Predictive regression models for biochemical methane potential tests of biomass samples: Pitfalls and challenges of laboratory measurements

This paper is a compilation of some experimental results published in peer-reviewed articles dealing with predictive regression models between biochemical methane potential tests and different chemical parameters characterizing the organic content of biomass samples. Results reviewed were focused on laboratory measurements with the main objective of bringing together the existing experience to evaluate pitfalls and challenges that could be generalized for future research using this kind of substrates. Firstly, BMP test measurements were briefly described for experimental approaches according to different factors such as inoculum, physical and chemical experimental conditions, inoculum to substrate ratio and gas measurement systems. A lot of information necessary when reporting BMP studies was not included in the description of most articles. It is also unexpectedly the lack of positive control tests as a way to check the reliability of the experimental results obtained. As consequence, BMP test results from different laboratories are normally inconsistent and irreproducible. Secondly, chemical parameters analysed in experimental research works such as moisture/dry matter, total chemical oxygen demand, carbohydrates, lipids, proteins and lignin were also reported in a comparative way. In fact, 70% of analytical determinations were covered in some degree, but the presence of a correct reference description was only occasional. Finally, general regression models were summarized. However, the development of one overall model that applies to all kind of samples is difficult to achieve. In order to be reliable and widely applicable, predictive regression models for methane production of biomass samples should be based on accurate laboratory measurements.