Improving Smart Farm Heating Performance and Energy Efficiency by Means of AI Data Center Waste-Heat Recovery

This study evaluated the seasonal heating performance of a single-span strawberry greenhouse maintained at a constant heating setpoint temperature using a conventional air-source heat pump (ASHP) or water-to-water heat pump (DCHP) connected to a data center. Python-based dynamic simulations were conducted to calculate the greenhouse’s energy balance, incorporating the factors of conduction, convection, and ventilation driven by weather inputs. The heat pump performance was estimated by means of the temperature-based coefficient of the performance correlations. The results indicate that integrating a greenhouse with a data center improves seasonal efficiency and reduces electricity consumption. The DCHP increased the seasonal performance factor from 3.40 (ASHP) to 3.94, resulting in an absolute gain of 0.54 (approximately 15.8%). Additionally, annual compressor electricity consumption was reduced from 38,800 kWh to 26,942 kWh, resulting in savings of 11,858 kWh (approximately 30.5%). These savings are particularly significant during the winter months, when a warmer and more stable heat source can enhance the performance of the heat pump. Considering the energy mix factor, the annual greenhouse gas emissions decreased from 18,550 kg CO₂-equivalent to 12,881 kg CO₂-equivalent, highlighting a clear carbon reduction benefit in addition to energy savings.