Modeling of real scale waste activated sludge anaerobic digestion process by anaerobic digestion model 1 (ADM1)

It is essential to dispose sewage sludges, known as biosolid that occur as results of water and wastewater treatment, with proper treatment processes which fulfill the required environmental healt criteria. Agricultural land application, landfilling and incineration are the final disposal methods of municipal sewage sludge. It is necessary to apply thickening, stabilization, conditioning and dewatering processes to municipal sewage sludge before its final disposal. Stabilization includes a series of biological and chemical processes that are implemented on sludge to reduce the organic content in the sludge (transforming the sludge to more stable or inert organic and inorganic substances), remove pathogens and toxicity and reduce the odor potential. Anaerobic digestion process is one of the most commonly used biological stabilization method. Anaerobic digestion provides electric and thermal energy production via biogas production as well as sludge stabilization. The aim of this thesis is modelling of real scale anaerobic digestion process treating municipal sewage sludge by using Anaerobic Digestion Model No:1, ADM1. As a detailed and a new model, ADM1 is developed by International Water Association (IWA). ADM1 is a structured model identified by a number of kinetic parameters with disintegration and hydrolysis, acidogenesis, acetogenesis and methanogenesis steps. Model includes 26 dynamic state concentration variables, 19 biochemical kinetic processes, 3 gas-liquid transfer kinetic processes and 8 implicit algebraic variables per liquid reactor. As a differential equation set, there are 32 dynamic state concentration variables and additional 6 acid-base kinetic processes per reactor. In this study, real scale wastewater treatment plant (Hurma WWTP) sludge digestion process was modeled with the purpose of generating the data to understand the process better by contributing to the prediction of the process operation contidions and yields which will function as a base to future anaerobic sludge stabilization process operations. In this context, Hurma WWTP sludge digestion plant process data was supplied for a year, obtained process variables were evaluated in terms of process efficiency, important kinetic parameters in ADM1 were estimated with parameter estimation method using Aquasim program, real scale plant process variables were predicted via ADM1 model by using the determined kinetic parameters and the obtained results were discussed. Hurma WWTP anaerobic sludge digestion unit is evaluated with regards to process and state variables (influent sludge flow, anaerobic reactor VS loading, pH, temperature, VFA and alkalinity concentrations, retention time etc.) and process yields (VS removal, methane production yield and methane production rate). Hurma WWTP anerobic sludge digestion unit VS removal, methane production yield and methane production rate were calculated as 46.4%, 0.48 m3CH4/kgVSremoved and 0.33 m3CH4/m3day, respectively. It is a common approach in the model studies that are performed with ADM1 to estimate the kinetic parameters with high impacts on model results and use the values of kinetic parameters which were suggested in literature with low impacts on model results. In this study, to estimate the variables of real scale sludge digestion process with high accuracy and to provide high model prediction power, the values of four kinetic parameters kdis (d-1): disintegration rate constant, khyd_ch (d-1): carbohydrate hydrolysis rate constant, khyd_pr (d-1): protein hydrolysis rate constant, khyd_li (d-1): lipid hydrolysis rate constant) that have a strong effect on structured ADM1 were calculated by using the parameter estimation modul in the Aquasim program. kdis, khyd_ch, khyd_pr and khyd_li values were found as 0.101 (d-1), 10 (d-1), 10 (d-1) and 9.99 (d-1), respectively. Other kinetic parameters which were accepted as they have low effect on ADM1 model, are chosen from the literature and they are used in the modelling. Selected target variables in the anaerobic sludge digestion system – produced biogas flow (Qgas), percentage of CH4 in biogas, produced methane flow (QCH4), anaerobic reactor pH values, anaerobic reactor total alkalinity (TA) and volatile fatty acid (VFA) concentrations, effluent VS concentrations - were modeled with ADM1 structured kinetic model by using the kinetic parameters which were obtained after parameter estimation. Correlation coefficient (R) and determination coefficient (R2) values between measured biogas flow rates and estimated are calculated as 0.76 and 0.58, respectively. Obtained high correlation coefficient shows that structured ADM1 model estimated the quantity of produced biogas with high accuracy. ADM1 model captured the increase and decrease trends of biogas production in general. However, it is observed that the biogas production is strongly affected by system operation performance and system influent characteristics. Correlation coefficient and determination coefficient between estimated and measured values of alkalinity concentration are calculated as 0.74 and 0.54, respectively. ADM1 structural model estimated the values of anaerobic reactor alkalinity concentration with the high prediction performance. Although the decreases on the value of pH in reactor were not experienced, ADM1 model estimated pH values with a medium level accuracy (R=0.51). The correlation coefficient between the estimated and measured values of VFA is found as 0.43. The obtained correlation coefficient shows that structured ADM1 model estimated the values of anaerobic reactor VFA with a low prediction capability. Total VFA (acetic, propionic, butyric and valeric acid) is expressed as mg HAc/L and it is thouhgt that the limited number of VFA types determined in the model could be the cause of low correlation coefficient. Anaerobic reactor effluent VS concentration values are also estimated with relatively low prediction performance (R=0.29) by ADM1. Performance of ADM1 increased when anaerobic sludge system operated at steady state conditions. Decreased ADM1 performance is observed for estimating the system varibles values when anaerobic sludge digestion system was operated under unsteady state condition in comparison to steady state conditions. However, acquired ADM1 performance is in acceptable level since the digestion process is dynamic and in a non-lineer nature. Within this scope, ADM1 model approach can be used as a leading tool to understand the process of anaerobic digestion, to evaluate system performance and for process optimisation and design. However, it is essential to determine kinetic coefficients specific to the waste used in model and test in real scale plants to minimize the errors by establishing the necessary modifications. KEYWORDS: ADM1, Anaerobic digestion, Modelling of anaerobic digestion process, Sewage sludge COMMITTEE: Assoc. Prof. Dr. N.Altınay PERENDECİ (Supervisor) Prof. Dr. Ayşe FİLİBELİ Prof. Dr. Ayşe MUHAMMETOĞLU