Co-digestion of surplus yeast with brewery wastewater is a potentially economical method for recovering energy, in the form of biogas, from this difficult to dispose of by-product. In this work a modified version of the ADM1 (Anaerobic Digestion Model No. 1) was calibrated for an anaerobic digester fed with thermally pre-treated brewery yeast surplus wastewater. The model could predict changes to reactor methane production and reduction of biodegradable matter when fed with both pre-treated and untreated yeast surplus wastewater at varying loading rates. Model calibration focused on low temperature thermal pre-treatment as experiments into a combined thermal-alkaline pre-treatment did not show any significant improvements. A low temperature pre-treatment of 60 °C for 30 minutes was sufficient for yeast inactivation and allowed for stable and more efficient operation of the high-rate anaerobic digester over a period of 232 days. The low temperature and time for pre-treatment also reduced the evaporation of easily biodegradable residual ethanol present in the influent, while still maintaining a low level of suspended matter. Inline measurements of gas composition, production and effluent chemical oxygen demand were sufficient for reliable model calibration of these same outputs. More detailed characterization of influent and effluent is required if organic acid concentrations for pH control are needed.
This study introduces the concept of computer modelling and simulation of complex bioprocesses and systems using an approach that combines the reference net formalism with machine learning and optimisation techniques. Reference nets are an extension of high level Petri Nets, which can be used as a central visualisation and modelling tool. The net-innet paradigm used by reference nets makes it possible to model complex processes, such as those found in the food and beverage industry. A plugin/interface system based on the java programming language allows implementation of advanced mathematical modelling techniques at specific points in entire system simulations. Separate optimisation tools can also run and modify existing reference net models for fast solutions to efficiency problems.We present an example system that simulates a specific section of a beer brewery using the reference net formalism, which is optimised using a genetic algorithm. We show in detail how the different software packages can be combined for a simulation based optimisation approach. The optimisation technique specifically addresses the wastewater pollution load in regard to its chemical oxygen demand. A beer brewery was chosen as an example for this study due to the constantly increasing requirements to lower energy and water consumption in this industry. One possibility to lower the energy and water demands is to effectively treat wastewater produced by the brewery, which can introduce cost savings by providing recycled water and biogas. Most approaches to wastewater treatment are end-of-pipe solutions that do not consider the brewery as a whole. A brewery contains many processes that can be running concurrently and interacting with one another (e.g. brewing, clean-inplace and bottling) with each process producing varying amounts of wastewater with different pollution loads. Optimisation of the scheduling of the different processes with respect to the wastewater production will allow for more effective wastewater treatment, and therefore cost and energy savings.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.