A series of tests were conducted in seven retail shops in a strip mall in Cocoa, Florida to examine how roof whitening would impact air conditioning load.  The roof of the strip mall was conventional: un-surfaced galvanized corrugated metal with R-11 insulation suspended by purlins underneath.  Below the insulation was the roof plenum which contained both the air handler and duct air distribution system.  Cooling was accomplished in each shop by a dedicated direct expansion air conditioning (AC) system between 3 and 4 tons in size.  Each had its air conditioning metered for half of the summer in a baseline condition.  Building temperatures as well as meteorological conditions were also obtained on a 15-minute basis.  The roof was then resurfaced white at mid summer using a commercially available roof coating product.  The measured roof surface reflectivity was altered from approximately 29% to 75%.

The tests were phased over a two summer period so that the impact of surface degradation could be measured in the second year of exposure.  The results for both phases of the project, showed a 25.3% average reduction (8.6 kWh) in summer space cooling energy (34.1 kWh/day to 25.5 kWh/day) in the seven shops with a range of savings of 13 – 48%. The percentage savings varied with the temperature maintained in the shops; those maintaining the lowest interior temperatures saved the least on a percentage basis, although the absolute space cooling energy reductions were more similar ranging from 6.4 to 13.4 kWh/day.  Total annual air conditioning in the seven monitored shops averaged 6,780 kWh; estimated savings averaged 1,670 kWh.  Impacts on space heating were not measured.

The space cooling energy demand reduction was concentrated during the summer afternoon and early evening hours between 11 AM and 7 PM as seen in Figure E-1.  During the utility coincident peak demand period (defined as 4 – 5 PM EST or 5 – 6 PM EDST) the overall electric demand reduction averaged 592 W (from 1469 to 877 W) or 40.3%.

We also used coincidental peak load factor (CPLF) analysis to examine the impact of the reduction on load shape.  Using the defined peak period, the coincidental peak load factor of space cooling in the aggregate sample of buildings was altered from 0.97 in the baseline configuration to 1.21 after the roof was made white.  This analysis indicates that the change to a white roof had a favorable impact on the building cooling load shape – the reduction during the utility peak demand hour was greater than the average reduction over the 24-hour cycle.

Finally, the application of the reflective roofing material was found to have a beneficial impact on the ability of the cooling systems in each of the monitored store fronts to maintain interior comfort during peak summer load conditions.  Most of the shop owners mentioned that after the application they found the interior much cooler when opening for business and two increased the interior thermostat setting to compensate.  Measured interior temperatures showed both less variation during the day as well as lower values in several shops during the late afternoon hours.

The estimated annual savings averaged $150/year per store front against a cost of application of $1,375.  However, the economics for an operating utility program would likely be much more favorable in targeting commercial facilities which are in the process of installing new roofing system or re-roofing.  In either case, the incremental costs for using white roofing materials can be negligible for white standing seam metal, white single ply membrane or white coatings over modified bitumen roofs.