Improvement of the mathematical model of the extraction efficiency of landfill gas

Abstract

Landfill gas, the main component of which is methane, is formed in places where municipal solid waste is buried. According to the degree of damage to the environment, methane is considered the second most harmful greenhouse gas emitted into the Earth's atmosphere. Therefore, improving the dependence of the efficiency of landfill gas extraction on the main impact parameters in order to increase the prevalence of the use of renewable energy sources, save fossil energy sources and simultaneously reduce the intensity of environmental pollution is an urgent scientific and technical task. The purpose of the study is to improve the mathematical model of predicting the efficiency of landfill gas extraction with the aim of increasing the use of renewable energy sources, saving fossil energy sources and simultaneously reducing the intensity of environmental pollution. The research was carried out by second-order experimental design using the Box-Wilson method using a rotatable central composite design using the developed software, which is protected by a copyright certificate. An improved mathematical model for forecasting the efficiency of landfill gas extraction in different countries was obtained, which, unlike the basic model, provides greater convergence with actual data, contains a significantly smaller number of coefficients. It was established that, according to Fisher's test, the hypothesis about the adequacy of the obtained regression model can be considered correct with 95 % confidence. The correlation coefficient was 0.99992, which indicates the high reliability of the obtained results. The resulting improved regression dependence can be used to increase the prevalence of renewable energy sources, save fossil energy sources, and simultaneously reduce the intensity of environmental pollution. The response surfaces of the target function - the efficiency of landfill gas extraction in different countries and their two-dimensional cross-sections in the planes of influence parameters are constructed, which allow you to visually display the obtained dependence and the nature of the simultaneous influence of several factors on the target function.