Professional Documents
Culture Documents
Quantifying the direct benefits of cool roofs in an urban setting: Reduced cooling
energy use and lowered greenhouse gas emissions
Tengfang Xua, *, Jayant Sathayea, Hashem Akbarib, Vishal Gargc, Surekha Tetalic
a
Lawrence Berkeley National Laboratory (LBNL), One Cyclotron Road, MS90R4000, Berkeley, CA 94720, USA
b
Department of Building, Civil, and Environmental Engineering, Concordia University, Montreal, Canada
c
International Institute for Information Technology (IIIT), Hyderabad, India
a r t i c l e i n f o a b s t r a c t
Article history: Cool roofs, cool pavements, and urban vegetation reduce cooling energy use in buildings, lower local air
Received 7 July 2011 pollution, and decrease greenhouse gas (GHG) emissions from urban areas. To promote widespread and
Received in revised form large-scale implementation of cool roofs to moderate urban heat-island phenomenon, more awareness
14 August 2011
and understanding of cool roof benefits at the local level is needed. As part of an investigation of
Accepted 19 August 2011
strategies to mitigate urban heat-island effects, field data gathered from a monitoring project on cool
roofs in India were reviewed. An innovative field-based analytical method was developed to quantify
Keywords:
cooling energy savings resulting from the installation of cool roofs on commercial buildings. For build-
Urban heat-island
Cool roof
ings monitored in the Metropolitan Hyderabad region, the measured annual energy savings from roof-
Energy savings whitening of previously black roofs ranged from 20 to 22 kWh/m2 of roof area, corresponding to
Greenhouse gas a cooling energy use reduction of 14e26%. The application of white coatings to uncoated concrete roofs
Mitigation measure resulted in annual savings of 13e14 kWh/m2 of roof area, corresponding to cooling energy savings of 10
Field-based analytical method e19%. The annual direct CO2 reductions associated with the reduced cooling energy use were estimated
to be 11e12 kg CO2/m2 of flat roof area. Additional field work on various building types and locations will
help to understand magnitudes of regional or global potential in energy savings and GHG emission
reductions from applying cool roofs. Knowledge about quantified cool roof benefits at both the local and
regional level may promote the formulation of new policies and programs throughout the world.
Ó 2011 Elsevier Ltd. All rights reserved.
1. Introduction (MW) for air-conditioning for buildings in the Los Angeles Basin [1].
Similar air temperature increases in urban areas are taxing the
1.1. Background ability of developing countries to meet urban electricity demand
while raising global greenhouse gas (GHG) emissions associated
Urban heat-island (UHI) is a common phenomenon where with energy use and power generation.
urban temperatures are significantly higher than those of its In fact, a number of past studies have documented that air
surrounding suburban and rural areas in summertime. Urban heat- temperatures in urban areas are higher and have increased faster
islands can affect communities by increasing summertime surface than those in surrounding suburban and rural areas. For example,
temperature of building envelopes and infrastructures; intensifying from 1930 to 1990, downtown Los Angeles recorded a growth of
thermal discomfort; elevating cooling energy use and peak energy 0.5 C per decade [1]. Greek researchers analyzed ambient air
demand; adding air pollution; and raising risks in heat-related temperature data from the National Observatory of Athens for the
illness or mortality. A higher air temperature tends to increase period between 1950 and 2004, and found that the number of
cooling needs and reduce working efficiency of cooling systems for hours with high air temperatures in the urban Athens area
built environments, resulting in higher power demand and energy dramatically increased in the last two decades [2]. Giannopoulou
use. For example, a study estimated that an increase of 1 C in air et al. [3] recently conducted a detailed statistical analysis of the
temperature would require the addition of about 500 megawatts heat-island characteristics and distribution in the greater Athens
area using temperature data from 25 stations. The central and
western parts of the city, which are more densely built and have
* Corresponding author. Tel.: þ 001 510 486 7810. less green coverage, are significantly warmer than surrounding
E-mail address: TTXu@LBL.Gov (T. Xu). areas during the summer. Hua et al. [4] used historical data on air
0360-1323/$ e see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.buildenv.2011.08.011
2 T. Xu et al. / Building and Environment 48 (2012) 1e6
Table 1
Monitoring periods and roof-surface reflectance.
Monitoring Date Measured outdoor air temperature West building roof East building roof
period and solar radiation (estimated reflectance) (estimated reflectance)
(Average from 9 a.m. to 5 p.m.)
Phase I JanuaryeMarch 29.1 C, 653 W/m2 Concrete roof (0.3) Concrete roof (0.3)
Pre-coating
Phase II MarcheJuly 32.4 C, 643 W/m2 White roof (0.7) Black roof (0.1)
Post-coating
Phase III AugusteDecember 27.2 C, 529 W/m2 White roof (0.7) White roof (0.7)
Post-coating
a field-based analytical method to quantify actual cooling energy 2.2. Review of pre- and post-coating field monitoring
savings induced by applying cool roofs on commercial buildings.
The magnitudes of actual cooling energy savings can then be The field study of cool roof impacts included baseline (i.e., pre-
compared with results from a calibrated simulation using the field coating) characterizations, applications of various roof coatings,
data. Next, with some informed assumptions, we illustrated a way and post-coating characterizations. These tasks were carried out in
to estimate potential GHG reductions induced by applying cool sequence, with similar building occupancies and operation, in the
roofs. two buildings selected.
Key parameters measured included indoor and outdoor air
2. Field study of cool roof effects: methods and results temperatures, roof-surface temperatures, roof heat fluxes, solar
radiation, and electricity usage of each building’s cooling systems.
In the following sections, we first briefly reviewed building and Details of instrumentation, monitoring process, and measurement
weather characteristics and field monitoring, and then analyzed uncertainties are available in a more comprehensive formal report
the field data, including comparisons of key performance metrics [17]. In particular, heat flux meters had a rated accuracy of 5% of
and examination of the correlations between concurrently the reading, while power meters had a rated accuracy of within
measured air-conditioning energy uses by each building. This was 0.5% of the reading. Table 1 summarizes the averaged outdoor air
followed by field-based analysis developed in this paper to quantify temperature and solar radiation monitored for each of the three
actual magnitudes of cooling energy savings corresponding to phases, each corresponding to specific roof coatings.
applying cool roofs on buildings.
2.3. Pre- and post-coating comparison of performance metrics
2.1. Summary of building and weather characteristics
Table 2 is a comparison between the east and west buildings of
Two commercial buildings with identical enclosures and performance metrics including maximum roof-surface tempera-
similar use in Hyderabad, India, were selected for this study. tures, heat flux through ceilings, and air-conditioning energy use.
Each building had a flat, 700 m2 concrete roof and the top floor Results are compared both before and after various roof coatings
(i.e., second floor) was air-conditioned by its own central system were applied to the roofs of the west and east buildings.
when occupied, while the first floor in each building had its own For the west building, the maximum roof-surface temperature
separate air-conditioning system. The both systems had the decreased from 54.7 C in Phase I to 41.2 C (a reduction of 13.5 C)
same design and control strategies. There was no shade on the in Phase II e even though the average outdoor air temperature
roofs caused by nearby structures or trees. Sealed single-pane increased from 29.1 C in Phase I to 32.4 C in Phase II (an increase
glass windows accounted for 7% of the wall area for both of 3.3 C, Table 1), after the white coating was applied to the original
buildings. Each building’s top floor was used as learning centers concrete roof. The maximum roof-surface temperature decreased
with similar operation and occupancies during normal working further down to 38.3 C in Phase III, during which the white coating
hours throughout the year, while the air-conditioned space on remained on the roof and the outdoor air temperature decreased to
the first floor was used as offices with similar occupancies. The
two buildings were next to each other, experiencing the same
weather conditions, which were changing over the course of this Table 2
Monitoring results for the west and east buildings (9 a.m.e5 p.m.).
study. Internal heat load from equipment in the east building
was similar to that of the west building. Various roof coatings Parameters and performance metrics West building East building
were applied to the roof of each building to create different Roof area (m2) 700 700
reflectance in different phases of the study that lasted for about Maximum roof-surface temperature ( C)
12 months. Phase I 54.7 54.7
Phase II 41.2 71.3
As the sixth most populous city and urban area in India, Phase III 38.3 39.6
Hyderabad has hot summers starting from late February to early Roof heat flux (W/m )2
June, monsoon seasons starting from late June to early October, and Phase I e Peak 12.8 11.0
mild winters from late October to early February. The climate of Phase II e Peak 12.6 21.9
Phase III e Peak 7.6 8.6
Hyderabad remains warm through most parts of the year. Given
Phase I e Average 2.2 2.8
that weather conditions change by the day, month, and season over Phase II e Average 3.6 9.7
the course of the year, it would be prohibitively difficult to quantify Phase III e Average 0.1 0.8
the direct energy impact of applying cool roofs by simply focusing Daily air-conditioning energy use in kWh/day
on a single building because the weather condition is uncontrol- Phase I 219 200
Phase II 285 280
lable. In the field measurements, the two buildings were monitored Phase III 215 187
concurrently yearlong.
4 T. Xu et al. / Building and Environment 48 (2012) 1e6
27.2 C on average. For the east building, the maximum roof-surface Phase II when it was the hottest. The rise of solar heat gains through
temperature increased from 54.7 C in Phase I to 71.3 C in Phase II the windows intensified the demand for more cooling energy use in
(an increase of 16.6 C), after the black coating was applied to the the west building compared to that of the east building, especially
original concrete roof; and decreased down to 39.6 C in Phase III, in the hotter season (e.g., Phase II).
during which the roof was coated white while the outdoor air In addition, the reduction of daily cooling energy use in Phase I
temperature decreased by 5.2e27.2 C on average. In Phase III, with compared to Phase III (from 219 to 215 kWh/day) seemed to be
both roofs having the similar white coating under the same smaller than expected, especially when corresponding to the
weather condition, the maximum roof temperatures of both roofs increased roof albedo (from 0.3 to 0.7), lowered outdoor air
were also similar. temperature (from 29.1 to 27.2 C), and reduced solar radiation
For the west building, the peak heat flux through the ceiling (653e529 W/m2). In fact, field observations discovered an increase
decreased slightly from 12.8 W/m2 in Phase I to 12.6 W/m2 in Phase in average internal gain due to operation of non-air-conditioning
II due to the application of white coating on the original concrete systems in both buildings from Phase I to Phase III. The increased
roof even though the average outdoor air temperature increased non-air-conditioning equipment use corresponded with increased
from 29.1 C in Phase I to 32.4 C in Phase II (an increase of 3.3 C). number of occupants, which also magnified the net increase in total
The peak heat flux decreased to 7.6 W/m2 in Phase III, during which internal heat generation within the buildings. The increased
the white coating remained on the roof and the weather was cooler. internal heat gains would demand higher daily cooling energy use
For the east building, the peak heat flux increased from 11.0 W/ if all other factors remained the same. As a net result, measured
m2 in Phase I to 21.9 W/m2 in Phase II, after the black coating was difference of daily cooling energy between Phase I and Phase III of
applied to the original concrete roof, coupled with an increase of each building was smaller than expected.
3.3 C in average outdoor air temperature; the peak heat flux In summary, the above comparisons show that while cooler
decreased further to 8.6 W/m2 in Phase III, during which the roof roofs contributed to lowering the maximal roof-surface tempera-
was coated white while the outdoor air temperature decreased by tures and moderating the heat flux through the roofs into the
5.2e27.2 C on average. In Phase III, with both roofs having the buildings, cooling energy use could increase nonetheless due to
similar white coating under the same weather condition, the peak warmer weather conditions. Such direct comparisons are insuffi-
heat flux through both roofs again was similar. cient to quantify the magnitudes of cooling energy savings attrib-
For the average heat flux between 9 a.m. and 5 p.m. in Phases I uted to reflectance changes, because actual cooling energy was also
and III, both buildings exhibited similar magnitudes, with lower affected by other concurrent factors that are uncontrollable from
values happening in Phase III e during which the average outdoor the field, e.g., outdoor air temperatures, heat gains, or solar radia-
air temperatures and solar radiation were both lower. In Phase II, tion intensity.
the west building, with its white roof, exhibited a higher average In the following section, we developed a new analytical method
heat flux during the daytime than that in Phase I or Phase III, largely based upon the concurrently measured field data gathered for both
because outdoor air temperatures were higher in Phase II than they buildings, and used the method to quantify the direct cooling
were in Phase I and Phase III. The observed variations in average energy savings that resulted from installation of cooler roofs.
heat flux from 2.2 to 3.6 W/m2 of roof area exhibited that the
impact of increasing outdoor temperatures (i.e., by 3.3 C) on the 2.4. Field-based analysis to quantify cooling energy savings
average heat flux is more dominant than the sole effect of applying resulting from cool roofs
white coating on the concrete roof (i.e., an increase of reflectance by
0.4). By contrast, in Phase II, the east building with the black coating For each building selected in this study, we anticipated that
applied to the roof exhibited much higher average heat flux during actual weather conditions would have a significant impact on the
daytime than that in Phase I and Phase III. For example, average energy use of each air-conditioning system. The comparison of
heat flux changed from 2.8 to 9.7 W/m2 of roof area from Phase I to monitored results of average daily energy usage by air-conditioning
Phase II. This indicates that having both a higher outdoor temper- systems during Phase I and Phase II confirmed such impacts. For
ature and a black roof (i.e., a decrease of reflectance by0.2) has example, in the case of the west building, even though its bare
a significant collective impact on increasing average heat flux concrete roof was given a white coating for Phase II, cooling energy
through roofs. use increased slightly from Phase I. Notably, the weather was cooler
The table also shows average daily air-conditioning electricity during Phase I and became warmer in Phase II. In the following, we
use in each building with various roof coatings applied sequentially first examine the correlation between concurrently measured air-
during the course of the study. For simplicity, we analyzed daily conditioning energy usages for each building, and then develop
cooling energy use of each building from 9 a.m. to 5 p.m. on the field-based analysis to quantify the magnitudes of energy
weekdays. One would expect that the average daily air- savings due to roof-surface reflectance changes.
conditioning energy use in the buildings may tend to increase While coating tasks were carried out in sequence with concur-
with the rise in outdoor temperatures and a decrease in roof- rent monitoring in both buildings, each building maintained its
surface reflectance. From Phase I to Phase II, the average outdoor normal occupancies and operations at both sites were similar.
temperatures increased from 29.1 to 32.4 C because of the seasonal Therefore, the impact analysis was performed by quantifying the
change, corresponding to an increase of average cooling energy use correlations of the concurrent energy metrics between the two
from 219 to 285 kWh/day e even though the roof-surface reflec- buildings, grouped by two phases of the study e Phase II (white
tance increased from 0.3 to 0.7 for the west building; while for the roof on the west building, black roof on the east building) and Phase
east building there was a smaller increase of average cooling energy III (both buildings with white roofs).
use from 200 to 280 kWh/day e even though the roof-surface Fig. 2 shows regressions of concurrent daily cooling energy use
reflectance decreased from 0.3 to 0.1 (Black). Apparently, there between the two buildings, grouped by Phase II and Phase III during
are additional factors affecting the daily cooling energy demand. which roof-surface reflectance was used as a controlled parameter.
For example, because the total area of the west-facing windows in The quantification of the correlations between the two buildings
the west building was larger than that of the east building, higher using the west building as the reference case can then be used to
solar heat gains into the air-conditioned space in the west building quantify cooling energy savings potentials attributable to the
were expected than that of the east building in all phases, including changes of roof-surface reflectance for the east building.
T. Xu et al. / Building and Environment 48 (2012) 1e6 5
building. It is interesting to note that Bhatia et al. [18] used the health impacts. Overall, applying cool roofs globally can be effective
experimental data from the field work reported in [17] to calibrate in reducing cooling energy use and the associated GHG emissions.
a building simulation to predict annual energy savings of applying The study also suggests that additional field studies maybe
white roof over gray-concrete roof (reflectance 0.3) of the west helpful to quantify practical scales of potential energy savings
building. The calibrated model predicted an annual cooling energy through the installation of cooler roofs in various locations
savings of 15 kWh/m2 due to cool roof application, which was throughout the world. Future field studies may build upon the
similar to the ranges (13e14 kWh/m2 of roof area) that we have success presented in the paper and benefit from the innovative
quantified through the field-based analysis in this paper. field-based analytical method developed from this study. Knowl-
Using GHG emissions factor of 820 kg CO2 per MWh for India edge about quantified cool roof benefits at both the local and
[19], we estimated that annual direct CO2 reductions associated regional levels may promote the formulation of new policies and
with the reduced energy use would be approximately 11e12 kg programs throughout the world.
CO2/m2 of flat roof area per year by changing a concrete roof to
white-coated roof. Acknowledgments
Because actual cooling energy use and potential energy savings
in buildings due to applying cool roofs is affected by numerous This work was sponsored by USAID. The authors would like to
factors, including but not limited to localized weather, building acknowledge the field work performed by individuals from LBNL
enclosures, design, operation, controls, system types, and occu- and IIIT included in [17], and a U.S. cool roof manufacturer that
pancies, the estimated ranges of energy savings and CO2 emission provided in-kind support for applying roof coatings. This paper
reductions presented above were based upon the field results from benefits from valuable comments from anonymous reviewers for
only one geological location and a specific building type. In order to the journal, and graphic assistance of Anthony Ma.
calculate the regional or national potential in energy savings, more
field work on representative building types in the region or country References
will help to further fill knowledge gaps in understanding magni-
tudes of potential energy savings and GHG emission reductions [1] Akbari H, Pomerantz M, Taha H. Cool surfaces and shade trees to reduce
energy use and improve air quality in urban areas. Solar Energy 2001;70(3):
across different regions and climates in the future. Such under- 295e310.
taking would make it possible to investigate energy use and savings [2] Livada I, Santamouris M, Assimakopoulos M. On the variability of summer air
potential by advancing the field-based method that can benefit temperature during the last 28 years in Athens. Journal of Geophysical
Research 2007;112:D12103. doi:10.1029/2006JD008140.
from further statistical experimenting.
[3] Giannopoulou K, Livada I, Santamouris M, Saliari M, Assimakopoulos M,
Caouris Y. On the characteristics of the summer urban heat island in Athens,
4. Conclusions Greece. Sustainable Cities and Society 2011;1:16e28.
[4] Hua L, Ma Z, Guo W. The impact of urbanization on air temperature across China.
Theoretical and Applied Climatology 2007;93(3e4):179e94. Springer. Retrieved
In the effort to reduce energy use in buildings in urban areas, from, http://www.springerlink.com/index/10.1007/s00704-007-0339-8.
lower local air pollutant concentrations, and decrease greenhouse [5] Ren G, Zhou Y, Chu Z, Zhou J, Zhang A, Guo J, et al. Urbanization effects on
gas emissions, we developed an innovative field-based analytical observed surface air temperature trends in north China. Journal of Climate
2008;21:1333e48. doi:10.1175/2007JCLI1348.1.
method to evaluate the effectiveness of applying cool roof coatings [6] Mohan M, Kikegawa Y, Gurjar B, Bhati S, Kandya A, Ogawa K. Assessment of
on roofs, and used this method to quantify their direct impact on urban heat island intensities over Delhi. The Seventh International Conference
cooling energy savings and the associated GHG emission on Urban Climate, 29 June - 3 July 2009, Yokohama, Japan;2009.
[7] Ojima T. Changing Tokyo metropolitan area and its heat island model. Energy
reductions. and Buildings 1990/1991;15e16:191e203.
Roofs with higher surface reflectance significantly lowered [8] Wong N, Yu C. Study of green areas and urban heat island in a tropical city.
roof-surface temperatures and reduced heat flux through the roofs Habitat International 2005;29:547e58.
[9] Jusuf S, Wong N, Hagen E, Anggoro R, Hong Y. The influence of land use on the
compared to roofs with lower surface reflectance. Using the field-
urban heat island in Singapore. Habitat International 2007;31:232e42.
based analytical method developed in this study, we quantified [10] Rosenfeld A, Akbari H, Romm J, Pomerantz M. Cool communities: strategies
energy saving effects of cool roofs on individual buildings. The for heat island mitigation and smog reduction. Energy and Buildings 1998;28:
51e62. doi:10.1016/S0378-7788(97)00063-7.
measured annual energy savings from roof-whitening of previ-
[11] Akbari H, Rosenfeld A, Taha H. Summer heat islands, urban trees, and white
ously black roofs ranged from 20 to 22 kWh/m2 of roof area in the surfaces. ASHRAE Transactions 1990;96(1). American Society for Heating,
Metropolitan Hyderabad region, corresponding to an air- Refrigeration, and Air Conditioning Engineers, Atlanta, Georgia.
conditioning energy use reduction of 14e26%. An estimated [12] Taha H. Modeling the impacts of increased urban vegetation on the ozone air
quality in the south coast air basin. Atmospheric Environment 1996;30(20):
annual savings of 13e14 kWh/m2 of roof area could be achieved by 3423e30.
applying white coatings to uncoated concrete roofs on buildings in [13] Konopacki S, Gartland L, and Akbari H. Demonstration of energy savings of cool
the region, corresponding to cooling energy savings of 10e19%. roofs. Lawrence Berkeley National Laboratory Report, LBNL-40673; 2000.
[14] Santamouris M, Synnefa A, Kolokotsa D, Dimitriou V, Apostolakis K. Passive
The annual direct CO2 reductions associated with the reduced cooling of the built environmentduse of innovative reflective materials to
energy use by whitening concrete roofs were estimated to be fight heat island and decrease cooling needs. International Journal Low
11e12 kg CO2/m2 of flat roof area. The analysis has found that Carbon Technologies 2008;3(2):71e82.
[15] Yu C, Wong N. Thermal impact of strategic landscaping in cities: a review.
a cool roof can significantly reduce overall cooling energy use in Advances in Building Energy Research 2009;3(1):237e60.
buildings, but the relative effect of roof-whitening on air- [16] Shashua-Bar L, Tsiros I, Hoffman M. A modeling study for evaluating passive
conditioning energy use is more pronounced on cooler, rather cooling scenarios in urban streets with trees e case study: Athens, Greece.
Building and Environment; 2010. doi:10.1016/j.buildenv.2010.06.008.
than warmer days. [17] Akbari H, Xu T, Taha H, Wray C, Sathaye J, Garg V, Tetali S, Babu M, Reddy K.
While the findings indicate significant potential in energy Using cool roofs to reduce energy use, greenhouse gas emissions, and urban
savings and reductions in GHG emissions (i.e., direct benefits) heat-island effects: Findings from an India experiment. Lawrence Berkeley
National Laboratory Report, LBNL-4746E; 2011.
resulting from the use of cool roofs on individual buildings studied,
[18] Bhatia A, Mathur J, Garg V, Akbari H. Determination of energy saving of cool
large-scale deployment of mitigation measures can bring addi- roof on an office building in Hyderabad, India using calibrated simulation.
tional indirect benefits to the urban area, such as cooling outside Solar Energy Society of India 2010;20. No. 1 & 2 June - December, 2010.
air; thereby improving the efficiency of cooling systems, reducing [19] Central Electricity Authority (CEA). CO2 baseline database for the Indian
power sector, user guide version 5.0. November 2009. Government of India,
GHG emissions, decreasing smog, and indirectly cutting pollution Ministry of Power, http://www.cea.nic.in/planning/c%20and%20e/user_guide_
from power plants e all improving environmental quality and ver5.pdf; 2009.