Predicting energy needs in mechanically ventilated Broiler houses in South Korea using a building energy simulation-driven algorithm

Accurately predicting energy needs in livestock farms is vital for efficient design and operation. A lack of research on energy requirements in livestock facilities has often resulted in a reliance on empirical knowledge rather than data. This study addresses this gap by using Building Energy Simulation to assess and predict the thermal characteristics of broiler houses. Numerical simulations under 17,280 experimental conditions were conducted to determine cooling and heating energy needs based on regional climate data, building specifications, orientation, insulation, and broiler occupancy. Meteorological and internal climate variables were used to compute sensible and latent heat production and ventilation requirements. The simulation model was validated against multiple broiler rearing periods using experimental data. Validation results showed a Normalized Mean Bias Error of 0.7806 %, a Coefficient of Variation of the Root Mean Square Error of 9.7510 %, and a Mean Absolute Percentage Error of 3.824 %. Regression models for energy needs were developed for nine administrative divisions, four orientations, and twelve insulation configurations. These findings reduce construction costs and improve operational efficiency. Additionally, a regression-based application was developed to predict heating and cooling energy needs, providing an intuitive tool for farm owners and consultants to support decision-making in broiler house construction and energy management. • BES-based model developed to estimate heating and cooling energy needs. • Model validated using field data from four rearing periods. • Regression equations derived for monthly and annual energy needs. • Predictive model integrated into a mobile tool for farm application.