The objective of this demonstration was to show that geothermal heat pump efficiency during summer air conditioning could be improved by removing heat from the ground loop in the wintertime using dry fluid coolers.  A secondary objective was to develop an Application Manual for this technology to help Department of Defense (DoD) facility managers determine if this technology could be applied to their facility, and to help them design the systems.

Geothermal heat pump systems use the ground as a heat source and heat sink to heat and cool buildings. These systems, also known as ground source heat pump systems, use reversible heat pumps to either extract or reject heat into a water loop (the ground loop) that runs through the building, and interacts with the subsurface through trenches, wells, or boreholes.

When the building is being cooled, heat from the building is absorbed by a liquid refrigerant in the evaporator, converting the liquid into a vapor. The cool low pressure vapor is compressed in an electrically driven compressor to convert it into a hot, high pressure vapor. A heat sink (the ground loop) is used to remove heat from the hot vapor causing it to condense back into a liquid. The liquid is then routed back to the evaporator to complete the cycle.  The basic principle of operation is that the building heat is transferred using the latent heat from vaporization of the refrigerant from the heat source to the heat sink. Several units of heat energy can be transferred per unit of electricity consumed by the compressor, making this method extremely efficient for heat transfer.

When the building is being heated, the refrigeration cycle is reversed, and heat is extracted from the heat source (the ground loop) to evaporate the liquid refrigerant. The refrigerant vapor condenses in a coil inside the building, releasing heat to the building.

There are three main types of ground loops: open loops using wells, closed loops using trenches, and closed loops that use boreholes. With an open loop, groundwater is pumped from a well through the heat pump system, and back into the ground through another well. This type of system can be very effective, but it requires access to a productive aquifer with associated permitting and water chemistry considerations.

Closed loop systems use sealed piping to move a mixture of water and antifreeze through the ground. Small household geothermal systems often use shallow trenches for these closed loops, but trenches become impractical for larger buildings, where the necessary length of the ground loop may be thousands of feet. The most common ground loop configuration for larger buildings consists of an array of vertical boreholes extending up to several hundred feet deep into the ground with a horizontal spacing of 20 feet or more.  These vertical boreholes are typically constructed by drilling a 6-inch diameter borehole. A high density polyethylene U-tube is installed in the borehole, and is grouted into place using a thermally conductive grout. This design isolates the ground loop fluid from the groundwater system, heat transfer between the ground loop, and the subsurface occurs by thermal conduction.

When the heat pump system is in air conditioning mode, the ground loop rejects the building heat into the ground loop, resulting in an increase in temperature in the subsurface. When the heat pump is in heating mode, heat is extracted from the ground loop, and delivered to the building, causing the ground to cool. 

An analysis of the data collected at this building showed if a dry fluid cooler was installed at the same time as the geothermal system, and operated efficiently, that it would eliminate the problem of ground loop temperature increase over time. The resulting savings in costs are large compared to the alternative of increasing the number of borehole heat exchangers to reduce the ground loop temperature. 

A primary performance objective for this demonstration was to show an increase in electrical efficiency of more than 15% compared to a conventional geothermal heat pump system without a dry fluid cooler. Using the same building and geothermal ground loop characteristics, a comparison was made between a system with and without a dry fluid cooler. During the first year of operation, the energy costs with and without the dry fluid were similar, because the system without a dry fluid cooler had not heated up much. After 10 years of operation, the annual energy costs were calculated to be about 12% lower for the system with the dry fluid cooler. This difference increased over time to about 19% by 30 years. 

Considering the capital cost for the dry fluid cooler, the payback period is calculated to be about 23 years (assuming an energy inflation rate of 5% and a general inflation rate of 2%). With a lower energy inflation rate of 2%, the payback period is 30 years or more. 

 A key assumption in these calculations was that the system without the dry fluid cooler would be able to operate for decades with very high ground loop water temperatures. There is a high likelihood that an external cooler would be required at some point in the near future simply to continue operation of the system. If this is the case, then it would be far better to install the cooler initially when the system is constructed and avoid the high ground loop temperatures from the start. The capital costs would be nearly the same, the energy and energy cost savings would be substantial.

The following guidelines are provided for application of wintertime cooling using dry fluid coolers in new systems. The first three steps are recommended for every geothermal system installed in cooling dominated areas.

1)   Calculate building heating and cooling loads using a building simulation tool such as eQUEST (Hirsch & Associates, 2016).

2)   Simulate a conventional geothermal heat pump system using a ground loop simulation tool such as GLHEPro (IGSHPA, 2016). There is a trade-off between increased ground loop size, and loop temperature. With the addition of a dry fluid cooler, a smaller ground loop can be used, but it must still be large enough to accommodate a reasonable flow rate for the building load.

3)   Using reasonable ground loop sizes, assess the degree to which the ground loop temperature will increase over the expected life of the system. If the average loop temperature increases by more than about 15°F, a dry fluid cooler would be beneficial.

4)   The dry fluid cooler should be sized to match the ground loop flowrate; the flowrate should fall within the cooler design range.

5)   The dry fluid cooler should be sized and operated so that it can reject an amount of heat equal to the yearly cooling load minus the yearly heating load. Using variable frequency drive fan motors, the fan speeds should be controlled by the temperature difference between the water entering the cooler and the outside air, reaching maximum fan speeds when the temperature difference is large (~20 degrees or more). Most of the loop heat will be rejected in the winter using this type of operation, and the heat rejection energy efficiency can be very high.

On a yearly average basis, the dry fluid cooler should be sized so that the heat rejection is on the order of 10-20% of the cooler rated capacity. For example, if the desired yearly cooler heat rejection is 600,000 kBTU, this is equivalent to an average rate of about 5.7 tons. This would be about 12% of the rated capacity of a 4-fan, 48-ton dry fluid cooler. During the peak cooling months of December and January, the cooler is likely to be operating at 25-35% of the rated capacity (averaged over the month).

6)   The ground loop system should include antifreeze to prevent damage to the dry fluid cooler during freezing temperatures.

For existing systems that are suffering from high ground loop temperatures, a retrofit following steps 4–6 can be used to stabilize and reduce the temperatures. In this case, the dry fluid cooler heat rejection during the first year or two of operation will be higher due to the higher ground loop temperatures.