The advantages of PV-assisted heat pumps are that the peak power production of a PV array often occurs and tracks with the periods of peak power demand of the heat pump in meeting the cooling load of the residence. This matching characteristic can be significant to the economics of the PV system, since the need for electric energy storage is greatly reduced while the use of the power output of the array is maximized. The need for power conditioning is also simplified if the heat pump is capable of working with the DC output of the array, as in modern, variable speed heat pumps.  A benefit can also be derived by the electric utilities, because the power provided by the PV array helps to reduce the peak demand.

Foster-Miller investigated the PV-assisted heat pump on an analytical and economic basis. The existing models ESPRE and PVCAD were used for this purpose. The performance of each system was investigated for the service territories of the participating utilities. Optimization of the array size, orientation, and operating strategy were done at this time.

The best size for the array was set at approximately 3 kW. This was determined by the number of modules needed to raise the voltage to the level required by the heat pump. The best orientation for the array is in a westerly direction, which tends to delay the peak output of the array to late afternoon, which produces better coincidence with the air conditioning peak demand. Optimum tilt for the array was found to be latitude minus 15 degrees, which resulted in higher array outputs in the summer months.  Several control strategies were investigated. The one seen most promising involved the use of set back temperatures to pre-cool the residence prior to the afternoon peak period. This strategy gave the greatest control over when array output was used to meet the air conditioning peak demand.

In parallel with the analysis work, a system design was performed to define the physical and electrical interface between the array and the heat pump. Issues of power conditioning, battery storage, and load sharing by the array and at the grid were addressed. Issues of safety were also examined, particularly for the high voltage array configuration and prevention of power feed to the grid during outage. The effects of harmonic distortion by the heat pump drive on the array, the residence, and the grid were also handled.

From the analysis and design tasks, the final system configuration for field test was specified.  The systems were then installed and instrumented at two test sites, one in Mesa, AZ and the other in Marfa, TX.  Remote data acquisition was employed to collect data files from these sites on a daily basis.  A test schedule was then carried out that allowed the testing of the PV-assisted heat pumps over a full range of ambient conditions experienced at each site.

The data collected provided the following:

• Complete power profiles and energy consumption information on the PV array, the heat pump and the residence.
• Determination of the cooling and heating loads.
• Characterization of the ambient conditions and insolation so that the test results can be generalized.

The field tests were conducted for one year at each site. The Marfa test site used considerably more energy for HVAC, due partly to the use of auxiliary electric heating and a lower set point for space cooling. The amount of solar energy utilized at Marfa was not proportionally greater that used in Mesa. This suggests that a large portion of the HVAC energy was used at night when output of the array was not available. The savings achieved at Marfa was lower, at 24.6 percent, than that seen at Mesa, 29.5 percent. 

Return to Robotics

Printable Version