You are on page 1of 23

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/282834336

Roof ponds as passive heating and cooling systems: A systematic review

Article  in  Applied Energy · December 2015


DOI: 10.1016/j.apenergy.2015.09.061

CITATIONS READS

22 1,922

2 authors:

Ayyoob Sharifi Yoshiki Yamagata


Hiroshima University National Institute for Environmental Studies
36 PUBLICATIONS   1,479 CITATIONS    224 PUBLICATIONS   3,692 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

coordinating lead author of Climate Change and Cities, Cambridge University Press, (in press) View project

Forest Carbon Monitoring System (FCMS) View project

All content following this page was uploaded by Ayyoob Sharifi on 31 March 2018.

The user has requested enhancement of the downloaded file.


Applied Energy 160 (2015) 336–357

Contents lists available at ScienceDirect

Applied Energy
journal homepage: www.elsevier.com/locate/apenergy

Review

Roof ponds as passive heating and cooling systems: A systematic review


Ayyoob Sharifi a,⇑, Yoshiki Yamagata b
a
Global Carbon Project – Tsukuba International Office, National Institute for Environmental Studies, 16-2 Onogawa, Tsukuba, Ibaraki Prefecture 305-8506, Japan
b
Center for Global Environmental Research, National Institute for Environmental Studies, 16-2, Onogawa, Tsukuba, Ibaraki Pref. 305-8506, Japan

h i g h l i g h t s

 Literature on roof pond cooling and heating systems has been reviewed.
 Nineteen roof pond cooling systems and 4 roof pond heating systems are identified.
 Some roof pond variants can provide year-round thermal comfort.
 Climatic conditions, pond depth, and roof material affect performance of roof ponds.
 Major research gaps are identified.

a r t i c l e i n f o a b s t r a c t

Article history: This paper systematically reviews literature on passive heating and cooling of buildings using roof ponds.
Received 16 June 2015 The main aims were to gain a detailed understanding of different roof pond configurations and their
Received in revised form 29 August 2015 performance, compare effectiveness of different roof pond variants, evaluate performance of different
Accepted 12 September 2015
roof pond variants relative to other passive design techniques, explore effects of climatic conditions
and various design configurations on the performance of roof ponds, and identify gaps in knowledge
and data. Overall, 19 roof pond cooling and 4 roof pond heating systems were identified. This review
Keywords:
suggests that, in some cases, roof ponds can provide year-round thermal comfort while reducing demand
Roof pond
Passive heating and cooling
for active heating and cooling systems. Therefore, they can be utilized in efforts towards the goal of
Passive design carbon–neutral design. Roof ponds with wet gunny bags, shaded roof ponds, ventilated roof ponds,
Energy saving and roof ponds with movable insulation proved to be more effective relative to other variants of roof
Carbon-neutral design pond cooling systems. Comparing performance relative to other passive strategies showed that, in many
cases, roof pond cooling systems are about equally effective in maintaining indoor thermal comfort.
Meteorological conditions, water depth, roof deck material, and thickness of the insulating panel
are the main factors affecting performance of roof ponds. Several areas of weakness in the literature
are identified. Future research should, among other things, provide more detailed knowledge on
performance of all roof pond variants, further explore suitability of roof ponds relative to other passive
design measures, examine effectiveness of roof ponds when combined with other passive design
measures, analyze life-cycle costs of roof ponds, and provide more knowledge on their real-world
application.
Ó 2015 Elsevier Ltd. All rights reserved.

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 337
2. Methods and materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 337
3. Basic principles of passive cooling and heating using roof ponds. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 338
4. Overview of the reviewed studies. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 338
4.1. Geographic distribution and climatic conditions of the studied areas. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 338
4.2. Methods of measuring thermal effectiveness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 338
5. Roof pond variants and their performance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 338

⇑ Corresponding author. Tel.: +81 29 850 2672, mobile: +81 90 6614 2920; fax: +81 29 850 2960.
E-mail addresses: sharifi.ayyoob@nies.go.jp, sharifigeomatic@gmail.com (A. Sharifi).

http://dx.doi.org/10.1016/j.apenergy.2015.09.061
0306-2619/Ó 2015 Elsevier Ltd. All rights reserved.
A. Sharifi, Y. Yamagata / Applied Energy 160 (2015) 336–357 337

5.1.Different variants of roof pond cooling systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 338


5.1.1. Open roof ponds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 338
5.1.2. Roof ponds with movable insulation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 339
5.1.3. Roof ponds with floating insulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 341
5.1.4. Walkable roof ponds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 343
5.1.5. Roof ponds with gunny bags . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 343
5.1.6. Shaded roof ponds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 345
5.1.7. Ventilated roof ponds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 345
5.1.8. Closed roof ponds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 346
5.2. Different variants of roof pond heating systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 347
5.2.1. Skytherm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 347
5.2.2. Roof-integrated water solar collector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 348
5.2.3. Walkable pond with water spraying during daytime hours . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 348
5.2.4. Open roof pond without sprays . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 349
6. Comparing performance of different variants of roof ponds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 349
7. Comparing performance of roof ponds and other passive techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 350
8. Factors affecting performance of roof ponds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 352
8.1. Effects of meteorological conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 352
8.2. Effects of water depth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 352
8.3. Effects of roof deck material and thickness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 353
8.4. Effects of thickness and emissivity of the roof pond cover . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 353
9. Summary and conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 354
Appendix A. Supplementary material . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 356
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 356

1. Introduction been published on this subject [18–20]. The review by Tiwari,


Kumar [20] was published in 1982 when literature on roof ponds
Buildings account for about 35–40% of final global energy con- was still scarce. Two more recent reviews have been conducted
sumption [1], and about 35% of world’s total GHG emissions by Givoni [18], and Spanaki, Tsoutsos [19]. The first paper reported
[2,3]. This significant share indicates that substantial reduction in on studies investigating potential of various passive cooling sys-
building energy consumption is required for climate stabilization tems, including radiant cooling, and indirect evaporative cooling
below a global average of 2 °C above pre-industrial conditions. by roof ponds for reducing indoor temperature. Findings from sev-
Over 90% of people’s lifetime is spent inside buildings and, not sur- eral studies focused on different variants of roof ponds (i.e. open,
prisingly, the largest share of energy consumed in buildings can be with movable insulation, ventilated, Cool-roof, and shaded) have
attributed to space heating, cooling, and ventilation provided by been presented in this work. The review by Spanaki, Tsoutsos
active and energy-intensive mechanisms such as air conditioning [19] provides a relatively comprehensive account of roof ponds
[4–6]. Shift to passive and less energy-intensive techniques is, designed for the purpose of passive cooling of buildings. After cat-
therefore, necessary for achieving the target of near-zero GHG egorizing different variants of roof ponds, they have compared per-
emission buildings. Passive techniques, once implemented, can formance of several roof pond variants and finally made some
provide indoor thermal comfort without, or with minimal, use of recommendations on choosing the desired roof pond system. This
electrical or mechanical means [7,8]. It should be noted that some review builds on previous research by providing statistical data on
of the techniques explained in this paper require using equipment the reviewed literature, presenting a more comprehensive account
such as fans, pumps, and mechanical control for their enhanced of different roof pond configurations and their performances,
operation. However, evidence shows that the amount of energy investigating both cooling and heating performances of roof ponds,
required for these purposes is very small and, therefore, these sys- providing more information on comparative performance of differ-
tems can be regarded as passive [9]. ent roof pond variants, comparing performance of roof ponds with
Among different elements of building envelope, roof is believed that of other passive measures, exploring effects of climatic condi-
to be the most important when it comes to developing passive tions and various design configurations, and identifying gaps in
measures [10]. This is because it is the most exposed part of a knowledge and data.
building to direct solar radiation and there is enough evidence
indicating that roof alone can be responsible for up to about 50%
of heat load in single or two story buildings during summer 2. Methods and materials
[11–15]. Shading the roof, increasing roof thickness, enhancing
albedo of roof, insulating the roof and providing false ceiling, veg- This is a desktop research that included content analysis of lit-
etating the roof, spraying and flowing water over the roof, and pro- erature related to roof ponds. Literature review was conducted
vision of roof ponds are several passive measures for regulating using the method outlined in Pullin and Stewart [21]. The specific
heat gain through the roof [10,16]. This study is focused on the lat- review questions to be addressed were: ‘‘what are the major roof
ter type. Using water as an ideal thermal mass (due to its large vol- pond variants studied in the literature?”, ‘‘what are the basic statis-
umetric heat capacity and the fact that it is cheap and nontoxic tics related to roof ponds examined in the reviewed studies?
[17]) roof ponds are capable of providing passive heating and cool- (Typology, configuration, geographic distribution, etc.)”, ‘‘how
ing. Since the invention of roof pond system by Harold Hay and his effective are different roof pond variants in providing passive heat-
colleagues in late 1960s [18,19], a vast body of work has been pub- ing and cooling?”, ‘‘how do roof ponds perform relative to other
lished on design and performance of different types of roof ponds. passive measures?”, and ‘‘what are the major research gaps that
The broad aim of this systematic review is to synthesize this scien- need to be addressed?”. Overall, over 80 studies were reviewed
tific literature. Over the past 4 decades, three review papers have for the purpose of this study (see the Online Supplementary
338 A. Sharifi, Y. Yamagata / Applied Energy 160 (2015) 336–357

Appendix for details on the search protocol). While mainly peer- plastic bags [39]. Since high levels of air humidity reduce the
reviewed journal articles, a few number of books, patents, and con- potential for evaporative [24] and radiative cooling [25], it is sug-
ference papers were also reviewed. Extracted data were recorded gested that wet and open roofs should be applied in dry climates.
on Microsoft Excel spreadsheets. These spreadsheets were later Since no evaporative cooling is involved in the operation of dry
analyzed and synthesized to address the review questions. roof ponds, their performance is less sensitive to air humidity. Roof
pond cooling systems reviewed in this paper have mainly focused
3. Basic principles of passive cooling and heating using roof on arid and warm-temperate climates. As for roof ponds designed
ponds for heating purposes, Table 3 indicates that they have mainly been
tested in temperate climates.
Roof ponds provide cooling benefits through indirect evapora-
tive cooling and/or radiant cooling [19,22–26]. In both processes 4.2. Methods of measuring thermal effectiveness
the roof acts as a heat exchanging element which is cooled by
evaporation on its surface, longwave radiation to the sky1, or both. The most common methods used for measuring effectiveness of
It then functions as a heat sink which absorbs indoor heat and the roof ponds are assessment based on the extent of change in indoor
heat penetrating into the building [18]. Since the ceiling is thermally temperature of the test cell compared to the indoor temperature of
coupled to the roof pond, the interior space is also cooled by radia- the control cell, extent of heat flux entering the building through
tion and convection [18,27,28]. Driving forces behind evaporation the roof with/without water pond, extent of change in indoor tem-
and radiation are respectively, ‘‘difference between vapor pressure perature relative to outdoor temperature, and magnitude of reduc-
at water surface temperature and vapor pressure of surrounding air” tion in the energy load, and cost savings attributable to application
and ‘‘difference between water surface temperature and effective of roof pond systems. The least used methods are extent of change
sky temperature” [19,22]. Since the roof acts as a heat exchanging in the number of cooling degree days, and extent of residents’ per-
element, roof pond cooling does not elevate the indoor moisture ception of indoor thermal comfort. For further details about these
content of the air. Readers interested in more details about the methods and indicators used for assessing thermal effectiveness of
underlying principles of radiation, evaporation, and convection roof ponds refer to Table 1 of the Online Supplementary Appendix.
may refer to Santamouris [17] and La Roche [27].
Roof ponds can also be used for passive heating during winter. 5. Roof pond variants and their performance
During the daytime, the pond is exposed to solar radiation and
heating storage capacity of water is utilized to store solar energy 5.1. Different variants of roof pond cooling systems
in the pond as sensible heat. This stored energy is then transferred
to the ceiling below by conduction, to be finally transmitted to the Overall, 19 different configurations were found in the literature.
indoor space by radiation and some convection [27,29]. As shown in Table 4, these are categorized into seven groups
Thermal functions of roof ponds may vary according to their according to their similarities. In the following subsections design
configurations and the time of day they are used. This and other configuration of the variants are described, followed by example
functions such as thermal insulation are explained in Section 5 (s) of their performance. More results on performance of roof
which is focused on different permutations of roof ponds. ponds can be found in Section 3 of the Online Supplementary
Appendix.
4. Overview of the reviewed studies
5.1.1. Open roof ponds
The overall number of studies providing experimental and/or An open roof pond is usually supported by a flat concrete roof, is
simulation results related to roof ponds was 70. A brief overview always exposed to the ambient environment, and can be with or
of the main characteristics of these studies is presented in Table 1. without sprays [19,39]. Open roof pond without water spraying
Major points to be mentioned are that majority of the studies are (Fig. 2a) is the most simple, and most used, roof pond configuration
related to roof pond cooling systems, only two studies have inves- discussed in this study. Minimum water depth recommended for
tigated application of roof ponds in high-rise buildings, and most this type of roof pond is 30 cm [39]. Depth of open ponds reviewed
studies have reported satisfying agreement between experimental in this study was between 5 and 50 cm, with an average depth of
and simulation results (see Table 2). 15.5 cm [39]. During daytime hours part of the heat gained from
solar irradiation is absorbed and stored into the water before get-
4.1. Geographic distribution and climatic conditions of the studied ting into the room. This heat absorption feature of water delays the
areas indoor peak temperature appearance [62]. Indoor temperature
peak appears in the late afternoon hours when the ambient tem-
Studies reviewed in this paper have been conducted in 17 dif- perature is already cooler making it easier to achieve thermal com-
ferent countries, with India, the US, and Israel having the largest fort using other passive techniques such as natural ventilation.
share (See Fig. 1). More locally specific information on the studied Open roof pond can also provide evaporative cooling benefits. As
areas and their climatic conditions can be found in the .kmz file the heat gain process by the water continues on, water tempera-
available as online Supplementary material to this paper. Table 3 ture increases and a certain amount of it will be evaporated caus-
provides detailed data on the climatic conditions of the areas ing cooling effects [62]. At night, outside ambient air temperature
wherein roof ponds have been developed and/or tested. Roof ponds is cooler than indoor temperature causing heat flux towards out-
can be classified into three broad categories, namely, dry, wet, and doors. Roof pond attenuates this process and prevents rapid heat
open [38] (cited in [39]). In dry roof ponds water is contained in loss [62]. Also, during nighttime hours the pond is cooled through
plastic bags. The roof pond is called wet if the plastic bags are radiant cooling. The following day this cooled pond functions as a
flooded or sprayed. In the open type water is not contained in heat sink that absorbs interior heat and heat from solar insolation
[22].
Open pond with sprays (Fig. 2b) has a structure and thermal
1
Radiant heat loss occurs all the time. However, negative radiant balance is only behavior similar to that of the open pond without sprays. The only
achieved during the nighttime hours. difference is that water is sprayed over the roof the entire day.
A. Sharifi, Y. Yamagata / Applied Energy 160 (2015) 336–357 339

Table 1
Main characteristics of the reviewed papers.

Description F % Description Description F %


Type of study Simulation 30 43 Indication of the amount of Yes 19 27
Experimental 14 20 the reduced load No 51 73
Both 20 29
Review/descriptive 6 8 Building height Low-rise 68 97
Agreement between simulation Yes 19 95 Both low- and high-rise 2 3
and experimental results Not mentioned 1 5 Building usage Residential 16 23
Services (edu, industrial, commercial) 5 7
Testing season Summer 50 71 Agricultural/GH 2 3
Winter 9 13 Test room 29 41
Both 11 16 Not mentioned (review etc.) 18 26
Concerned load Cooling 53 76 Consideration of building Yes 14 20
Heating 7 10 occupancy No 56 80
Both 10 14

Table 2
Degree of agreement between predicted and monitored values.

Type Description Ref.


Correlation coefficients between monitored Correlation coefficients between monitored and predicted data ranged 0.85–0.96 [30]
and predicted values The coefficient of correlation between predicted and experimental data ranging [13]
from 0.94 to 0.98 and standard error 1.76 and 2.80
The correlation coefficient was 0.88 [31]
Correlation between predicted and experimental data and standard error [32]
values of 0.94 and 1.17, respectively
Maximum deviation between monitored Maximum deviation of 4.05% was observed [33]
and predicted values (%) Maximum deviation of 7% was observed [34]
Maximum deviation of 5% was observed [35]
Maximum deviation between monitored The deviation ranges between ±1.5 °C [36]
and predicted values (°C) Agreement of better than +1.5 °C was observed [37]

Fig. 1. Geographic distribution of countries where studies have been conducted.

Spraying water enhances heat absorption and evaporative cooling. roof pond without spray under the warm-temperate climatic
Yannas and Erell [39] recommend droplet radius between 0.05 and conditions of Stellenbosch, South Africa. Indoor peak temperature
0.1 cm, spray water flow rate of 1.0–1.5 Vol/h, and spray height of of the test cell was reduced from 33.97 to 28.41 °C after it was
at least 50 cm. An important disadvantage of this system is its equipped with roof pond. This resulted in 51.07% and 40.62%
relatively high water consumption. According to Yannas and reductions in the peak cooling load of the room and total heat
Erell [39], although water consumption can be reduced by energy transferred to the room, respectively.
operating spraying only at night, maintaining a stable water
temperature in ponds with depth less than 30 cm requires 5.1.2. Roof ponds with movable insulation
continuous spraying. These ponds are placed atop rectangular roofs and covered with
Open roof ponds have been effective in providing thermal com- movable insulation panels made of different materials such as
fort in arid and warm-temperate climates. Vorster and Dobson [41] polystyrene [31,35,42], polyurethane [19,63], or Styrofoam [64].
conducted a simulation study to examine effectiveness of open Corrugated metal roof deck is favorable in the sense that it
340 A. Sharifi, Y. Yamagata / Applied Energy 160 (2015) 336–357

Table 3
Climatic conditions of the locations where the studies have been conducted (according to Köppen Geiger climate classification system [40]).

Cooling Heating Both Total


No. % No. % No. % No %
Equatorial climates 10 13 – – – – 10 13
Aw 8 10 – – – – 8 10
As 2 3 – – – – 2 3
Arid climates 24 32 2 3 4 5 30 40
BWh 8 10 1 1 3 4 12 16
BWk 2 3 – – – – 2 3
BSh 12 16 – – – – 12 16
BSk 2 3 1 1 1 1 4 5
Warm-temperate climates 21 27 6 8 8 11 35 46
Csa 4 5 – – 1 1 5 7
Csb 3 4 2 3 2 3 7 9
Cwa 13 17 3 4 2 3 18 23
Cfa 1 1 1 1 3 4 5 7
Snow climates – – 1 1 – – 1 1
Dfa – – 1 1 – – 1 1
Total 55 72 9 12 12 16 76 100
Main climates Precipitation Temperature
Köppen Geiger climate classification key
A: equatorial W: desert h: hot-arid
B: arid S: steppe k: cold-arid
C: warm-temperate f: fully humid a: hot summer
D: snow s: summer dry b: warm summer
w: winter dry

Table 4
Different configurations of roof ponds for passive cooling.

Type Frequency No. % Cum. % Climate Examples


Open roof ponds Open pond without sprays 18 22 26 Bsk, Bsh, BWh, Cfa, Csb, Cwa, [19,41–47]
Open pond with sprays 3 4 Csb, Cwa [19,39]
Roof ponds with movable insulation Pond with sprays (continuous spray) 1 1 30 Cs, Cw, Cf [19,39]
(covered daytime) Pond with sprays (nighttime spray) 1 1 Cs, Cw, Cf [39]
Pond without sprays 9 12 Bsh, Bwh, [10,12,19]
Skytherm 13 16 Aw, Bsh, Bsk, BWh, Csa, Csb, Cfa, Cwa, [19,48–51]
Roof ponds with floating insulation Energy-roof 1 1 6 Csa [19,39,52]
Roof pond with floating insulation (Cool-roof) 4 5 Aw, Csb, Bsh [18,53]
Walkable roof ponds Roof pond with insulation embedded within the pond 1 1 2 BWh [19,50]
Walkable roof pond with night water circulation 1 1 BWh [54,55]
Roof ponds with gunny bags Roof pond with wet gunny bags 5 6 7 Bsh, Csa [34]
Shaded roof pond with wet gunny bags 1 1 Bsh [42]
Shaded roof ponds Shaded pond with water enclosed in watertight bags 1 1 8 Csb [56]
Shaded pond with water spray 1 1 Csb [56]
Shaded pond without water spray 3 4 Csb, Cwa [57]
Cool-pool 2 2 Bsk, Cwa, Csa [58–60]
Ventilated roof ponds 12 15 15 Bsh, Bsk, Csa, As, Aw, BSh, BWh [18,33,42]
Closed roof ponds Evapo-reflective roof pond 1 1 5 Bwk [61]
Roof pond with additional insulation layer 3 4 Bsh, Bwh, Bwk, Cfa [42]
Total – 81 100 100 – –

provides better thermal coupling with the interior space and of the ponds ranged between 6 and 25 cm with an average of
enhances nocturnal radiative cooling [19]. The insulation panels 16.5 cm.
are either folding [31] or sliding [63] and are recommended to be Four different permutations of roof pond with movable insula-
made of a highly reflective opaque material that isolates the pond tion were drawn from the literature. These are pond with movable
from solar radiation and prevents the water in the pond from over- insulation and continuous water spray (Fig. 3a and b for day-and
heating during daytime hours [13,19,31,39]. They also help reduc- night-time operation, respectively), pond with movable insulation
ing fluctuation in the pond temperature. and nighttime-only water spray (Fig. 3c and d), pond with movable
The top insulation layer is removed at night to facilitate noctur- insulation and no water spray (Fig. 3e and f), and Skytherm (Fig. 3g
nal radiative cooling [13]. This exposure to the ambient air also and h). Skytherm is different from the other three variants in the
lowers the building temperature by heat loss through evaporation sense that in this system water is enclosed in transparent poly-
and convection [22] . An air gap separates the water surface from ethylene [65] or watertight polyvinylchloride (PVC) bags [48,63].
the covers and functions as an additional insulation layer. Depth Cooled water bags (cooled by nocturnal radiation) of the Skytherm
A. Sharifi, Y. Yamagata / Applied Energy 160 (2015) 336–357 341

Fig. 2. (a) open roof pond without sprays; (b) open roof pond with sprays. (all the illustrations are drawn based on information provided in the reviewed papers, particularly
the following references: [13,19,27,39]). Dominant heat transfer processes are shown by arrows.

transfer ‘‘coolness” into the building via the roof [18]. An optional experiment, through the vapor leakage, this system facilitates both
additional evaporative water layer can be added over the bags to sensible heat absorption and latent heat evaporation.
improve the cooling performance. Simulations for humid climate
of Atlanta showed that this can improve cooling performance of 5.1.3. Roof ponds with floating insulation
the system by up to 100% [49]. This category includes two different roof pond assemblages that
Very few results have been reported on the performance of roof utilize floating insulation panels for regulating thermal exchange
ponds with movable insulation and water spray. Daytime spraying with the exterior environment. Energy-roof (Fig. 4a and b) was
is desirable when water temperature is significantly above ambi- invented by Pittinger [52] in 1976. The system includes a
ent Wet Bulb Temperature (WBT) [39]. Likewise the case of open corrugated metal roof that functions as a waterproof base for the
roof ponds with water spray, droplet radius of 0.5–1.0 mm and shallow primary water chamber. A shallow floating insulation tray,
water flow rate of 1.0–1.5 pond volumes per hour are recom- made of ‘‘cellular foamed plastic or other suitable materials” exists
mended [39]. Simulating cooling performance of roof ponds with above the primary water chamber. This is called the secondary
movable insulation and continuous water spraying, Yannas and water chamber in this article. The secondary water chamber is
Erell [39] showed that, compared with conventional air- covered by a thin and transparent plastic film [52]. A retainer
conditioned buildings with a set-point temperature of 25 °C, cap is placed above the vertical side walls of the system to limit
energy savings of 50–100% can be achieved in different European the maximum upward movement of the secondary floating water
locations. When spraying is limited to night time only, energy chamber. The secondary chamber is connected to the water
savings achieved in different locations would be 30–100%. chamber below through a piping system. A pump is placed at the
Simulation and experimental studies conducted under hot-dry corner of the primary chamber (or external to the chamber) to
conditions of Sde Boker, Israel, showed that indoor temperature transfer a thin layer of water into the secondary water chamber
of test room featuring ‘‘roof pond with movable insulation and at summer nights and cool the system through nocturnal long-
no water spraying” is about 6 °C lower than that of the control wave radiation. The next day, water contained in the secondary
room with bare roof [10]. chamber is drained, through valves, into the primary chamber
Study by Givoni [18], under hot-dry conditions of Atascasdero, and provides indoor cooling by free convection and heat absorp-
showed that indoor temperature of a building with Skytherm is tion from the space below. The floating insulation helps to main-
about 6 and 8 °C lower than that of a conventional building in July tain indoor thermal comfort temperature by reducing heat gain
and October, respectively. Higher difference in October is from solar radiation [52]. Further cooling can be achieved by spray-
explained by the fact that nights, when cooling process takes place, ing water over the upper surface of the transparent plastic film.
are longer in that month [18]. It was found that, when the Equipment including a spray head, a valve, and sensors are incor-
maximum ambient outdoor temperature was about 34 °C, porated into the system to spray water over the surface according
maximum indoor temperature of the room with Skytherm was to the outside air temperature and incoming solar radiation.
21 °C. Skytherm was capable of maintaining indoor temperature Cool-roof (Fig. 4c and d) was patented by Richard Bourne and
within the range of 20–23.5 °C [18]. David Springer in 1992. The pond is placed above a concrete roof
A slightly different variant of Skytherm system was introduced deck covered with a waterproof membrane of materials such as
and tested in an experiment by Ahmad [64] under climatic condi- PVC. Recommended depth of water contained in the pond is
tions of Baghdad, Iraq. Results showed that a Skytherm (with vapor 9 cm [53]. An 8 cm thick insulating panel, made of materials such
leakage when covered) is effective in stabilizing daily indoor tem- as polystyrene, floats above the water and covers the water cham-
perature at around 25 °C. The maximum outdoor temperature was ber pond from edge-to-edge. Bourne and Springer [53] recommend
about 44 °C during the experimental period. The experiment that the exterior surface of the insulating panel should be white to
showed that leakage of vapor from the roof pond, during daytime reflect solar radiation. A pump, with inlet in the pond and outlet
hours, enhances the performance of the system. According to this facing the external environment, sprays water onto the exterior
342 A. Sharifi, Y. Yamagata / Applied Energy 160 (2015) 336–357

Fig. 3. (a and b) Roof pond with movable insulation and continuous water spray; (c and d) roof pond with movable insulation and nighttime water spray; (e and f) roof pond
with movable insulation, without water spray; (g and h) Skytherm.

surface of the floating insulation when desired. Drains are included temperature of the experimental cell was always about 8 °C lower
in the insulation panel in order to transfer the cooled water back to than that of the control cell. It is argued that this system can pro-
the pond during daytime hours and provide indoor thermal com- vide favorable results even in lightweight buildings. An illustration
fort through heat exchange with the interior space. of the system developed by La Roche [27] is shown in Fig. 1 of the
La Roche [27] reports on the performance of an Energy-roof that Online Supplementary Appendix.
was developed and tested at the Lyle Center for Regenerative Under hot-humid conditions of Veracruz, Mexico, Garcia and
Studies at California State Polytechnic University, Pomona Givoni [66] conducted an experiment to investigate the cooling
(warm-temperate climate). Results showed that maximum indoor potential of a Cool-roof at night. A test cell equipped with
A. Sharifi, Y. Yamagata / Applied Energy 160 (2015) 336–357 343

Fig. 4. (a and b) Energy-roof; (c and d) Cool-roof.

Cool-roof was compared with a control cell with a conventional this top layer to ensure that the water level does not exceed about
roof. It was found that average, maximum and minimum indoor 3 cm above the insulation panels. The lowest water container
temperatures of the test cell are lower than the corresponding includes drains that are used to remove water from the pond in
temperatures of the control cell. Average indoor temperature of winter [50].
the control cell was 1.1 °C higher than that of the test cell. Walkable roof pond with night water circulation is another type
However, these values were always higher than the corresponding of walkable roof ponds (Fig. 5c and d). This has also been developed
values of the water in the roof pond (2.3 °C higher for the by Givoni and its configuration is similar to that of the previous
maximum temperature). This is explained by the fact that heat system. Unlike roof pond with insulation embedded within the
gain through uninsulated walls is more than the cooling achieved pond, this system has only one layer of pebbles that fills the water
through the roof pond [66]. Based on this, the authors argue that pond. An insulation layer is placed above the levelled layer of
the best cooling potential of this roof pond configuration can be pebbles. Thin tiles are placed over the insulation, making the roof
achieved if the walls are insulated. suitable for normal use. During night hours water is circulated over
the tiles to enhance evaporation [54,67].
Conducting an experiment under warm-temperate conditions
5.1.4. Walkable roof ponds of Beit She’an, Israel, Givoni [50] found that, during the 17-day
These types of roof ponds allow walking over the roof surface experimental period, indoor maximum temperature of the room
without disturbing the functionality of the roof pond system cooled by ‘‘roof pond with embedded insulation” was always about
below. Two different variants are classified under this category, 2 °C lower than that of the control room with conventional roof.
namely, ‘‘roof pond with insulation embedded within the pond”, With outdoor maximum air temperature of about 38 °C, indoor
and ‘‘walkable roof pond with night water circulation”. Both vari- maximum temperature was always about 30 °C.
ants are applied to buildings with flat roofs made of reinforced Al-Hemiddi [55], cited in Givoni [54] tested performance of
concrete. The first variant (Fig. 5a and b) was developed and tested walkable roof pond with night water circulation in hot-arid condi-
by Givoni in Beit She’an, Israel [50]. The proposed system is com- tions of Riyadh, Saudi Arabia. When the outdoor temperature was
posed of seven layers. The pond is supported by a slightly tilted in a range between 28 and 42 °C, the indoor temperature of the
reinforced concrete flat roof capable of carrying a load of up to experimental room was fluctuating between 27 and 29 °C and its
400 kg/m2 . The slanted roof is required to drain winter rainwater. average was about 28 °C. This was about 7 °C lower than the
A water-proof membrane layer covers this supporting roof. The average indoor temperature of the control room with bare roof.
third layer is the lower water container, about 2 cm thick and com-
posed of pebbles. Five-to ten-centimeter thick insulation panels,
made of material such as polystyrene and with gaps allowing 5.1.5. Roof ponds with gunny bags
‘‘thermosyphonic circulation”, are placed above this first layer of Roof pond with wet gunny bags (RPWGB) and shaded roof pond
pebbles. These insulation panels act as a dividing layer between with wet gunny bags (shaded RPWGB) are two roof pond permuta-
the lower water container and the upper one, which is again com- tions to be discussed here (see Fig. 6). An insulated roof pond, with
posed of pebbles and about 12 cm thick. A float is incorporated into materials such as polystyrene panels, is placed over the roof and
344 A. Sharifi, Y. Yamagata / Applied Energy 160 (2015) 336–357

Fig. 5. (a and b) Walkable roof pond with insulation embedded within the pond; (c and d) walkable roof pond with night water circulation.

Fig. 6. (a and b) RPWGB; (c and d) Shaded RPWGB.


A. Sharifi, Y. Yamagata / Applied Energy 160 (2015) 336–357 345

filled with water up to a certain level (average water depth in the a cooler indoor environment. In one of the configurations spaying
reviewed papers was 15 cm). The roof can be made of either con- was applied between 8 pm and 6 am to further enhance evapora-
crete or metal. Gunny bags or other fabrics are placed over a light, tive cooling. Results showed that shaded pond (with exposed
water-resistant grid such as a perforated PVC plate. This grid is water surface) performs better than the shaded pond with water
supported by polystyrene strips, or other floatable materials, that enclosed in watertight bags. Among the three variants of shaded
are attached to its underneath side and keep it afloat pond (with exposed water surface), the one with horizontal spray-
[10,19,34,42]. The pond and the soaked gunny bags not only inter- ing performed slightly better.
cept solar radiation, but also dissipate part of it through evapora- Sodha and Singh [59] examined performance of a Cool-pool sys-
tion, convection and nocturnal radiation [10,42]. Shaded RPWGB tem under hot climatic conditions of New Delhi, India. When max-
(Fig. 6c and d) is different in the sense that a shading insulation imum indoor temperature of the control room (without any
panel (e.g. polystyrene) is placed in a certain distance above the treatment) was about 40 °C, maximum indoor temperature of the
gunny bags so that air can flow freely between the shading panel room with Cool-pool was about 35.4 °C. Average indoor tempera-
and the surface of gunny bags. tures achieved in buildings with light and heavy roofs were about
Performance of RPWGB has been explored in several studies 33.3 and 33.7 °C, respectively. The similar performance of heavy
[10,34,42,68]. The lower the temperature of the pond bottom, the and light roofs is explained by the large thermal storage of the
more effective the system will be in reducing indoor cooling walls of the building.
energy demand [34]. Simulations under warm-temperate condi-
tions of Chania, Crete, Greece, showed that thickness of gunny bags
does not impact the performance of the system [34]. Performance 5.1.7. Ventilated roof ponds
of the roof pond is, however, effected by the optical properties of This variant consists of a roof pond placed over an either con-
the gunny bags. Study by Spanaki and Kolokotsa [34] showed that crete or metallic roof slab (Fig. 8). [6,9]. Average water depth of
50% reduction of emissivity (from 0.93 to 0.47) drops the pond bot- ponds investigated in the reviewed papers was 9 cm. The pond is
tom temperature by 2.38 °C [34]. In a simulation study conducted covered by a permanent insulation layer separated from the water
under hot-arid conditions of Sde Boker, Israel, Tang and Etzion [10] surface by a ventilated layer of air. There should be at least a 0.3 m
found that compared with a reference building with conventional gap above the water surface to optimize air flow [24]. The cover is
dry roof, interior space of a test building with RPWGB was about usually made of highly reflective metallic sheets that are insulated
7 °C cooler. Under hot-arid climatic conditions of Sde Boker, with a thin layer of materials such as EPS, Styrofoam, and polystyr-
Shaded RPWGB has been tested by Runsheng and Etzion [42]. ene [9,30,69]. Although wind can flow freely in the space between
Results of their study showed that this assemblage performs even water level and insulation panel, the system can be supplemented
better than RPWGB. This will be further explained in Section 6. with fan(s) installed on one or both sides (inlet and outlet vents)
that operate intermittently to reduce water temperature and
5.1.6. Shaded roof ponds enhance evaporation [6,69]. As the system is covered, cooling is
Roof pond variants classified under this category are covered by only achieved through convection and evaporation [17,70]. Water
louver plates, separated from the water surface by a ventilated air contained in the roof is drained in winter and the vents are closed
layer, that intercept solar radiation. These louvers are permanently to constitute an additional insulation layer inhibiting excessive
integrated into the roof pond structure and can be either fixed or heat loss [18,19,33,69]. Main advantage of this system is that it
operable. The operable ones are closed during daytime hours to can also be applied to non-flat roofs [71].
minimize solar heat gain and opened at night to provide evapora- Results of various studies indicate that ventilated roof pond is
tive and radiative cooling. If fixed, the louver plates are placed in a effective in providing thermal comfort by lowering indoor temper-
position that allow some nocturnal radiation while blocking diur- ature and reducing the amplitude of indoor temperature oscillation
nal direct solar radiation. Water in the pond can either be exposed in both dry and humid climates. Experimental and simulation
or enclosed in watertight bags. In the latter case nighttime cooling study conducted by Spanaki and Kolokotsa [33], under warm-
is only achieved through radiation [27,56]. Shaded pond either has temperate conditions of Crete, Greece, showed that, compared to
only a flat water pond over the roof (Fig. 7a through f) or else the bare roof, ventilated roof pond was capable of reducing maximum
flat water pond may be coupled to a vertical water column con- indoor temperature and variation in daily indoor temperature fluc-
nected at one end with the flat water pond and at the other end tuation by approximately 30% during the 32-day period (4.8–8.2 °C
with the space below (Fig. 7g and h) [19,59,60]. The latter structure and 1.6–3.1 °C for bare roof and roof pond, respectively). Results of
is called Cool-pool and, in addition to the above mentioned ther- a simulation study conducted for 411 Brazilian location featuring
mal processes, provides cooling benefits through further heat different climates showed that in all cases, the system is capable
exchange with the indoor environment. Compared with other vari- of keeping indoor conditions within the thermal comfort range.
ants of shaded roof ponds, Cool-pool has higher maintenance costs However, better performance was achieved in the arid locations
[19]. However, its water consumption is relatively low [59]. [30].
Under warm-temperate conditions of New Delhi, India, Yadav Performance of fan-assisted ventilated roof pond was also
and Rao [57] conducted a simulation study to investigate thermal examined in hot-arid climate of Baghdad, Iraq. Experimental
effectiveness of shaded roof ponds without water spray results showed that when the mean outside temperature is over
(Fig. 7e and f). Results showed that for a hot day in May when 45 °C, the system provides an average indoor temperature of
the maximum ambient temperature exceeded 40 °C, maximum 31.8 °C. It is argued that with a fan air movement of around
indoor temperatures of rooms with shaded roof ponds on concrete 0.5 m/s and residents adjusting their behavior by wearing appro-
deck and metal deck were about 30 and 28 °C, respectively. Under priate outfit, this level of temperature can be regarded as satisfy-
warm-temperate conditions of Southern California, Marnich, LaR- ing. The system was also proved effective in delaying occurrence
oche [56] conducted experiments to compare cooling effectiveness of indoor peak temperature. While the maximum outdoor temper-
of three different configurations of shaded ponds, namely shaded ature reached 51.9 °C in the afternoon, the maximum indoor tem-
pond with water sprays (Fig. 7c and d); shaded pond without water perature did not exceed 33.4 °C and this was reached later than the
spray, and shaded pond with water enclosed in watertight bags outdoor peak time. Another important achievement was reduction
(Fig. 7a and b). Comparing performances of these shaded pond in the indoor temperature fluctuation from 7 °C to a maximum
variants to that of a control cell, they found that all of them provide value of 2.3 °C [69]. For thermostat setting of 26 °C, it was found
346 A. Sharifi, Y. Yamagata / Applied Energy 160 (2015) 336–357

Fig. 7. (a and b) Shaded pond with water enclosed in watertight bags; (c and d) shaded pond with water spray; (e and f): shaded pond without water spray; (g and h) Cool-
pool.

that this system, coupled with wall cooling, can reduce annual developed and tested by Bencheikh [61,72]. This roof pond is
cooling load from 2133 to 256 kW [69]. supported by a concrete deck over which a water pool, filled with
small pieces of rock, is placed (Fig. 9a and b). An unventilated air
5.1.8. Closed roof ponds layer separates water pool from the upper roof which is a flat
These roof ponds are always covered by an insulation layer aluminum plate ‘‘painted with titanium-based pigment” to
(Fig. 9). One type is the ‘‘evapo-reflective roof pond” that was enhance its reflective properties. The system reflects incoming
A. Sharifi, Y. Yamagata / Applied Energy 160 (2015) 336–357 347

Fig. 8. Ventilated roof pond.

Fig. 9. (a and b) Evapo-reflective roof pond; (c and d: roof pond with additional insulation layer.

solar radiation during daytime hours. At night, temperature of the of combining roof pond cooling with other passive measures such
water pool is higher than that of the aluminum plate. As this is a as nocturnal natural ventilation.
closed roof pond, the water vapor does not leave the system. The
vaporized water condenses and falls back into the water pool. This
5.2. Different variants of roof pond heating systems
way, heat is transferred outside the system. Radiation between the
two humid internal surfaces of the system enhances heat exchange
Different configurations of roof pond heating systems are
[61].
shown in Table 5. Comparing this table with Table 4 shows that
Roof pond with additional insulation layer was introduced by
relatively fewer studies have been conducted on roof ponds for
Runsheng and Etzion [42] (Fig. 9c and d). In their experiment the
passive heating.
pond was covered by a 2.75 cm polystyrene panel. Authors argue
that no evaporative cooling is achieved by this system. However,
the pond acts as an additional insulation layer on the roof [62]. 5.2.1. Skytherm
Experiments conducted by Bencheikh and Bouchair [61] In the Design configuration of this roof pond variant is identical to
hot-arid climate of Laghouat, Algeria, showed that evapo- that of the Skytherm for cooling purpose. The only difference is
reflective roof pond can reduce and delay indoor heat peak arrival. that the insulation panels are opened during daytime to absorb
Reduction values ranged between 6 and 10 K. Applying natural solar heat and closed at night to prevent heat loss. This is the most
ventilation at night (8 pm–9 am) provided 2–3 °C further reduc- commonly practiced roof pond heating system. Some of the
tions in the indoor temperature [72]. This indicates importance studies mentioned here and in the corresponding section of
348 A. Sharifi, Y. Yamagata / Applied Energy 160 (2015) 336–357

Table 5
Different configurations of roof ponds for passive heating.

Type No. % Climate Examples


Skytherm 10 62.5 Bsk, BWh, Cwa, Csa, Csb, Dfa [15,29,48,49,63,65,73–76]
Roof-integrated water solar collector 2 12.5 Cfa, Csb [77,78]
Walkable roof pond (with day water circulation) 1 6.25 BWh [36]
Open roof pond without sprays 3 18.75 Cfa, Cwa [29,79,80]
Total 16 100 – –

Online Supplementary Appendix have used slightly different


configurations. However, they can still be categorized as Skytherm.
As a pioneering study, Hay and Yellott [74] developed Skytherm
as a system to be used in the arid climate of Phoenix, Arizona.
Results of their experiments showed that when the outdoor ambi-
ent temperature was within the range of 5 and 21 °C (with an aver-
age of 11 °C), the indoor temperature in the room covered with
Skytherm was between 21 and 27 °C and the average was 24 °C.
It is argued that this system is capable of providing comfort condi-
tions for non-freezing conditions. Under freezing conditions, sup-
plementary heating would be required. Raeissi and Taheri [48]
compared winter performances of a bare roof and two types of Sky-
therm (metal and concrete deck) under the arid climatic conditions
of Shiraz, Iran. Both variants were able to significantly reduce heat-
ing load in winter. Reductions of heating load, relative to the bare
roof, were 48.6% and 86.2% for concrete and metal Skytherm,
respectively.
Under freezing conditions, conventional Skytherm may not be Fig. 10. Roof-integrated water solar collector [78].
capable of providing sufficient heating. Creating a non-convective
zone in the water pond improves the performance of the system.
This can be achieved by using viscosity additives such as natural rial such as polyethylene that is less expensive, more elastic, and
gums and polymers that create a 10–30 cm thick clear layer that has better mechanical properties. During winter days the system
has the same transmission characteristics as water. Simulation is uncovered and the water chamber below the lower glass layer
study conducted in Boulder, Colorado, where ambient air temper- is filled with water. This system collects solar energy and can warm
ature was always near or below zero, showed that a non- up indoor space by radiation. At winter nights the awning is
convective Skytherm (including a non-convective water zone, a extended over the system and the water chamber is drained. The
convective water zone, and a concrete deck) was effective in water that is heated during daytime hours is sent down to the stor-
always keeping the water at the bottom of the pond above 16 °C. age tank and recirculated through the standard heat water system.
[73]. The three air chambers established above the roof provide night-
A slightly different variant of Skytherm has been proposed by time thermal insulation. It is argued that this system performs bet-
Tavana and Kammerud [49]. What makes this system different is ter than conventional Skytherm because it can adjust the amount
an air gap and a movable insulation below the steel support deck. of thermal exchange with the indoor environment by modifying
Likewise the regular Skytherm, during heating season the top insu- the volume of water pumped into the roof. Among other advan-
lation is closed at night to prevent heat loss. When conditions are tages of this variant is its ability to alleviate the mismatch between
more severe, the bottom insulation can also be closed to further indoor thermal demand and roof pond maximum heating capacity
reduce the heat loss [49]. In terms of cooling and heating potentials [77]. Juanicó has introduced a new configuration of this system
of the system, simulations showed that it is capable of fulfilling all suitable to be installed above inclined roofs. Simulations showed
the heating needs of a typical residential building in Phoenix, Ari- that installing such a system on a roof with a 60 degree slope
zona. It was, however, not able of meeting the needs of a typical can collect 225 kW h energy, three times as much as the horizontal
residential building is Sacramento, Atlanta, and Washington DC roof. This is enough to meet the heating demand of a low-quality
[49]. house in Bariloche (temperate climate). Savings achievable by this
system are significant. This equals annual amount of 900 m3 of nat-
5.2.2. Roof-integrated water solar collector ural gas for both inclined and horizontal roof in Buenos Aires, and
This system has a structure similar to that of Skytherm and was 2000 and 6000 m3 for horizontal and inclined roofs, respectively, in
developed by Juanicó in 2006 (Fig. 10) [77]. Although not com- Bariloche [78].
pletely passive, roof-integrated water solar collector can perform
better than Skytherm. It includes a water pond (about 5 cm deep) 5.2.3. Walkable pond with water spraying during daytime hours
that is placed above a water proof metallic base. A first glass layer This variant has a structure identical to that described, in Sec-
covers this pond and provides a water-tight chamber. Above this tion 5.1.4, for walkable pond with night water circulation. How-
layer is an air chamber which is, in turn, covered by another layer ever, in this case water is sprayed over the tiles during the day
of glass. The system also includes a storage tank located near the to absorb heat which is then stored in the water pond below the
house and a hot-water circulating system (pump and boiler) to insulation layer. This stored heat is then transferred to the indoor
drain or fill the water chamber if needed. Finally, there is a rolling environment [36]. Thermal performance of this system was exam-
awning that can be extended or rolled back according to the season ined under cold winter conditions of Riyadh, Saudi Arabia. Experi-
and time of day. The glass layers can be replaced with plastic mate- ments showed that without spraying, temperature of the test room
A. Sharifi, Y. Yamagata / Applied Energy 160 (2015) 336–357 349

was lower than that of the control room (without any treatment) 6. Comparing performance of different variants of roof ponds
and therefore not successful in heating the room. However, when
water was sprayed during sunny hours it was observed that the A considerable number of the reviewed papers (13), have pro-
system was capable of heating the room. For minimum outdoor vided information on the comparative performance of roof pond
dry bulb temperature (DBT) of 11 °C, indoor DBTs of the control variants. Main findings are described here and more information
room and the test room were 18 and 19.5 °C, respectively. Average can be found in Section 5 of the .docx file available as online
indoor temperature of the test room was 2 and 2.5 °C higher than Supplementary material. Table 6 provides information on the
indoor temperature of the control room and the ambient outdoor pairwise comparisons, the purpose of the experiments (heating/
temperature, respectively [36]. The author argues that the system cooling), number of times each pairwise comparison has been
is capable of meeting heating needs when it is not severely cold. To conducted, and related references. The column titled ‘‘P” displays
gain better results, it is recommended that more conductive mate- performance level of variants mentioned under the first column
rials should be used instead of concrete. compared to that of variants mentioned under the second one.
As can be seen, these pairwise comparisons are mainly related
to roof ponds designed for passive cooling purpose. This is
5.2.4. Open roof pond without sprays expected given the limited number of reviewed papers focused
This roof pond variant has a structure identical to the one on roof pond heating systems (See Table 5). Skytherm provides
described in Section 5.1.1. It is always exposed to the ambient better heating benefits than open roof pond [29,76]. Under mild
environment. Evaporative and radiative heat losses undermine winter conditions of Delhi, India, performance of Skytherm was
the effectiveness of this system. Experimental study in New Delhi, compared with that of an open roof pond without water spray.
India showed that it can provide limited heating benefits. When By preventing heat loss at night, Skytherm provided a better per-
the ambient outdoor temperature dropped to 0.5 °C, minimum formance. Indoor air temperature achieved by Skytherm was
temperature observed in the test room with open roof pond was always about 2 °C warmer than that achieved by open roof pond,
14.6 °C. This was 1.3 °C higher and 1.9 °C lower than minimum indicating that nighttime insulation increases heating efficiency
indoor temperatures of the control room and the test room with of the system [29]. Overall, Skytherm offers a reasonably well per-
Skytherm [29]. Using non-convective roof pond can improve ther- formance with moderate water requirements. However this sys-
mal performance of the system. Adding thickeners such as polymer tem needs supplementary mechanical control and has
oil/water gels makes water non-convective. In New Delhi, India, it comparatively high maintenance costs [19].
was found that when depth of the non-convective zone is higher It can be seen from Table 6 that open roof pond only performs
than 6 cm the system is capable of meeting all heating demands better than covered pond with an additional insulation layer [42].
of a test building in a typical winter day [80]. Another option is Study conducted under hot-arid climatic conditions of Sde Boker
to drain the pond water at night. Here water is pumped into an indicates that cooling provided by Shaded RPWGB and RPWGB is
open roof pond placed above a roof with skylight. The warmed comparatively better than that provided by the other variants
water flows into the heat storage unit under the floor and is used [42]. Cooling performance of the RPWGB was similar to that of
for warming the interior space. [79]. the pond with movable insulation and without water spray.

Table 6
Comparative performance of different roof pond variants.

Roof pond type P C/H Clim F Ref.


Open pond without water spray Roof pond with additional insulation layer U C Bsh 1 [42]
Pond with movable insulation and without water spray Roof pond with additional insulation layer U C Bsh 1 [42]
Open pond without water spray U C Bsh, BWh 2 [37,42]
Ventilated roof pond U C Bsh 1 [42]
Pond with movable insulation and nighttime water spray Open pond with water spray U C Csb 1 [39,56]
Pond with movable insulations and continuous spraying  C EU 1 [39]
Skytherm Open roof pond without sprays U C BWh, Cwa 2 [37,57]
Open roof pond without sprays U H Cwa 2 [29,76]
Cool-roof Ventilated roof pond =, P C Bsh 2 [18,70]
RPWGB Open roof pond without sprays U C Bsh 1 [42]
Pond with movable insulation and without water spray =, P,  C Bsh 3 [10,42]
Ventilated roof pond U C Bsh 1 [42]
Roof pond with additional insulation layer U C Bsh 1 [42]
Shaded RPWGB Open roof pond without sprays U C Bsh 1 [42]
Pond with movable insulation and without water spray U C Bsh 1 [42]
Roof pond with additional insulation layer U C Bsh 1 [42]
RPWGB U C Bsh 1 [42]
Ventilated roof pond U C Bsh 1 [42]
Shaded pond without water spray Open pond without water spray U C Cwa 1 [57]
Skytherm U C Cwa 1 [57]
Shaded pond with water enclosed in watertight bags U C Csb 1 [56]
Shaded pond with water spray Shaded pond without water spray P C Csb 1 [56]
Shaded pond with water enclosed in watertight bags U C Csb 1 [56]
Cool-pool Open pond without sprays U C Cwa 1 [59]
Ventilated roof pond Skytherm = C Bsh 1 [70]
Open roof pond without sprays U C Bsh, BWh, Bsk 3 [42,81,82]
Roof pond with additional insulation layer U C 1 [42]

Better: U; Slightly better: P Within 1 °C; Similar: =; Worse: ; Slightly worse: 6 Within 1 °C.
350 A. Sharifi, Y. Yamagata / Applied Energy 160 (2015) 336–357

Although the water temperature near the bottom of RPWGB was making them more desirable for arid areas which are usually water
about 2 °C higher than that of the roof pond with movable insu- stressed.
lation, it facilitated better daytime heat loss from the interior Not much is available on the comparative performance of
building into the pond. However, roof pond with movable insula- shaded roof ponds. Under warm-temperate conditions of New
tion (no spray) may perform better under different building con- Delhi, Yadav and Rao [57] performed simulations to compare the
ditions. In a more recent study conducted in Sde Boker, cooling cooling performance of an open roof pond, a shaded roof pond
performances of RPWGB and roof pond with movable insulation without water spray (metal and concrete deck), and a Skytherm
were compared in three different buildings featuring different (metal and concrete deck). It was found that the first system was
thermal mass, insulation level, and roof material [10]. It was not very effective. The Skytherm and shaded roof pond (both metal
found that the best performance in terms of cooling and temper- deck) were the most effective. When the ambient outside temper-
ature stabilization is achieved in the well-insulated building with ature was over 40 °C, simulated indoor temperatures for shaded
massive walls and a metal-deck roof. Unlike the case of poorly roof pond and Skytherm were about 28 and 26.8 °C, respectively
insulated building, RPWGB performed better and indoor temper- [57]. More information on the comparative performance of differ-
ature achieved by it was always about 0.95 °C lower. When these ent shaded roof pond variants is provided in Section 5.1.6. Shaded
two roof pond variants were applied to a poorly insulated build- roof pond variants (except for the one with water enclosed in
ing with a metal-deck roof, cooling performance was considerably watertight bags) are water intensive. Those using operable louvers
less effective and the building experienced higher indoor temper- also need mechanical operation and have comparatively high
ature swing. In this case roof pond with movable insulation per- maintenance costs.
formed better than RPWGB (indoor temperature was always Only one of the reviewed papers has discussed comparative
about 2.3 °C lower). This can be explained by higher thermal con- performance of Cool-pool system [59]. Simulation study conducted
ductivity of the gunny bags that results in transfer of more solar in warm-temperate climate of New Delhi showed that, compared
heat gain into the pond compared to roof pond with movable to an open roof pond, Cool-pool system provides a better indoor
insulation [10]. thermal comfort. The Cool-pool system provided the maximum
In hot-arid conditions of Sde Boker, RPWGB also performs bet- amount of cooling at 16 h, when cooling produced by the open roof
ter than ventilated roof pond. Only during early night hours water pond was minimum. This shows the importance of covering the
temperature at the bottom of the ventilated roof pond was lower roof during sunshine hours. Average indoor temperatures achieved
than that of the bottom of RPWGB. This is explained by the inverse by the open roof pond, and Cool-pool systems on a light roof were
thermal stratification in the RPWGB during early night hours due 35.3, and 33.3 °C, respectively. Better performance of Cool-pool
to the limited openness of the supporting PVC grid that resists system is because, unlike open roof pond wherein some incident
instant sinking of cooled water into the pond. Although due to this solar insolation is added to the cooling load of the building, in
phenomenon, heat flux (from interior building) into the RPWGB the Cool-pool system almost all the heat loss through convection,
was lower compared to the ventilated roof pond during the period evaporation, and radiation is from the indoor space and additional
from late afternoon to early evening, the overall diurnal heat loss load from solar radiation is negligible. As a result, water require-
was about 20% higher in the RPWGB. This is explained by the fact ments of the Cool-pool are also relatively lower. Water consump-
that fixed shading eliminates long wave nocturnal radiation into tion of the Cool-pool system was almost half that of open roof
the sky, reducing the cooling potential of ventilated roof pond pond. Additional advantage of this system is its suitability for
[42]. Even better performance can be achieved by placing an insu- buildings with limited roof area [59]. Compared to other systems
lation layer within a certain distance (about 20 cm) over RPEGB. Cool-pool has higher mechanical and maintenance requirements.
During daytime hours temperature near the bottom of shaded Further studies on needed on the comparative performance of this
RPWGB was about 4 °C lower than ambient WBT. The total daily system with respect to other effective variants such as RPWGB,
heat loss from building into shaded RPWGB was about 50% higher Skytherm, and ventilated roof pond.
than that into the ventilated roof pond, indicating the significant
improvement achievable by shading RPWGB [42]. Major advan-
tages of RPWGB and shaded RPWGB (over other systems such as 7. Comparing performance of roof ponds and other passive
skytherm, pond with movable insulation, Cool-roof and Energy- techniques
roof) are that they don’t require supplementary mechanical equip-
ment and have low operational and maintenance requirements As Table 7 indicates, such comparisons have been made in 18
[19]. Main drawback is that this system uses a large amount of papers. Overall, six different roof pond variants have been com-
water [12]. pared with other passive cooling and heating measures. Again,
Ventilated roof pond has also proved to be a viable option given the main focus has been on passive cooling techniques. A summary
its performance and cost effectiveness. Cooling performance of of the main results is provided here and more explanations can be
this system (in arid and temperate climates) is comparable to that found in Section 6 of the online Supplementary appendix.
of Skytherm, pond with movable insulation (no spray), and Under humid continental climate of Muncie, Indiana, passive
Cool-roof [18,42,70]. Comparison of the cooling performance heating performance of a Skytherm (placed under a pitched roof)
of ventilated roof pond and Cool-roof, under arid and warm- was compared to that of other passive measures, namely, ‘‘direct
temperate climates, showed that both systems were able to main- gain”, ‘‘Trombe wall”, ‘‘water wall”, and ‘‘Sunspace” [75]. Promising
tain indoor temperature between about 23–24 °C and this was results on the effectiveness of Skytherm were obtained from the
around 3 °C cooler than outside ambient temperature [18,70]. Very experiments. Arguing that less than 10 °C of temperature swing
similar results were found when ventilated roof pond was com- is necessary for providing thermal comfort, the author found that
pared with Skytherm under the same climatic conditions [70]. except for the Sunspace, the other strategies met this criterion.
Much the same as what was said about RPWGB, ventilated roof However, the best performance was gained by Skytherm followed
pond does not need mechanical operation and has minimum by Trombe wall, and water wall. The highest and lowest variations
maintenance requirements [19]. However, since cooling is mainly of operative temperature were for direct gain and Skytherm,
produced by evaporation, it consumes a large amount of water respectively. Minimum and maximum temperature swings for
[70]. Skytherm, Cool-roof, and pond with movable insulation with- these two passive measures were 7.8–10.3 °C and 1.2–1.4 °C,
out water spray consume comparatively less amount of water [70], respectively.
A. Sharifi, Y. Yamagata / Applied Energy 160 (2015) 336–357 351

Table 7
Comparative performance of roof pond variants with respect to other passive measures.

Roof pond type Other passive measure P C/H Clim F Ref.


Open pond without sprays Water spray on the roof =, 6 =, , C Cwa, 7 [20,41,44,47,59,83,84]
6, 6, = Csb
Water flow over the roof P; P; U C Cwa, 3 [20,44,84]
Wetted gunny bags  C Cwa 1 [83]
Night flushing U C Csb 1 [41]
Active mass cooling 6 C Csb 1 [41]
Vegetated pergola + thin water layer over the roof 6 C Cwa, 1 [85]
Vegetated pergola U C Cwa, 1 [85]
Shaded roof (shaded bare roof) U C Bsk 1 [82]
Roof garden = C Cwa, 1 [85]
Inverted earthen pots = C Cwa, 1 [85]
Removable canvas U C Cwa, 1 [85]
Open pond with high-albedo ‘‘Vault + high Albedo”; ‘‘Ventilated vault + albedo + air gap”; ‘‘Flat roof + albedo”; 6 C BWh 1 [67]
roof coating ‘‘Double roof + air gap + albedo”; ‘‘Double roof + albedo”
‘‘Vent vault + albedo”; ‘‘Flat roof + air gap + albedo”; ‘‘Dome + albedo”; ‘‘Ventilated P C BWh 1 [67]
vault + air + insulation”; ‘‘Ventilated vault”; ‘‘Ventilated vault + air gap”; ‘‘Double
roof”; ‘‘Ventilated vault + insulation + albedo”; ‘‘Ventilated vault + insulation”
Roof pond with movable Water spray on the roof U C Bsh 1 [13]
insulation, without water Well-insulated roof U C Bsh, 3 [12,13,86]
spray Pieces of white glazed tiles stuck over the roof  C Bsh 1 [12]
Wetted gunny bags  C Bsh 2 [12,86]
White-painted roof U, = C Bsh, 2 [12,86]
‘‘Sania over the roof”; ‘‘cement–vermiculite thermal insulation over the roof” U C Bsh 1 [12]
Skytherm High-albedo roof coating U C Aw 1 [31]
‘‘Direct gain”; ‘‘Trombe wall”; ‘‘water wall”; ‘‘Sunspace” U H Dfa 1 [75]
Cool-pool Water flow over the roof U C Cwa 1 [59]
RPWGB Wetted gunny bags U C Bsh 1 [68]
Ventilated roof pond High thermal mass building U C Bsh 1 [9]
Shaded roof (shaded bare roof) U C Bsk 1 [82]

Better: U; Slightly better: P Within 1 °C; Similar: =; Worse: ; Slightly worse: 6 Within 1 °C.

Many studies have compared cooling performance of open roof conducted experiments to compare the outside roof temperature
pond with that of other passive techniques. It can be seen from achieved by applying open roof pond and wetted gunny bags (day-
Table 7 that ‘‘open pond without spray”, roof garden”, and time only) systems. Temperature reductions (as compared to the
‘‘inverted earthen pots over the roof” have exhibited similar levels bare roof) were in the order of 23 and 27 °C, respectively. In terms
of performance under arid and temperate conditions. Performance of heat flux through the roof, it was found that roof pond can
of ‘‘water spray on the roof” is similar or slightly better than that of reduce peak heat gains in the order of 85% and wetted gunny bags
‘‘open pond without spray”. There is, however, one study indicating can even reverse heat flux [83].
better performance by the former [41]. Given the similarities For the hot-arid conditions of Al Maadi, Egypt, Dabaieh et al.
between the studies in terms of parameters such as relative [67] performed simulations to investigate thermal performance
humidity, roof deck material, building type, etc., this can be of various combinations of different roof shapes and materials. Per-
explained by the lower depth of the open pond used in that study. formances of 16 different combinations were compared to that of a
As will be explained in Section 8.2, deeper open ponds provide bet- flat bare roof without any treatment (See Fig. 3 of the online
ter performance. Other reason may be differences in water spray Supplementary appendix for different roof shapes used in the
type (continuous/intermittent) and spray flow rate. These informa- study). Results showed that an open pond combined with a
tion is not provided in the concerned studies. Under temperate con- high-albedo roof coating provides one of the best performances
ditions, open roof pond has been found to perform slightly better (ranked fifth after ‘‘vault + albedo”, ‘‘ventilated vault + albedo”,
that ‘‘water flow over the roof”. However, performance by ‘‘active ‘‘flat roof + albedo” and ‘‘ventilated vault + air gap + albedo”.
mass cooling” and a combination of ‘‘vegetated pergola” and ‘‘thin Roof pond with movable insulation, without water spray has
water layer over the roof” has been slightly better than open roof performed better than measures such as ‘‘water spray on the roof”,
pond. Better performance has been achieved compared to ‘‘night and ‘‘well-insulated roof”. However, performances achieved by
flushing”, ‘‘vegetated pergola over the roof”, and ‘‘shaded roof”. measures such as placing ‘‘pieces of white glazed tiles stuck over
In a simulation study conducted under warm-temperate condi- the roof” and covering the roof with ‘‘wetted gunny bags” have
tions of Stellenbosch, Vorster and Dobson [41] compared cooling proved to be better. In a simulation and experimental study com-
performance of an open roof pond with other cooling systems paring cooling performance of two identical test cells with different
namely, roof spray, active mass cooling, and night flushing. It roof configurations in the hot-humid climate of Maracaibo, Vene-
was found that the open roof pond can reduce peak temperature zuela, Rincon and Almao [35] found that, taking 25 °C as the refer-
from 33.97 to 28.41 °C. This resulted in 51.07% and 40.62% reduc- ence comfort temperature, average daily thermal load reductions
tions in the peak cooling load of the room and total heat energy offered by roof pond with movable insulation with respect to the
transferred to the room, respectively. Corresponding values for reference cell (with well-insulated roof) were 41% and 66% in
roof-spray, active-mass cooling, and night flushing techniques August and January, respectively. Under hot-arid conditions of
were, respectively, 59.36% and 72.24%, 51.39% and 58.58%, and Jodhpur, India, Nahar, Sharma [12] conducted experiments to com-
5.50% and 31.67%. This indicates that open roof-pond provides pare a roof pond with movable insulation to other passive measures
better performance than the night flushing technique. It should, applied to the roof namely, Sania over the roof, cement–vermiculite
however, be noted that the former is more costly and requires thermal insulation over the roof, roof painted with white cement,
maintenance. Under hot-dry conditions, Jain and Rao [83] roof provided with air void insulation, pieces of white glazed tiles
352 A. Sharifi, Y. Yamagata / Applied Energy 160 (2015) 336–357

stuck over the roof, and roof covered with wetted gunny bags. It humidity is desirable for efficient functioning of cooling roof pond
was found that only the two latter measures performed better than systems [23,44,59,82,85,88]. For an open roof pond designed for
the roof pond. However, the roof covered with wet gunny bags has cooling purpose in the warm-temperate conditions of New Delhi,
considerable water requirements (50 l/m2 per day) [12]. Sodha and Kumar [47] found that increasing relative humidity
Under hot-humid climatic conditions of Maracaibo, González results in larger amount of heat flux into the room. Simulation
and Givoni [31] compared cooling performance of a reference test results showed that for each 0.1 increase in the humidity rate,
cell with a highly reflective roof to that of a Skytherm. Simulation hourly heat flux into the room was increased by about 8 W/m2 .
and experimental results showed that cooling performance of the Similar results were found in another simulation study investigat-
test cell with Skytherm is better and the indoor temperature pro- ing effects of increasing relative humidity on cooling efficiency of
vided by it is always about 2 °C cooler. Again it should be noted an open roof pond in New Delhi, India [20]. In another simulation
that Skytherm has higher construction and maintenance costs study conducted in New Delhi, increasing relative humidity levels
which have not been considered in the study. deteriorated cooling performance of shaded pond and Skytherm
Cool-pool system has been compared with water flow over the systems. For relative humidity over 50%, both system were more
roof. Results of simulated study conducted under warm-temperate or less ineffective [57]. Based on these results it is argued that roof
conditions of New Delhi, India, showed that it was more effective ponds are mainly suitable for arid climates. However, there are
in reducing indoor room temperature. Average indoor tempera- studies showing that Skytherm [31], Cool-roof [66], and fan-
tures achieved by water film, and Cool-pool systems on a light roof assisted ventilated roof pond [30] can be effective even under
were 35.3, 33.3 °C, respectively. For heavy roofs, the corresponding humid climates. In Skytherm water is enclosed in watertight bags,
values were 35.6, and 33.7 °C, respectively. The similar perfor- reducing the impact of relative humidity. Effectiveness of the other
mance of heavy and light roofs can be explained by the large ther- two systems could be explained by intermittent night-time water
mal storage of the walls of the building [59]. spray over the roof, and enhancement of evaporation by using fans,
As expected, RPWGB has performed better than wetted gunny respectively.
bags over the roof. Due to creation of thermal stratification in the Wind enhances evaporative and also convective cooling [22].
pond, the thermal comfort provided by the RPWGB is always However, it reduces nocturnal radiation to sky [35]. All roof pond
better. Thermal stratification is important because it eliminates cooling systems, expect those where water is enclosed in water-
heat convection from the top of the pond to its bottom and the tight bags, utilize evaporative cooling. Therefore, increased wind
conduction is also very slow because of the poor heat conductivity velocity is expected to enhance their thermal performance. Simu-
of water. As a result, thermal stratification minimizes heat transfer lation study conducted by Tiwari and Kumar [20] in New Delhi,
to the bottom of the pond and keeps water at the bottom at a low India showed that increasing wind velocity significantly improves
temperature, facilitating better heat loss from the interior building thermal performance of an open roof pond system. In the hot-
into the pond [87]. Another effective factor is that RPWGB is less humid climate of Maracaibo, Venezuela, Rincon and Almao [35]
sensitive to increasing absorptivity of the gunny bags compared found that increasing wind velocity reduces nocturnal radiation
with the roof covered with wetted gunny bags [68]. provided by Skytherm. As nocturnal radiation accounts for the lar-
Finally, ventilated roof pond has performed better than massive gest share of cooling potential in the Skytherm system, this has
buildings and shaded roofs. Kruger and Cruz [9] compared cooling negative impact on the performance of the system. However,
potential of a ventilated roof pond with that provided by a high despite its negative impact on the cooling potential of the Sky-
thermal mass building in the hot-arid climate of Sde Boker Israel. therm, increasing wind velocity enhances convection through the
Simulation results showed that ventilated roof pond provide walls and therefore decrease the overall internal temperature [35].
greater maximum temperature depression. Lower efficiency of
the high mass building in terms of reducing the maximum temper- 8.2. Effects of water depth
ature was explained by the lower nocturnal local wind speed and
the considerable time needed for a high-mass building to cool Roof pond type, type of the roof deck (metal or concrete), local
down at night. In the room with ventilated roof pond the cooling climatic conditions, and purpose of use (heating or cooling) are the
degree days were almost eliminated as no indoor temperature main factors that should be taken into account when determining
was registered above the maximum adaptive thermal comfort optimum water depth for roof ponds.
limit. The high-mass building was also effective in reducing the Pond depth values between 5 and 75 cm are recommended for
number of cooling degree days. In terms of total indoor comfort, different roof pond systems. Results show that values approaching
the high mass building performed better due to the fact that its the lower limit are more appropriate for covered ponds. Upper
indoor minimum temperature is almost always above the lowest limit is recommended for uncovered ones. In a simulation study
adaptive comfort limit as prescribed by ASHRAE Standard 55 for conducted in the hot-arid climate of Jodhpur, India, Jain [13] found
Sde Boker’s climatic conditions. Since using ventilated roof pond that, for a roof pond with moveable insulation, water depth of 5 cm
alone drops the minimum indoor nighttime temperature below is sufficient and increasing the depth does not have significant
the lower adaptive comfort limit in Sde Boker (in summer), it is impact on the performance of the system. This was explained by
suggested that ventilated roof pond and high thermal mass should the fact that additional volume of water increases heat storage
be combined to counter this problem. capacity of the pond, thereby reducing evaporative heat transfer
due to reduced water temperature. Similar results have been found
in other studies [82,84]. In their simulation study of a ventilated
8. Factors affecting performance of roof ponds
roof pond under hot-arid conditions of Shiraz, Iran, Raeissi and
Taheri [82] found that increasing water depth decreases amount
Performance of roof ponds is affected by various internal and
of evaporation per unit volume of water and therefore the overall
external factors, several of which will be discussed in this section.
daily cooling load requirements are increased. According to their
study, better performance of low-depth ponds can be attributed
8.1. Effects of meteorological conditions to smaller mass of heat-storing water in the pond that makes it
easier to heat up the water to a degree farther from the WBT and
Increased humidity in the atmosphere reduces the potential for therefore achieve better evaporative cooling. A more recent study
evaporative and radiative cooling [24,25]. Therefore, lower relative by the same authors recommends 10 cm as the minimum required
A. Sharifi, Y. Yamagata / Applied Energy 160 (2015) 336–357 353

depth to obtain thermal (both cooling and heating) benefits from a pond with movable insulation in three buildings featuring different
Skytherm [48]. Similar results were found in a simulation study thermal mass, insulation levels, and roof materials. Simulation and
conducted in Phoenix, Arizona [49]. Experimental and simulation experimental results showed that the best performance in terms of
studies conducted on a roof pond with movable insulation, under cooling and temperature stabilization is achieved in the well-
hot-arid climatic conditions of Kuwait, showed that shallow roof insulated building with massive walls and metal-deck roof.
pond of about 5 cm depth can provide optimum results and reduce When passive heating is desired, increasing absorptivity of roof
overall costs of design [37]. Higher values of pond depth are rec- surface can improve thermal performance of the system. Sodha
ommended for uncovered roof ponds. In a simulation study con- and Singh [76] found that when absorptivity of roof surface of a
ducted under hot-arid climatic conditions of Asyut, Egypt, it was Skytherm is increased from 0.2 to 0.8, more heat flux enters the
observed that increasing water depth, of an open pond without interior space. On average, each 0.2 increase in the absorptivity
sprays, from 5 to 75 cm enhances net total passive cooling effects resulted in about 15 kJ/hm2 increase in the infiltrating heat flux.
of water (changed from 3.6 to 22.26 MJ/m2) [23]. Thickness of roof slab is another parameter that can affect per-
Effects of roof pond depth may vary depending on roof deck formance of roof ponds. Effects may vary depending on the type of
type. The cooling potential of RPWGB was investigated in the roof pond. For open roof ponds designed for cooling purpose
hot-arid climate of Sde Boker, Israel. It was found that, in the build- thicker slabs are desirable. The effect of increasing the roof thick-
ing with concrete roof, increasing water depth up to 20 cm ness (concrete) on the cooling performance of an open roof pond
improves the thermal stratification effect and therefore the cooling system was examined in the warm-temperate climate of New
potential of the system [68]. Similar results were found in another Delhi, India. It was found that increasing thickness of the roof from
study investigating impacts of pond depth on the cooling potential 0.15 to 0.30 cm decreases heat flux into building during all hours of
of a RPWGB under warm-temperate conditions of Crete, Greece the day [47]. This was, however, not true when open roof pond was
[34]. In buildings with metal-deck roof, however, the upper thresh- designed for heating purpose. Study conducted under arid climatic
old for water depth was found to be 5 cm, above which the cooling conditions of Boulder, Colorado demonstrated that best heating
effect of the RPWGB was deteriorated [68]. performance of an open roof pond was achieved when thickness
Optimum water depth may also vary depending on the roof approached zero [46]. Experimental study on the performance of
pond’s purpose of use (heating or cooling). Effect of water depth a RPWGB under warm-temperate conditions of Crete, Greece
was investigated in an open roof pond with water, concrete, and showed that changing the thickness of the concrete slab does not
(substrate) insulation layers. Numerical calculations showed that have considerable impacts on the performance of the system
ascending order of thickness of these layers provides the best com- [34]. Raeissi and Taheri [48] investigated the effect of roof thick-
bination for summertime use in New Delhi. For wintertime use, ness on the thermal performance of two variants of Skytherm sim-
descending order of thickness was desirable. However, for the cold ulated for the arid conditions of Shiraz, Iran. It was observed that
winters of Boulder, Colorado, ascending order of thickness proved for a Skytherm with a metal deck roof, increasing roof thickness
suitable [46]. Investigating the effect of increasing pond depth on does not noticeably affect the amount of energy transfer. This is
the heating efficiency of a simulated Skytherm in Phoenix, Arizona, explained by the high conductivity of the metallic roof. However,
it was found that changing the depth from 5 to 15 cm will increase for the concrete Skytherm, reducing roof thickness, reduced the
heat gain by 15%. Further increase of the depth did not result in energy loads (heating and cooling) showing that thinner concrete
considerable additional heat gain. As for the cooling mode, it was roof is desirable.
found that increasing water depth only resulted in slight improve-
ment of the cooling performance. It, however, reduced daily fluctu-
8.4. Effects of thickness and emissivity of the roof pond cover
ations of incoming and outgoing heat flux, meaning that more
‘‘stable comfort conditions” can be achieved by having a deeper
Information provided on these parameters is limited. Simulating
pond [49].
thermal performance of a Skytherm in arid and warm-temperate
For the tropical climate of Northern Thailand, Chu and Boonlong
climates, Tavana et al. [49] found that increasing thickness of the
[22] demonstrated that climatic conditions such as ambient tem-
top insulation cover enhanced cooling and heating performance
perature, relative humidity, and wind velocity also influence the
of the system. For the dry climate of Phoenix, Arizona, the opti-
optimum water depth of a roof pond with movable insulation.
mum thickness was found to be 3 cm, above which no further
Accordingly, average values of 9, 17 and 16 cm were obtained for
improvements would be achieved. Similar results were obtained
January, April and October, respectively. Water depth affected the
in a simulation and experimental study examining thermal behav-
lowest temperature that could be obtained in the roof pond. For
ior of a Skytherm under temperate conditions of New Delhi,
tropical climate of Northern Thailand the authors argue that the
India [76]. Simulation study by Raeissi and Taheri [48], in the arid
maximum temperature drop, below the ambient temperature, is
climate of Shiraz, Iran, showed that increasing thermal resistance
1 °C and roof pond depth of about 10 cm is needed for this purpose.
(from 0.009 to 2.822 m2 C/W) of the insulating panel of the
Skytherm reduces cooling and heating loads in typical summer
8.3. Effects of roof deck material and thickness
and winter days, respectively. Cooling load reductions were
10.5% and 3.2% for the metal and concrete Skytherms, respectively.
Roof deck facilitates thermal coupling between the pond and
Effects of increasing thermal resistance were more significant at
the interior space. Optimum configuration varies depending on
the winter. Heating load reductions achieved as a result of this
the type and purpose of the roof pond. Simulation study conducted
increase were 24.6% and 5.5% for metal and concrete Skytherms,
in the warm-temperate climate of New Delhi, India demonstrated
respectively. Emissivity of the cover surface and solar absorptance
that, when ambient outside temperature exceeded 40 °C, maxi-
of the opaque cover had negligible effects on the thermal perfor-
mum indoor temperatures for shaded pond (concrete), shaded
mance of the system [19,39,76].
pond (metal), Skytherm (concrete), and Skytherm (metal) were
about 30.8, 28, 29.6, and 26.8 °C, respectively [57]. Similar results
were found when summer and winter performances of two types 9. Summary and conclusions
of Skytherm (metal and concrete deck) were compared under
hot-arid climate of Shiraz, Iran [48]. In Sde Boker, Israel, Tang Implementation of passive design techniques can contribute to
and Etzion [10] compared cooling performance of RPWGB and roof climate stabilization by providing significant savings on energy
354 A. Sharifi, Y. Yamagata / Applied Energy 160 (2015) 336–357

consumption in the built environment. Roof ponds are relatively variants, future research should also take these factors into account
simple passive design elements capable of minimizing energy and provide quantitative and/or qualitative information related to
use in buildings. A vast body of work has been published on roof each of them.
ponds, indicating the availability of a substantial amount of knowl- Research on comparing roof pond systems with other passive
edge on their structure and performance. Reviewed studies have techniques is also limited. Of the roof pond heating variants, Sky-
employed either analytical, experimental, or both methods to therm is the only one that has been compared with other common
determine the effectiveness of roof ponds. Good agreement passive heating systems such as direct gain, Trombe Wall and Sun-
between the results indicates validity of the models used in these space. It has proved more effective than these systems. As for the
studies. This systematic review listed different variants developed cooling variants, only a limited number have been included in
for cooling and heating purposes. Table 8 summarizes the main the comparisons (see Tables 7 and 8). Comparing performance of
features of these variants. A summary of general advantages and RPWGB, Cool-pool, ventilated roof pond, and Skytherm to other
disadvantages of roof ponds is also presented in Table 9. passive strategies showed that, in many cases, these roof pond
Overall, 19 different roof pond cooling variants were discussed. variants perform better or are about equally effective in maintain-
Open roof ponds, ponds with movable insulations, and ventilated ing indoor thermal comfort. These findings are promising. How-
roof ponds are major roof pond cooling types that have received ever, it would be rash to conclude that roof pond systems always
comparatively higher attention in the literature. Little information perform better than other passive techniques. Such a conclusion
exists on performance of Energy-roof, walkable roof ponds, shaded would only be supported if other issues such as cost, maintenance,
ponds, and shaded RPWGB. More research is needed on these vari- resource consumption, and applicability to different climates were
ants. Results showed that all variants (except Roof pond with addi- also considered in the comparative analyses. This is a major gap in
tional insulation layer) are capable of enhancing thermal comfort knowledge that needs to be addressed.
in arid and warm-temperate climates. It was found that Skytherm, This review also discussed several factors affecting performance
Cool-roof, and ventilated roof pond can even be effective in humid of roof ponds. Increased humidity has negative impacts on perfor-
climates. Effectiveness of Skytherm can be explained by the fact mance of roof pond cooling systems. It was argued that keeping
that water is enclosed in watertight bags and therefore its perfor- water in watertight bags (dry roof ponds) and enhancing evapora-
mance is less affected by relative air humidity. This implies that tion by using fans to blow air over the water surface can be used to
‘‘shaded pond with water enclosed in water tight bags” may also mitigate the impact of humidity. Wind enhances evaporative cool-
be suitable for use in humid climates. Future research is needed ing. However, it reduces longwave radiation to sky. It is therefore
to verify this assumption empirically. Cool-roof uses nighttime necessary to work on strategies for minimizing its negative
intermittent water spray over the roof to enhance evaporation. impacts on longwave radiation.
This can explain its effectiveness in humid climates. However, Regarding the optimum water depth, values between 5 and
since this system has only been examined in two humid climates, 75 cm were recommended for different roof pond systems. Results
this result should be treated with caution and further research is showed that values approaching the lower limit are more appro-
necessary to better understand the performance of Cool-roof in priate for covered ponds. Upper limit is recommended for uncov-
humid climates. Effectiveness of ventilated roof pond in humid cli- ered roof ponds. Regarding appropriate roof deck thickness and
mate can be attributed to its use of fans to enhance water material, these recommendations can be made: For covered ponds
evaporation. designed for heating and cooling purposes metal deck performs
Unlike roof pond cooling systems, the literature on roof pond better. Due to high conductivity of metal, deck thickness does
heating systems is limited. This is probably because of their limita- not have a significant impact. For covered ponds that have concrete
tions compared to other passive heating techniques. In addition to deck, thinner deck is desirable (both heating and cooling). For open
the cost and maintenance issues, roof ponds are argued to only be roof ponds designed for cooling and heating purposes, thicker and
capable of providing sufficient heating for climates where outside thinner concrete slabs are recommended, respectively.
air temperature does not fall below 5 °C [27]. Other passive tech- About the effects of thickness and emissivity of the roof pond
niques such as direct gain and Sunspace need less maintenance cover, existing evidence indicates that cooling and heating effec-
and can be applied in any climate and to any building, regardless tiveness of Skytherm improves with increasing thickness up to
of its height [89]. Of the four different variants mentioned in the about 3 cm, after which it remains constant. Effects of emissivity
study, Skytherm and ‘‘roof-integrated water solar collector” proved have been found to be negligible. Effects of these parameters on
effective in temperate climates. It was also argued that creating the performance of other covered roof pond variants should also
non-convective zone in the water pond can make the Skytherm be studied in future research.
applicable to freezing conditions as well. This review also highlights a number of other key knowledge
This review indicates that limited evidence on the comparative gaps that need to be addressed:
performance of different roof ponds is available in the literature.
Several studies include comparisons between two or more roof – Lack of scientific understanding of the magnitude of energy savings
pond cooling systems. However, variants such as Energy-roof, possible by large scale application of roof ponds. Existing studies
walkable roof pond, Cool-roof, and Cool-pool have not received have provided information on the magnitude of energy load
due attention. Results show that, under arid and warm- reduction from applying these systems to individual buildings.
temperate conditions, RPWGB, shaded roof ponds, ventilated roof There is a need to also explore energy saving potential of
ponds, and roof ponds with movable insulation provide better per- deployment at neighborhood and urban scales.
formance than other types (descending order of performance). – Little information about the effects of neighboring buildings. Most
However, as can be seen from Table 8, there are also other factors of the studies reviewed in this paper have used stand-alone test
such as initial costs, maintenance and functioning costs, need for cells in their experiments and numerical analyses. Since in real-
mechanical support, water consumption, and type of target build- world conditions buildings may not be exposed to the ambient
ing that should be taken into account. For instance, high water environment on all sides, the effects of neighboring buildings
requirement of RPWGB, shaded roof ponds, and ventilated roof should be taken into account in the heat transfer analyses. If
ponds undermines the viability of using them in water-stressed applied to buildings that share side walls, roof ponds may pro-
areas. When making comparisons between different roof pond vide further thermal benefits.
Table 8
The main features of different roof pond variants mentioned in this review (Adapted, with additions and changes, from [19]).

Open/ Open/ Movable Movable Movable Skytherm Energy- Cool- Insulation Walkable RPWGB Shaded Shaded Shaded Shaded Cool- Ventilated Evapo- Additional Roof-
no with insulation/ insulation/ insulation/ roof roof embedded with RPWGB with with without pool roof pond reflective insulation integrated
sprays sprays continuous nighttime without within the water closed water water layer water
spray spray sprays pond circulation water spray spray solar
bags collector
Type (thermal load C B C C C B C C C B C C C C C C C C C B
concerned)
Simplicity of the system  U U U U U     U U     U U U 
Widespread know-how U U   U U  U   U    U  U   
Low-moderate initial U U U U U U     U U U U U  U U U 
costs
Low-moderate U        U  U U U U U  U U U 
maintenance and
functioning costs

A. Sharifi, Y. Yamagata / Applied Energy 160 (2015) 336–357


Low-moderate water U    U U   U    U   U  U U U
consumption
Low contamination       U U U U U U     U U U U
Absence of mechanical U        U  U U H H H H H U U 
support (fan, pump,
mechanical control)
Water spray over the roof  U U U   U U  H    U      
Removable cover   U U U U              U
Fixed cover         U U  U H H H H U U U 
Floating insulation       U U            
Availability of air gap or   U U U U      U U U U U U U  U
ventilated air layer in
the system
Water enclosed in      U       U       H
watertight bags
Relatively high  U U U U U U U U U U U  U U U U U  U
effectiveness
Relative effectiveness in      U  U     U    U   
humid climates
Applicability to concrete U U U U U U  U U U U U U U U U U U U 
roof
Applicability to metal   U U U U U U   U U U U U U U   U
deck roof
Applicability to non-flat                 U   U
roofs
Integrated into building       U      U U U U U   
envelope
Walkability         U U          
Effectiveness in medium       U         U    U
to high-rise buildings
Applicability to buildings                U    
with limited roof area
Comparable or better  – – – U U – – – – U – – – – U U – – –
performance with
respect to other
passive techniques

 no compliance; U compliance; – no information available; H can be either yes or no; C cooling; B both cooling and heating.

355
356 A. Sharifi, Y. Yamagata / Applied Energy 160 (2015) 336–357

Table 9
General advantages and disadvantages of roof pond systems.

Advantages
– A great deal of knowledge is available on design and operation of roof pond cooling systems
– These system have proved to be useful in arid, as well as temperate climates. Some variants are also applicable to humid climates
– Roof ponds can also provide co-benefits in terms of stormwater management and the stored water can also be used for fire-extinguishing purposes
– Unlike many other passive measures, performance of roof ponds is not affected by building orientation
– Unlike direct evaporative cooling systems, roof pond cooling systems do not increase indoor humidity to an undesirable level
Disadvantages
– Roof ponds increase the weight of the building and are not suitable for buildings constructed according to lightweight construction standards. Retrofitting such
buildings to make them more resistant may impose substantial cost premiums
– Roof pond systems can only be installed atop flat roofs (only ventilated roof pond is an exception)
– Unlike other passive techniques such as green roof and high-albedo coating, roof ponds affect the accessibility of roof for other uses (walkable roof pond is an
exception)
– In case of improper maintenance, leakage and water contamination issues may occur
– Most variants can only provide indoor comfort in one-or two-story buildings
– It is argued that the most suitable climate for implementing roof pond systems is arid climate. These are regions, where in many cases, water availability is a serious
issue. Using grey water may, partially, resolve this conflict
– Unless coupled with other passive measures, roof ponds may not be able to completely eliminate energy loads of buildings

– Lack of understanding of cumulative effects of combining roof [10] Tang R, Etzion Y. Cooling performance of roof ponds with gunny bags floating
on water surface as compared with a movable insulation. Renew Energy
ponds with other passive measures. In most cases energy loads
2005;30:1373–85.
of buildings could only be eliminated if roof pond techniques [11] Vijaykumar KCK, Srinivasan PSS, Dhandapani S. A performance of hollow clay
are employed in conjunction with other passive measures. tile (HCT) laid reinforced cement concrete (RCC) roof for tropical summer
Future studies should examine how roof pond techniques could climates. Energy Build 2007;39:886–92.
[12] Nahar NM, Sharma P, Purohit MM. Performance of different passive techniques
be combined with other passive strategies to eliminate thermal for cooling of buildings in arid regions. Build Environ 2003;38:109–16.
loads while providing indoor comfort conditions. Also, further [13] Jain D. Modeling of solar passive techniques for roof cooling in arid regions.
attention to the importance of residents’ acclimatization and Build Environ 2006;41:277–87.
[14] Pearlmutter D, Rosenfeld S. Performance analysis of a simple roof cooling
behavioral changes is needed. system with irrigated soil and two shading alternatives. Energy Build
– Lack of information on the life cycle costs of roof ponds. Analysis of 2008;40:855–64.
life cycle costs and payback period should be conducted to help [15] Juanico LE. A new design of configurable solar awning for managing cooling
and heating loads. Energy Build 2009;41:1381–5.
developers make more informed decisions regarding suitability [16] Pisello A, Pignatta G, Castaldo V, Cotana F. Experimental analysis of natural
of these systems. This will also provide insights on affordability gravel covering as cool roofing and cool pavement. Sustainability-Basel
of roof pond systems. 2014;6:4706–22.
[17] Santamouris M. Advances in passive cooling. Buildings, energy, solar
– Little knowledge on the real-world application of roof ponds. Last, technology. London, Sterling, VA: Earthscan; 2007.
but not the least, future work needs to be done to provide more [18] Givoni B. Indoor temperature reduction by passive cooling systems. Sol Energy
knowledge on the extent of implementation of roof ponds. 2011;85:1692–726.
[19] Spanaki A, Tsoutsos T, Kolokotsa D. On the selection and design of the proper
roof pond variant for passive cooling purposes. Renew Sustain Energy Rev
2011;15:3523–33.
[20] Tiwari G, Kumar A, Sodha M. A review—cooling by water evaporation over
Appendix A. Supplementary material roof. Energy Convers Manage 1982;22:143–53.
[21] Pullin AS, Stewart GB. Guidelines for systematic review in conservation and
environmental management. Conserv Biol 2006;20:1647–56.
Supplementary data associated with this article can be found, in [22] Chu SCK, Boonlong P. Design optimization of passively cooled room. J Energy
the online version, at http://dx.doi.org/10.1016/j.apenergy.2015. Eng – ASCE 1992;118:18–37.
09.061. These data include Google maps of the most important [23] Ali AHH. Passive cooling of water at night in uninsulated open tank in hot and
areas. Energy Convers Manage 2007;48:93–100.
areas described in this article.
[24] Erell E. Evaporative cooling. In: Santamouris M, editor. Advances in passive
cooling. London: Earthscan; 2007. p. 228–61.
[25] Erell E. Radiative cooling. In: Santamouris M, editor. Advances in passive
References cooling. London: Earthscan; 2007. p. 262–96.
[26] Dimoudi A. Passive cooling of buildings. In: Asimakopoulos D, Santamouris M,
[1] Santamouris M, Kolokotsa D. Passive cooling dissipation techniques for editors. Passive cooling of buildings. Oxon OX.: Earthscan; 2013. p. 35–55.
buildings and other structures: the state of the art. Energy Build [27] La Roche P. Carbon-neutral architectural design. Boca Raton, FL: CRC Press;
2013;57:74–94. 2012.
[2] La Roche P, Berardi U. Comfort and energy savings with active green roofs. [28] Alvarez S, Molina JL. Cooling by natural sinks. In: Santamouris M, editor. Solar
Energy Build 2014;82:492–504. thermal technologies for buildings: the state of the art Oxon. Earthscan; 2013.
[3] Saadatian O, Sopian K, Lim CH, Asim N, Sulaiman MY. Trombe walls: a review p. 140–63.
of opportunities and challenges in research and development. Renew Sustain [29] Chandra S, Kaushik DD, Kaushik SC. Passive solar heating of a non-
Energy Rev 2012;16:6340–51. airconditioned building with movable insulation over the roof pond. Appl
[4] Saman WY. Towards zero energy homes down under. Renew Energy Energy 1984;17:235–49.
2013;49:211–5. [30] González Cruz E, Krüger E. Evaluating the potential of an indirect evaporative
[5] Brager G, Zhang H, Arens E. Evolving opportunities for providing thermal passive cooling system for Brazilian Dwellings. Build Environ 2015.
comfort. Build Res Inform 2015;1–14. [31] González E, Givoni B. Radiative and radiative/evaporative passive cooling
[6] Bravo G, Gonzalez E. Thermal comfort in naturally ventilated spaces and under systems for a hot humid climate–Maracaibo. PLEA2004.
indirect evaporative passive cooling conditions in hot-humid climate. Energy [32] Ghosal MK, Tiwari GN, Srivastava NSL. Modeling and experimental validation
Build 2013;63:79–86. of a greenhouse with evaporative cooling by moving water film over external
[7] van Hooff T, Blocken B, Hensen JLM, Timmermans HJP. On the predicted shade cloth. Energy Build 2003;35:843–50.
effectiveness of climate adaptation measures for residential buildings. Build [33] Spanaki A, Kolokotsa D, Tsoutsos T, Zacharopoulos I. Assessing the passive
Environ 2014;82:300–16. cooling effect of the ventilated pond protected with a reflecting layer. Appl
[8] Stevanović S. Optimization of passive solar design strategies: a review. Renew Energy 2014;123:273–80.
Sustain Energy Rev 2013;25:177–96. [34] Spanaki A, Kolokotsa D, Tsoutsos T, Zacharopoulos I. Theoretical and
[9] Kruger E, Cruz EG, Givoni B. Effectiveness of indirect evaporative cooling and experimental analysis of a novel low emissivity water pond in summer. Sol
thermal mass in a hot arid climate. Build Environ 2010;45:1422–33. Energy 2012;86:3331–44.
A. Sharifi, Y. Yamagata / Applied Energy 160 (2015) 336–357 357

[35] Rincon J, Almao N, Gonzalez E. Experimental and numerical evaluation of a [63] Niles PW. Thermal evaluation of a house using a movable-insulation heating
solar passive cooling system under hot and humid climatic conditions. Sol and cooling system. Sol Energy 1976;18:413–9.
Energy 2001;71:71–80. [64] Ahmad I. Improving the thermal performance of a roof pond system. Energy
[36] Al-Hemiddi N. The thermal performance of a Roof-Pond integrated to a Convers Manage 1985;25:207–9.
building for heating during Cold-Winter Desert climate conditions in Saudi [65] Ayoob A, Attalage R. Theoretical and experimental analysis of a simply
Arabia. Int J Ambient Energy 1999;20:45–52. designed solar retrofitting heating and cooling system. Renew Energy
[37] Al-Jamal K, Alameddine O, Al-Shami H, Shaban N. Passive cooling evaluation of 1991;1:513–8.
roof pond systems. Solar Wind Technol 1988;5:55–65. [66] Garcia C, Givoni B. Cooling by roof pond with floating insulation in the hot
[38] Marlatt W, Murray KA, Squier SE. Roof pond systems. Canoga Park, CA humid climate of Veracruz, Mexico. PLEA. 2007.
(USA): Energy Technology Engineering Center; 1984. [67] Dabaieh M, Wanas O, Hegazy MA, Johansson E. Reducing cooling demands in a
[39] Yannas S, Erell E, Molina JL. Roof cooling techniques : a design hot dry climate: a simulation study for non-insulated passive cool roof thermal
handbook. London; Sterling, VA: Earthscan; 2006. performance in residential buildings. Energy Build 2015;89:142–52.
[40] Kottek M, Grieser J, Beck C, Rudolf B, Rubel F. World map of the Koppen-Geiger [68] Tang R, Etzion Y. On thermal performance of an improved roof pond for
climate classification updated. Meteorol Z 2006;15:259–63. cooling buildings. Build Environ 2004;39:201–9.
[41] Vorster J, Dobson R. Sustainable cooling alternatives for buildings. J Energy [69] Kharrufa SN, Adil Y. Upgrading the building envelope to reduce cooling loads.
Southern Africa 2011;22:48–66. Energy Build 2012;55:389–96.
[42] Runsheng T, Etzion Y, Erell E. Experimental studies on a novel roof pond [70] Givoni B. Performance and applicability of passive and low-energy cooling
configuration for the cooling of buildings. Renew Energy 2003;28:1513–22. systems. Energy Build 1991;17:177–99.
[43] Kaushik SC, Chandra S. Solar thermal modelling of a building with a roof pond [71] Manzan M, Saro O. Numerical analysis of heat and mass transfer in a passive
and ventilation control systems. Build Environ 1982;17:273–84. building component cooled by water evaporation. Energy Build
[44] Dhiman N, Kumar A, Tiwari G. Cooling by evaporation of flowing water over a 2002;34:369–75.
hollow roof. Appl Energy 1982;10:107–20. [72] Bencheikh H. Full-scale experimental studies of a passive cooling roof in hot
[45] Sodha M, Singh U, Srivastava A, Tiwari G. Experimental validation of thermal arid areas. Prog Exergy, Energy Environ. Springer; 2014. p. 507–15.
model of open roof pond. Build Environ 1981;16:93–8. [73] Kaushika ND, Rao SK. Non-convective roof pond with movable insulation for
[46] Sodha M, Kaushik S, Nayak J. Performance of trombe walls and roof pond passive solar space heating in cold climates. Build Environ 1983;18:9–17.
systems. Appl Energy 1981;8:175–91. [74] Hay HR, Yellott JI. International aspects of air conditioning with movable
[47] Sodha M, Kumar A, Singh U, Tiwari G. Periodic theory of an open roof pond. insulation. Sol Energy 1969;12:427–38.
Appl Energy 1980;7:305–19. [75] Fernández-González A. Analysis of the thermal performance and comfort
[48] Raeissi S, Taheri M. Skytherm: an approach to year-round thermal energy conditions produced by five different passive solar heating strategies in the
sufficient houses. Renew Energy 2000;19:527–43. United States Midwest. Sol Energy 2007;81:581–93.
[49] Tavana M, Kammerud R, Akbari H, Borgers T. A simulation model for the [76] Sodha M, Singh U, Tiwari G. A thermal model of a roof pond system with
performance analysis of roof pond systems for heating and cooling. Lawrence moveable insulation for heating a building. Build Environ 1982;17:135–44.
Berkeley National Laboratory; 1980. [77] Juanico L. A new design of roof-integrated water solar collector for domestic
[50] Givoni B. Passive and low energy cooling of buildings. New York: Van Nostrand heating and cooling. Sol Energy 2008;82:481–92.
Reinhold; 1994. [78] Juanico LE. New design of solar roof for household heating and cooling. Int J
[51] Alp AV. Vernacular climate control in desert architecture. Energy Build Hydrogen Energy 2010;35:5823–6.
1991;16:809–15. [79] Thomason HE, Thomason HJL. Solar houses/heating and cooling progress
[52] Pittinger AL. Energy roof. Google Patents; 1976. report. Sol Energy 1973;15:27–39.
[53] Bourne RC, Springer DA. Energy-saving protected roof systems. Google [80] Kaushik ND, Rao SK. Non-convective roof pond for passive space heating. Appl
Patents; 1992. Energy 1982;12:21–36.
[54] Givoni B. Climate considerations in building and urban design. New York: Van [81] Kharrufa SN, Adil Y. Roof pond cooling of buildings in hot arid climates. Build
Nostrand Reinhold; 1998. Environ 2008;43:82–9.
[55] Al-Hemiddi NA. Passive cooling systems applicable for buildings in the hot-dry [82] Raeissi S, Taheri M. Cooling load, reduction of buildings using passive roof
climate of Saudi Arabia. Los Angeles: University of California; 1995. options. Renew Energy 1996;7:301–13.
[56] Marnich R, LaRoche P, Yamnitz R, Carbonnier E. Passive cooling with self- [83] Jain S, Rao K. Experimental study on the effect of roof spray cooling on
shading modular roof ponds as heat sink. 2010. unconditioned and conditioned buildings. Build Sci 1974;9:9–16.
[57] Yadav R, Rao D. Ditigal simulation of indoor temperatures of buildings with [84] Chandra S, Kaushik S, Bansal P. Thermal performance of a non-air-conditioned
roof ponds. Sol Energy 1983;31:205–15. building for passive solar air-conditioning: evaluation of roof cooling systems.
[58] Kasaeian A, Sameti M, Eshghi AT. Simplified method for night sky radiation Energy Build 1985;8:51–69.
analysis in a cool-pool system. Sustain Energy 2014;2:29–34. [85] Nayak J, Srivastava A, Singh U, Sodha M. The relative performance of different
[59] Sodha M, Singh S, Kumar A. Thermal performance of a cool-pool system for approaches to the passive cooling of roofs. Build Environ 1982;17:145–61.
passive cooling of a non-conditioned building. Build Environ 1985;20:233–40. [86] Nahar NM, Sharma P, Purohit MM. Studies on solar passive cooling techniques
[60] Bainbridge DA, Haggard KL. Passive solar architecture : he ating, cooling, for arid areas. Energy Convers Manage 1999;40:89–95.
ventilation, daylighting, and more using natural flows. White River Junction, [87] Tang RS, Etzion Y, Erell E. Experimental studies on a novel roof pond
Vt.: Chelsea Green Pub.; 2011. configuration for the cooling of buildings. Renew Energy 2003;28:1513–22.
[61] Bencheikh H, Bouchair A. Passive cooling by evapo-reflective roof for hot dry [88] Sodha M, Khatry A, Malik M. Reduction of heat flux through a roof by water
climates. Renew Energy 2004;29:1877–86. film. Sol Energy 1978;20:189–91.
[62] Yang W, Wang Z, Zhao X. Experimental investigation of the thermal isolation [89] Givoni B. Characteristics, design implications, and applicability of passive solar
and evaporative cooling effects of an exposed shallow-water-reserved roof heating systems for buildings. Sol Energy 1991;47:425–35.
under the sub-tropical climatic condition. Sustain Cities Soc 2015;14:293–304.

View publication stats

You might also like