b'T&R UPDATETHE REAL DIRT ON REFLECTIVE ROOFSfor all 90 roofs are displayed. The information in Figure 2 should generally impress, in that the current in-service inventory of roofs is returning respectable numbers. The average reflectivity of all roofs and all readings for this study was 63.6%. Consider that from the 1980s through 2010 the dominant roof membrane in the United States was black EPDM; which will return a solar reflectivity between 7-10%. Figure 2 Plot of sample means of fractional solar reflectivityclimate zone, we arrive atTable 2. for all roof membranes in this study, against their reported age.Table 2. A break down of Membrane Class versus Climate ZoneThe CRRC-S100 test is in the authors opinion an idealized aging test. The membrane samples are aged outdoors, but under ideal conditions. These samples are not challenged by numerous deleterious conditions found on any common in-service roof. These deleterious conditions include, but are not limited to:1.Climate zone 2.Drainage ImpedimentFigure 2 represents the measured sample mean reflectivity3.Roof Slopefor each roof sampled. All but one membrane in this study4.Predominant Wind Currentshas a CRRC three-year-aged solar reflectivity recorded in their publicly accessible database. If we take the difference5.Aerodynamics of Buildingbetween the three-year-aged value recorded in the CRRC6.Building Occupancy database and the sample mean, from this study, for each of7.Local Pollutants these membranes, we arrive at Figure 3. 8.Membrane Surface Condition9.Biological Growth10. Place of Manufacture11. Insulation Amount12. Air Handling Unit Exhaust13. Overhanging Foliage14. Dust StormsAny or all of these conditions may be present on a roof. The Figure 3 Plot of normalized fractional difference betweeneffect of climate zones is apparent in Table 2.The buildup roof sample mean and CRRC database reported three-year- of dirt and organic matter in poorly drained roof areas, aged value. Drainage Impediment, can easily be seen from satellite The pattern produced by the data in Figure 3 isimagery as easily as on the roof level. The impact of Roof enlightening. The most basic observation that should beSlope has been investigated by the NRCA. The Place of made is that the solar reflectivity value returned by theManufacture was seen on several roofs in this study. In one CRRC-S100 test method appears to not to correctly predictcase the same roof membrane, from the same manufacturer, actual in-service roof reflectivity. If it did, trend line wouldbut manufactured at two separate plants produced a at least pass through 0 as a differential at three years.Itdifferential of 0.04 between two adjacent membrane sheets does not. Statistical measures of this data confirms theon the same roof, installed at the same time. Of the most visual observation of Figure 3, that the CRRC-S100 roofinteresting issues was the Dust Storms that local contractors membrane aging methodology does not match in-servicein Phoenix, AZ referred to as a Haboob. The dust brought roofs.to the roof by the Haboob was evident on the PVC roofs If we break the data down further, say into membrane versusand generally not present on the TPO roofs, in Phoenix. 34 www.mrca.orgMidwest Roofer'