Evaluation and machine learning prediction on thermal performance of energy walls in underground spaces as part of ground source heat pump systems

Abstract

Energy walls (i.e. wall heat exchangers) exhibit more complex thermal performance than do energy piles and borehole heat exchangers in ground source heat pump systems, due to their individual geometry. This study conducted field tests of energy wall thermal performance to analyze the influence and interaction of the pipe burial depth, flow rate, heating load, and operation mode. To evaluate the energy wall thermal performance in multiple dimensions, three novel evaluation indices (energy utilization ratio, thermal saturation, and effectiveness) were defined. A thermal performance SVR prediction model of an energy wall was established based on 227 field-measured samples. The heat exchange power, energy utilization ratio, and effectiveness of the ground side pipe are approximately 15%, 8% and 64% greater than those of the excavated side pipe, respectively. Considering the interaction between the flow rate and heating load, blindly increasing the heating load is not the most effective way to improve the thermal performance. The influences of heating loads and operation modes on heat exchange power differs from that on the energy utility ratio and both side’s thermal saturations, indicating that there is a complex interaction between the heating loads and operation modes. The cross-verification and prediction results show that the optimized thermal performance of the SVR model exhibited satisfactory generalizability and prediction performance.