Linearized Kinetic Models for the Simulation of the Mesophilic Anaerobic Digestion of Pre-hydrolyzed Woody Wastes

Abstract

The most common kinetic models available in the literature to describe the COD consumption during batch anaerobic digestion are checked and compared in their linearized forms in the case of prehydrolyzed mixtures simulating the composition of woody wastes. The introduction of the COD value at the end of batch digestion in both the First Order and the Monod models allows one to take into account the deviations due to the progressive increase in the fraction of difficultly-digestible substances present in the medium. The so-modified version of the First Order model as well as that proposed by Singh and co-workers not only show the best fits to the experimental data but also demonstrate that they are the simplest and easiest ones for immediate use.

Efficient Biogas Production through Process Simulation

“…8,[10][11][12][13][14] A smaller number of models uses various mathematical methods and control engineering theories, such as generic algorithms, fuzzy logic, neural networks, or stochastic methods. [15][16][17][18][19] A subgroup of models can be characterized by its specialization on particular substrates, [20][21][22] reactor types, [23][24][25] and specific modeling purposes, [26][27][28][29] respectively. Generally, one has to weigh high modeling complexity against the necessary effort for model parameterization.…”

“…Whereas some models − are of comparably basic nature with respect to the number of modeled state variables and process steps, other models simulate the biogas process almost entirely, including a wide variety of intermediate products. − Another distinction can be drawn with respect to modeling techniques. A large group of models is mainly based on mass and energy balances as well as different growth and product formation/degradation kinetics. ,− A smaller number of models uses various mathematical methods and control engineering theories, such as generic algorithms, fuzzy logic, neural networks, or stochastic methods. − A subgroup of models can be characterized by its specialization on particular substrates, − reactor types, − and specific modeling purposes, − respectively.…”

Development and application of an automatic feeding control to manage anaerobic co-digestion of winery wastes

“…Their parameters have a physical interpretation and are adjustable, for instance by a parameter estimation procedure (Lauwers et al, 2013). On the other hand, they can't describe the degradation of complex substrates, characterized by simultaneous presence of compounds with different degradability, because a higher number of parameters should be considered (Converti et al, 1999). For this reason, the most common grey-box models proposed for anaerobic digestion (Monod, first order, Contois, Singh) are inadequate to represent the actual situation inside the digester fed by complex substrates (Cecchi et al, 1990a;Converti et al, 1999).…”

“…On the other hand, they can't describe the degradation of complex substrates, characterized by simultaneous presence of compounds with different degradability, because a higher number of parameters should be considered (Converti et al, 1999). For this reason, the most common grey-box models proposed for anaerobic digestion (Monod, first order, Contois, Singh) are inadequate to represent the actual situation inside the digester fed by complex substrates (Cecchi et al, 1990a;Converti et al, 1999). Step-diffusional model represents the evolution of grey-box models because takes into account the nature and chemical characteristics of compounds present in the substrates, including the extent to which they are removed.…”

“…In fact, it considers four specific groups of compounds on basis of different utilization rate, clearly identified observing the plotting of biogas production rates (GPR) versus time during the semi-continuous degradation. The stepdiffusional model was mainly applied to AD of organic fraction of municipal solid waste (Cecchi et al, 2013(Cecchi et al, , 1997, but recently this model has been applied also to AD of prehydrolysed woody wastes (Converti et al, 1999) and to winery wastes (Da Ros et al, 2014).…”