. Fluorescence in situ hybridization (FISH) using existing 16S rRNA-targeted oligonucleotide probes showed common nitrifying bacteria in the low pH chalk reactor. The prevalent nitrifying bacteria were identified in the Nitrosomonas oligotropha, Nitrosomonas europeae/eutropha, Nitrosospira and Nitrospira related groups, all well known nitrifiers. Microelectrode measurements showed that the pH in the biofilm was low and similar to that of the bulk pH. Finally, reactor performance using a non-chalk biofilm carrier (sintered glass) with the same bacterial inoculum also showed high rate nitrification below pH 5. The results suggest that inhibition of nitrification at low pH is highly overestimated.     

Modelling the kinetics of UV/H2O2 oxidation of dichloroacetic acid.

The intrinsic reaction kinetics of the decomposition of dichloroacetic acid (DCA) using UV/H2O2 was studied. A complete mathematical model, including the effect of the absorbed radiation intensities and H2O2 concentration was developed. The results of the kinetic measurements were analysed using a complete mathematical model of the experimental device that was used for the laboratory operation (a differential reactor inside a recycle). In this way it was expected to obtain intrinsic kinetic parameters. Experimental data agree well with theoretical predictions esmploying just two kinetic parameters derived from the proposed reaction mechanism.     

Fermentative hydrogen production in a system using anaerobic digester sludge without heat treatment as a biomass source.

Hydrogen produced from anaerobic fermentation of organic matter is a sustainable energy source. Anaerobic hydrogen-producing systems have been typically seeded with heat-treated inocula to eliminate hydrogen-consuming methanogens. This can be both energy- and economically-intensive. In this work, operational parameters were modified to determine if operating a reactor at low pH (5.5) and low SRT (10 hours) would result in a hydrogen-producing system free of methanogens using anaerobic digester sludge with no heat treatment as an inoculum. Initially, the reactor exhibited a hydrogen productivity of approximately 7.9% when fed glucose. After purging was begun with 10% CO2/90% N2, the hydrogen productivity increased to > 20% for the first day. Hydrogenotrophic methanogens then established themselves in the reactor, reducing the hydrogen productivity in the second non-purged phase by 80%. The operational controls examined were not sufficient to eliminate hydrogen-consuming methanogens for longer than approximately one week, and thus further methods must be developed.     

Use of an anaerobic sequencing batch reactor for parameter estimation in modelling of anaerobic digestion.

The model structure in anaerobic digestion has been clarified following publication of the IWA Anaerobic Digestion Model No. 1 (ADM1). However, parameter values are not well known, and uncertainty and variability in the parameter values given is almost unknown. Additionally, platforms for identification of parameters, namely continuous-flow laboratory digesters, and batch tests suffer from disadvantages such as long run times, and difficulty in defining initial conditions, respectively. Anaerobic sequencing batch reactors (ASBRs) are sequenced into fill-react-settle-decant phases, and offer promising possibilities for estimation of parameters, as they are by nature, dynamic in behaviour, and allow repeatable behaviour to establish initial conditions, and evaluate parameters. In this study, we estimated parameters describing winery wastewater (most COD as ethanol) degradation using data from sequencing operation, and validated these parameters using unsequenced pulses of ethanol and acetate. The model used was the ADM1, with an extension for ethanol degradation. Parameter confidence spaces were found by non-linear, correlated analysis of the two main Monod parameters; maximum uptake rate (k(m)), and half saturation concentration (K(S)). These parameters could be estimated together using only the measured acetate concentration (20 points per cycle). From interpolating the single cycle acetate data to multiple cycles, we estimate that a practical "optimal" identifiability could be achieved after two cycles for the acetate parameters, and three cycles for the ethanol parameters. The parameters found performed well in the short term, and represented the pulses of acetate and ethanol (within 4 days of the winery-fed cycles) very well. The main discrepancy was poor prediction of pH dynamics, which could be due to an unidentified buffer with an overall influence the same as a weak base (possibly CaCO3). Based on this work, ASBR systems are effective for parameter estimation, especially for comparative wastewater characterisation. The main disadvantages are heavy computational requirements for multiple cycles, and difficulty in establishing the correct biomass concentration in the reactor, though the last is also a disadvantage for continuous fixed film reactors, and especially, batch tests.     

Sensitivity analysis of a biofilm model describing mixed growth of nitrite oxidisers in a CSTR.

A simple kinetic model has been developed for describing nitrite oxidation by autotrophic aerobic nitrifiers in a CSTR reactor, in which mixed (suspended and attached) growth conditions are prevailing. In this work, a critical dimensionless parameter is identified containing both biofilm characteristics and microbial kinetic parameters, as well as the specific (per volume) surface of the reactor configuration used. Evaluation of this dimensionless parameter can easily provide information on whether or not wall attachment is critical, and should be taken into account either in kinetic studies or in reactor design, when specific pollutants are to be removed from the waste influent stream. The effect of bulk dissolved oxygen (DO) concentration on the validity of this model is addressed and minimum non-limiting DO concentrations are proposed depending on the reactor configuration.     

Multivariate statistical analysis of a high rate biofilm process treating kraft mill bleach plant effluent.

This study reports on a multivariate analysis of the moving bed biofilm reactor (MBBR) wastewater treatment system at a Canadian pulp mill. The modelling approach involved a data overview by principal component analysis (PCA) followed by partial least squares (PLS) modelling with the objective of explaining and predicting changes in the BOD output of the reactor. Over two years of data with 87 process measurements were used to build the models. Variables were collected from the MBBR control scheme as well as upstream in the bleach plant and in digestion. To account for process dynamics, a variable lagging approach was used for variables with significant temporal correlations. It was found that wood type pulped at the mill was a significant variable governing reactor performance. Other important variables included flow parameters, faults in the temperature or pH control of the reactor, and some potential indirect indicators of biomass activity (residual nitrogen and pH out). The most predictive model was found to have an RMSEP value of 606 kgBOD/d, representing a 14.5% average error. This was a good fit, given the measurement error of the BOD test. Overall, the statistical approach was effective in describing and predicting MBBR treatment performance.