GROUND-SOURCE (GEOTHERMAL) HEAT PUMPS

by John W. LUND

1. INTRODUCTION
When a geothermal resource temperature falls below the 40-50°C range, it is frequently
impractical to use the fluid directly for most applications. A heat pump can be used to transfer heat
from a low-temperature resource to that of a high-temperature reservoir, thus providing the higher
temperature needed for space heating. The process can also be reversed by removing heat from
a high-temperature resource and rejecting it to a lower temperature reservoir, thus providing
cooling to a space.
Air-source heat pumps have been used for many years for both space heating and
cooling; however, their efficiency is influenced by the variation in outside air temperature. When
heat is most needed, the outside air is cooler, thus often requiring backup electric resistance
heating during the coldest days. Similarly, cooling is most needed during the hottest days,
requiring the equipment to work at low efficiencies.
Ground-source heat pumps, often referred to as geothermal heat pumps, overcome the
problem of resource variations, as ground temperatures remain fairly constant throughout the
year. Depending upon the soil type and moisture conditions, ground (and groundwater)
temperatures experience little if any seasonal variations below about 10 m.
The ground-source or geothermal heat pumps (GSHP or GHP), thus have several
advantages over air-source heat pumps. These are: (1) they consume less energy to operate, (2)
they tap the earth or groundwater, a more stable energy source than air, (3) they do not require
supplemental heat during extreme low outside temperature, (4) they use less refrigerant, (5) they
have a simpler design and consequently less maintenance, and (6) do not require the unit to be
located where it is exposed to weathering.
The main disadvantage is the higher initial capital cost, being about 30 to 50% more
expensive than air source units. This is due to the extra expense and effort to bury heat
exchangers in the earth or providing a well for the energy source. However, once installed, the
annual cost is less over the life of the system, resulting in a net savings. The savings is due to the
coefficient of performance (COP), averaging over 3 for GSHP as compared to 2 for air-source heat
pumps. A corresponding improvement is obtained in the cooling mode, as measured by the
energy efficiency ratio (EER). These terms are defined later in this paper.

NOMENCLATURE (Kavanaugh, undated)

Two basic types of ground-source heat pumps exist: ground-coupled and water-source.
There are a variety of names for the ground-coupled heat pumps. These include ground-source
heat pumps, earth-coupled heat pumps, earth energy systems, ground source systems,
geothermal heat pumps, closed-loop heat pumps, and solar energy heat pumps. Much of the
confusion arises from local marketing needs. Sales people may wish to connect GCHPs to
renewable energy sources (solar, geothermal), dissociate them from air hear pumps (GS systems),
or connect them to environmental awareness (earth energy). A generally (although not universally)
recognized nomenclature is shown in Fig. 1.
Ground-coupled heat pumps are a subset of ground-source heat pumps (GSHPs).
GSHPs also include groundwater and lake water heat pumps (water source). The distinguishing
feature of GCHPs is that they are connected to a closed-loop network of tubing that is buried in the
ground. The most common method of ground-coupling is to bury thermally-fused plastic pipe
either vertically or horizontally. A water or antifreeze solution is circulated through the inside of the
tubing and heat is released to or absorbed from the ground. No water enters the system from the
ground. Water-to-air heat pumps are located inside the building and are connected to the water
loop with a circulator pump. This type of system is referred to as a secondary fluid GCHP since
there is an intermediate liquid between the refrigerant and the ground.

149
 Figure 1: Ground-source heat pump types
A less frequently used system is referred to as a direct expansion (DX) GCHP.
Refrigerant lines are buried in the ground in either a vertical or horizontal arrangement. Thus, the
intermediate heat exchanger and fluid are eliminated. The possibility of higher efficiency than
secondary fluid GCHPs does exist. However, larger charges of refrigerant are required and
system reliability is compromised. Therefore, the future of DX GCHP is not clear because of
environmental concerns.
A variety of ground-coupled heat exchanger designs have recently been proposed and
demonstrated as cost-cutting alternatives. These include the Geo-Bag (a large plastic bag buried
in the ground and filled with water), a large diameter borehole with spiral coils, and the horizontal-
placed slinky coil (Fig. 2).

150
 Figure 2: Slinky coil ground-coupled heat pump system
2. DETAILS ON THE COMMON GROUND-SOURCE HEAT PUMPS SYSTEMS (LUND, 1989)
Two major types exist: ground-coupled (closed loop) or water source (open loop). The
ground-coupled uses a buried earth coil with circulating fluid in a closed loop of horizontal or
vertical pipes to transfer thermal energy to and from the earth. The water-source uses a well or an
open pond to provide an energy source or sink. Ground-coupled systems have been used in
northern Europe for many years, but were not used on a commercial scale in the U.S. until 1980.
Ground coupling is used where insufficient well water is available, where the quality of the well
water is a problem, where drilling and casing of wells are expensive, or where disposal of well
water is restricted.
In the horizontal mode of the ground-coupled system, pipes are buried in trenches spaced
a minimum of 1.5 m apart and from 1.2 to 1.8 m deep. This allows for minimum thermal
interference between pipes; however, this system is affected by solar radiation. Solar radiation will
cause a cycling of soil temperatures, that lags in time and decreases with depth due to the
insulating properties of the soil as shown in Fig. 3; however, the temperature is much more stable
than for air- source units. Moist soil will have greater temperature swings than dry soil. The loops
can be placed in a double layer as shown in Fig. 4. Vertical installation (Fig. 5) of the coils are
used where land space is limited or trenching would disturb the surface landscape, and drilling
costs are reasonable. Holes are drilled approximately 45 m deep and 4.5 to 6.0 m apart.

Figure 3: Depth vs. annual ground temperature variation for Tuscaloosa, Alabama

151
Figure 4: Two-pipe horizontal ground heat exchanger (source: Oklahoma State University):
Earth coil type: Horizontal - two-layer;
Water flow: Series;
Typical pipe size: 3.8-5.1 cm;
Practical length
Double loop: 18.2-26.0 m/kW;
Single loop: 36.4-52.0 m/kW;
Burial depth: 1.2-1.8 m

Figure 5: Series vertical ground heat exchanger (source: Oklahoma State University):
Earth coil type: Vertical - Single U-Bend;
Water flow: Series;
Pipe sizes: 2.5, 3.8 & 5.1 cm;
Bore length: 14.3-17.3 m/kW;
Pipe length (single pipe): 19.9-34.7 m/kW

152
 Computer programs have been developed (Dexheimer, 1985) to calculate the length of
horizontal earth coils for heating and cooling. Polyethylene pipes are the most popular in use, and
along with socket-fusion joining, are usually guaranteed for over 50 years.
Whereas, horizontal loops are affected by solar radiation, rain and wind, the vertical loops
are controlled by the mean-annual temperature of the area and the geothermal gradient and thus,
have a more stable temperature environment. Water wells are usually used where one is already
available, such as for domestic water supply. Normally, a minimum diameter of 15 cm and a
production of about 3.23 L/min/kW of heat pump capacity is required. 10.5 kW, a typical residential
load, requires about 34.1 L/min. The 15 cm diameter well casing is required to place the pump and
return line (Fig. 6). The fluid can either be returned to the well by the return line, placed in an
injection well, or disposed on the surface such as for irrigation. Pipes have also been anchored to
the bottom of surface ponds (minimum depth of 1.8 m); however, the heating and cooling
capacities are affected by solar radiation and other surface weather factors similar to the horizontal
loops. Installation is cheaper and heat transfer is more efficient; however, ponds do not maintain
a constant temperature as wells do and the pipes are more vulnerable to accidental damage.

3. HEAT PUMP OPERATION (KAVANAUGH, 1991)

The operation of the heat pump unit is the same for air-source and ground-source
configuration. The main difference is that the air-source requires an outside unit (accumulator and
fan) which may frost up in cold weather, requiring frequent defrosting. They also require a
backup heating source (electric or gas) when outside temperatures are too low for efficient
operation. The operation and cycle in both cooling and heating mode of the heat pump are shown
in Figs. 7 and 8 (Oklahoma State University, 1988).

Figure 6: Cross-section view of geothermal well (source: Water-Source Heat Pump Book)

Figure 7; Cooling cycle (source: Oklahoma State University)

153
 Figure 8: Heating cycle (source: Oklahoma State University)
As seen from these figures, the basic components of a standard heat pump are an electric
motor-driven compressor, a reversing valve, an expansion device, and two heat exchangers. A
desuperheater can be added to heat domestic hot water as shown on the left side of the figures.
One heat exchanger transfers heat between the heat pump and the environment, and the second
transfers heat to and from the interior of the building, referred to as the condenser or evaporator
(depending on which mode is used).
Refrigerant enters the compressor shell as a low-temperature, low-pressure gas. It
passes around the motor and is heated before entering the intake of the compression chambers.
The compression process elevates both the pressure and temperature of the refrigerant gas. In
the cooling mode (Fig. 7), this gas enters the reversing valve and is routed to the heat exchanger
in contact with the environment (ground heat exchanger). Since the gas is at a high temperature,
relatively cool air or water from the environment (geothermal source) can be used to remove
heat from the refrigerant in the heat exchanger. Removal of heat results in the cooling and
condensing of the refrigerant. Pressure loss is usually small in the condenser; therefore, the
refrigerant exits the condenser as a liquid with a temperature slightly above the environment's.
The liquid then experiences a drop in pressure across the restriction in the expansion
device. This causes a rapid decrease in temperature. The temperature inside the building is
much
warmer than the refrigerant entering the indoor heat exchanger. Therefore, the liquid is
evaporated, and in the process, heat is removed from the building air in the evaporator. Thus, we
have the desired cooling effect. The evaporated gas is then passed through the reversing valve
before returning to the compressor.
In the heating mode (Fig. 8), the solenoid in the reversing valve is switched so that the hot
compressor discharge gas is routed to the indoor heat exchanger. This exchanger is now used to
condense the hot refrigerant with the relatively cool indoor air; therefore, the desired heating effect
is carried out on the building. The condensed liquid enters the expansion device in a reverse
direction. The pressure loss results in a temperature decline so that the environment (ground)
transfers heat to the cold refrigerant. This causes the refrigerant to evaporate. The low-
temperature, low-pressure gas returns to the compressor through ports in the reversing valve.
Actual heat pumps may have additional components such as a multiple expansion
devices, fans, pumps, additional heat exchangers, auxiliary heat sources, accumulators, and
control and safety mechanisms. However, the basic means of “pumping heat" between the
building and the environment (geothermal source) is essentially the process shown in Figures 7
and 8.

154
4. EXAMPLES OF GROUND- AND WATER-SOURCE HEAT PUMP TYPES
(KAVANAUGH, 1991)
Figure 9 shows four common ground and water-source heat pumps. The home on the far
left utilizes a vertical closed-loop and ground-coupled heat pump. Water is circulated through
a water-to-refrigerant heat exchanger (condenser in cooling, evaporator in heating) in the heat
pump. Upon leaving out of the house and into underground headers buried 0.6-1.2 m below the
surface. In this installation, the flow is split into three vertical U-Bend ground couplings. Bore
depth is typically 13.0-21.7 m/kW of heat pump, pipe size of 1.9-3.8 cm, and bore diameter of 10-
15cm. Obviously, the driller in this installation stopped when the rock strata was encountered.
Deeper or more shallow bore depths are possible as long as the required bore per kW requirement
is maintained. Note also that water flow is split into three circuits. This is often necessary to
minimize water pump size.

Figure 9: Ground-coupled, groundwater and two lake-water heat pump systems

in a typical hydro-geological formation.
The next house is heated and cooled by a groundwater heat pump system. A well (10 cm
casing minimum) is drilled into the shallow aquifer. A well screen is placed on the bottom section
of the casing through which groundwater is extracted. A submersible pump located slightly above
the screen delivers water to the heat pump in the house. In this installation, the water is obviously
disposed of at the surface; however, it could also be reinjected if aquifer water levels are critical.
Two important requirements in these systems are a clean, well-developed well, and that the pump
should be located about a meter below the lowest seasonal static water level. It would not be
uncommon to drill to a deeper aquifer; although, the shallow aquifer is preferred in terms of
operating and first cost.
Lakes can be used in two manners. The home on the left shore utilizes an open-loop heat
pump system. Water is pumped from a submersible pump near (but not on) the lake bottom and
through the heat pump, and is returned near the surface of the lake. Above surface pumps can
also be used. They require greater power input, freeze precautions and some method of
preventing loss of suction during off-periods.
The home on the right lake front is heated and cooled with a closed-loop heat pump.
Much like the ground-coupled system, water (or a water/antifreeze mixture) is circulated through a
copper or thermally-fused plastic pipe coil and the heat pump. These systems do not require water
filtration like the open systems. However, they often do require freeze protection and protection
from damage, and performance improvements are moderate in shallow lakes.

4.1 Specific Examples of GCHP Use

Worldwide, GSHP account for 12% of the geothermal energy used for direct applications,
amounting to approximately 16,500 TJ (4,580 GWh) annually. Present estimates indicate that
there are 150,000 groundwater and 250,000 ground-coupled (55% vertical) heat pump installations

155
in the USA (updated from Ellis, 1989). The annual growth rate is estimated at 10%. Ground-
source heat pumps account for 63% of the geothermal direct use in the USA, amounting to 12,000
TJ (3,340 GWh) annually. A typical installation, which would be for a single-family residence,
consists of a 10.5 kW (3 tons) using about 30 L/min with a 6°C temperature drop in the circulating
fluid. This would shave about 5 kW off winter-peak heating demand and about 2.5 kW from
summer demand. Thus, 200,000 homes using GSHPs would avoid constructing a new 1,000 MW
power plant. Although the incremental cost of the ground-coupled closed loop adds about US$
3,000 to the cost of a residential heating system, payback occurs in 3-5 years from money saved
on utility bills. Currently, the main GCHP uses in the USA is in the mid-western and southeastern
states, where many utilities were offering rebates of US$ 500-2000 to homeowners to install GSHP
in order to take advantage of the peak shaving (by increasing load leveling for the utility, referred to
as demand side management) (Lund, 1988; Lienau and Lund, 1992).
The largest GSHP installation in the United States is the Galt House East Hotel and
Waterfront Office Building in Louisville, Kentucky (Pinckley, 1995). Heating and air conditioning is
provided for 600 hotel rooms, 100 apartments, and 89,000 square meters of office space, totaling
161,650 square meters. The system can extract 177 L/s of groundwater from four wells at 14°C
and can either remove energy from the well water for heating or add heat to the well water from the
air conditioning. The water is then discharged into a storm-water system. The system provides
15.8 MW of cooling and 19.6 MW of heating. The hotel complex energy use is approximately 53%
of a similar non-GSHP system in an adjacent unit, for a monthly savings of approximately US
$25,000. The emission of CO2, SO2, HOx and particulates are also reduced.
There have also been increased utilization of GSHP in Europe, especially in Germany,
Austria and Switzerland. In Switzerland, more than 20,000 borehole heat exchangers (BHE) have
been installed and have been operating reliably for decades (Rybach and Sanner, 1999). A similar
number are also installed in Germany. A typical single-dwelling house has a capacity demand of
about 10 kW; however, the BHE system is 30 to 40% higher in cost in comparison with a
conventional oil-fired- system. Environmental awareness, enforced by a governmental subsidy
(US$ 200 per heat pump kW), is the main incentive for the BHE installations in Switzerland.

5. EQUIPMENT (KAVANAUGH AND RAFFERTY, 1991)

In the commercially available size range, equipment is in two basic configurations:
positive displacement and centrifugal. Centrifugal machines are used for the largest applications
with positive displacement equipment for smaller capacities. The following sections briefly discuss
each of these types of equipment.

5.1 Positive Displacement

Reciprocating compressor heat pumps, the most common positive displacement type, are
available as standard units in sizes generally below 0.88 MW heating output (McQuay, 1986). This
equipment employs a 1, 4, 6, or 8-cylinder compressor on smaller equipment and multiple 4 or 6-
cylinder compressors on larger units (Carrier Corp., 1981). These units are also used for the
smaller residential applications of around 10.5 kW.
Capacity control is accomplished by suction cut-off type cylinder unloading down to −15-
20% capacity below which hot gas by-pass must be employed. As a result, it is important,
particularly for space heating applications, that equipment selection considers off-peak operation.
The number of control steps is dependent upon the number of compressor cylinders, with 4-step
control available on the smaller units and up to 8-step control on the large units.
Table 1 illustrates the off-peak performance of small, medium and large (with respect to
the 0.88 MW capacity) reciprocating heat pumps.
The increase in efficiency at part load is because of the nature of capacity control
employed by the manufacturer from which data were taken. This is a result of a special unloading
arrangement and the part load operation of two separate refrigerant circuits on the heat pump.
This increases the amount of heat transfer area available in the evaporator and condenser relative
to load requirements, thus increasing efficiency.
The refrigerant employed is a function of the temperatures between which the machine is
working. Table 2 presents a summary of refrigerant temperature limitations.

156
 Table 1. Off-Peak Performance of Reciprocating Heat Pumps (McQuay, 1986)
Small Medium Large
Capacity kW Capacity kW Capacity kW
(%) (%) (%) (%) (%) (%)
100 100 100 100 100 100
86 81 92 88 94 90
68 60 84 76 87 80
24 20 72 64 75 62
- - 60 54 59 46
- - 31 27 51 37
- - - - 42 29
- - - - 23 14

Evaporators are the shell-and-tube type with water generally on the shell side. However,
one major manufacturer produces equipment with water on the tube side. Condensers are also
shell-and-tube with water on the tube side. Reciprocating machines do not generally include a
separate liquid sub-cooling heat exchanger, though sub-cooling is addressed in condenser circuitry
(Carrier Corp., 1987).
Table 2: Reciprocating Heat Pump Refrigerant Temperature Limitations
(McQuay, 1986; Carrier Corp., 1981)
Maximum Condenser Minimum/Maximum Evaporation
Refrigerant Leaving Water Temperature Leaving Water Temperature
(°C) (°C)
R-22 54 6/32
R-500 66 4/38
R-12 77 4/38
R-114 104 21/49

Packaged reciprocating heat pumps are supplied from the factory with all safety and
operating controls for the machine including, in most cases, compressor starters. The machines
need only to be interfaced with system controls and a power source.
Other positive displacement models are rotary (rolling piston, rotary valve, single screw
and twin screw), and orbital (scroll and trochoidal). Their performance is similar to the
reciprocating compressor. See the ASHRAE Equipment Guide for more details.

5.2 Centrifugal
Centrifugal heat pumps are available in capacities ranging from − 0.29-7.32 MW in a
single unit (McQuay, 1983). The equipment features a single or dual compressor, depending upon
the size. One large manufacturer of this equipment in the U.S. employs a high-speed wheel,
driven by a hermetically-sealed squirrel cage motor through a single helical-gear couple. Motor
cooling is provided by controlled liquid refrigerant injection. A second manufacturer employs a 3-
stage compressor operating at motor speed. Refrigerant temperature limitations are similar to
those shown in Table 2 for reciprocating equipment.
Stable part-load operation is maintained by inlet guide vanes with the assistance of an
adjustable diffuser block at the wheel exhaust (McQuay, 1983). Construction of the balance of the
machine is similar to that of the reciprocating machine with the exception that the source water in
the evaporator flows through the tubes, rather than the shell as in reciprocating equipment. This
configuration permits the use of alternate tube construction materials to accommodate (without the
use of a heat exchanger loop) aggressive fluids in certain applications.

6. PERFORMANCE (LIENAU, ET AL., 1995)

The energy performance of a GSHP system can be influenced by three primary factors:
the heat pump machine, the circulating pump or well pumps, and the ground-coupling or
groundwater characteristics.

157
 The heat pump is the largest single energy consumer in the system. Its performance is a
function of two things: the rated efficiency of the machine and the water temperature produced by
the ground-coupling (either in the heating or cooling mode). The most important strategy in
assembling an efficient system is to start with an efficient heat pump. It is difficult and expensive to
enlarge a ground-coupling to improve the performance of an inefficient heat pump.
Water-source heat pumps are currently rated under one of three standards by the
American Refrigeration Institute (ARI). These standards are ARI-320, ARI-325, and ARI-330.
The standard intended for ground-coupled systems is ARI-330 entitled "Ground Source Closed
Loop Heat Pump Equipment." Under the standard, ratings for cooling EER and heating COP are
published. It's important to consider that these are single-point ratings rather than seasonal values
as in the case of the air-source equipment. Cooling EER values are based on an inlet water
temperature of 25°C. Heating COP values are based on a heating inlet temperature of 0°C.
These values are characterizations of a northern climate.
The current ARI directory contains equipment with EER ratings of less than 10 to a high of
18.6. COP values range from 2.8 to 3.6. It is evident that there is a wide range of equipment
performance at the standard rating conditions. Based on these values, it is evident that the
performance of the equipment can vary by as much as 100% according to the quality of heat pump
purchased.
GSHP Performance
5.0

4.5

4.0
COP

3.5

3.0

2.5

2.0
30 40 50 60 70
Entering Water Temperature, F

1994 1987

Figure 10: GSHP performance improvements from 1987 to 1994 for heating mode
GSHP Performance
20

18

16
EER

14

12

10

8
50 55 60 65 70 75 80 85
Entering Water Temperature, F
1994 1987

Figure 11: GSHP performance improvements from 1987 to 1994 for cooling mode
In recent years, there has been a substantial increase in the efficiency of GSHP
equipment. Based on performance reported in the ARI directory for 1987 and 1994, the increase
in EER ranges from 26 to 56 percent, and in COP from 35 to 50 percent depending on the entering
water temperature. Figures 10 and 11 show this increase in performance for a typical machine
based on average values of EER and COP as a function of entering water temperature. Based
on improvements in performance of GSHPs from 1987 to 1994, the date of a GSHP installation
should be noted.

158
 The actual performance of the equipment is a function of the water temperature
produced by the ground-coupling. The values discussed above are based on standard rating
conditions (25°C cooling and 0°C heating). The actual temperatures are a function of the local
ground temperatures and the design of the ground-coupling. For example, in a region where the
local ground temperature is 16°C and the ground loop is designed for the customary -7 to -4°C
aboveground temperature, a heat pump rated at an EER of 16.8 would actually operate at an EER
of 14.2 under peak-load conditions. A poorly designed loop, which forces the unit to operate at -
1°C aboveground temperature, would reduce the value to less than 13.0. These examples are for
cooling operation which is the dominant load in commercial applications. The same relationship
holds for heating operations, however.
EER vs EWT
24

22

20

18
EER

16

14

12

10
50 60 70 80 90 100 110
Entering Water Temperature, (F)

Figure 12: EER for a 10.5 kW GSHP

Figures 12 and 13 show EER and COP as a function of inlet temperatures for a 10.5 kW
machine designed for ground-coupled systems suggests the following guidelines for pumping
power for commercial ground-coupled systems.

COP vs. EWT

5.5

5.0

4.5
COP

4.0

3.5

3.0
30 40 50 60 70 80
Entering Water Temperature, (F)

Figure 13: COP for a 10.5 kW GSHP

The system energy performance is also influenced by the pumping energy required to
circulate the fluid through the heat pump and the ground loop. One author (Kavanaugh, 1991) in
the design of ground-coupled systems suggests the following guidelines for pumping power for
commercial ground-coupled systems:
1. Efficient systems: <50 watts/ton (<14 watts/kW)
2. Acceptable systems: 50-100 watts/ton (14-28 watts/kW)
3. Inefficient: >100 watts/ton (>28 watts/kW)

159
 To put these values into perspective, consider an office building with a 175 kW cooling
load and heat pump units selected to operate at an EER of 14 under peak conditions.
With an efficient circulating pump design (10 watts/kW}, the energy demand of the
circulating pump would amount to 175 kW x 10 watts/kW = 1750 watts. Combining the pump
demand with the heat pump unit demand results in a system EER of 13.5.
The same building and equipment coupled to a poorly designed pumping system
consuming 34 watts (6,000 watts pumping power) per kW would yield a system EER of only 12.2;
thus, compromising the premium paid for the higher-efficiency equipment. As indicated above,
coupling this system to an inadequate ground-coupling could easily reduce the system EER to
between 10 and 11.
In summary, it is necessary when evaluating a ground-coupled system to consider the
efficiency of the machine, the adequacy of the ground-coupling and the nature of the pumping
design to fully understand the efficiency of the system.

6.1 Basic formulas

Energy Efficiency Ratio (cooling mode):
Qc coolingc apacity (kW)
EER = =
Qe electric power input (compressor) (kW)
Coefficient of Performance (heating mode):
Qh heating capacity (kW)
COP = =
Qe electric power input (compressor) (kW)
Usually the heat extraction rate for heating is less than the heat rejection for cooling; thus,
the GCHP should be designed to accommodate the larger value.

7. HEAT PUMP UNITS (KAVANAUGH, 1991)

Currently in the U.S., the water-source heat pump unit most widely used is the packaged
water-to-air system. Split systems, air-to-water and water-to-water are offered on a more limited
scale. Component variations in water-to-air units occur primarily in the type of expansion device
and water-to-refrigerant coils. Before 1975, most units had copper (or copper-nickel) tube-in-tube
coils with capillary tube expansion devices that are intended for use with groundwater with
temperatures above 13°C. The market today includes units capable of handling inlet solution
temperatures between -4 and 38°C. This is accomplished with better heat exchangers, expansion
devices, and compressors. Two commonly used water-to-refrigerant exchangers are modified
tube-in-shell (water in tube side) and coaxial tube-in-tube (water on inner tube refrigerant side in
the annulus). Both have extended surfaces on the refrigerant side to compensate for the lower film
coefficients. Manufacturers are selecting heat exchangers with lower water-side pressure losses
to minimize pumping requirements.

Figure 14: Typical water-to-air heat pump arrangement

160
 Expansion valves permit a much wider acceptable range of refrigerant evaporation and
condensation temperatures. This device is especially suited to ground-coupled, lake water, and
closed-loop water systems in which temperature fluctuations are experienced.
In hot, humid climates, the addition of a heat exchanger in the high side of the refrigerant
loop for heating domestic water is almost always recommended. This device is typically a
desuperheater that uses compressor heat to generate hot water either in the cooling mode or with
excess heating capacity (available in southern climates) in the heating mode. Units are now
available that have larger heat exchangers and control mechanisms that permit the full condensing
capacity of the refrigerant circuit to be used for heating water.
Figures 14 and 15 show a typical arrangement of a packaged water-to-air heat pump.
The desuperheater water heater and pump are typically available either as part of the package
or as a field-installed option. The pump for the primary water-to-refrigerant coil is usually not part
of the package because its size and type vary significantly.

Figure 15: Water-source heat pump packages

7.1 Unit performance

Not all "water-source" heat pumps are alike. Some are optimized for heating in colder
climates and some are not intended for low water temperatures. Many have quality, well-designed
components; but, some contain "bottom-of-the-line" equipment. The quality of the various units
can not be elaborated upon in this paper. However, the rated performance and resulting
characteristics can be discussed. Table 3 is included to show the wide variation in the
performance of four similarly sized water-to-air heat pumps. Total cooling (TC) and unit energy
efficiency ratios (EER) are given at two different entering water temperatures (EWT). Note that the
EERs are comparable at 29°C (85oF) for the non-scroll compressor models; but, Brand X has a
substantially lower value at 18°C (65oF) EWT. The most recent advances have come with the use
of scroll compressors (Brand C). They have outstanding heating and cooling efficiencies.
However, work remains to be done in order to enhance heating capacities with lower entering
water temperatures.

161
 The differences are even more important in the heating mode. Heating capacity in
3
10 Mbtu/hr (−MJ/hr) and COP are shown for two values of EWT. Note that Brand A has
substantially greater capacity and COP with a 7°C (45°F) EWT than either Brands B or X. While
Brand X performance may be acceptable with groundwater, its performance will be poor with a
ground-coupled or lake water system. Brand C costs more; but, the improvements in cooling
efficiency warrant use in cooling-dominated climates. In heating-dominated climates, Brand A may
be a better choice.
Table 3: Water-to-Air Heat Pump Capacities and Efficiencies 9 gpm, 1200 cfm,
80°/67°F EAT (34 L/min, 34 m3/min, 27°/19°C)
Cooling Heating Water coil head loss
TC TH
(1000 Btu/h) EER* (1000 Btu/h) COP*
EWT (°F) 65 85 65 85 45 65 45 65 (ft H2O)
Brand A 42 37 14.6 11.2 43 54 4.1 4.5 14
Brand B 38 34 14.6 11.5 32 44 3.3 3.9 15
Brand C (scroll) 42 39 18.7 14.8 34 43 4.1 4.7 7.4
Brand X 37 36 13.0 11.3 21 29 2.3 2.9 20
* If it includes compressor and fan power, deduct 6 to 8% for closed-loop system efficiency and 15 to 25%
for open-loop system efficiency

Another important aspect of water-source heat pumps is pressure drop (or head loss)
across the water-to-refrigerant coil. When this value is high, additional or larger pumps will be
required. The recommended 6 to 8 percent efficiency penalty may substantially increase in closed-
loop systems. Brands A and B have marginally high losses and Brand X's is unacceptable. Brand
C has very good head loss.

8. COMMERCIAL (LARGE-SCALE) GROUND-SOURCE HEAT PUMP SYSTEMS

(RAFFERTY AND KNIPE, 1988; RAFFERTY, 1995; KAVANAUGH & RAFFERTY, 1997)
Unitary ground-source heat pump systems for commercial buildings can be installed in a
variety of configurations. The oldest and, until recently, most widely used approach was the
groundwater system. In this design (Fig. 16), groundwater from a well or wells is delivered to a
heat exchanger installed in the heat pump loop. After passing through the heat exchanger (where
it absorbs heat from or delivers heat to the loop), the groundwater is disposed of on the surface or
in an injection well. The use of an injection well is desirable in order to conserve the groundwater
resource.

Circulating Pump

Heat Pump

Heat Exchanger

Production Injection
Well Well

Figure 16: Groundwater heat pump system

162
 A second and increasingly popular design is the ground-coupled heat pump system. In
this approach (Fig. 17), a closed loop of buried piping is connected to the building loop. For most
larger commercial applications, the buried piping is installed in a grid of vertical boreholes 30-90 m
deep. Heat pump loop water is circulated through the buried piping network absorbing heat from
or delivering heat to the soil. The quantity of buried piping varies with climate, soil properties and
building characteristics, but is generally in the range of 13-22 m/kW of system capacity. Borehole
length requirements are almost always dictated by heat rejection (cooling mode) duty for
commercial buildings.

Circulating Pump
Heat Pump

Vertical Bore
Ground Loops

Figure 17: Ground-coupled heat pump system

A third design for ground-source systems in commercial buildings is the "hybrid" system.
This approach (Fig. 18) may also be considered a variation of the ground-coupled design. Due to
the high cost associated with installing a ground loop to meet the peak cooling load, the hybrid
system includes a cooling tower. The use of the tower allows the designer to size the ground loop
for the heating load and use it in combination with the tower to meet the peak cooling load. The
tower preserves some of the energy efficiency of the system, but reduces the capital cost
associated with the ground loop installation.

Circulating Pump

Heat Pump

Vertical Bore
Ground Loops

Closed Circuit
Cooling Tower

Figure 18: Hybrid ground-coupled heat pump system

163
 In addition to the three designs discussed above, ground-source systems can also be
installed using lake water, standing column wells and horizontal ground-coupled approaches. The
first three schemes have wide use and broad potential application.
In the USA, it is common in the ground-source heat pump industry to refer to the costs for
the ground-source portion of the system on cost per ton (3.5 kW) basis. In keeping with this
practice, most cost data presented is expressed in terms of cost per ton. It is important to note,
however, that the cost per ton refers to the actual load imposed on the ground-source portion of
the system. This is not the same as the installed capacity of the equipment. Due to load diversity,
the peak load imposed upon the heat rejection equipment is always less than the total installed
capacity. The load used for cost calculations is frequently referred to by engineers as the block
load.

8.1 Results
Figure 19 presents the results for the 16°C (60°F) groundwater case assuming the use of
a single production/ injection well pair to serve the system. The four curves shown indicate cost (in
$/ton) for four different groundwater well depths: 60, 120, 180 and 240 m (200, 400, 600 and 800
ft). In all cases, the values shown include costs for the production wells, well flow testing,
production well pump, pump variable-speed drive, buried piping for transport of the groundwater to
the building, heat exchanger to isolate the groundwater from the building loop, heat exchanger
controls, injection well, injection well flow testing, and a 15% contingency factor. As indicated, the
depth requirement for the wells has a substantial impact upon the installed cost. In addition, the
unit cost for small systems (50-100 tons [176-352 kW]) is often higher by a factor of 3 compared to
costs for larger systems (300-500 tons [1055-1758 kW]).
16

14

12

1 ton = 3.52kW
10
Cost in $/ton

1$/ton = 0.284$/kW
Hundreds

0
0 1 2 3 4 5
Capacity in Tons
Hundreds

200' 400' 600' 800'

Figure 19: Groundwater system cost (1 well - 60oF [15.5oC] water)

For ground-coupled systems, actual project costs rather than calculations were used.
Costs for these systems are a function of two values: the number of feet of borehole necessary
per ton of heat rejection, and the cost per foot for completing the vertical bore and installing the
piping. To arrive at a cost per ton, a value of US $5 per foot (US $16/m) of bore has been used.
Although some recent projects have been the beneficiary of cost as low as US $3.75 per ft (US
$12/m), and one as low as US $3 per ft (US $10/m), many areas of the country are still reporting
costs of as much as US $15 per ft (US $50/m). Cost estimates for this comparison are based on a
line cost of US $5 per ft and 200 ft/ton = US $1000/ton (US$ 285/kW).
Hybrid systems include both a ground loop and a cooling tower. The ground loop is sized
to meet the heating load and, it along with the tower, is used to meet the cooling heat rejection
load. As a result, hybrid system costs are a combination of ground loop costs and cooling tower
costs. Using the US $5 per ft (US $16/m) value for the hybrid ground loop portion and vendor
quotes for the cooling tower, Fig. 20 shows the cost per ton for the hybrid system based on 16°C

164
(60°F) soil temperature. The four curves shown for the hybrid system reflect costs for different
ratios of heating loop length versus cooling loop length. As indicated, hybrid systems enjoy more
favorable economics as the heating ground loop length decreases as percentage of the cooling
ground loop length requirement. This is because the cost per ton of the cooling tower is less than
the cost per ton of the ground loop.

12

10

8
Cost in $/ton
Hundreds

0
0 1 2 3 4 5
Capacity in Tons
Hundreds

0.3 0.4 0.5 0.6 GC 200/5

Figure 20: Hybrid system cost - 60oF (16oC) soil

Generally, the hybrid system is attractive in situations where ground loop costs per ton
are high, and where the heating loop length requirement is low relative to the cooling loop length
requirement.
16

14

12
Cost in $/ton
Hundreds

10

2
0 1 2 3 4 5
Capacity in Tons
Hundreds

GW: 800 ft/1 wells GW: 600 ft/2 wells Hybrid @ 0.5 ratio

Hybrid @ 0.6 ratio GC: 200 ft/ton @ $5/ft

12

10

8
Cost in $/ton
Hundreds

0
0 1 2 3 4 5
Capacity in Tons
Hundreds

GW: 200 ft wells/1 well GW: 200 ft wells/2 wells Hybrid @ 0.3 ratio

Hybrid @ 0.4 ratio GC: 200 ft/ton @ $5/ft

Figure 21: Ground-source system cost, 60oF (16oC) water or soil:

a) Low case and b) High case

165
 Figure 21 presents a comparison of the three types of systems for 60°F (16°C) soil. The
ground-coupled system cost line is based upon US $5 per ft (US $16/m) and 200 ft per ton (17.3
m/kW) (US $1000 per ton = $285/kW). The two hybrid system curves in Fig. 21a are based upon
loop length ratios (heating ÷ cooling) of 0.30 and 0.40; whereas, Fig. 21b used loop ratio lengths
0.5 and 0.6. The former are the most favourable conditions for hybrid systems. The two
groundwater curves are based upon 200 ft (60 m) wells and one production/ injection well pairs
(upper curve). Again, these are the most favorable conditions calculated for groundwater systems
in this paper. It is clear that, based on these conditions, the ground-water system enjoys
substantial capital cost advantage over the remaining two systems over the entire range of
capacity covered. The loop ratios in Fig. 21b are the least favorable for the hybrid system.
As indicated (Fig. 21b) at system capacities of 100-175 tons (350-610 kW) and above, the
groundwater system has the capital cost advantage over hybrid and ground-coupled systems.
Below this range, the hybrid system is the most attractive. It is only under conditions of less than
100 tons (350 kW) with well depths of 800 ft (240 m), that the groundwater system capital cost
exceeds that of the ground-coupled system.
This discussion addresses only system capital cost. In the process of system selection,
other issues should be considered as well. These would include operating costs such as electricity
for pumps and fans, water treatment costs (tower), and regulatory issues with respect to
groundwater. As a result, system capital cost provides only a portion of the information required
for informed decision making.

9. GROUND-COUPLED HEAT EXCHANGER DESIGN (BRAUD ET AL., 1988)

9.1 Heat Exchanger Configurations

Concentric Pipes. The concentric pipe heat exchanger consists of a closed pipe casing
with an inner pipe for return flow, Fig.22. The pipes are at the axis of an earth cylinder of radius
equal to the radius of thermal influence of the heat effect. Heat transfer between the circulating
fluid, as it flows down the annual space, and the surrounding earth is the useful heat transfer. As
the fluid returns up the inner pipe, it experiences heat gain due to heat flow across the inner pipe
wall, and crossover heat flow is detrimental to the heat exchange process. It can be reduced to
practical low values by proper selection of material for the inner pipe.

Figure 22: Concentric pipe configuration for vertical heat exchanger

U-Bend. The U-bend configuration consists of two pipes side-by-side and connected to
each other with a 180° fitting at the bottom (Fig. 23). Fluid flows down one pipe and returns up the
166
other. As the fluid descends, the temperature difference in the fluid and the earth mass causes
heat flow. Because of temperature differences in the two fluid streams, there is also some
deterious crossover heat flow. In this configuration, there are two walls of pipe material to impede
the crossover heat flow; but, the presence of each pipe interferes with the heat loss to earth of the
other.

Figure 23: Two pipe side-by-side configuration for vertical heat exchanger
Equation for Heat Flow. The equation for rate of heat transfer from the fluid in the heat
exchanger to the earth mass is:
Q
= U × ∆T (1)
L
where:
Q = Rate of heat transfer, Btu/hr (W) for the whole heat exchanger length,
L = Length of the heat exchanger, m (ft),
U = Conductance rate for heat transfer from the circulating fluid to the earth, Btu/hr °F
(temperature difference per ft of length)(W/°C/m) for the operating conditions,
∆T = (T2 + T1)/2 - To, the difference in average fluid temperature in the pipes (T2 + T1)/2,
and the earth temperature T,
T2 = Fluid exit temperature, °F (°C),
T1 = Fluid entry temperature, °F (°C),
T0 = Earth temperature, °F (°C).

As given in many heat transfer texts, the conductance term U for heat flow from fluid in the
heat exchanger to the earth can be estimated with the conductance coefficient for composite
cylinders. The impedance to heat flow is caused by the thermal resistance of the pipe wall and the
soil cylinder around the casing. Fluid surface resistance films are small relative to the other terms
and are encompassed in the two resistance terms. We can express U as:
2π 2π
U= = (2)
Soi resistance + Pipe resistance Rs + R p
The thermal influence of the pipe walls which separate the two fluid streams affects the
temperature change in the fluid passing through the heat exchanger for given operating conditions,
and it is manifest in the magnitude of the U-value as defined first equation. The casing wall

167
resistance Rp can be calculated with sufficient accuracy from handbook values. The earth thermal
resistance term Rs values can be solved for in the second equation. In fact, the purpose of field
testing is to quantity the Rs value under different operating conditions and heat exchanger designs.
Field testing was thus performed for the two heat exchanger configurations in Louisiana.

9.2 Results
Earth Conductance. The rate at which the earth would absorb heat was relatively high at
the beginning of a run and declined as time went on. With on-off injection of heat, the
instantaneous conductance values were always higher than with constant-on rate as shown in Fig.
24, where conductance values for 25%, 50% and 100% on-time are given. Asymptote values are
the best estimates of the conductance (U-value) derived from regression curve fits to the test data,
Fig. 25. The constant-on conductance rate (i.e., 100% value) (continuous heat injection for 48 or
more hours) was found to be U = 4.86 W/°C/m ; for 50% on time, the effective U-value was 7.51
W/°C/m; and for 25% duty cycle, U was 11.88, see Table 4. By use of Equation 2, the earth
resistance term Rs was also calculated, Table 4. With the steel casing, the thermal resistance of
the steel was negligible relative to the surrounding earth resistance term (Rp 0).
Most of the runs were made with SCH 40 PVC inner return pipe. Runs with thin wall SDR
26 inner pipe exhibited less temperature drop of the circulating water. This effect was manifest in
effective conductance values of U = 4.24 W/°C/m in 100% duty cycle. Wall thickness increased
the heat transfer rate by only 5% over SCH 40.
Measured comparable heat exchange rates to earth with a steel casing with U-values
range from 5.19 W/°C/m to 3.46.

Figure 24: Typical runs of heat injection to earth with three-duty cycle

Table 4: Earth Heat Exchange Rate with Steel Casing and SCH 40 PVC Inner Pipe
Percent Run Time 100 50 25
U: W/oC/mt 4.86 7.51 11.88
Earth Resistance Rs oC m/W 1.29 0.889 9.526

168
 Figure 25: Earth absorption rate, Q/L plotted vs. average water-to-earth
temperature difference for 25, 50 and 100% run time (duty cycle)
PVC Pipe Casing. Present day pipe costs dictate that plastic pipe is cheaper pipe than
metal. Besides, PVC is widely used for water wells. Earth heat transfer rates with PVC plastic
pipe can be calculated with second equation, the earth resistance term, Rs, Table 4, and the known
thermal conductivity and wall thickness for PVC pipe. The U-value is given in Table 5.
Table 5: Heat Exchanger Parameters for PVC Casing and Inner Pipe
Percent run time 100 50 25
U: W/°C/m 3.58 4.85 6.37
PVC pipe resistance Rpo C m/W 0.461 0.461 0.461
Thin wall PVC casing and SCH 40 PVC inner pipe are the most cost-effective combination
A residential heat pump using a PVC concentric pipe heat exchanger was monitored with
good agreement to the experimental values. U-values ranged from 2.9 to 5.1 depending on run-
time for the particular test day in June and July, 1981. Heating mode operation during severe cold
weather in January, 1982, caused circulation water to drop to a minimum of 15°C. This value is a
safe temperature and indicates that circulation water does not need antifreeze protection in South
Louisiana.
Conductance values in Table 4 agree with other field tests of vertical earth-coupled heat
exchangers. Bose, et al. (1980) determined the long-term conductance to earth in a 12.7 cm PVC
casing with 3.2 cm SCH 40 PVC inner pipe. The U-value was 2.95.
U-Bend. Twelve runs of heat injection were made with the 81 m (265 ft) polyethylene pipe
U-bend heat exchanger. Values of earth conductance are given in Table 6.
Table 6: Conductance to Earth with Polyethylene U-bend Heat Exchanger
Percent run time 100 50 25
U: W/°C/m 3.46 4.71 11.60
These values are close to those for PVC pipe except for 25% duty cycle. Because the
earth resistance to heat flow is so much greater than the pipe wall resistance, the two designs
(concentric pipe vs. U-bend) provide about the same performance.

169
9.3 Heat Exchanger Size (Length)
Besides the energy parameters for the heat pump, one must know the highest supply
water temperature acceptable for cooling mode and the lowest temperature acceptable for heating
mode. These values along with local earth temperature establish the design temperatures for
sizing an earth-coupled heat exchanger.
Example: Find the heat exchanger length needed for a heat pump with 7,032 W cooling
capacity at 35°C in an area with earth temperature T = 21°C. The heat pump duty cycle is
estimated to be 50% run time during warmest summer days. Manufacturer specifications give a
high temperature limit of 35°C for entering water. The heat pump discharge water will be 5.6°C
warmer than entry. Total heat rejection of the heat pump = 4,747 W per ton or 9,493 W. In the
heating mode, the heat pump has a heat of absorption value of 7,325 W at 7.2°C. Discharge water
will be 3.2°C cooler than entry. Low temperature limit for entering water is 7.2°C. Design with
PVC pipe, concentric pipe configurations.

Solution:
Cooling mode:
1. Find the design water-to-earth temperature difference, ∆T:
35 + (35 + 5.6)
∆T = − 21 = 16.8
2
2. In Table 5.5 read the effective conductance rate for PVC pipe casing with 50% duty cycle,
U = 4.85 W/°C/m)
3. Solve for L in the Equation 1:
9,483 W
L= = 117 m
4.85 W/ °C/m × 16.8°C

Heating mode:
1. Find design water-to-earth temperature difference, ∆T:
(7.2 − 3.2)* 7.2
∆T = 21 − = 15.4°C
2
2. Assume that the heat pump will run 12 h in 24 h in coldest weather. The 50% duty cycle value
for U then applies.
U = 4.85 W/°C/m
3. Solve for L in the Equation 1:
7,325 W
L= = 97 m
4.85 W/ °C/m × 15.6°C
As this example shows, the heat exchanger length needed is the larger value which is
about 117 m of PVC casing for the heat pump in cooling mode.

10. ADDITIONAL INFORMATION ON GROUND-COUPLED INSTALLATIONS

Ground-loop installed cost currently represents a large portion of system cost. The initial
key to simulating market interest in GSHPs is bringing down those costs, according to an Electric
Power Research Institute (EPRI) report, "Ground-Source and Hydronic Heat Pump Market Study."
The report provides detailed information on the residential GSHP market, emphasizing the
requirements for performance improvement and cost reduction.
The size and installed cost of the ground loop depends on the thermal conductivity of
surrounding soil or rock. See Fig. 26, in which thermal resistivity (the inverse of conductivity) is
plotted versus moisture content with soil type as the parameter. Thermal conductivity is a sensitive
function of the type of soil or rock and its moisture content. The ground-loop designer must be
able to identify the soil or rock, choose a design value for the minimum moisture content, and
arrive at a design value for thermal conductivity. Until recently, there was no rational approach to
this task. EPRI research shows that four soil texture classes (sand, silt, clay and loam) may be
used to characterize soils and provides design values of thermal conductivity for each. The range
of values for identified rock types has also been developed. This information and its rationale are
presented in a report, "Soil and Rock Classification According to Thermal Conductivity."

170
 A companion volume, "Soil and Rock Classification for the Design of Ground-Coupled
Heat Pump System: Field Manual," published in cooperation with the National Rural Electric
Cooperative Association (NRECA), Oklahoma State University (OSU), and the International
Ground-Source Heat Pump Association (IGSHPA), describes simple field procedures for
identifying soil and rock types, and provides appropriate thermal property design values. The soil
and rock field manual is scheduled for incorporation in future editions of the authoritative, "Closed-
Loop/Ground-Source Heat Pump Systems Installation Guide," published by NRECA, OSU and
IGSHPA.

Figure 26: Ground loops function more efficiently in wetter soils

where thermal resistivity is lower.

11. GROUT FOR GROUND-COUPLED SYSTEMS

Recent work by Brookhaven National Laboratory (Allan, 1998) studied the thermal
conductivity of various cement-sand grouts used in GCHP. Their research investigated the
suitability of this material for grouting vertical boreholes. The role of mix variables such as
water/cement ratio, sand/cement ratio and super-plasticizers dosage were measured. The work
summarized the thermal conductivity, permeability and bond strength results for selected grout
mixes. Addition of sand and use of super-plasticizers to reduce water/cement ratio improved grout
performance compared with neat cements. Field trials have been undertaken.

REFERENCES

1. Allan, M. (1998). “Cementitious Grouts for Geothermal Heat Pump Systems,” Transactions,
Vol. 22, Geothermal Resources Council, Davis, CA, pp. 419-423.
2. Bose, J. E. et al. (1980). Earth-Coupled and Solar-Assisted Heat Pump Systems. 5th Annual
Heat Pump Technology Conference, Oklahoma State University, Stillwater, OK.
3. Braud, H. J.; Oliver, J. and H. Klimkowski (1988). Earth-Source Heat Exchanger for Heat
Pump. Geo-Heat Center Quarterly Bulletin, 5-1, 12-15.
4. Carrier Corp. (1981). Application Data - The Heat Machine. Carrier Corporation, Syracuse,
NY.
5. Carrier Corp. (1987). Packaged Hermetic Reciprocating Liquid Chillers. Carrier Corporation,
Syracuse, NY.
6. Dexheimer, R. D. (1985). Water-Source Heat Pump Handbook. National Water Well
Association, Worthington, OH.

171
7. Ellis, D. (1989). Personal communication. President of the International Ground-Source Heat
Pump Association, Stillwater, OK, and Vice President of Marketing, Water Furnace
International, Inc., Ft. Wayne, IN.
8. Kavanaugh, S. (undated). Ground-Coupling with Water-Source Heat Pumps. University of
Alabama, Tuscaloosa, AL, 14 pp.
9. Kavanaugh, S. (1991). Ground and Water-Source Heat Pump - A Manual for the Design and
Installation of Ground-Coupled, Groundwater and Lake Water Heating and Cooling Systems in
Southern Climates. The University of Alabama, Tuscaloosa, AL, 163 pp.
10. Kavanaugh, S. and K. Rafferty (1997). Ground-Source Heat Pumps - Design of Geothermal
Systems for Commercial and Institutional Buildings. American Society of Heating,
Refrigeration and Air-Conditioning Engineers, Inc., Atlanta, CA, 167 p.
11. Lienau, P. J.; Boyd, T. L. and R. L. Rogers (1995). Ground-Source Heat Pump Case Studies
and Utility Programs. Geo-Heat Center, Klamath Falls, OR, 82 pp.
12. Lienau, P. J. and J. W. Lund (1992). Geothermal Direct Use. Testimony presented at the
House Subcommittee on Environment, July 30. Geo-Heat Center, Klamath Falls, OR.
13. Lund, J. W. (1988). Geothermal Heat Pump Utilization in the United States. Geo-Heat
Center Quarterly Bulletin, 11-1, 507.
14. Lund, J. W. (1989). Geothermal Heat Pumps - Trends and Comparisons. Geo-Heat Center
Quarterly Bulletin, 12-1, 1-6.
15. McQuay (1983). Heat Recovery Water Heaters, Catalog 1210. McQuay Inc., Minneapolis,
MN.
16. McQuay (1986). Templifier Reciprocating Water Chillers, Catalog 1200-2. McQuay Inc.,
Minneapolis, MN.
17. Oklahoma State University, Division of Engineering Technology (1988). Closed-
Loop/Ground-Source Heat Pump Systems - Installation Guide. International Ground-Source
Heat Pump Association, Stillwater, OK.
18. Pinckley, M. E. (1995). Galt House East Hotel and Waterfront Office Building. Proceedings
of the World Geothermal Congress 1995, Florence, Italy, 3, 2277-2279.
19. Rafferty, K. (1995). A Capital Cost Comparison of Commercial Ground-Source Heat Pump
Systems, Geo-Heat Center Quarterly Bulletin, 16-2, 7-10.
20. Rafferty, K. and E. Knipe (1988). Some Considerations for Large Water-Source Heat Pumps.
Geo-Heat Center Quarterly Bulletin, 11-1.
21. Rybach, L. and B. Sanner (1999). “Ground-Source Heat Pump Systems - The European
Experience,” Proceedings of the International Geothermal Days - Oregon 1999, Geo-Heat
Center, Klamath Falls, OR, pp.159-170.

172