promoted by


Performance has been documented during a heating
season for a full-scale prototype of a novel heat pipeaugmented solar heating system that significantly
improves heating performance relative to other
conventional passive solar heating systems. The system
utilizes heat pipes to provide effective one-way heat
transfer into storage, and incorporates insulation that
reduces heat losses from the room compared to
conventional systems. Simulations performed for four US
locations representing a range of winter temperatures and
available insolation exhibited higher performance for the
heat pipe system than for common direct gain and
concrete and water wall indirect gain systems, particularly
in cold climates with low insolation. A small-scale
laboratory model was constructed and tested under
controlled conditions to confirm system component
performance and to test a range of component variations.
Measured system efficiency was 85.1% for essentially
normal incidence. A full-scale prototype was constructed,
installed and instrumented. The vertical glazing of the
prototype was faced 10.5 degrees east of south, contained
aluminum absorbers, and received partial shading from an
overhang from March to October and in the mornings
from a short offset in the south wall. Results from the
recent heating season (January, 2010 – March, 2010)
showed an average daily peak efficiency of 61.4%, with a
maximum daily peak efficiency during the heating season
of 83.7%. The system thermal mass proved to be
appropriately sized, with minimal reduction of average.

hourly room gains between periods of insolation and
periods of no insolation. The average hourly room gain
during periods of insolation was 90 W/m
, and the
calculated average hourly room gain during periods of no
insolation was 77 W/m
; resulting in only a 14% reduction
in room gains.
All three common types of contemporary passive solar space
heating systems (direct, indirect and isolated gain) present
limitations related to either thermal performance or
configuration requirements. Direct gain systems, which use
south-facing apertures to allow solar radiation directly into
the room, have significant heat losses back through the same
aperture. Also, because the thermal mass is located within
the room, a larger thermal mass is required relative to the
other systems to moderate room temperature swings.
Indirect gain systems, in which solar radiation strikes a
thermal storage material between the exterior of the building
and the room, reduce temperature swings within the room,
but have similarly large heat losses. In contrast, isolated
gain systems, which utilize natural convection to transfer
heat from a solar collector to the room or to a thermal
storage unit, need not increase losses from the building,
since natural convection automatically stops when the
collector is colder than the building. Natural convection,
however, typically requires that the solar collector be located
below the room, which limits the application of isolated gain
systems to buildings for which this configuration is practical.
More info at CCRES http://solarserdar.blogspot.com.
By incorporating heat pipes (Fig. 1) into an isolated gain
system, phase change provides effective heat transfer with
minimal elevation difference required between collector and
storage (1). In the heat pipe augmented system, solar
radiation is transmitted through a transparent cover to an
absorber. The absorber, which can conveniently incorporate
a selective surface to allow more efficient gains of solar
energy compared to typical direct and indirect gain systems,
heats up and conducts that energy to the evaporator section
of the heat pipe. A phase change fluid boils in the evaporator
and the vapor rises through a sloped adiabatic section that
runs through the insulated wall. The vapor then enters the
slightly higher condenser section of the heat pipe where it
condenses; transferring the energy through the wall of the
pipe to the thermal mass, which in turn transfers energy to
the room via free convection and radiation. The condensed
liquid then falls by gravity back to the evaporator end of the
heat pipe, completing the cycle. The small elevation
difference necessary between the evaporator and condenser
is easily incorporated within the south wall of the building.
When the absorber is not receiving solar energy, the fluid
remains liquid in the evaporator and essentially no heat
transfer takes place along the heat pipe. The primary path
for heat loss is thus through the insulation, which can be
made as thick as necessary to minimize such losses. The heat
pipe system has losses during nighttime and cloudy days that
are an order of magnitude smaller than those of conventional
direct gain and indirect gain systems. The heat pipe system
provides the high efficiency of isolated gain systems, without
requiring that the collectors be located below the building.
Fig. 1: Scematic of a heat pipe.
More info at CCRES http://solarserdar.blogspot.com.

This paper reports the results of computer simulations and
small-scale and full-scale experiments conducted to assess
the performance of heat pipe augmented passive solar
systems, and further optimize the design and performance of
the system.

Using MATLAB software, a set of programmed thermal
networks were used to simulate the performance of several
conventional passive solar heating systems, including
direct gain, concrete wall and water wall indirect gain,
and that of the heat pipe system (2). For typical regional
weather conditions, including temperature, wind speed
and insolation, TMY3 weather data was used. To
investigate sensitivity of heat pipe system performance,
computer simulations were run for ranges of system
parameters, including glazing characteristics, selective
surfaces, absorber thicknesses, insulation properties, and
the number and material of heat pipes.

Simulations were performed for Louisville, KY,
Albuquerque, NM, Rock Springs, WY and Madison, WI to
represent a range of winter temperatures and available
insolation. The baseline conventional passive solar
systems and load requirement for the comparison was
constant among climates tested. A load to collector ratio of
10 W/m
K was chosen to produce a solar fraction of
approximately 50% for the heat pipe system in Louisville.
An experimental model was created to test the overall
efficiencies and conductance of the heat pipe passive solar
system under controlled conditions (Fig. 2). The model
was built using a single copper heat pipe, the evaporator
section of which was soldered to a copper absorber plate.
The absorber was plated with black chrome. An aluminum
frame was constructed to hold a low-iron glass cover and
the absorber plate was attached to an insulating wall.
Behind the insulated wall, 189 liters (50 gallons) of water
in a plastic tank served as a thermal storage device. The
water tank was insulated to limit thermal losses to the
room, so that energy gains could be readily calculated.
Insolation of about 700 W/m
was simulated with three 1
kW metal halide bulbs, which approximate the spectrum
of the sun. Four pyranometers were used to calibrate the
lighting system and to monitor its output during the
experiments. SUVA 120 was used as the phase change
fluid within the heat pipe.
solar simulator
evaporator end
of heat pipe
absorber plate
adiabatic section
of heat pipe
condenser end
of heat pipe
storage tank
cover glass
Fig. 2: Schematic of bench-scale experiment.

Twenty-eight T-type thermocouples were used to monitor
absorber, heat pipe, water and ambient temperatures. The
rate of increase of tank temperature was used to determine
the power output of the system, while the pyranometers
determined power input. System performance was
calculated for variations in heat pipe fill fraction and
insulation level on the adiabatic section.
A full-scale prototype was constructed and installed on the
south-facing wall of a classroom on the university’s
campus (3). The system used a low-iron glass cover, black
chrome plated aluminum absorbers, and five copper heat
pipes and water tanks. The vertical glazing of the
prototype was faced 10.5 degrees east of south, contained
aluminum absorbers, and received partial shading from an
overhang from March to October and in the mornings
from a short offset in the south wall. The system was
instrumented with a pyranometer and 31 thermocouples,
eight each on the absorber, heat pipe and water tank of the
central heat pipe unit, five in the remaining four water
tanks, one for ambient temperature and one in the room.
Thermal efficiency, η, for the prototype was calculated for
each hourly period of available insolation for each day of
the heating season as
η ===
qs +++ qr
where IT is the incident solar radiation received by the
system per unit of collector area (measured by the
pyranometers) and qs
is the rate of energy gain in the
thermal mass
s ===

where m is the mass of the thermal storage per unit of
collector area, Cp is the specific heat of the thermal
storage, and t is the time over which the temperature
increase T is measured. The heat delivered to the room
from storage qr
was found using
qr ===
Ts −−− Tr
where Ts
is the storage temperature, Tr
is the room
temperature, and Rtotal
is the total thermal resistance from
the thermal mass to the room air (Fig. 3). (qr
for the wellinsulated tank in the bench-scale system was negligibly
Fig. 3: Thermal resistance network from the thermal mass
to the room.
Computer simulation results showed that the direct gain
system performed well in cool and sunny Albuquerque, but
produced a net loss in cold and cloudy Madison (Fig. 4).
The indirect gain systems performed better than direct
gain in all locations except Albuquerque, where direct
gain and water wall performances were nearly identical
and both performed better than the concrete wall system.
The water wall system provided greater gains than the
concrete wall in all climates. The heat pipe system
performed significantly better than all other systems in all
climates. In Louisville, for instance, the solar fractions
were 20.3%, 29.0%, 37.4% and 50.6% for direct gain,
concrete wall indirect gain, water wall indirect gain and
heat pipe systems, respectively. These performance values
were quite similar to those in Rock Springs, which is
sunnier but colder, and considerably better than Madison,
which is colder but only slightly cloudier.
More info at CCRES http://solarserdar.blogspot.com.
Fig. 4: Performance comparison of passive heating
For the bench-scale experiment, performance was best
with a fill of 120% of the evaporator volume and with an
insulated adiabatic section. The addition of external
condenser fins did not significantly improve performance.
System efficiency was 85.1 ± 0.72%.
For the full-scale prototype, the maximum daily peak
efficiency (using Equation 1) was 83.7%, and the average
daily peak efficiency was 61.4%. An alternate approach in
assessing system performance was also investigated. This
approach involved plotting the total solar input to the
system vs. the useful gains for every hourly period for
every day of analysis during the heating season for a
specified daily time frame of 9am – 5pm (Fig. 5).
Fig. 5: Heating season plot of average hourly solar input
vs. average hourly useful gains.
The trendline shown in Fig. 5, with a quite satisfactory
correlation coefficient of 0.927, represents the typical
performance of the prototype relating solar input, IT, to
useful gains per unit collector area, qu, as
u === 0.64I
T −−− 52.3 (4)
Equation 4 suggests that useful gains won’t occur until
solar input on the prototype is equivalent to 197 W, or 83
of collector area. The collector area may influence.

the threshold solar flux at which useful gains commence,
by varying thermal losses out the edges of the collector.
(The collector area of the prototype was 2.41 m
) The
slope of Equation 4 represents an asymptotic prototype
thermal efficiency for high IT of 64%, which is comparable
to the calculated system average peak efficiency of 61.4%.
The correlation of the curve fit of Equation 4 without
dependence of the coefficients on system temperatures is
remarkable. One might expect the intercept of Equation 4
to depend on, for instance, absorber temperature, storage
temperature and/or ambient temperature. Further analysis
of the data is needed to determine the sensitivity of the
correlation to these parameters.
Results for the full-scale prototype also illustrate the
effectiveness of the thermal mass in continuing to deliver
heat to the room during periods of no insolation. The
water-filled tanks both store thermal energy and deliver
energy passively to the room during periods of insolation
above the threshold for useful gains, reaching their
maximum temperature near sunset. During periods of
insolation below the threshold for useful gains (nighttime
and cloudy days), the tanks release stored energy to the
room. Calculated over the entirety of the heating season,
the average hourly rate of energy delivered from storage to
the room during periods of insolation was 90 W/m
, while
the average hourly delivery during periods of no insolation
was 77 W/m
This small 14% reduction demonstrates the .
effectiveness of the thermal mass in moderating day and
nighttime heating from the system.
Computer simulations showed that the heat pipe system
performed better than all other conventional passive space
heating system for all locations investigated. A system
efficiency of 85.1% was achieved with heat pipe
augmented bench-scale experiments. A full-scale
prototype of the heat pipe system achieved a peak thermal
efficiency of 83.7% and an average daily peak efficiency
of 61.4% during the heating season. Thermal mass sizing
for the prototype proved to be adequate and effective, with
a reduction in average hourly room gains of only 14%
during periods of no insolation versus periods of

(1) Susheela N & Sharp MK, “A heat pipe augmented
passive solar system for heating of buildings,” J
Energy Eng 127:1:18-36, 2001.

(2) Albanese M, Development of an integrated solar heat
pipe system for improving building energy efficiency.
MEng thesis, Department of Mechanical
Engineering, University of Louisville, June 2009.
(3) Chmielewski N, Design, construction, and
experimentation of a heat pipe augmented solar wall.
MEng thesis, Department of Mechanical
Engineering, University of Louisville, June 2009.

Brian S. Robinson, M. Keith Sharp
Renewable Energy Applications Laboratory
University of Louisville
200 Sackett Hall
Louisville, KY 40292
More info at CCRES http://solarserdar.blogspot.com.