Comparative Performance of Two Prototypes of a Passive Solar Heat Pipe System

Abstract

Thermal performance of an improved passive solar heat pipe system was directly compared to that of a previous prototype. Simulated and experimental results for the first prototype established baseline performance. Subsequently, potential improvements were simulated, and a second prototype was built and tested along side the first. The system uses heat pipes for high rates of heat transfer into the building, and limited losses in the reverse direction. The evaporator section of each heat pipe is attached to a glass-covered absorber on the outside of a south wall, and the slightly elevated condenser section is either immersed in water in a thermal storage tank or exposed to air in the room. Two-phase flow occurs in the heat pipe only when the evaporator is warmer than the condenser, creating a thermal diode effect. Computer simulations showed that system performance could be improved by using thicker insulation between the absorber and the storage tanks, and by switching from a copper to a rubber adiabatic section, which both reduced heat losses to ambient from the storage tanks. Early morning heating was improved by exposing one of five condensers directly to room air, which also improved overall system efficiency. A copper solar absorber soldered to the copper evaporator section improved heat conduction compared to the previous aluminum absorber bonded to the copper evaporator. Together these modifications improved simulated annual solar fraction by 20.8%. The new prototype incorporating these changes was tested along side the previous prototype in a two-room passive solar test facility during January through February of 2013. Temperatures were monitored with thermocouples at multiple locations throughout the systems, in each room and outdoors. Insolation was measured by four pyranometers attached to the building. Results showed that the design modifications implemented for the new model increased thermal gains to storage and to the room, and decreased thermal losses to ambient. The load-to-collector ratio for the experiments was 2.7 times lower than for the simulations, which decreased the potential for experimental improvements compared to the simulated improvements. However, average daily peak efficiency for the new system was 85.0%, compared to 80.7% for the previous system. Furthermore, the average storage temperature over the entire testing period for the new model was 13.4% higher than that of the previous model, while the average room temperature over the same period was 24.6% greater for the new system.