Model selection, identification and validation in anaerobic digestion: a review

Soft sensor for substrate characterization through the reverse application of the ADM1 model for anaerobic digestion plant operations

Accurately characterizing the substrate used in anaerobic digestion is crucial for predicting the biogas plant's performance. This issue makes particularly challenging the application of modeling in codigestion plants. In this work, a novel methodology called substrate prediction module (SPM) has been developed and tested, using virtual codigestion data. The SPM aims to estimate the inlet properties of the substrate based on the reverse application of the anaerobic digestion model n1 (ADM1). The results show that, while the SPM can estimate some properties of the substrate based on certain output parameters, there are limitations in accurately determining all required variables.

Modeling key intermediates during anaerobic digestion of lipid rich kitchen waste with an extended ADM1

Quantitative dynamics of the key intermediates, gases and carbohydrates during anaerobic digestion of different lipid rich kitchen waste and lipid rich model kitchen waste were modeled. Six batch reactors loaded with 25 gVS  l- 1 ( ∼ 39  g O 2  l- 1 ) kitchen waste and model kitchen waste during a batch experiment were considered in simulation. Observed dynamics of carbohydrates, volatile organic acids and gases were described by an extended benchmark simulation model no. 2 (BSM2). In this study the extended BSM2 included a more detailed β -oxidation for prediction of caproic acid. Furthermore, the extensions included carbohydrate digestion with an additional intermediate before propionic acid was released. In addition, a novel simplification approach for initial pH estimation was successfully applied. For parameter estimation a Markov Chain Monte Carlo method was used to obtain parameter distributions. With the presented model it was possible even with no calibrated data to predict point of times of intermediates maxima and propionic acid with relative stable concentration over several days for kitchen waste.

Applying machine learning to anaerobic fermentation of waste sludge using two targeted modeling strategies

Anaerobic fermentation is an effective method to harvest volatile fatty acids (VFAs) from waste activated sludge (WAS). Accurately predicting and optimizing VFAs production is crucial for anaerobic fermentation engineering. In this study, we developed machine learning models using two innovative strategies to precisely predict the daily yield of VFAs in a laboratory anaerobic fermenter. Strategy-1 focuses on model interpretability to comprehend the influence of variables of interest on VFAs production, while Strategy-2 takes into account the cost of variable acquisition, making it more suitable for practical applications in prediction and optimization. The results showed that Support Vector Regression emerged as the most effective model in this study, with testing R2 values of 0.949 and 0.939 for the two strategies, respectively. We conducted feature importance analysis to identify the critical factors that influence VFAs production. Detailed explanations were provided using partial dependence plots and Shepley Additive Explanations analyses. To optimize VFAs production, we integrated the developed model with optimization algorithms, resulting in a maximum yield of 2997.282 mg/L. This value was 45.2 % higher than the average VFAs level in the operated fermenter. Our study offers valuable insights for predicting and optimizing VFAs production in sludge anaerobic fermentation, and it facilitates engineering practice in VFAs harvesting from WAS.

Comparison of anaerobic digestion of starch- and petro-based bioplastic under hydrogen-rich conditions

To identify an economically viable waste management system for bioplastics, thermoplastic starch (TPS) and poly(butylene adipate-co-terephthalate) (PBAT) were anaerobically digested under hydrogen (H2)/carbon dioxide (CO2) and nitrogen (N2) gas-purged conditions to compare methane (CH4) production and biodegradation. Regardless of the type of bioplastics, CH4 production was consistently higher with H2/CO2 than with N2. The highest amount of CH4 was produced at 307.74 mL CH4/g volatile solids when TPS digested with H2/CO2. A stepwise increased in CH4 yield was observed, with a nominal initial increment followed by accelerated methanogenesis conversion as H2 was depleted. This may be attributed to a substantial shift in the microbial structure from hydrogenotrophic methanogen (Methanobacteriales and Methanomicrobiales) to heterotrophs (Spirochaetia). In contrast, no significant change was observed with PBAT, regardless of the type of purged gas. TPS was broken down into numerous derivatives, including volatile fatty acids. TPS produced more byproducts with H2/CO2 (i.e., 430) than with N2 (i.e., 320). In contrast, differential scanning calorimetry analysis on PBAT revealed an increase in crystallinity from 10.20 % to 12.31 % and 11.36 % in the H2/CO2- and N2-purged conditions, respectively, after 65 days of testing. PBAT surface modifications were characterized via Fourier transform infrared spectroscopy and scanning electron microscopy. The results suggest that the addition of H2/CO2 can enhance the CH4 yield and increase the breakdown rate of TPS more than that of PBAT. This study provides novel insights into the CH4 production potential of two bioplastics with different biodegradabilities in H2/CO2-mediated anaerobic digestion systems.

Innovative approach for predicting biogas production from large-scale anaerobic digester using long-short term memory (LSTM) coupled with genetic algorithm (GA)

An artificial neural network (ANN) model called long-short term memory (LSTM), coupled with a genetic algorithm (GA) for feature selection, was used to predict biogas production of large-scale anaerobic digesters (ADs) of Tehran South Wastewater Treatment Plant (Iran), with a biogas production of approximately 30,000 Nm3/d. In order to employ the real conditions, the hydraulic retention time (HRT) of the ADs (21 days) was considered as the LSTM look-back window. To evaluate the model predictions, three different scenarios were defined. In the first scenario, the model predicted the produced biogas by using raw wastewater characteristics and reached the coefficient of determination of R2 = 0.84. The GA selected four out of eleven parameters of raw wastewater, including loads of BOD5, COD, TSS, and TN (kg/d), as the most informative data for the model. In the second scenario, the model predicted the produced biogas by employing the data of the thickened sludge streams entering the ADs and yielded a higher accuracy (R2 = 0.89). In this scenario, GA selected two out of six parameters of the sludge streams, including total flow rate (m3/d) and average solids content (w/w%). Finally, in the third scenario, by putting the parameters of the two previous scenarios together, the model's prediction accuracy increased slightly (R2 = 0.90). The results demonstrated that the GA-LSTM modeling technique could achieve reliable performance in predicting biogas production of large-scale ADs by including HRT in modeling procedure. It was also found that the raw wastewater characteristics severely affect AD behavior and can be successfully used as the input data of the AD models.

Bayesian uncertainty quantification for anaerobic digestion models

Uncertainty quantification is critical for ensuring adequate predictive power of computational models used in biology. Focusing on two anaerobic digestion models, this article introduces a novel generalized Bayesian procedure, called VarBUQ, ensuring a correct tradeoff between flexibility and computational cost. A benchmark against three existing methods (Fisher's information, bootstrapping and Beale's criteria) was conducted using synthetic data. This Bayesian procedure offered a good compromise between fitting ability and confidence estimation, while the other methods proved to be repeatedly overconfident. The method's performances notably benefitted from inductive bias brought by the prior distribution, although it requires careful construction. This article advocates for more systematic consideration of uncertainty for anaerobic digestion models and showcases a new, computationally efficient Bayesian method. To facilitate future implementations, a Python package called 'aduq' is made available.

Advances, trends and challenges in the use of biochar as an improvement strategy in the anaerobic digestion of organic waste: a systematic analysis

A recently strategy applied to anaerobic digestion (AD) is the use of biochar (BC) obtained from the pyrolysis of different organic waste. The PRISMA protocol-based review of the most recent literature data from 2011-2022 was used in this study. The review focuses on research papers from Scopus® and Web of Knowledge®. The review protocol used permits to identify 169 articles. The review indicated a need for further research in the following challenges on the application of BC in AD: i) to increase the use of BC in developing countries, which produce large and diverse amounts of waste that are the source of production of this additive; ii) to determine the effect of BC on the AD of organic waste under psychrophilic conditions; iii) to apply tools of machine learning or robust models that allow the process optimization; iv) to perform studies that include life cycle and technical-economic analysis that allow identifying the potential of applying BC in AD in large-scale systems; v) to study the effects of BC on the agronomic characteristics of the digestate once it is applied to the soil and vi) finally, it is necessary to deepen in the effect of BC on the dynamics of nitrogen and microbial consortia that affect AD, considering the type of BC used. In the future, it is necessary to search for new solutions in terms of the transport phenomena that occurs in AD with the use of BC using robust and precise mathematical models at full-scale conditions.

Machine learning for quantile regression of biogas production rates in anaerobic digesters

Anaerobic digestion is a well-established tool at wastewater treatment plants for processing raw sludge; it can also be used to generate renewable energy by harvesting biogas in anaerobic digesters. Operational parameters, such as temperature, are usually set by plant operators according to expert knowledge. To completely utilize the potential of operational management, in this study, we calibrated a novel Temporal Fusion Transformer based on six years of life-scale time series data together with categorical features such as public holidays. The model design allows for the interpretability of the output in contrast to traditional data-driven techniques, using multi-head attention. In addition to forecasting the median biogas production rates for the following seven days, our model also yields quantiles, making it less prone to strong fluctuations. We used three well-known statistical techniques as benchmarks. The mean absolute percentage error of our forecasting approach is below 8 %.

Dynamic modeling and parameter estimation of biomethane production from microalgae co-digestion

This work proposes a dynamic modeling procedure applied to biomethane production from microalgae residual co-digestion. A two-stage anaerobic digestion representation is selected, considering acidogenesis and methanogenesis as main reaction pathways. Based on the experimental database generated in the University of Mons Laboratories, several candidate models, assuming the presence or absence of biomass dynamics, are suggested, and parametric structural and local identifiability studies are performed. An original parameter estimation procedure is applied to a data-set partition used for model direct validation. The remaining experiment data are dedicated to cross-validation. The results point out how these dynamic models may serve as advanced monitoring software tools such as digital twins, even in the presence of incomplete process data.

Dynamic simulation and optimization of anaerobic digestion processes using MATLAB

Time series-based modeling provides a fundamental understanding of process fluctuations in an anaerobic digestion process. However, such models are scarce in literature. In this work, a dynamic model was developed based on modified Hill's model using MATLAB, which can predict biomethane production with time series. This model can predict the biomethane production for both batch and continuous process, across substrates and at diverse conditions such as total solids, loading rate, and days of operation. The deviation between literature and the developed model was less than ± 7.6%, which shows the accuracy and robustness of this model. Moreover, statistical analysis showed there was no significant difference between literature and simulation, verifying the null hypothesis. Finding a steady and optimized loading rate was necessary to an industrial perspective, which usually requires extensive experimental data. With the developed model, a stable and optimal methane yield generating loading rate could be identified at minimal input.

Sewage-Water Treatment and Sewage-Sludge Management with Power Production as Bioenergy with Carbon Capture System: A Review

Sewage-water treatment comprehends primary, secondary, and tertiary steps to produce reusable water after removing sewage contaminants. However, a sewage-water treatment plant is typically a power and energy consumer and produces high volumes of sewage sludge mainly generated in the primary and secondary steps. The use of more efficient anaerobic digestion of sewage water with sewage sludge can produce reasonable flowrates of biogas, which is shown to be a consolidated strategy towards the energy self-sufficiency and economic feasibility of sewage-water treatment plants. Anaerobic digestion can also reduce the carbon footprint of energy sources since the biogas produced can replace fossil fuels for electricity generation. In summary, since the socio-economic importance of sewage treatment is high, this review examined works that contemplate: (i) improvements of sewage-water treatment plant bioenergy production and economic performances; (ii) the exploitation of technology alternatives for the energy self-sufficiency of sewage-water treatment plants; (iii) the implementation of new techniques for sewage-sludge management aiming at bioenergy production; and (iv) the implementation of sewage-water treatment with bioenergy production and carbon capture and storage.

Model Predictive Control: Demand-Orientated, Load-Flexible, Full-Scale Biogas Production

Biogas plants have the great advantage that they produce electricity according to demand and can thus compensate for fluctuating production from weather-dependent sources such as wind power and photovoltaics. A prerequisite for flexible biogas plant operation is a suitable feeding strategy for an adjusted conversion of biomass into biogas. This research work is the first to demonstrate a practical, integrated model predictive control (MPC) for load-flexible, demand-orientated biogas production and the results show promising options for practical application on almost all full-scale biogas plants with no or only minor adjustments to the standardly existing measurement technology. Over an experimental period of 36 days, the biogas production of a full-scale plant was adjusted to the predicted electricity demand of a “real-world laboratory”. Results with a mean absolute percentage error (MAPE) of less than 20% when comparing biogas demand and production were consistently obtained.

Application of machine learning in anaerobic digestion: Perspectives and challenges

Anaerobic digestion (AD) is widely adopted for remediating diverse organic wastes with simultaneous production of renewable energy and nutrient-rich digestate. AD process, however, suffers from instability, thereby adversely affecting biogas production. There have been significant efforts in developing strategies to control the AD process to maintain process stability and predict AD performance. Among these strategies, machine learning (ML) has gained significant interest in recent years in AD process optimization, prediction of uncertain parameters, detection of perturbations, and real-time monitoring. ML uses inductive inference to generalize correlations between input and output data, subsequently used to make informed decisions in new circumstances. This review aims to critically examine ML as applied to the AD process and provides an in-depth assessment of important algorithms (ANN, ANFIS, SVM, RF, GA, and PSO) and their applications in AD modeling. The review also outlines some challenges and perspectives of ML, and highlights future research directions.

Best Operating Conditions for Biogas Production in Some Simple Anaerobic Digestion Models

We consider one-step and two-step simple models of anaerobic digestion that are able to adequately capture the main dynamical behaviour of the full anaerobic digestion model ADM1. We do not consider specific growth functions. We only require them to satisfy certain qualitative assumptions. These assumptions are satisfied for concave growth functions, but they are also satisfied for a large class of growth functions found in many applications. We consider the maximisation of the biogas production with respect to the operating parameters of the model, which are the dilution rate and the substrate input concentration. We give the best operating conditions and we describe them as a subset of the set of operating parameters. Our models incorporate biomass decay terms, corresponding to maintenance. Numerical plots with specified growth functions and biological parameters illustrate the obtained results.

Optimising the Flux Enhancer Dosing Strategy in a Pilot-Scale Anaerobic Membrane Bioreactor by Mathematical Modelling

Flux enhancers (FEs) have been successfully applied for fouling mitigation in membrane bioreactors. However, more research is needed to compare and optimise different dosing strategies to improve the filtration performance, while minimising the use of FEs and preventing overdosing. Therefore, the goal of this research is to develop an optimised control strategy for FE dosing into an AnMBR by developing a comprehensive integrated mathematical model. The integrated model includes filtration, flocculation, and biochemical processes to predict the effect of FE dosing on sludge filterability and membrane fouling rate in an AnMBR. The biochemical model was based on an ADM1, modified to include FEs and colloidal material. We developed an empirical model for the FE-induced flocculation of colloidal material. Various alternate filtration models from the literature and our own empirical models were implemented, calibrated, and validated; the best alternatives were selected based on model accuracy and capacity of the model to predict the effect of varying sludge characteristics on the corresponding output, that is fouling rate or sludge filterability. The results showed that fouling rate and sludge filterability were satisfactorily predicted by the selected filtration models. The best integrated model was successfully applied in the simulation environment to compare three feedback and two feedforward control tools to manipulate FE dosing to an AnMBR. The modelling results revealed that the most appropriate control tool was a feedback sludge filterability controller that dosed FEs continuously, referred to as ∆R20_10. Compared to the other control tools, application of the ∆R20_10 controller resulted in a more stable sludge filterability and steady fouling rate, when the AnMBR was subject to specific disturbances. The simulation environment developed in this research was shown to be a useful tool to test strategies for dosing flux enhancer into AnMBRs.

Cultivation of sulfide-driven partial denitrification granules for efficient nitrite generation from nitrate-sulfide-laden wastewater

Sulfide partial denitrification (SPD) is an alternative pathway for nitrite production accompanied with elemental sulfur (S⁰) production for nitrate removal from wastewater with anammox. In this study, the SPD granular sludge was cultivated for the first time in an upflow anaerobic sludge bed (UASB) reactor to reach the efficacy of maximum nitrate-to-nitrite transformation ratio of 92% and an in-situ maximum NO₃^--N reduction rate of 2.46 kg-N/m³-d, both much higher than literature results. Mature granules had an average particle size of 2.52 mm and hold smooth surface with excess rod bacteria. The elements Ca and S, and proteins in extracellular polymeric substances contributed to granule structure's stability. Enriched Thiobacillus genus was proposed to accumulate nitrite at moderate HRT (2-6 h). The immobilized functional strains assist efficient partial nitrification reactions to be realized with formed S⁰ as byproduct.

Reactor Designs and Configurations for Biological and Bioelectrochemical C1 Gas Conversion: A Review

Microbial C1 gas conversion technologies have developed into a potentially promising technology for converting waste gases (CO₂, CO) into chemicals, fuels, and other materials. However, the mass transfer constraint of these poorly soluble substrates to microorganisms is an important challenge to maximize the efficiencies of the processes. These technologies have attracted significant scientific interest in recent years, and many reactor designs have been explored. Syngas fermentation and hydrogenotrophic methanation use molecular hydrogen as an electron donor. Furthermore, the sequestration of CO₂ and the generation of valuable chemicals through the application of a biocathode in bioelectrochemical cells have been evaluated for their great potential to contribute to sustainability. Through a process termed microbial chain elongation, the product portfolio from C1 gas conversion may be expanded further by carefully driving microorganisms to perform acetogenesis, solventogenesis, and reverse β-oxidation. The purpose of this review is to provide an overview of the various kinds of bioreactors that are employed in these microbial C1 conversion processes.

Modeling the Performance of Full-Scale Anaerobic Biochemical System Treating Deinking Pulp Wastewater Based on Modified Anaerobic Digestion Model No. 1

The deinking pulp (DIP) is a main resource for paper making, and the wastewater from DIP process needs to be treated. Anaerobic biochemical technique has been widely applied in DIP wastewater treatment, due to the remarkable capability in reducing high chemical oxygen demand (COD). In this study, a mathematical simulation model was established to investigate the performance of a full-scale anaerobic biochemical system for treating DIP wastewater. The model was based on Anaerobic Digestion Model No. 1 (ADM1), which was modified according to the specific anaerobic digestion process for DIP wastewater treatment. The hydrodynamic behavior of a full-scale anaerobic biochemical system was considered in this model. The characteristics of the influent DIP wastewater were assessed, and then, the substrate COD proportion was divided successfully for the necessity of ADM1 applying. The Monte Carlo technique was implemented to distinguish the most sensitive parameters that influenced the model output indicators comprising effluent COD and biogas production. The sensitive parameters were estimated and optimized. The optimized value of k_{_m_pro} is 12.02, K_{_S_pro} is 0.35, k_{_m_ac} is 4.26, K_{_S_ac} is 0.26, k_{_m_h2} is 16.62, and K_{_S_h2} is 3.21 × 10^–5. The model was calibrated with 150 days operation values measured in the field. The subsequent 100 days on-site values were used to validate the model, and the results obtained by the simulations were in good agreement. This study provides a meaningful and theoretical model guidance for full-scale wastewater anaerobic biochemical treatment simulation.

Linking population dynamics to microbial kinetics for hybrid modeling of bioelectrochemical systems

Mechanistic and data-driven models have been developed to provide predictive insights into the design and optimization of engineered bioprocesses. These two modeling strategies can be combined to form hybrid models to address the issues of parameter identifiability and prediction interpretability. Herein, we developed a novel and robust hybrid modeling strategy by incorporating microbial population dynamics into model construction. The hybrid model was constructed using bioelectrochemical systems (BES) as a platform system. We collected 77 samples from 13 publications, in which the BES were operated under diverse conditions, and performed holistic processing of the 16S rRNA amplicon sequencing data. Community analysis revealed core populations composed of putative electroactive taxa Geobacter, Desulfovibrio, Pseudomonas, and Acinetobacter. Primary Bayesian networks were trained with the core populations and environmental parameters, and directed Bayesian networks were trained by defining the operating parameters to improve the prediction interpretability. Both networks were validated with Bray-Curtis similarly, relative root-mean-square error (RMSE), and a null model. A hybrid model was developed by first building a three-population mechanistic component and subsequently feeding the estimated microbial kinetic parameters into network training. The hybrid model generated a simulated community that shared a Bray-Curtis similarity of 72% with the actual microbial community at the genus level and an average relative RMSE of 7% for individual taxa. When examined with additional samples that were not included in network training, the hybrid model achieved accurate prediction of current production with a relative error-based RMSE of 0.8 and outperformed the data-driven models. The genomics-enabled hybrid modeling strategy represents a significant step toward robust simulation of a variety of engineered bioprocesses.

A simple kinetic model applied to anaerobic digestion of cow manure

A simple model of anaerobic degradation in a continuous stirred digester is presented. The hydrolysis of cow manure was modelled as consisting of two fractions, one rapidly degradable and the other more slowly degradable, and both processes were represented by first-order kinetics in a two-substrate first-order (TSFO) model. The fractions were separated by water flushing. Biomethane potential (BMP) tests were performed to determine the hydrolysis constant and biodegradability of each fraction. The hydrolysis constants of the rapidly and slowly degradable fractions were 0.278 and 0.069 d^-1, respectively. Coupled with a simple anaerobic digestion model, the TSFO model was used to simulate the digester behaviour and predict methane production. Experiments in a 3.0 L digester were used to determine the decay constant and yield values and to validate the model. Two solid loads (2.9 and 4.4 gVS/L.d) were applied to the digester, and the dynamics of both biodegradable fractions, the non-biodegradable fraction and the microorganism concentration were reproduced by the model. These results approximate the actual biodegradable solids removal to within 85%. A parametric sensitivity study was performed, and the results show that the hydrolysis constant mainly influences the biodegradable fractions and that the decay and yield parameters mainly influence the microorganism concentration.

Systematic simplification of the Anaerobic Digestion Model No. 1 (ADM1) - Laboratory experiments and model application

Due to a limited number of available measurements on agricultural biogas plants, established process models, such as the Anaerobic Digestion Model No. 1 (ADM1), are rarely applied in practise. To provide a reliable basis for model-based monitoring and control, different model simplifications of the ADM1 were implemented for process simulation of semi-continuous anaerobic digestion experiments using agricultural substrates (maize silage, sugar beet silage, rye grain and cattle manure) and industrial residues (grain stillage). Individual model structures enable a close depiction of biogas production rates and characteristic intermediates (ammonium nitrogen, propionic and acetic acid) with equal accuracy as the original ADM1. The impact of different objective functions and standard parameter values on parameter estimates of first-order hydrolysis constants and microbial growth rates were evaluated. Due to the small number of required model parameters and suitable system characteristics, simplified model structures show clear advantages for practical application on agricultural biogas plants.

Optimization of biochemical sulfide potential (BSP) assay for anaerobic biodegradability assessment

The anaerobic biodegradability assessment (biodegradation extent and kinetics) of organic wastes is critical for optimum design and evaluating treatment efficiencies for anaerobic treatment technologies. The biochemical sulfide potential (BSP) assay has previously demonstrated the advantages of its time efficiency and measurement accuracy for biologically assessing substrate degradability, while its application is limited by undefined operational parameters. In this study, the BSP assay was further optimized through a systematic investigation of a critical parameter, inoculum-to-substrate ratio (ISR), and the applicable kinetic model to unravel the potential use of BSP assays for anaerobic waste treatment. Under two series of experimental scenarios, the common ISR ranges of 0.5-4.0 (based on the traditional BMP assay) and extreme ISRs (as low as 0.1) were studied, in which the advantage of a BSP assay on extreme ISRs was highlighted. Meanwhile, the underlying cause and mechanism for biodegradability discrepancies under different ISRs (0.1-6.0) were further investigated. The extracellular polymeric substance (EPS) characterization of residual organics and the two-substrate first-order hydrolysis model analyses revealed that the hydrolysis process of slowly-biodegradable organics fraction was hindered under improper ISR conditions. Furthermore, the Cone model was evaluated as more appropriate for biodegradation kinetics analysis in BSP assays among the five common kinetic models (i.e., Exponential, Fitzhugh, Cone, Transference, and modified Gompertz models). Overall, the results provide fundamental guidance on designing consistent BSP assays and put a step forward in standardizing the BSP assay for anaerobic biodegradability assessments.

Multivariable Robust Regulation of Alkalinities in Continuous Anaerobic Digestion Processes: Experimental Validation

A multivariable adaptive feedback control for highly uncertain continuous anaerobic digestion processes is proposed to regulate the volatile fatty acids (VFA) concentration, the strong ions concentrations, and the total and intermediate alkalinities. The multivariable control scheme includes a Luenberger observer to estimate both the unmeasured variables (i.e., VFA) and unknown microbial growth kinetics. The control approach is designed using an exponential Lyapunov function to resemble the typical exponential biological growth of the involved microbial consortia. Taking into account physicochemical equilibrium, alkalinities are represented as a function of the state variables. As a result, the control problem becomes a regulation problem on alkalinities, and in turn, a tracking control problem on the state variables, with two manipulated variables—the dilution rate and the feed rate of a strong alkali solution—while the state variables’ set-points are given as a function of pH. The implementation of this multivariable control scheme was experimentally tested and validated in a 0.982 m3 pilot plant treating agro-industrial wastewater and demonstrated to be robust in the face of unknown microbial growth kinetics. Results showed the potential for practical application and optimization of industrial digesters.

Predicting the performance of anaerobic digestion using machine learning algorithms and genomic data

Modeling of anaerobic digestion (AD) is crucial to better understand the process dynamics and to improve the digester performance. This is an essential yet difficult task due to the complex and unknown interactions within the system. The application of well-developed data mining technologies, such as machine learning (ML) and microbial gene sequencing techniques are promising in overcoming these challenges. In this study, we investigated the feasibility of 6 ML algorithms using genomic data and their corresponding operational parameters from 8 research groups to predict methane yield. For classification models, random forest (RF) achieved accuracies of 0.77 using operational parameters alone and 0.78 using genomic data at the bacterial phylum level alone. The combination of operational parameters and genomic data improved the prediction accuracy to 0.82 (p<0.05). For regression models, a low root mean square error of 0.04 (relative root mean square error =8.6%) was acquired by neural network using genomic data at the bacterial phylum level alone. Feature importance analysis by RF suggested that Chloroflexi, Actinobacteria, Proteobacteria, Fibrobacteres, and Spirochaeta were the top 5 most important phyla although their relative abundances were ranging only from 0.1% to 3.1%. The important features identified could provide guidance for early warning and proactive management of microbial communities. This study demonstrated the promising application of ML techniques for predicting and controlling AD performance.

Design of profitable networks of biogas plants in Chile

Locations for profitable biogas plant networks for the generation of electric power were investigated. A population-based location algorithm was developed. Using information regarding biogas power and plants cost found in the literature, together with data on the location and number of dairy cows of 572 farms in southern Chile, economic analises were made. Each farm was evaluated as a potencial location for a biogas plant, both alone and as part of different combinations. The algorithm starts with a population of one-plant combinations and generates new iterations by inserting additional biogas plants as long as the NPV increases. In each iteration, only a diverse subset of the combinations is selected. This approach is compared to the application of the ArcGIS Network Analyst (NA) function for the set cover problem, usually employed in the literature for biogas plant network location problems. Using the technical parameters for the operation of biogas plants found in the literature, we show that the proposed algorithm gives better results than the NA algorithm in terms of profit. Biogas plant networks were obtained for different scenarios of biomass availability (40–80%), energy sale prices (90–100 USD/MWh) and transport costs (0.3–0.4 USD/tkm). The results indicate that 18 of the 572 farms would be good candidates to site a biogas plant in at least one of the scenarios. A maximum of eight farms appear in a scenario with 80% of biomass availability and a transport cost of 0.3 USD/tkm. This solution reaches a NPV of USD 3,538,394, which exceeds by more than 70% the USD 2,075,057 obtained with the solution retrieved by the NA function. As the algorithm presented obtained better results than the ArcGIS network analyst, it can be used as an appropriate design tool for biogas plant networks.

Biogas Plants in Renewable Energy Systems—A Systematic Review of Modeling Approaches of Biogas Production

Biogas production is a relevant component in renewable energy systems. The paper addresses modeling approaches from an energy system, as well as from a process optimization, point of view. Model approaches of biogas production show different levels of detail. They can be classified as white, gray, and black box, or bottom-up and top-down approaches. On the one hand, biogas modeling can supply dynamic information on the anaerobic digestion process, e.g., to predict biogas yields or to optimize the anaerobic digestion process. These models are characterized by a bottom-up approach with different levels of detail: the comprehensive ADM1 (white box), simplifications and abstractions of AD models (gray box), or highly simplified process descriptions (black box). On the other hand, biogas production is included in energy system models. These models usually supply aggregated information on regional biogas potentials and greenhouse gas emissions. They are characterized by a top-down approach with a low level of detail. Most energy system models reported in literature are based on black box approaches. Considering the strengths and weaknesses of the integration of detailed and deeply investigated process models in energy system models reveals the opportunity to develop dynamic and fluctuating business models of biogas usage.

Exploration and verification of the feasibility of sulfide-driven partial denitrification coupled with anammox for wastewater treatment

Anaerobic ammonia oxidation (anammox) is a well-developed biotechnology for treating high-strength ammonium wastewaters. Recently, partial denitrification has been considered as an alternative to supply anammox with the required nitrite. In this study, a process of sulfide-driven partial denitrification and anammox (SPDA) was developed and operated continuously in an upflow anaerobic sludge blanket (UASB) reactor for 392 days. This reactor was fed with synthetic wastewater containing 100 mgN/L nitrate, 80 mgN/L ammonium and 20-80 mgS/L sulfide. After 160 days of operation, the reactor reached stable performance, and the nitrogen removal efficiency and rate were maintained at 80% and 0.29 kgN/(m³•d), respectively. The estimated nitrogen removal via anammox and sulfide-driven denitrification were 87.2% and 12.8%. Additional batch experiments were conducted to investigate the effects of sulfide on anammox and the mechanisms of nitrogen removal in the SPDA system. The following results were obtained: (1) sulfide had an inhibitory effect on the specific anammox activity with IC₅₀ of 9.7 mgS-H₂S/L. (2) The rapid oxidation of sulfide by sulfur-oxidizing bacteria (SOB) could relieve the toxic effects of sulfide on the anammox in the SPDA system. (3) Sulfide bio-oxidation was a two-step reaction with biologically produced elemental sulfur (BPS⁰) as the intermediate, and the second step using BPS⁰ as the electron donor, can efficiently produce nitrite via partial denitrification (NO₃^- → NO₂^-) as a supply for anammox. Finally, a high-throughput sequencing analysis identified Thiobacillus and Sulfurimonas as the dominant genera of SOB in the SPDA system, and Candidatus Kuenenia as the dominant anammox bacteria. Overall, this research gives the foundation for the practical application of sulfide-driven partial denitrification and anammox process in the future.

Co-digestion of palm oil mill effluent with chicken manure and crude glycerol: biochemical methane potential by monod kinetics

In Thailand, the palm oil industry produces a huge amount of palm oil mill effluent (POME), mostly used for electricity generation through biogas production. Co-digestion with other waste can further improve biogas yield and solve waste management problems. Most previous studies relied on biochemical methane potential (BMP) assay or batch co-digestion to obtain the optimal mixing ratio, ignoring the kinetic part or treat it for sole discussion of the results. This work directly uses mechanistic models based on Monod kinetics to describe the experimental results obtained from the co-digestion of POME (40 ml, BMP = 281.2 mlCH₄/gCODadded)) with chicken manure (CM) (0–50 g) and crude glycerol (Gly) (0–10 ml). The best mixing ratio between CM and POME was 5 gCM: 40 mlPOME (BMP = 276.9 mlCH₄/gCODadded). The best ratio for Gly and POME was 2 mlGly: 40 mlPOME (BMP = 211.9 mlCH₄/gCODadded). Adding Gly only 2 mlGly/40 mlPOME doubled the amount of biogas. Hence, crude glycerol is a good substrate for on-demand biogas output. The co-digestion increases the methane output but with a decreased yield. A multi-substrate Monod model was developed based on the levels of digestion difficulty. A partial-least squared fitting was used to estimate its main parameters. All parameters included in the model passed the significant tests at a 95% confidence level. The model can describe the experimental results very well, predict observable state variables of batch co-digestion, and allow a simple extension for continuous co-digestion dynamics. A limited continuous experiment was conducted to confirm the applicability of the model parameters of POME digestion obtained from BMP tests to predict a continuous AD. The results show good potential but must be carefully interpreted. It is generally possible and practical to directly obtain design and operational parameters from BMP assays based on only accumulated biogas curves and initial and final COD/VS.

Modeling and Simulation of Biogas Production in Full Scale with Time Series Analysis

Future biogas plants must be able to produce biogas according to demand, which requires proactive feeding management. Therefore, the simulation of biogas production depending on the substrate supply is assumed. Most simulation models are based on the complex Anaerobic Digestion Model No. 1 (ADM1). The ADM1 includes a large number of parameters for all biochemical and physicochemical process steps, which have to be carefully adjusted to represent the conditions of a respective full-scale biogas plant. Due to a deficiency of reliable measurement technology and process monitoring, nearly none of these parameters are available for full-scale plants. The present research investigation shows a simulation model, which is based on the principle of time series analysis and uses only historical data of biogas formation and solid substrate supply, without differentiation of individual substrates. The results of an extensive evaluation of the model over 366 simulations with 48-h horizon show a mean absolute percentage error (MAPE) of 14–18%. The evaluation is based on two different digesters and demonstrated that the model is self-learning and automatically adaptable to the respective application, independent of the substrate’s composition.

Effect of temperature on the production of a recombinant antivenom in fed-batch mode

In the pharmaceutical industry, nanobodies show promising properties for its application in serotherapy targeting the highly diffusible scorpion toxins. The production of recombinant nanobodies in Escherichia coli has been widely studied in shake flask cultures in rich medium. However, there are no upstream bioprocess studies of nanobody production in defined minimal medium and the effect of the induction temperature on the production kinetics. In this work, the effect of the temperature during the expression of the chimeric bispecific nanobody CH10-12 form, showing high scorpion antivenom potential, was studied in bioreactor cultures of E. coli. High biomass concentrations (25 g cdw/L) were achieved in fed-batch mode, and the expression of the CH10-12 nanobody was induced at temperatures 28, 29, 30, 33, and 37°C with a constant glucose feed. For the bispecific form NbF12-10, the induction was performed at 29°C. Biomass and carbon dioxide yields were reported for each culture phase, and the maintenance coefficient was obtained for each strain. Nanobody production in the CH10-12 strain was higher at low temperatures (lower than 30°C) and declined with the increase of the temperature. At 29°C, the CH10-12, NbF12-10, and WK6 strains were compared. Strains CH10-12 and NbF12-10 had a productivity of 0.052 and 0.021 mg/L/h of nanobody, respectively, after 13 h of induction. The specific productivity of the nanobodies was modeled as a function of the induction temperature and the specific growth rates. Experimental results confirm that low temperatures increase the productivity of the nanobody.Key points• Nanobodies with scorpion antivenom activity produced using two recombinant strains.• Nanobodies production was achieved in fed-batch cultures at different induction temperatures.• Low induction temperatures result in high volumetric productivities of the nanobody CH10-12.

Not Just Numbers: Mathematical Modelling and Its Contribution to Anaerobic Digestion Processes

Mathematical modelling of bioprocesses has a long and notable history, with eminent contributions from fields including microbiology, ecology, biophysics, chemistry, statistics, control theory and mathematical theory. This richness of ideas and breadth of concepts provide great motivation for inquisitive engineers and intrepid scientists to try their hand at modelling, and this collaboration of disciplines has also delivered significant milestones in the quality and application of models for both theoretical and practical interrogation of engineered biological systems. The focus of this review is the anaerobic digestion process, which, as a technology that has come in and out of fashion, remains a fundamental process for addressing the global climate emergency. Whether with conventional anaerobic digestion systems, biorefineries, or other anaerobic technologies, mathematical models are important tools that are used to design, monitor, control and optimise the process. Both highly structured, mechanistic models and data-driven approaches have been used extensively over half a decade, but recent advances in computational capacity, scientific understanding and diversity and quality of process data, presents an opportunity for the development of new modelling paradigms, augmentation of existing methods, or even incorporation of tools from other disciplines, to ensure that anaerobic digestion research can remain resilient and relevant in the face of emerging and future challenges.

Symmetry structures in dynamic models of biochemical systems

Understanding the complex interactions of biochemical processes underlying human disease represents the holy grail of systems biology. When processes are modelled in ordinary differential equation (ODE) fashion, the most common tool for their analysis is linear stability analysis where the long-term behaviour of the model is determined by linearizing the system around its steady states. However, this asymptotic behaviour is often insufficient for completely determining the structure of the underlying system. A complementary technique for analysing a system of ODEs is to consider the set of symmetries of its solutions. Symmetries provide a powerful concept for the development of mechanistic models by describing structures corresponding to the underlying dynamics of biological systems. To demonstrate their capability, we consider symmetries of the nonlinear Hill model describing enzymatic reaction kinetics and derive a class of symmetry transformations for each order of the model. We consider a minimal example consisting of the application of symmetry-based methods to a model selection problem, where we are able to demonstrate superior performance compared to ordinary residual-based model selection. Moreover, we demonstrate that symmetries reveal the intrinsic properties of a system of interest based on a single time series. Finally, we show and propose that symmetry-based methodology should be considered as the first step in a systematic model building and in the case when multiple time series are available it should complement the commonly used statistical methodologies.

Existing Empirical Kinetic Models in Biochemical Methane Potential (BMP) Testing, Their Selection and Numerical Solution

Biochemical Methane Potential (BMP) tests are a crucial part of feasibility studies to estimate energy recovery opportunities from organic wastes and wastewater. Despite the large number of publications dedicated to BMP testing and numerous attempts to standardize procedures, there is no “one size fits all” mathematical model to describe biomethane formation kinetic precisely. Importantly, the kinetics models are utilized for treatability estimation and modeling processes for the purpose of scale-up. A numerical computation approach is a widely used method to determine model coefficients, as a replacement for the previously used linearization approach. However, it requires more information for each model and some range of coefficients to iterate through. This study considers existing empirical models used to describe biomethane formation process in BMP testing, clarifies model nomenclature, presents equations usable for numerical computation of kinetic parameters as piece-wise defined functions, defines the limits for model coefficients, and collects and analyzes criteria to evaluate and compare model goodness of fit.

Exploitation of the ADM1 in a XXI century wastewater resource recovery facility (WRRF): The case of codigestion and thermal hydrolysis

The aim of this study is to test the capability of the anaerobic digestion model n1 (ADM1) to reproduce data from full-scale digesters operated in a wastewater resource recovery facility (WRRF) where both thermal hydrolysis and codigestion with industrial waste are carried out. Furthermore, the potential uses of the model in a WRRF are also described, with particular relevance for plant engineers/operators. The model capability was calibrated and validated with data from full-scale digesters from the Mapocho-Trebal WRRF (Biofactoría) in Santiago, Chile. A success simulation rate, defined as the percentage of experimental values of a certain variable that lies within the simulation band given by a simulation tolerance established by the user/operator, was established to test the capability of the model as objectively as possible. Regarding the full-scale digester fed with thermally pretreated mixed sludge, success rates of 65% for biogas production and 60-100% for other variables were achieved. Regarding the full-scale digester in codigestion mode, the model had a success rate of approximately 60% for predicting the biogas flow for the whole evaluation period, while for the other variables, values between 70 and 100% were attained. The lowest success rates were observed for the volatile fatty acid (VFA) concentration in the digestate. Despite the lack of available data and the number of assumptions that had to be made, the model was demonstrated to be capable of reproducing the behavior of the full-scale reactors. A proper, up-to-date, calibrated and validated model can aid in the decision-making process in a WRRF, for instance, in determining some unmeasured inlet conditions, in improving the resilience of the process and in managing the incorporation of a new cosubstrate into the plant, among others.

Bacteriostatic effects of nanometer silver disinfectant on the biofilms in dental unit water lines

Background/purpose

Dental unit water lines (DUWLs) may be contaminated by aerobic bacteria in clinical settings and comprehensive disinfecting methods should be considered without delay. Herein, this study aims to investigate the timeliness and dynamic bacteriostatic effects of different forms of nanometer silver (NMS) disinfectant on bio-film in DUWLs.

Materials and methods

Bacterial DUWLs samples were respectively treated with different NMS forms, including liquid phase and solid phase at the concentrations of 0.25%, 0.5%, 1% and 2% and their bacteriostatic effects were observed at the 1st, 4th, 7th, 14th, ^28th day.

Results

The bacteriostatic effects of liquid phase NMS at all concentrations were unsatisfactory and the bacteriostatic rate was only 20% at the 1st day. However, there appeared massive bacteria growth at the 4th, 7th, 14th, 28th day. Comparatively, no bacteria growth was found at the 1st, 4th, 7th, 14th, ^28th day after sterilizing with different concentrations of solid phase NMS and the bacteriostatic rate was 100%.

Conclusion

Microbial contamination in DUWLs can be disinfected by different NMS forms, among which solid phase NMS is more bactericidal against bacteria bio-films, demonstrating significant roles of solid phase NMS in preventing DUWL contamination.

An Unstructured Model for Anaerobic Treatment of Raw Cheese Whey for Volatile Fatty Acids Production

The whey is a byproduct of the dairy industry that, if not treated properly, can cause serious environmental pollution problems. Anaerobic treatment is an alternative for its recovery, since, in addition to reducing the organic load. it allows the generation of value-added products such as volatile fatty acids (VFA) and biogas. However, the process is very complex and requires specific operating conditions that guarantee its stability and favor the production of value-added compounds. In this work, an unstructured mathematical model is proposed to evaluate the dynamic behavior of the stages of the anaerobic degradation process of the whey (i.e., hydrolysis, acidogenesis, acetogenesis and methanogenesis). The proposed model considers the dynamic variation in pH during the experiment. To validate the model, an experimental set was carried out at pH and temperature conditions that favor the production of VFAs. Experimental results show that the anaerobic treatment of the raw cheese whey favors pH = 5.5; for T = 40 °C, the maximum VFA production is obtained (30.71 gCOD L−1), and for T = 35 °C, a 45.81% COD degradation is reached. The proposed model considers the effect of pH and temperature and it is validated in the region where the experimental tests were carried out. The model parameters were estimated using the Levenberg–Marquardt method, obtaining coefficients of determination R2 > 0.94. The proposed model can describe the dynamic behavior of the key variables in the anaerobic treatment of raw cheese whey at different pH and temperature conditions, finding that VFA production is favored at pH ≥ 7, while the highest COD removal results in acidic conditions

Interpretable machine learning for predicting biomethane production in industrial-scale anaerobic co-digestion

The objective of this study is to apply machine learning models to accurately predict daily biomethane production in an industrial-scale co-digestion facility. The methodology involved applying elasticnet, random forest, and extreme gradient boosting to input-output data from an industrial-scale anaerobic co-digestion (ACoD) facility. The models were used to predict biomethane for 1-day, 3-day, 5-day, 10-day, 20-day, 30-day, and 40-day time horizons. These models were fit on four years of operational data. The results showed that elastic net (a model with assumptions of linearity) was clearly outperformed by random forest and extreme gradient boosting (XGBoost), which had out-of-sample R²values ranging between 0.80 and 0.88, depending on the time horizon. In addition, feature importance and partial dependence analysis demonstrated the marginal and interaction effects on biomethane of selected biowaste inputs. For instance, food waste co-digested with percolate were shown to have strong positive interaction effects. One implication of this study is that XGBoost and random forest algorithms applied to industrial-scale ACoD data provide dependable prediction results and may be a useful complement for experimental and mechanistic/theoretical models of anaerobic digestion, especially where detailed substrate characterization is difficult. However, these models have limitations, and suggestions for deriving additional value from these methods are proposed.

Self-optimizing attainable regions of the anaerobic treatment process: Modeling performance targets under kinetic uncertainty

Despite the advantage of model-based design, anaerobic digesters are seldom designed using biokinetic models due to lack of reliable kinetic coefficients and/or systematic approaches for incorporating kinetic models into digester design. This study presents a systematic framework, which couples practical identifiability, uncertainty quantification and attainable region (AR) concepts for defining process performance targets, especially when reliable kinetic coefficients are unavailable. Within the framework, we introduce the concept of self-optimizing ARs, which define performance targets that results in near optimal operation in spite of variations in kinetic coefficients. Using the case of modified Hill model, only 3 out of the 6 model parameters (unidentifiable set) are responsible for the model prediction uncertainty. The uncertainty bands (mean, 10th percentile and 90th percentile) on the model states has been computed using the Monte Carlo Simulation procedure and attainable regions for the different levels of uncertainty has been constructed and the boundaries interpreted into digester structures. The self-optimizing attainable regions have been defined as the intersection region of the attainable regions corresponding to the mean, 10th percentile and 90th percentile. Incorporating uncertainty significantly reduces performance targets of the process but increases self-optimality in defining performance targets. Unlike the attainable region, which represents the limits of achievability for defined kinetics, the self-optimizing attainable region represents the set of all possible states attainable by the system even in cases of kinetic uncertainty. In summary, the concept of self-optimizing ARs provides a systematic way of defining process performance targets and making economic decisions under conditions of uncertainty.

Review of Anaerobic Digestion Modeling and Optimization Using Nature-Inspired Techniques

Although it is a well-researched topic, the complexity, time for process stabilization, and economic factors related to anaerobic digestion call for simulation of the process offline with the help of computer models. Nature-inspired techniques are a recently developed branch of artificial intelligence wherein knowledge is transferred from natural systems to engineered systems. For soft computing applications, nature-inspired techniques have several advantages, including scope for parallel computing, dynamic behavior, and self-organization. This paper presents a comprehensive review of such techniques and their application in anaerobic digestion modeling. We compiled and synthetized the literature on the applications of nature-inspired techniques applied to anaerobic digestion. These techniques provide a balance between diversity and speed of arrival at the optimal solution, which has stimulated their use in anaerobic digestion modeling.

ADM1 approach to the performance optimisation and biogas H2S prediction of a large-scale anaerobic reactor fed on sugarcane vinasse

In this paper, we present extensions to the Anaerobic Digestion Model No. 1 (ADM1) to simulate hydrogen sulphide in biogas and solids retention efficiency. The extended model was calibrated and validated against data from a large-scale covered in-ground anaerobic reactor (CIGAR), processing sugarcane vinasse. Comparative scenarios and set-ups of a CIGAR with and without a settling tank unit (settler) were simulated to investigate the reactor's performance. Biogas flow, methane content, and yield with settler were 15,983 Nm³/d, 57%, and 0.198 Nm³CH₄/kgCOD, respectively, which were 9.4%, 1.8%, and 11.64%, higher than without the settler. Improvements are combination of influent flow rate 116% higher and increased solids retention time by using a settler. The optimised modelled reactor, the volume of which was reduced by 50%, was able to produce 83% more methane per volume of reactor with half the retention time. After model calibration and validation, we assessed the quality of predictions and its utility. The overall quality of predictions was assessed as high accuracy quantitative for CH₄ and medium for H₂S and biogas flow. A practical demonstration of ADM1 to industrial application is presented here to identify the potential optimisation and behaviour of a large-scale anaerobic reactor, reducing, consequently, expenditure, risk, and time.

Assessing practical identifiability during calibration and cross-validation of a structured model for high-solids anaerobic digestion

High-solids anaerobic digestion (HS-AD) of the organic fraction of municipal solid waste (OFMSW) is operated at a total solid (TS) content ≥ 10% to enhance the waste treatment economy, though it might be associated to free ammonia (NH₃) inhibition. This study aimed to calibrate and cross-validate a HS-AD model for homogenized reactors in order to assess the effects of high NH₃ levels in HS-AD of OFMSW, but also to evaluate the suitability of the reversible non-competitive inhibition function to reproduce the effect of NH₃ on the main acetogenic and methanogenic populations. The practical identifiability of structural/biochemical parameters (i.e. 35) and initial conditions (i.e. 32) was evaluated using batch experiments at different TS and/or inoculum-to-substrate ratios. Variance-based global sensitivity analysis and approximate Bayesian computation were used for parameter optimization. The experimental data in this study permitted to estimate up to 8 biochemical parameters, whereas the rest of parameters and biomass contents were poorly identifiable. The study also showed the relatively high levels of NH₃ (i.e. up to 2.3 g N/L) and ionic strength (i.e. up to 0.9 M) when increasing TS in HS-AD of OFMSW. However, the NH₃ non-competitive function was unable to capture the acetogenic/methanogenic inhibition. Therefore, the calibration emphasized the need for target-oriented experimental data to enhance the practical identifiability and the predictive capabilities of structured HS-AD models, but also the need for further testing the NH₃ inhibition function used in these simulations.

Simulation of two-dimensional attainable regions and its application to model digester structures for maximum stability of anaerobic treatment process

Unlike high-rate anaerobic digesters that employ some mechanism to retain microbial sludge mass, low-rate systems use sufficiently long hydraulic retention times to ensure process stability, which becomes economically unattractive for treating large quantities of waste. This study presents the use of attainable region to develop a new strategy to enhance the stability of low-rate digesters. By considering three digestion cases, diary manure only (batch 1) or diary manure with granular (batch 2) or lagoon (batch) sludge as innoculum, the following findings were obtained. (1) For a given concentration of volatile acids in an anaerobic digester, higher concentrations of methanogenic archae can be attained using a digester structure (combination of different digesters) as opposed to single digester. (2) For a given digested substrate, a change in the source of inoculum results in a change in the limits of achievability by the system (attainable limits for batches 1, 2 and 3 were 46.486(g/L)², 5.562(g/L)² and 0.551(g/L)², which resulted in performance improvements of 118.604%,175.627% and 200.436% respectively), and hence optimal digester structure. The evidence from this study suggests that the technique can be used to simultaneously improve process stability, define performance targets and propose digester structures required to achieve a given target.

Integration of bioenergetics in the ADM1 and its impact on model predictions

In this work, the integration of dynamic bioenergetic calculations in the IWA Anaerobic Digestion Model No. 1 (ADM1) is presented. The impact of bioenergetics on kinetics was addressed via two different approaches: a thermodynamic-based inhibition function and variable microbial growth yields based on dynamic Gibbs free energy calculations. The dynamic bioenergetic calculations indicate that the standard ADM1 predicts positive reaction rates under thermodynamically unfeasible conditions. The dissolved hydrogen inhibition approach used in ADM1 is, however, deemed as adequate, offering the trade-off of not requiring dynamic bioenergetics computation despite the need of hydrogen inhibition parameters. Simulations of the model with bioenergetics showed the low amount of energy available in butyrate and propionate oxidation, suggesting that microbial growth on these substrates must be very limited or occur via alternative mechanisms rather than dissolved hydrogen.