Since 2000 research has been conducted at the University of Kansas (KU) to evaluate the thermal performance of building walls fitted with phase change materials (PCMs). The purpose of the investigations was to assess peak air conditioning demand reductions, thermal load shifting and energy savings. PCMs work by storing relatively large amounts of heat energy when melting. This heat is released upon solidification of the PCM when the temperature surrounding the PCM drops to below the PCM solidification point.
For building applications PCM phase changes are predominantly of solid-liquid transitions. The PCM can be organic, for example paraffins, waxes and oils, or inorganic, for example, hydrated salts.
Both have advantages and disadvantages. For example, organic-based PCMs are flammable and comparatively expensive, while inorganic-based PCMs are non-flammable and relatively inexpensive. However, these are also corrosive and prone to supercooling. There are also PCMs that use mixtures of organic and inorganic compounds and some that are contained within hydrophilic silica powders. The PCMs used in the KU research are shown in Figure 2.
In the present research PCMs were integrated into two common types of walls used in North American residential construction. These are frame walls and structural insulated panel (SIP) walls. These types of systems are shown in Figure 3.
The thermal performances of walls fitted with PCMs were evaluated using two identical 1.83m × 1.83m × 1.22m (6ft × 6ft × 4ft) test houses as shown in Figure 4, where one house was used as a control house and the other as an experimental house.
The roof was a built-up roof with gray asphalt shingles, 6.8kg (15lb felt, and 1.27cm (1/2 inch) plywood sheathing. The wall assemblies were 1.11cm (7/16 inch) plywood siding, 5.08cm × 10.16cm (2 inch × 4 inch) studs and 1.27cm (1/2 inch) plywood board or gypsum wallboard.
Insulation (fibreglass and cellulose) with a resistance of 1.94m2K/W (R-11) was used for both the ceiling and the walls. In each test house a window with an area of 0.32m2 (3.4ft2) was placed in the south-facing walls.
For the cooling system a chilled water system was developed, which is depicted in Figure 5. The chilled water system included a water tank, a drop-in coil water chiller, a temperature controller and a set of water pumps. The temperature controller was connected to the chiller to regulate the chilled water temperature in the tank, which was set at around 12.8°C ± 2.8°C (55°F ± 5°F).
The chilled water was circulated from a 265L (70 gallon) insulated plastic tank to a fan-coil-unit (FCU) located inside each house. The pumps and the electromagnetic valves were controlled by low voltage thermostats to maintain the test houses' indoor air temperatures at approximately 21.5°C ± 0.5°C (70.7°F ± 0.9°F). Fan coil units were installed inside each house next to the east-facing walls. Monitoring systems were installed to measure and collect space cooling loads, wall heat fluxes, air and surface temperatures and relative humidity.
During the tests, the indoor air temperatures of both houses were well controlled and maintained almost identical to less than 1.5°C difference. The PCMs were encapsulated either in copper pipes, arranged horizontally in the stud walls and placed next to the interior plywood wallboard or in reflective foil sheets located in various places within the insulation cavity of the wall. In other cases, not reported in this article, the PCMs were mixed directly with the cellulose insulation.
PCM concentrations of 10%, 15% and 20% were investigated. The concentrations were based on the mass of the interior sheathing. This was done to be able to compare these results to previous results from other researchers in which the wallboards had been saturated with PCM.
Type T thermocouples (T/C) were installed to measure indoor and outdoor air and wall surface temperatures. For air temperature measurements, the
T/Cs were shielded with aluminum tape to minimise radiation exchange effects.
For surface temperatures the T/Cs were covered and painted with a thin film of the same colour and texture as the surface, the temperature of which it measured. Each wall was instrumented with several T/Cs arranged in parallel grids as shown in Figures 6 and 7. This arrangement gave a representative wall temperature, which was the average of the measured points.
The accuracy of the T/C was within ±0.6°C (±1°F) of the true value of the measurements. Flat Thermal Flux Meters (TFMs) were attached to the interior wall surfaces to measure heat fluxes across the walls (Figure 7). The accuracy of these sensors was 1% in departure of reading over the repeatable range of the sensor.
Relative humidity (RH) was measured with relative humidity transducers and a tripod weather station was installed, which had a wind speed sensor, a pyranometer and temperature and relative humidity probes. Year-round outdoor weather conditions were monitored and measured.
Figures 8 - 11 show the schematics of the PCM configurations used in the KU research.
It was necessary to perform calibration tests before any retrofit. For this, the thermal performances of the two houses were compared and recorded as a baseline. Indoor air temperatures, wall temperatures and heat fluxes were measured and compared to verify their similarity in thermal performance. This is shown in Figures 12, 13 and 14.
During the calibration period, the indoor air temperatures were controlled to a high precision of 0.05°C (0.1°F) difference between both test houses (Figure 12). The control house was kept at an average indoor air temperature of 24.17°C (75.5°F), while the soon-to-be-retrofit house was kept at an average temperature of 24.22°C (75.6°F).
Figure 13 shows the similarity in temperature of the outside surface temperatures of the south facing walls. Figure 14 shows the heat transfer rate across the south facing walls. The average difference in heat transfer rate in the walls of both houses was in the range of 3%.
Results and discussion
Figure 15 shows the comparison in the heat transfer rates between a standard frame wall and a wall in which PCM had been added via pipe encapsulation in the configuration of Figure 8.
The results of Figure 15 are for the summer. These results showed that the average peak heat flux through the walls which were outfitted with PCMs decreased by as much as 41.7% when 20% PCM was applied. The PCM used was a commercially available highly crystalline n-paraffin-based-PCM with a melting point of 26°C (78ºF).
Structural Insulated Panels (SIPs) are composite walls made from three layers, two of which are oriented strand board (OSB), which sandwich a layer of expanded polystyrene (EPS). These panels are predicted to gain popularity in the North American construction market because of their energy-efficiency and ease of handling during construction. This stage of the KU research was to assess the thermal performance PCM-enhanced SIPs.
Experiments were carried out using the same two test houses. A similar highly-crystalline n-paraffin was used. Summer results showed that when a concentration of 10% was applied, the peak heat fluxes across the walls decreased by an average of 37% in a south-facing wall. When a concentration of 20% PCM was applied, the peak heat fluxes decreased by an average of 62%, also for south-facing walls. This is shown in Figures 16 and 17.
Later experiments explored the heat transfer rate across walls fitted with hydrated salt PCMs. The PCM used was calcium hexahydrate with a melting point of 29°C (82.4°F). The peak heat transfer rates during the calibration tests were also nearly identical, less than 3% different on average.
The difference in peak heat transfer rates between the control walls and the walls with PCMs at 10% PCM concentration was approximately 27%. The average reductions in peak heat transfer rates as a result of using PCMs in frame walls in the north, south, east, and west walls were 33.7%, 25.6%, 24.3% and 24.6%, respectively. Figures 18 - 21 show the comparisons. In terms of cooling load shifting to off-peaks times, it was found that the shift was about one hour.
Figure 22 depicts how the walls outfitted with PCMs were able to keep a more constant inside wall surface temperature and a narrower temperature swing than the standard wall. The segment in the left of each wall in the figure represents the data of the calibration period, while the segment in the right represents the data of retrofit period at 10% PCM concentration. Each segment shows indoor surface temperatures for a standard wall and for a wall outfitted with PCMs.
For example, for the north walls the indoor surface temperature of the control house was on average 23.6°C (74.5°F) while the surface temperature of the wall
outfitted with PCMs was 22.5°C (72.5°F). The temperature swing in the standard wall was 2.22°C (4.0°F) while it was 1.06°C (1.9°F) for the wall outfitted with PCMs.
Table 1 summarises the findings related to the reductions in inside wall surface temperatures and in the daily temperature swings. As stated above, it was observed that the walls with PCMs were able to not only lower the inside wall surface temperature of the walls, but also their daily temperature fluctuations. The average reduction of inside wall surface temperature and daily temperature fluctuations were 1.44°C (2.6°F) and 1.44°C (2.6°F), respectively. These results could translate to human comfort and to the increase in the life of comfort equipment.
The same analyses were performed for a PCM concentration of 20% PCM. The difference in peak heat transfer rates and indoor surface temperatures between the control walls and the walls outfitted with PCMs were nearly the same as the values for the 10% PCM concentration. This means that doubling the amount of PCM did not produce significant improvement. Figures 23 - 26 show the heat transfer rates across the north, south, east and west walls.
The average reduction in peak heat transfer rates when using PCMs in the north, south, east and west walls were 27.1%, 29.2%, 25.7%, and 27.2%, respectively. From these results and aside from the north-facing wall, it was seen that doubling the quantity of PCM improved the performance by
3.6%, 1.4 %, and 2.6 % for the south, east and west walls respectively.
Table 2 summarises the indoor surface temperatures in the control and the PCM-outfitted walls for a PCM concentration of 20%. The average indoor surface
temperature of the four control walls was 23.9°C (75.0°F) while the indoor surface temperatures in the walls fitted with PCMs was 22.4°C (72.3°F).
The average temperature swing in the control walls was 3.17°C (5.7°F) while it was 1.72°C (3.1°F) in the walls outfitted with PCMs. The surface temperature of the walls outfitted with PCMs was also more constant than those for the control walls. This is shown in Figure 27.
Later research investigated the influence of using paraffin-based PCMs encapsulated in reflective aluminium sheets (PCM Shield). This configuration proved to be not only the most practical, but also the one that yielded the largest reduction in heat transfer rates. A sample of this data is shown in Figure 28. This research also showed that an optimal location for the PCM shield exists.
|Average surface temperature in °C (°F)||Difference in °C (°F)||Average daily temperature swing in °C (°F)||Difference in °C (°F)|
|North||23.6 (74.5)||22.5 (72.5)||11.1 (2.0)||2.22 (4.0)||1.06 (1.9)||1.16 (2.1)|
|South||24.1 (75.3)||22.8 (73.0)||1.28 (2.3)||3.72 (6.7)||1.89 (3.4)||1.83 (3.3)|
|East||24.4 (76.0)||21.8 (71.2)||2.67 (4.8)||4.22 (7.6)||1.44 (2.6)||2.78 (5.0)|
|West||23.9 (75.0)||22.9 (73.3)||0.72 (1.3)||3.44 (6.2)||3.39 (6.1)||0.05 (0.1)|
|Average||24.0 (75.2)||22.6 (72.6)||1.44 (2.6)||3.39 (6.1)||1.94 (3.5)||1.44 (2.6)|
Table 1: Reductions in inside wall surface temperatures and reductions in temperature fluctuations from using PCMs at a 10% PCM concentration.
|Average surface temperature in °C (°F)||Difference in °C (°F)||Average daily temperature swing in °C (°F)||Difference in °C (°F)|
|North||23.6 (74.5)||22.3 (72.2)||1.28 (2.3)||2.17 (3.9)||0.83 (1.5)||1.33 (2.4)|
|South||24.2 (75.5)||22.7 (72.8)||1.50 (2.7)||4.17 (7.5)||2.06 (3.7)||2.11 (3.8)|
|East||24.1 (75.3)||21.7 (71.1)||2.33 (4.2)||3.28 (5.9)||1.06 (1.9)||2.22 (4.0)|
|West||23.8 (74.8)||22.9 (73.2)||0.89 (1.6)||3.00 (5.4)||2.94 (5.3)||0.06 (0.1)|
|Average||23.9 (75.0)||22.4 (72.3)||1.50 (2.7)||3.17 (5.7)||1.72 (3.1)||1.44 (2.6)|
Table 2: Reductions in inside wall surface temperatures and reductions in temperature fluctuations from using PCMs at a 10% PCM concentration.
A computer model used to predict heat transfer rates across residential walls fitted with PCMs was developed. The model was a transient heat transfer with phase change model. It was verified against experimental data. A verified model would translate the experimental results to full-scale buildings in any location. Simulated results versus experimental data of heat transfer rates are shown in Figures 29 - 31.
The cumulative difference between the model and experimental data was about 13% and was 2.1% for peak times.
Heat transfer rates through several kinds of wall arrangements were investigated in well-controlled test houses under full weather conditions. The arrangements included frame walls and structural insulated panels.
For the frame walls, PCMs were encapsulated in pipes and in reflective aluminum sheets. Two kinds of PCMs were used: paraffin-based and hydrated salt-based PCMs. In the structural insulated panels, only the pipe-encapsulated PCMs were tested.
It was found that PCMs produce reductions in heat transfer rate, both in total heat transfer and peak heat transfer. The reductions in heat transfer rates varied depending on the wall type, PCM type, PCM amount and encapsulation method. For the frame walls, the PCM encapsulated within reflective foil sheets yielded the highest reductions of 52.4% (peak) and 35.6% (total) for a PCM concentration of about 15%.
Shield location was found to be relevant. For the structural insulated panels, a paraffin based PCM produced a peak reduction of 37% at 10% PCM concentration and a peak reduction of 62% at 20% PCM concentration. PCMs produced more stable overall wall temperatures.