Developments in Feedstock Analysis and Process Control for Biogas Production

Abstract: A sustainable treatment of organic wastes, generation of renewable energy, reduction of green house gas emissions, limiting the eutrophication and closing the nutrient cycles are several environmental challenges for which anaerobic digestion may be a suitable solution. There is today a large amount of available organic waste materials that could serve as potential feedstock for biogas production. However, their potential to produce biogas may vary significantly and in order to utilise them efficiently it is important to analyse them. One alternative to identify the quality of the feedstock is by performing a biochemical methane potential (BMP) test, a simple batch assay carried out under well-controlled conditions. In order to improve the reliability of such a test, the work described in this thesis investigated how a large number of environmental and experimental factors influenced the methane potential and degradation rate obtained from BMP tests of different material types. For example, correction of the gas volume for ambient temperature and pressure was found to be essential, and neglecting this factor can introduce errors up to 80% at the most extreme conditions when comparing results from different parts of the world. The solubility of both methane and carbon dioxide has been shown to introduce underestimations of gas flow measurements at flows below 1 and 192 L/day, respectively, when a volumetric method based on liquid displacement is used. Furthermore, important factors such as the inoculum to substrate ratio and substrate concentration have been shown to significantly influence both the methane potential and degradation rate, and the optimal settings for these may differ substantially between different inocula and substrate types. A drawback with BMP tests is the long time needed to follow the biogas production (often >30 days). Generating faster results would make it possible to use the test for alternative applications, e.g. quality checks and pricing of the feedstock. One approach to speed up the result generation is through the use of a kinetic model that predicts the methane potential at an earlier stage. For this purpose, a method was developed that could estimate the BMP of 188 samples within 10% relative root mean squared error just six days after initiating the test. Besides the limited knowledge of the feedstock, there is a need to improve the operation of anaerobic digesters. They are often operated far below their capacity as a consequence of an inadequate process monitoring and control. Several control strategies have been shown to improve the process operation, but as they often differ in their structure and methodology, there is a need to compare them as well as investigate their compatibility with a certain process. With this in mind, a simulation platform was developed and used to evaluate three previously suggested control strategies. It was found that these control strategies differed significantly in performance and that a control algorithm based on proportional control of pH using a cascade structure was best at both rejecting disturbances and maximising the gas production.