1 s2.0 S0360132322003006 Main
1 s2.0 S0360132322003006 Main
1 s2.0 S0360132322003006 Main
A R T I C L E I N F O A B S T R A C T
Keywords: Green roof systems have become an important part in creating sustainable cities. They can provide a wide range
Cooling effect of economic, environmental, and social benefits. The goal of this research was to quantify the thermal perfor
Green roof mance improvements from a green roof with mineral wool substrate installed on a school building in a humid
Heat flux
subtropical climate. In-situ measurements during a summer period included heat fluxes through the green and
Pearson correlation
Thermal performance
reference roof, vertical temperature profile through both roofs, the local air temperature above roofs, and local
Mineral wool meteorological parameters. Furthermore, the summer thermal performance of green and reference roof and the
green roof cooling effect were evaluated concerning meteorological parameters using the Pearson correlation
analysis. The results indicate that the green roof layers have improved thermal performance of the roof with
respect to reduced conductive heat flow by 57% and delayed heat transfer. The maximum and averaged refer
ence roof to green roof outdoor surface temperature difference was 27.5 ◦ C and 5.5 ◦ C, respectively. It was found
that ambient temperature and relative humidity have a dominant role on the thermal performance of the green
and reference roof, while solar radiation and ambient temperature present the key meteorological determinants
of the green roof cooling effect.
cm) which can support a greater variety of plant species, even trees, and
1. Introduction can sustain walking paths and benches. However, they have higher
initial costs and require an irrigation system and more maintenance.
Over the last decades, the increased urbanization and economic The green roof substrate or growing medium is vitally important for
growth have resulted in less urban green spaces. Dense urban areas with successful vegetation growth on the green roof. It is usually a blend of
a large number of high-rise buildings can provide green space on roofs. soils, which needs to be lightweight, chemically stable, well-aerated,
Green roof systems have become an important part in creating sustain free-draining, and able to supply water and nutrients to plants [7].
able cities. They can provide a wide range of economic, environmental, The hydrophilic mineral wool in extensive green roof systems is used as
and social benefits which include the reduction of air pollution [1], a substitute for traditional soil substrate. Mineral wool is used as water
water run-off [2], urban heat island effects [3], noise [4], and an in retention and drainage layer as well as surface growth media. This type
crease in urban biodiversity [5]. Green roofs can significantly reduce of substrate rapidly absorbs rainwater and drains when material ap
building cooling and heating energy demands through evapotranspira proaches saturation [8].
tion, photosynthesis, solar shading, and thermal insulation [6]. Many studies have been performed over the past years in different
There are two main classifications of green roofs: extensive green fields about green roofs. According to an experiment conducted in the
roofs and intensive green roofs, depending on weight, substrate layer, Mediterranean climate [9], a green roof reduces the roof surface tem
plant type, maintenance, irrigation, and cost. Extensive green roofs have perature by an average of 12 ◦ C as compared to a black bituminous roof
a thin substrate layer (<15 cm) with a limited variety of plants. These in summer, while during the winter difference between bituminous and
green roof systems are lightweight, easy to install, and low maintenance. greens roofs was almost 4 ◦ C. In addition, the green roof reduces 100%
They are widely used in practice for retrofit public and commercial of thermal energy entering through the building’s roof in summer and
buildings. Intensive green roofs have a thicker substrate layer (20–200 reduces from 30% to 37% of the outgoing thermal energy in winter. The
* Corresponding author.
E-mail address: [email protected] (D. Kostadinović).
https://doi.org/10.1016/j.buildenv.2022.109061
Received 10 January 2022; Received in revised form 2 March 2022; Accepted 1 April 2022
Available online 5 April 2022
0360-1323/© 2022 Elsevier Ltd. All rights reserved.
D. Kostadinović et al. Building and Environment 217 (2022) 109061
2
D. Kostadinović et al. Building and Environment 217 (2022) 109061
Belgrade (44◦ 49′ 14′′ N, 20◦ 27′ 44′′ E) the capital city of Serbia. New reference roof and green roof are 0.41 W/m2K and 0.27 W/m2K
Belgrade is one of the central urban business districts in Serbia and one respectively, which were calculated using the estimated method [28].
of the major ones in Southeast Europe as well as the second-most The layer of gravel was removed and a green roof was installed on PVC
populous municipality of Serbia. Belgrade lies at an average elevation foil. From bottom to top, the green roof consists of four layers including
of 117 m above sea level and has a humid subtropical type of climate. waterproof layer, drainage layer with a water reservoir, rock mineral
This climate is considered to be Cfa, according to the Köppen-Geiger wool as a growing substrate, and vegetation mat. The mineral wool
climate classification [27]. Belgrade has four seasons, with warm to hot growing substrate is a porous lightweight layer made of mainly volcanic
summers and cold winters. The warmest month is July with an average materials which represent good media for plants rooting and growing.
maximum temperature of 23 ◦ C. January is the coldest month, with an The vegetation mat consists of the following eight Sedum species: Sedum
average temperature of 1.4 ◦ C. Belgrade is moderately sunny with 2.112 acre, Album coral carpet sedum, Sedum wall album, Sedum hybridum,
average sunny hours during the year. Sedum sexangulare, Sedum hispanicum, Sedum floriferum, and Sedum
Kamtschaticum. The cross-sectional view of the green roof and reference
2.2. Experimental set-up roof experimental plot and position of sensors are shown in Fig. 2.
Below the green and reference roof, there are locker rooms that were
The field experiment was conducted in a school building built in not occupied due to school summer break. In both rooms, the indoor air
1975 with a three-story reinforced concrete structure. The selected temperature of 24 ◦ C was maintained by air conditioners. At the time of
building represents a typical school building with regards to the archi the measurement period, the green roofs had existed for more than a
tectural characteristics and roof type and material, from the period of year hence the roof was uniformly covered with vegetation and Sedum
most intensive construction of larger school buildings in urban areas of plants were well established. The experimental green roof has not an
Serbia. In November 2018, a lightweight extensive green roof with a irrigation system.
surface of 25 m2 was installed on the flat roof of a building wing at a
height of approximately 4 m from the ground level. The adjacent 2.3. Experimental methods
reference roof plot with the same surface area has been chosen to be
compared with the green one. Fig. 1 presents locations of the experi In this paper the empirical method was used, with quantitative data
mental green roof and the reference roof plots at the school building. collection through experimental measurements. A statistical analysis
From bottom to top, the structural roof consists of a concrete slab, was also conducted to explore the influence of meteorological parame
vapor barrier, bitumoperlite, polyurethane, hydro isolation, and con ters on summer thermal performance of green roof and reference roof
crete slabs (30 × 30 cm) with a total thickness of 45 cm. The roof has a and green roof cooling effect. An experimental measurement was con
5% slope for rainfall water management. In 2017, new hydro isolation ducted from 3 to 6 July 2020. Experimental equipment has been placed
was added on top of the roof which includes bitumen with fiberglass on both roof plots to compare the thermal performance of the green roof
mesh, PVC foil, and gravel with a total thickness of 5 cm. The U value for and the reference roof. Both roofs were equipped with the heat flux
Fig. 1. a) The experimental site located at the School building, Belgrade, Serbia. b) Locations of the green roof experimental plot (GR) and the reference roof plot
(RR) on the top of the School building. c) The experimental green roof plot with environmental monitoring sensors. d) The reference roof plot with environmental
monitoring sensors. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)
3
D. Kostadinović et al. Building and Environment 217 (2022) 109061
high solar radiation and ambient temperature, clear sky condition, and
an absence of precipitations. The measured ambient temperature, rela
tive humidity, and solar radiation are shown in Fig. 3, while wind speed
is shown in Fig. 4. During the hottest time of the day, the solar radiation
was around 900 W/m2, the ambient temperature was around 35 ◦ C, and
the relative humidity was between 35% and 45%. The wind speed
ranged from 0.2 m/s to 4.4 m/s.
The following parameters were compared on both roofs: outdoor and
indoor roof surface temperature, the local air temperature above roofs,
and heat flux through roofs. When it comes to the graphs of tempera
tures, the given measured values are the means of measured values at
three measuring points. The first analysis compared the outdoor surface
temperature of the green roof and reference roof shown in Fig. 5. As
expected, the temperature of the outdoor surface of reference roof has
significantly higher values compared to the green roof and ambient
temperature, with large daily temperature fluctuations. It can be
observed that the highest reference roof-green roof surface temperature
differences are between 13:00 and 15:00. The maximum and averaged
Fig. 2. Schematic cross-sectional view of the green roof system and reference reference roof to green roof outdoor surface temperature difference was
roof with positions and heights of the installed sensor. (For interpretation of the
27.5 ◦ C and 5.5 ◦ C, respectively. Similar findings were reported in NYC
references to color in this figure legend, the reader is referred to the Web
[29], where daily maximum and averaged summer black roof to green
version of this article.)
roof membrane temperature difference was 33 ◦ C and 9.9 ◦ C, respec
tively. This large temperature reduction is caused by solar shading,
sensor attached to the ceiling to record the heat flow through the green
thermal insulation, and latent heat loss through evapotranspiration. A
roof (qGR) and reference roof (qRR). To obtain temperature distribution
small amount of solar radiation is reflected from the reference roof,
along the vertical direction thermocouples have been installed at three
while a large part is absorbed leading to high surface temperature.
points on the green and reference roof plots. The eighteen thermocou
During the observed period the outdoor surface temperature of the green
ples were applied on the green roof to measure indoor (Ts,in GR) and
roof was below 30 ◦ C.
outdoor (Ts,out GR) green roof surface, drainage layer (Td), growing
As can be seen in Fig. 5 outdoor surface temperature of the reference
substrate layer (Tgs), vegetation layer (Tv) temperature, and local air
roof increases rapidly and reaches its peak around noon when solar ra
temperature at 10 cm above vegetation (T10 GR). The outdoor surface
diation is strong. High temperatures on the roof cause thermal stresses in
temperature of the green roof was measured at the interface between the
materials and have a negative impact on the durability of the roof
waterproof layer and the structural layer. The reference roof was
membrane [30]. The reference roof outdoor surface is directly exposed
equipped with nine thermocouples to measure indoor (Ts,in RR) and
to solar radiation, while the energy that the green roof received needs to
outdoor (Ts,out RR) roof surface temperature, and local air temperature at
be transferred through the conductive heat flux within the green roof
10 cm above the reference roof (T10 RR).
Four meteorological parameters were monitored to determine
background weather conditions, including ambient temperature (Ta),
relative humidity (RH), incoming solar radiation (SR) and wind speed
(WS) at 150 cm above reference roof. All sensors readings were averaged
over 10 min periods, and the data were stored in the data acquisition
system. All monitoring sensors were placed at the center of the roof plots
to avoid edge and shadow effects. The technical specifications of the
sensors with measured parameters are summarized in Table 1. Cali
bration of the sensors was done before installation.
Table 1
Technical specifications of the sensors.
Sensor Model Parameters Measuring range Resolution Accuracy
4
D. Kostadinović et al. Building and Environment 217 (2022) 109061
Fig. 6. The indoor surface temperature of the green roof and reference roof.
(For interpretation of the references to color in this figure legend, the reader is
referred to the Web version of this article.)
5
D. Kostadinović et al. Building and Environment 217 (2022) 109061
Fig. 9. The heat flux through the green roof and reference roof in relation to
Fig. 8. The vertical temperature profile through the green roof. (For interpre solar radiation. (For interpretation of the references to color in this figure
tation of the references to color in this figure legend, the reader is referred to legend, the reader is referred to the Web version of this article.)
the Web version of this article.)
follows the temperature trend within the roofing systems. The green roof
10:00 a.m. to 6:00 p.m. during which the vertical temperature sequence experienced negative heat flow from 1:00 a.m. to 00:30 p.m., while
was: vegetation layer > growing substrate layer > drainage layer > positive heat flow during the remaining time. The reference roof heat
outdoor roof surface, indicating solar heat gain and thermal damping. flux was almost always positive indicating that heat was transferred
The vegetation surface has the highest averaged temperature of from the roof structure to the interior due to the strong solar radiation
25.8 ◦ C, while the average temperature of the growing substrate was and high outdoor roof surface temperature. In general, during the
25.3 ◦ C. According to Ref. [33], 27% of the solar radiation is reflected, observed period more heat is transferred into the building than trans
60% is absorbed by the vegetation and 13% is transmitted to the roof ferred out of the building through both roofs.
structure. During the day from 10:00 p.m. to 4:00 p.m., the temperature The obtained results show lower heat flux through the green roof
of vegetation layer surface is higher than the ambient temperature, both day and night compared to the reference roof. The maximum heat
suggesting that the vegetation mat is effective in absorb and retaining flux was 7.6 W/m2 for the reference roof and 3.3 W/m2 for the green
heat. This can be explained by the following reasons: (1) the dark color roof. The green roof significantly decreased heat flow through the roof
and lower albedo value of vegetation contribute to heat absorption and by 57% and delayed heat transfer for 3 h and 40 min due to increased
incur higher temperature of vegetation layer surface; (2) non-uniform thermal mass, shading effect of vegetation, and evapotranspiration. This
heights of sedum plants and different surface roughness trap heat and is in agreement with the study [35] which compared a low insulated flat
retard the ventilation of the air layer closest to the surface. The vege roof to green roofs and calculated a potential reduction of 60% of heat
tation surface cooled rapidly and has a minimum temperature at night. flux during the summer period. The temperature in rooms below green
The Sedum tissues had restricted temperature buffering capacity, tend and reference roof was kept constant; therefore any heat flow between a
ing toward a high daytime maximum and low nighttime minimum [34]. building and its environment creates energy demand for space condi
During day time temperature of rock mineral wool substrate was tioning. The green roofs could reduce the power consumption of the air
lower than the ambient temperature. The vegetation layer acts as a conditioning system while maintaining the same indoor air temperature
thermal insulator and significantly affects the mineral wool tempera [36].
ture. Shading by vegetation layer shields mineral wool substrate from During the night, heat flux through the reference roof reached a
incoming solar radiation. The outdoor roof surface temperature of the minimal value of − 0.08 W/m2 when the temperature of the outdoor
green roof reduced significantly because of the presence of vegetation, reference roof surface drops to a minimum daily value. In the early
growing substrate, and drainage layer. Substrate and drainage layer morning, the minimal value of − 3.4 W/m2 was recorded for heat flux
temperatures have compressed daily regime and delayed peaks through the green roof.
compared to the ambient temperature. There is an evident correlation between the heat flux through the
The delay in the temperature pick of substrate and drainage layer reference roof and solar radiation with a small-time lag (50 min) for the
concerning ambient temperature is respectively 40 min and 1h and 10 daytime heat flux peak, indicating the low thermal inertia of the refer
min. Since the water reservoirs of the drainage layer are most of the time ence roof. The experimental measurements shown that main effect of
filled with air which has a low thermal conductivity, the plastic drainage green roofs was to reduce daily fluctuations of building roof temperature
layer has a good thermal insulation effect. The highest fluctuation in and correspondingly to decrease fluctuations of conductive heat flux
daily temperatures of green roof layers was noted for the vegetation through building roof. The daily fluctuations of heat flux in and out of
layer 18.2 ◦ C and the lowest for the drainage layer 8.2 ◦ C. The fluctu the building were lower for the green roof than for the reference roof.
ation in the daily temperature of the mineral wool substrate is 12.2 ◦ C. In The average heat flux amplitude for the green roof was 5.1 W/m2 while
terms of temperature fluctuations, the outdoor surface temperature of 6.3 W/m2 for the reference roof. The experimental measurements
the reference roof is more strongly affected by the solar radiation indicate that the lightweight green roof have improved thermal per
compared to those measured on the outdoor surface of the green roof formance of the roof with respect to reduced conductive heat gain.
and mineral wool substrate surface.
The values of solar radiation and heat flux measured at the interface
between the indoor surface of green and reference roof and indoor air 3.2. Correlation analysis
are shown in Fig. 9. The heat flux sensors were calibrated such that a
positive reading shows heat flow into the building, while a negative Meteorological parameters play an important role in the proper
reading represents heat flow outside of the building. The heat flux evaluation of the thermal insulation and evapotranspiration effects of
the green roof. The Pearson correlation analyses were performed to
6
D. Kostadinović et al. Building and Environment 217 (2022) 109061
7
D. Kostadinović et al. Building and Environment 217 (2022) 109061
8
D. Kostadinović et al. Building and Environment 217 (2022) 109061
[22] B.-S. Lin, C.-C. Yu, A.-T. Su, Y.-J. Lin, Impact of climatic conditions on the thermal symposium on heating, ventilation and air conditioning (ISHVAC) and the 3rd
effectiveness of an extensive green roof, Build. Environ. 67 (2013) 26–33, https:// international conference on building energy and environment (COBEE), Procedia
doi.org/10.1016/j.buildenv.2013.04.026. Eng. 121 (2015) 15971604, https://doi.org/10.1016/j.proeng.2015.09.186.
[23] S.W. Tsang, C.Y. Jim, Theoretical evaluation of thermal and energy performance of [33] P.M. Krusche, D. Althaus, I. Gabriel, M. Weig-Krusche, Ökologisches Bauen
tropical green roofs, Energy 36 (5) (2011) 3590–3598, https://doi.org/10.1016/j. Herausgegeben Vom Umweltbundesamt, 1982, https://doi.org/10.1007/978-3-
energy.2011.03.072. 322-84893-2. Bauverlag GmbH, Wiesbaden & Berlin.
[24] Y. He, H. Yu, N. Dong, H. Ye, Thermal and energy performance assessment of [34] C.Y. Jim, Assessing climate-adaptation effect of extensive tropical green roofs in
extensive green roof in summer: a case study of a lightweight building in Shanghai, cities, Landsc. Urban Plann. 138 (2015) 54–70, https://doi.org/10.1016/j.
Energy Build. 127 (2016) 762–773, https://doi.org/10.1016/j. landurbplan.2015.02.014.
enbuild.2016.06.016. [35] N.H. Wong, Y. Chen, C.L. Ong, A. Sia, Investigation of thermal benefits of rooftop
[25] L.L.H. Peng, X. Yang, Y. He, Z. Hu, T. Xu, Z. Jiang, L. Yao, Thermal and energy garden in the tropical environment, Build. Environ. 38 (2003) 261–270, https://
performance of two distinct green roofs: temporal pattern and underlying factors in doi.org/10.1016/S0360-1323(02)00066-5.
a subtropical climate, Energy Build. 185 (2019) 247–258, https://doi.org/ [36] W. Yang, Z. Wang, J. Cui, Z. Zhu, X. Zhao, Comparative study of the thermal
10.1016/j.enbuild.2018.12.040. performance of the novel green (planting) roofs against other existing roofs,
[26] D. Dimitrijević Jovanović, D. Kostadinović, P. Živković, D. Ranđelović, Green Sustain. Cities Soc. 16 (2015) 1–12, https://doi.org/10.1016/j.scs.2015.01.002.
living roofs as a part of green infrastructure, in: Proceedings of 3rd International [37] M. Santamouris, Cooling the cities - a review of reflective and green roof mitigation
Conference on Urban Planning - ICUP2020, 2020, pp. 171–178, 12-13 November. technologies to fight heat island and improve comfort in urban environments, Sol.
[27] M. Kottek, J. Grieser, C. Beck, B. Rudolf, F. Rubel, World map of the Köppen-Geiger Energy 103 (2014) 682–703, https://doi.org/10.1016/j.solener.2012.07.003.
climate classification updated, Meteorol. Z. 15 (3) (2006) 259–263, https://doi. [38] C.Y. Jim, L.L.H. Peng, Weather effect on thermal and energy performance of an
org/10.1127/0941-2948/2006/0130. extensive tropical green roof, Urban For. Urban Green. 11 (2012) 73–85, https://
[28] ISO 6946:2017-Building Components and Building Elements-Thermal Resistance doi.org/10.1016/j.ufug.2011.10.001.
and Thermal Transmittance-Calculation Methods, International Organization for [39] P. Bevilacqua, The effectiveness of green roofs in reducing building energy
Standardization: (Geneva, Switzerland). consumptions across different climates, A summary of literature results, Renew.
[29] S.R. Gaffin, C. Rosenzweig, J. Eichenbaum-Pikser, R. Khanbilvardi, T. Susca, Sustain. Energy Rev. 151 (2021) 111523, https://doi.org/10.1016/j.
A Temperature and Seasonal Energy Analysis of Green, White, and Black Roofs, rser.2021.111523.
Columbia University, Center for Climate Systems Research, New York, 2010. [40] C.Y. Jim, Effect of vegetation biomass structure on thermal performance of tropical
Technical Report. green roof, Landsc. Ecol. Eng. 8 (2012) 173–187, https://doi.org/10.1007/s11355-
[30] B. Schindlera, L. Blausteina, A. Vasla, G. Kadasb, M. Seifanc, Cooling effect of 011-0161-4.
Sedum sediforme and annual plants on green roofs in a Mediterranean climate, [41] M. Eksi, D.B. Rowe, I.S. Wichman, J.A. Andresen, Effect of substrate depth,
Urban For. Urban Green. 38 (2019) 392–396, https://doi.org/10.1016/j. vegetation type, and season on green roof thermal properties, Energy Build. 145
ufug.2019.01.020. (2017) 174–187, https://doi.org/10.1016/j.enbuild.2017.04.017.
[31] A. Solcerova, F. van de Ven, M. Wang, M. Rijsdijk, N. van de Giesen, Do green roofs [42] J. Franzaringa, L. Steffana, W. Ansel, R. Walker, A. Fangmeier, Water retention,
cool the air? Build. Environ. 111 (2017) 249–255, https://doi.org/10.1016/j. wash-out, substrate and surface temperatures of extensive green roof mesocosms-
buildenv.2016.10.021. Results from a two year study in SW-Germany, Ecol. Eng. 94 (2016) 503–515,
[32] Y. Hea, H. Yua, M. Zhaoa, Thermal performance study of extensive green roof in https://doi.org/10.1016/j.ecoleng.2016.06.021.
Shanghai district: a case study of lightweight building in winter, 9th international