Co-composting of Biowaste: Simultaneous Optimization of the Process and Final Product Quality Using Simulation and Optimisation Tools

Abstract

To improve the financial sustainability and business management of composting facilities, mostly small-scale and manually operated facilities, the simultaneous optimization of the process and quality of the final product must be prioritized. This study used artificial neural networks (ANNs) and particle swarm optimization (PSO) as tools to evaluate the composting of biowaste (BW) with different cosubstrates (i. star grass (SG) and ii. a SG and sugarcane filter cake (SFC) mixture). The simulation aimed to maximize product quality in the shortest processing time by varying the mixing ratio and turning frequency. The simulation showed optimal conditions with a turning frequency of twice per week with the following mixtures: (i) BW:SG (72.9:27.1), 76 days of processing; and (ii) BW:SFC:SG (60:16:24), 88 days of processing. The results showed the effect of the type of carbon source in the cosubstrates on the retention time, which may imply the need for a larger area in composting facilities. On the other hand, the findings show that a minimum time is required to achieve a product that meets quality standards, although a longer processing time reduces the agricultural value of the compost. This model can be used to define design criteria and operating conditions and select a cosubstrate that can contribute to improving the agricultural quality of the final product. Graphic Abstract: [Figure not available: see fulltext.].