Article
Evaluation of the solar reflectance of highly reflective roofing sheets installed on building roofs
Journal of Building Physics 37(2) 170–184 Ó The Author(s) 2012 Reprints and permissions: sagepub.co.uk/journalsPermissions.nav DOI: 10.1177/1744259112459263 jen.sagepub.com
Jihui Yuan1, Kazuo Emura2 and Hideki Sakai3
Abstract There is a strong need to prevent the heat island effect. One of the important countermeasures to the heat island effect is to reduce thermal storage in buildings. The highly reflective material on outer walls and roofs is a significant way to prevent heat from penetrating indoors; such materials have been installed on the surface of buildings in Osaka, Japan. To evaluate the solar reflectance of such highly reflective material, we measured the solar reflectance of reflective roofing sheets installed on the building roofs in Osaka, Japan, over a span of time from February 2010 until September 2011, and examined the change in solar reflectance during this time. Keywords Heat island, building roofs, highly reflective roofing sheets, solar reflectance
Introduction Background Heat island effect has been documented (Santamouris, 2001). Important research has been carried out to document its influence on the urban climate (Akbari et al.,
1
Doctor Course, Graduate School of Life and Science, Osaka City University, Sugimoto, Sumiyoshi-ku, Osaka, Japan
2
Professor, Graduate School of Life and Science, Osaka City University, Sugimoto, Sumiyoshi-ku, Osaka, Japan
3
Associate Professor, Graduate School of Life and Science, Osaka City University, Sugimoto, Sumiyoshi-ku, Osaka, Japan Corresponding author: Jihui Yuan, Doctor Course, Graduate School of Life and Science, Osaka City University, 3-3-138 Sugimoto, Sugimoto, Sumiyoshi-ku, Osaka 558-8585, Japan. Email:
[email protected]
Yuan et al.
171
1999; Santamouris, 2007). The intensity of heat island effect may rise up 10°C in hot climates (Livada et al., 2002; Mihalakakou et al., 2002; Santamouris et al., 1999), resulting in increased discomfort and higher pollution levels while it also has a serious impact on the cooling energy consumption of buildings (Hassid et al., 2000). The temperature of Osaka rises 2.1°C every 100 years, which is higher than the national average temperature of 1.0°C. The difference of 1.1°C is thought to be caused by heat island effect (Osaka District Meteorological Observatory, 2003).
Scope of this research In summer, heat emitted from building accounts for about half of the heat emitted from urban areas (Ministry of Environment Japan, 2004). A high solar reflectance helps to reflect sunlight away from a building, reducing roof temperatures. Therefore, to increase the solar reflectance of a building’s exterior surface (Sakai et al., 2009), we need to characterize the solar reflectance of highly reflective material (HRM) after installation. The solar reflectance of roofing sheets installed on the roofs of 20 primary and middle schools in February 2010 was measured in November 2010, March 2011, and September 2011.
Target buildings and measurement method Target buildings In February 2010, the highly reflective roofing sheets (HRRSs) were installed on about 70 primary and middle schools in Osaka. We selected 20 schools with an even spatial distribution. Figure 1 shows the locations of the schools.
HRRSs In total, five types of HRRSs have been installed on the roofs of the 20 schools. Those HRRSs have high solar reflectance in the near-infrared (NIR) range regardless of the reflectance in the visible (VIS) range. Table 1 shows the types of HRRSs and their specifications.
Measurement items and measurement method Small portable spectrophotometers (see Table 2) were carried to measure the solar reflectance in VIS range (400–780 nm) and the solar reflectance in NIR range (780– 1700 nm) separately. Then, using the solar reflectance in VIS and NIR ranges, the solar reflectance over the entire range of those wavelengths (VIS + NIR: 400– 1700 nm) was calculated. In this field measurement, based on the measurement method JIS K 5602 of Japanese Industrial Standards (JIS K 5602, 2008), we have adopted the wavelength range from 400 to 1700 nm because of the limit of the used spectrophotometers.
172
Journal of Building Physics 37(2)
Figure 1. Locations of field measurements (20 primary and middle schools in Osaka, Japan).
The method is close to ASTM C1549-09, which determines solar reflectance of flat opaque materials in a laboratory or in the field at four wavelengths in the solar spectrum: 380, 500, 650, and 1220 nm using a commercial portable solar reflectometer. In this article, the determination of solar reflectance is calculated as that relative to a standard white plate as measured with the spectrophotometer. The relative reflectance was corrected to the absolute reflectance using the absolute reflectance of the standard white plate. Finally, the solar reflectance was determined using the absolute reflectance and the spectral distribution of the solar radiation. The accuracy of the measurement results is considered to be about 0.01 because there was a difference of 60.005 during calibration of a standard white plate from a spectrophotometer. The functions for determination of solar reflectance measured by the spectrophotometer are Ð rðlÞEðlÞdðlÞ ð1Þ re = Ð EðlÞdðlÞ
Yuan et al.
173
Table 1. Five types of HRRSs installed at the 20 primary and middle schools. Type
School no.
District name
Specifications
A (white gray)
No. 5 No. 6 No. 7 No. 12 No. 13 No. 17 No. 20 No. 1 No. 3 No. 4 No. 8 No. 9 No. 10 No. 14 No. 2 No. 15 No. 19 No. 11 No. 16
Fukushima Fukushima Konohana Minato Minato Naniwa Suminoe Kita Miyakojima Miyakojima Konohana Nishi Nishi Minato Kita Taisho Abeno Nishi Tennoji
Solar reflectance is about 0.7 (model no: cool(#801)) Conforms to JIS A 5759
No. 18
Naniwa
B (pearl gray)
C (silver gray) D (white gray) E (gray)
Solar reflectance in near-infrared range (780–2500 nm) is about 0.662 (model no: V-10) Conforms to JIS K 5602
Solar reflectance in near-infrared range (800–2100 nm) is about 0.8 (model no: HR) Company internal measurement Solar reflectance in near-infrared range (780–2100 nm) is about 0.75 (model no: reflective sheet) Solar reflectance in near-infrared range (780– 2100 nm) is about 0.55 (model no: BSN-54) Company internal measurement
HRRS: highly reflective roofing sheet. JIS A 5759, 2008: Adhesive films for glazings; JIS K 5602, 2008: Determination of reflectance of solar radiation by paint film.
Table 2. List of measurement equipment. Measuring equipment
Specifications
Spectrophotometer 1 Spectrophotometer 2 Light source Light fiber Thermometer Infrared camera
Wavelength range: 320–1000 nm; accuracy: 60.005 Wavelength range: 900–1700 nm; accuracy: 60.005 Wavelength range: 360–2000 nm; power output: 6.5 W; White Wavelength range: 400–2500 nm; probe ferrule: 6.35 mm Temperature range: 240°C–110°C; temperature accuracy: 60.3°C Temperature range: 240°C–120°C; temperature accuracy: 61°C
Ð rðlÞ =
rwðlÞr0 ðlÞ 100
ð2Þ
where ‘‘re’’ is the solar reflectance, ‘‘r(l)’’ is the spectral reflectance, ‘‘rw(l)’’ is the absolute reflectance of the standard white plate, ‘‘r#(l)’’ is the relative reflectance
174
Journal of Building Physics 37(2)
Figure 2. A roof field measurement (at No. 10 school); the order of points 1–5 and 6–10 is from the center of the roof to the edge of the roof. 10 points in total.
of HRRSs measured in the field, and ‘‘E(l)’’ is the spectral distribution of hemispherical solar irradiance specified in ISO 9845-1 (ISO 9845-1, 1992). At each school, the measurement points were from the center toward the edges of the roof, two rows in the direction of the roof gradient, a total of 10 measurement points with five measurement points for each row. The average value of the 10 measurement points was used as the solar reflectance of the school. Figure 2 shows the example of a measurement field at the field measurement No. 10 school. Figure 3 shows the layout of the measurement points. In the field measurement in November 2010, March 2011, and September 2011, when measuring a reflectance, no surface treatment, such as cleaning, was done in November 2010 and March 2011, but the surface of measurement points was cleaned with distilled water in September 2011. The installation method of the spectrophotometer is shown in Figure 4. An artificial light source through an optical fiber is aimed at an incident angle of 45° onto the measurement point. From the measurement point, the reflection through optical fiber at the same angle of 45° goes to the spectrophotometer and is processed by a laptop computer on site. The air temperature was measured with a thermometer (Table 2 and Figure 2), and surface temperature of the HRRSs was taken with an infrared camera (Table 2 and Figure 5).
Yuan et al.
175
Figure 3. Layout of the measurement points; the distance between rows (1–5 and 6–10) is 1.15 m, and the distance between points in each row is 1.0 m.
Figure 4. Installation method of the spectrophotometer; both the angle of incident light and the angle of reflected light are at 45°. The diameter of the measurement point is 9 mm. PC: personal computer.
Result of field measurement An overview The solar reflectance of HRRSs was measured in November 2010 (about 9 months after construction), in March 2011 (about 13 months after construction), and in
176
Journal of Building Physics 37(2)
Figure 5. Installation method of infrared camera.
Figure 6. Solar reflectance values in VIS range of HRRSs installed on the roofs of the 20 primary and middle schools. HRRS: highly reflective roofing sheet; VIS: visible.
September 2011 (about 19 months after construction). The values of the solar reflectance in different wavelengths (VIS, NIR, and VIS + NIR) of HRRSs installed on the roofs of the 20 schools are shown in Figures 6 to 8. One of the 20 schools, No. 18 school, differed greatly from the others. The building management chose a gray color (Type E in Table 1) to reduce ‘‘light pollution.’’ Its catalog value of solar reflectance in NIR (780–2100 nm) is 0.55. To account for this, Table 3 shows the basic statistics of field measurement results, both with and without No. 18 school.
Yuan et al.
177
Figure 7. Solar reflectance values in NIR range of HRRSs installed on the roofs of the 20 primary and middle schools. NIR: near infrared; HRRS: highly reflective roofing sheet.
Figure 8. Solar reflectance values in all wavelengths (VIS + NIR) range of HRRSs installed on the roofs of the 20 primary and middle schools. VIS: visible; NIR: near infrared; HRRS: highly reflective roofing sheet.
The average solar reflectance of the 20 schools, excluding the No. 18 school, is about 0.45 for all wavelength (VIS + NIR) ranges. In most cases, the solar reflectance in the NIR is about 0.1 higher than that in the VIS range. The reason is that the HRRSs installed on the roofs are mainly designed with white/gray and gray in color specification, and with low reflectance in the VIS range and high reflectance in the NIR range. Comparing the values in March 2011 with November 2010, the
NIR (780–1700 nm) 0.54 ! 0.55 ! 0.51(0.58) 0.55 ! 0.56 ! 0.52(0.59) 0.60 ! 0.62 ! 0.60(0.68) 0.60 ! 0.62 ! 0.60(0.68) 0.34 ! 0.34 ! 0.33(0.42) 0.48 ! 0.50 ! 0.45(0.49) 0.06 ! 0.06 ! 0.07(0.07) 0.03 ! 0.04 ! 0.05(0.06)
VIS (400–780 nm)
0.43 ! 0.44 ! 0.40(0.46) 0.44 ! 0.45 ! 0.41(0.46) 0.51 ! 0.54 ! 0.50(0.57) 0.51 ! 0.54 ! 0.50(0.57) 0.17 ! 0.17 ! 0.18(0.20) 0.39 ! 0.39 ! 0.31(0.33) 0.07 ! 0.08 ! 0.08(0.09) 0.04 ! 0.05 ! 0.07(0.07)
Wavelength range
Average
0.48 ! 0.49 ! 0.45(0.51) 0.49 ! 0.50 ! 0.46(0.52) 0.55 ! 0.57 ! 0.54(0.61) 0.55 ! 0.57 ! 0.54(0.61) 0.24 ! 0.24 ! 0.24(0.30) 0.43 ! 0.45 ! 0.38(0.40) 0.06 ! 0.06 ! 0.07(0.08) 0.03 ! 0.04 ! 0.06(0.07)
All wavelengths (400–1700 nm)
VIS: visible; NIR: near infrared. Accuracy of the measurement results is 0.01. Lower row of statistics in each cell excludes the No. 18 school, and the value in braces is the value of the solar reflectance after cleaning.
Standard deviation
Minimum
Maximum
Solar reflectance (—) (November 2010 ! March 2011 ! September 2011)
Item
Table 3. Basic statistics of the solar reflectance at VIS, NIR, and VIS + NIR ranges.
178 Journal of Building Physics 37(2)
Yuan et al.
179
Figure 9. Solar reflectance values in the NIR range of five types of HRRSs: Type A includes seven schools, Type B includes seven schools, Type C includes three schools, Type D includes two schools, and Type E includes one school (as shown in Table 1). NIR: near infrared; HRRS: highly reflective roofing sheet.
average value of solar reflectance in 20 schools is almost unchanged. Comparing the values in September 2011 with the values in March 2011 and November 2010, the average value showed a clear decrease of about 0.04.
Comparison among the reflectance of five types of HRRSs Comparing the solar reflectance in NIR for five types of HRRSs, Type A is about 0.58 with the highest solar reflectance, and the order of the others is Type C (about 0.57), Type D (about 0.56), Type B (about 0.52), and Type E (about 0.34). As explained earlier, Type E is much lower than the other four types. Comparing the solar reflectance in NIR (see Figure 9) measured in March 2011 with the solar reflectance measured in November 2011, the greatest change was a 0.02 increase for Type A, with the others changing even lesser. The solar reflectance measured in September 2011 compared with the solar reflectance values measured in March 2011 and November 2010 decreased to about 0.02 for Type A, about 0.05 for Type B, about 0.04 for Type C, about 0.08 for Type D, and about 0.003 for Type E.
Comparison among samples of five types of HRRSs Samples of new HRRS material were purchased from the manufacturer, and their solar reflectance were measured using the same technique as the field measurements
180
Journal of Building Physics 37(2)
Figure 10. Solar reflectance values for all wavelengths (VIS + NIR) range at different measurement positions from the roof center to the roof edge (November 2010, March 2011, and September 2011). VIS: visible; NIR: near infrared.
in the laboratory. Because the instrument covers the measurement point completely blocking all outside light, there is no difference in light conditions between the field measurements and the laboratory measurements. The results are included in Figure 9. A sample of Type D was not obtained, so its catalog value is used as the sample value in Figure 9. The results show that the solar reflectance of field measurement is lower than the sample value of HRRSs. The difference ranges from 0.15 to 0.27. Moreover, except for the darker gray Type E, the measured solar reflectance values of the samples and solar reflectance values in the catalog in NIR are almost the same for all types, about 0.7. The lowest sample’s solar reflectance in NIR is 0.57 for Type E.
Effect of dirt and degradation on the surface of HRRSs Effect of measurement position As an example of the relationship between the measurement position and the solar reflectance, Figure 10 shows the solar reflectance for all wavelength (VIS + NIR) ranges and measurement positions from the roof center to the roof edge for No. 15 school. The gradient of measurement points 1–5 and 6–10 tends to decrease from the center to the edge. There is a possibility that because of dirt accumulation on the edges caused by rainfall, the solar reflectance of the edges might decrease. However, in this research, we find no significant trend of solar reflectance decreasing from center to edge. Furthermore, no significant change due to the difference in the wavelength (VIS + NIR) range was found.
Yuan et al.
181
Figure 11. Surface conditions of HRRSs: (a) the condition before cleaning, (b) the condition after cleaning, and (c) the convex part of surface after installation. HRRS: highly reflective roofing sheet.
Dirt on the surface of HRRSs Figure 11 shows the condition of one HRRSs. Figure 11(a) is the condition before cleaning the surface of HRRSs, Figure 11(b) is the condition after cleaning the surface of HRRSs, and Figure 11(c) is a convex part of the HRRS resulting from installation. Dirt can easily accumulate between the convex parts. As shown in Figure 9, comparing the solar reflectance in the NIR range after cleaning the surface with the solar reflectance of the field measurements in September 2011, there was an overall increase of about 0.1. The solar reflectance after cleaning the surface approached the solar reflectance of the samples, but a difference between 0.09 and 0.20 remains, and this is probably related to the installation method and the exposure to environmental conditions after construction. During the installation of HRRSs on the roof of No. 18 school, when mounting them firmly, many bulges (see Figure 11(c)) were produced. This can be considered one of the reasons that dirt accumulated easily.
Degradation on the surface of HRRSs From the fact that the solar reflectance of HRRSs was largely restored after cleaning the surface with distilled water, the deposition of dirt on the surface is considered as the main cause for the decrease in solar reflectance. There is no degradation on the surface of the HRRSs in this study.
Comparison of the temperature of the roof installed with HRRSs and concrete material In order to investigate the effect of temperature reduction after construction of HRRS, we measured the surface temperature of No. 6 school where both HRRS and concrete material existed on the same building in September 2011, as shown in
182
Journal of Building Physics 37(2)
Figure 12. Surface temperatures of HRRSs and concrete material: the average temperature is 20.6°C in figure (a) and the average temperature is 28.9°C in figure (b) (using infrared camera as shown in Table 2): (a) roof coated with HRRSs and (b) roof coated with concrete material. HRRS: highly reflective roofing sheet.
Figure 12. With an air temperature of 16.2°C, a solar radiation of 803 W/m2, and an average wind velocity of 2.1 m/s measured at the school, the average temperature of the HRRS is about 20.6°C (see Figure 12(a)), about 8.3°C lower than the temperature of the concrete roof surface without HRRS, which was about 28.9°C (see Figure 12(b)). It is considered that the effect of temperature reduction will be larger in the summer.
Conclusion The solar reflectance of HRRSs installed on the roofs of 20 primary and middle schools was measured, leading to the following conclusions: 1.
2.
3.
The average solar reflectance in all wavelength (VIS + NIR) ranges is about 0.45, and the solar reflectance in the NIR range is about 0.50 for the 20 schools excluding the No. 18 school, which used darker color sheets. Furthermore, the solar reflectance in NIR is about 0.1 higher than the solar reflectance in the VIS range. Except for the darker gray Type E, the measured solar reflectance values of the samples and those in the catalog in NIR were almost the same for all types, about 0.7. The solar reflectance in field measurements was between 0.15 and 0.27 lower than the measured solar reflectance values of the sample and the values in the catalog. The solar reflectance of field measurements is extremely low at the No. 18 school, it is only 0.34 in NIR, the cause is presumed that the solar
Yuan et al.
4.
5.
6.
7.
183
reflectance in NIR range is 0.57 for HRRSs is adopted by building management of Type E. Comparing the values measured in November 2010 with the values measured in March 2011, the solar reflectance of field measurement in 20 schools is almost unchanged, but comparing the values measured in September 2011 with the values measured in November 2010 and March 2011, the solar reflectance shows a clear decrease of about 0.04. Comparing the solar reflectance after cleaning the surface with the solar reflectance of the field measurement in September 2011, the reflectance increased about 0.1. The solar reflectance after cleaning the surface approached the solar reflectance of samples, but a difference between 0.09 and 0.20 remains, and this is probably related to the installation method and exposure to environmental conditions. From the fact that the solar reflectance of HRRSs was largely restored after cleaning the surface with distilled water, dirt is considered as the main reason for the decrease in solar reflectance. Surface degradation was not examined in this article. Compared with the exposed concrete material, the temperature of the roof with HRRSs installed was about 8.3°C lower.
Acknowledgements The authors are sincerely grateful to the Foundation of Construction Materials Industrial Promotion Tostem Corporation, the Architecture Research Foundation of Takenaka Scholarship Foundation of Japan, and the Urban Studies of Osaka City University (representative researcher: Prof. K. Emura). We are also grateful to the Department of Urban Development Department and Education Committee of Osaka for measuring.
Funding This study was supported by the Foundation of Construction Materials Industrial Promotion Tostem Corporation, the Architecture Research Foundation of Takenaka Scholarship Foundation of Japan, the Urban Studies of Osaka City University, and the Department of Urban Development Department and Education Committee of Osaka.
References Akbari H, Konopacki S and Pomerantz M (1999) Cooling energy savings potential of reflective roofs for residential and commercial buildings in the United States. Journal of Energy 24: 391–407. Hassid S, Santamouris M, Papanikolaou N, et al. (2000) The effect of the Athens heat island on air conditioning load. Energy and Buildings 32: 131–141. ISO 9845-1:1992 (1992) Solar energy, reflectance solar spectral irradiance at the ground at different receiving conditions, part 1: direct normal and hemispherical solar irradiance for air mass 1.5.
184
Journal of Building Physics 37(2)
JIS A 5759 (2008) Adhesive films for glazings. Japanese Industrial Standards Committee. Available at: http://www.jisc.go.jp/ JIS K 5602 (2008) Determination of reflectance of solar radiation by paint film. Japanese Industrial Standards Committee. Available at: http://www.jisc.go.jp/ Livada I, Santamouris M, Niachou K, et al. (2002) Determination of places in the great Athens area where the heat island effect is observed. Theoretical and Applied Climatology 71: 219–230. Mihalakakou P, Flokas H, Santamouris M, et al. (2002) Application of neutral networks to the simulation of the heat island over Athens, Greece using synoptic types as a predictor. Journal of Applied Meteorology 41: 519–527. Ministry of Environment Japan (2004) Reports on the countermeasures of suppressing urban heat island effect in 2003. Available at: http://www.jma-net.go.jp/osaka/ Osaka District Meteorological Observatory (2003) Ministry of the environment report: research on the environmental impact of business considering the heat island phenomenon in 2002. Available at: http://www.jma-net.go.jp/osaka/ Sakai H, Emura K, Igawa N, et al. (2009) Reduction of reflected heat of the sun by retroreflective materials. In: Proceedings of the second international conference on countermeasures to urban heat islands (SICCUHI), Berkeley, CA, 21–23 September, CDROM 211450, pp. 1–6. Available at: http://heatisland2009.lbl.gov/docs/211450-sakaippt.pdf Santamouris M (2001) Energy and Climate in the Urban Built Environment. London: James and James Science Publishers. Santamouris M, Mihalakakou G, Papanikolaou N, et al. (1999) A neural network approach for modeling the heat island phenomenon in urban areas during the summer period. Geophysical Research Letters 26: 337–340. Santamouris M, Paraponiaris K and Mihalakakou G (2007) Estimating the ecological footprint of the heat island effect over Athens, Greece. Climatic Change 80: 265–276.