Moisture content prediction in poultry litter using artificial intelligence techniques and Monte Carlo simulation to determine the economic yield from energy use.

The objective of this study is to determine the economic return of poultry litter combustion in boilers to produce bioenergy (thermal and electrical), as this biomass has a high-energy potential due to its component elements, using fuzzy logic to predict moisture and identify the high-impact variables. This is carried out using a proposed 7-stage methodology, which includes a statistical analysis of agricultural systems and practices to identify activities contributing to moisture in poultry litter (for example, broiler chicken management, number of air extractors, and avian population density), and thereby reduce moisture to increase the yield of the combustion process. Estimates of poultry litter production and heating value are made based on 4 different moisture content percentages (scenarios of 25%, 30%, 35%, and 40%), and then a risk analysis is proposed using the Monte Carlo simulation to select the best investment alternative and to estimate the environmental impact for greenhouse gas mitigation. The results show that dry poultry litter (25%) is slightly better for combustion, generating 3.20% more energy. Reducing moisture from 40% to 25% involves considerable economic investment due to the purchase of equipment to reduce moisture; thus, when calculating financial indicators, the 40% scenario is the most attractive, as it is the current scenario. Thus, this methodology proposes a technology approach based on the use of advanced tools to predict moisture and representation of the system (Monte Carlo simulation), where the variability and uncertainty of the system are accurately represented. Therefore, this methodology is considered generic for any bioenergy generation system and not just for the poultry sector, whether it uses combustion or another type of technology.