Unlocking advanced waste management models: Machine learning integration of emerging technologies into regional systems

Abstract

The waste management sector requires specialized systems analysis tools to facilitate decision-making and make waste management sustainable and efficient. While integrated systemic approaches exist for assessing conventional waste management systems, the integration of emerging technologies such as gasification, pyrolysis, and methane dry reforming remains largely overlooked. In this work, these three technologies have been integrated into a conventional regional waste management model by abstracting rigorous simulation models into machine-learning surrogate models. The resulting technology-rich waste management model incorporates material flow analysis and life-cycle assessment as tools for supporting policy and decision-making. The model was tested by assessing the environmental impacts and landfill rates for three technology implementation scenarios. Overall, the inclusion of these emerging technologies led to an environmental performance improvement compared to a reference system. For example, a 116.5 % reduction of the carbon footprint in the most optimistic scenario. Nevertheless, the mere addition of these technologies was not enough to achieve landfill rates below 10 %, reaching 37.6 % in the most optimistic scenario. Therefore, properly sizing capacity was found to be a key factor in minimizing both environmental impact and landfill rate.