Energy and Buildings 112 (2016) 198–208

Contents lists available at ScienceDirect

Energy and Buildings

journal homepage: www.elsevier.com/locate/enbuild

A review of research and development on water wall for building

applications
Ting Wu, Chengwang Lei ∗
School of Civil Engineering, The University of Sydney, Sydney, Australia

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

Article history: Heating, ventilating, and air-conditioning (HVAC) systems consume almost half of the total energy con-
Received 15 April 2015 sumption in buildings. In order to reduce the energy consumption of HVAC systems in buildings, different
Received in revised form 16 October 2015 passive heating and cooling strategies and thermal energy storage systems have been adopted. Water
Accepted 4 December 2015
wall is an excellent short-term thermal energy storage system which can be used to maintain thermal
Available online 9 December 2015
comfort in buildings while reducing the heating and/or cooling load of the buildings. A significant amount
of research work on water wall has been carried out since the 1980s. This paper presents a comprehensive
Keywords:
review of the previous studies of water wall systems, focussing mainly on the performance of various
Water wall
Thermal energy storage
configurations of water wall systems and the different research approaches adopted in the literature. The
Thermal performance advantages and disadvantages of the different research approaches have been highlighted. It is revealed
that the existing water wall research has been based mainly on an energy balance model and experimen-
tal modelling, but CFD (computational fluid dynamics) modelling has attracted little attention. Moreover,
this review has identified several gaps in the water wall research for further attention.
© 2015 Elsevier B.V. All rights reserved.

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199
2. A brief history of water wall . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199
3. Configurations of water wall . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 200
3.1. Water wall with an opaque building envelope . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 200
3.1.1. Performance comparisons with other strategies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 200
3.1.2. Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 201
3.2. Water wall with a semi-transparent building envelope . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 201
3.2.1. Performance comparisons with other strategies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 201
3.2.2. Optimization of thermal performance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 202
3.3. Water wall with phase change materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203
3.4. Water wall in conjunction with other passive technologies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203
3.4.1. With a sunspace . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203
3.4.2. With a solar chimney . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 204
3.4.3. With evaporative cooling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 204
3.5. Summary of water wall research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 204
4. An overview of methodologies for water wall research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205
4.1. Heat balance model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205
4.1.1. Steady state analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205
4.1.2. Transient analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205
4.1.3. Building energy simulation programs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205

∗ Corresponding author. Tel.: +61 2 9351 2457; fax: +61 2 9351 3343.
E-mail address: [email protected] (C. Lei).

http://dx.doi.org/10.1016/j.enbuild.2015.12.003
0378-7788/© 2015 Elsevier B.V. All rights reserved.
 T. Wu, C. Lei / Energy and Buildings 112 (2016) 198–208 199

4.2. Experimental modelling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206

4.3. CFD modelling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206
4.4. Summary of methodologies for water wall research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206
5. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206
Acknowledgement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207

1. Introduction Sadineni et al. [8] presented an overview of major components of

a building envelope relevant to building energy efficiency, which
At the present time, the energy consumption by residential and include different types of energy efficient walls (such as Trombe
commercial buildings accounts for nearly one quarter of the total walls, ventilated walls, and transwall, etc.), various types of fen-
worldwide consumption of delivered energy [1], and it is predicted estrations, roof thermal insulation materials and thermal mass,
that the building energy consumption will increase by approxi- air tightness, and infiltration of building envelopes, etc. Saadatian
mately 1.6% per year from 2010 to 2040. Almost half of the build- et al. [9] reviewed the application of Trombe walls in buildings for
ing’s energy consumption is for operating heating, ventilating, and heating and ventilation. Nine different types of Trombe walls and
air-conditioning (HVAC) systems [2]. Under this circumstance, sav- the effects of various important accessories on the performance of
ing energy from the HVAC systems in buildings becomes critical for Trombe walls were discussed. More recently, Stevanović [10] gave
the combat against the ongoing energy crisis and climate change. a comprehensive review of passive solar design strategies using dif-
Thermal energy storage (TES) is a technology that stores ther- ferent simulation software (e.g., EnergyPlus, TRNSYS, Fluent, etc.)
mal energy over a particular period of time and releases the stored and the optimization of building design in terms of its shape and ori-
energy at a later time. It has a great potential for reducing the entation, insulations, roof and floor construction, windows-to-wall
energy consumption in buildings. Over the past several decades, area, glazing type, and shading, etc.
growing research interest has been devoted to this area [3]. The It is clear from the above-mentioned reviews that water wall as
TES technologies for buildings can be classified into sensible heat a short-term TES system for passive heating, cooling and/or natu-
storage and latent heat storage according to the media used for ral ventilation has not received much attention. Since water wall
heat storage. In latent heat storage systems, a large amount of has unique advantages over other passive building technologies,
heat is stored or released during a phase change process (e.g., and a significant amount of research work has been carried out on
melting/solidification or gasification/liquefaction) of the storage this particular technology, the present review will focus on water
medium [4]. Phase change materials (PCMs) offer a high density for wall as a short-term TES strategy for maintaining thermal comfort
TES and require a smaller temperature difference between storing and reducing energy consumption in buildings. The purpose of the
and releasing heat than sensible heat storage systems. However, review is also to identify further research needs in this area.
PCMs also have some drawbacks. Inorganic PCMs are usually cor-
rosive, unstable, and tend to delay solidification (supercooling), 2. A brief history of water wall
whereas organic PCMs are flammable and have relatively low vol-
umetric latent heat storage capacity and low thermal conductivity The first water wall reported in the open literature was probably
[4,5]. Moreover, no PCM is suitable for the full temperature range the one built in 1947 at the Massachusetts Institute of Technology
of all climate conditions even in a fixed location [6]. by Hoyt Hottel and his students [11]. Their water wall comprised an
In sensible heat storage systems, heat is stored or released with array of cans which were painted black and placed behind double-
the temperature change of the storage medium [3]. Sensible heat pane glass. Unfortunately, the water wall performed poorly because
storage materials with high heat capacities, such as concrete, brick of its poor design including limited direct solar gain through win-
and water, have been used as thermal mass in dwellings in order dows, bad insulation design, and inadequate placement of curtains.
to moderate the temperature of internal spaces and to reduce the Water wall was adopted again in an innovative passive solar
need for summer cooling and winter heating. Sensible heat stor- house in New Mexico designed by Steve Baer in 1972 [11]. Stacked
age materials can be used for both short-term TES and long-term water drums were used in his Corrales house to provide thermal
TES. Short-term TES means that thermal energy is stored during mass for the passive design. The walls of the house were made of
the day and released during the night, whereas long-term TES often adobe and insulated aluminium sandwich panels. In total 90 water
works over a yearly cycle of heat storage and release [3]. A water filled barrels of 55-gallon each were stacked to 4–5 feet high in
wall acts as a short-term TES medium and can be used to maintain metal support frames. The ends of the barrels were painted black
thermal comfort in buildings while reducing energy consumption and faced south. Heat was collected from direct solar gain through
in buildings. The water wall system has unique advantages over glass. The system worked well and contributed about 85% of the
other short-term TES technologies due to the abundance, low cost, total space heating requirement.
and high heat capacity of water. Further, the heat stored in water Since then a variety of water walls have emerged, and many
may be redistributed by convection, and thus a water wall provides water wall projects have been constructed in both residential and
quicker heat exchange than a concrete or brick wall. commercial buildings (see e.g., [12,13]). A significant amount of
A large body of literature exists on passive solar technologies research has also been carried out to investigate the thermal per-
with short-term TES using different configurations and various formance of water wall systems (see for example [14–16]). Never-
storage media under different climate conditions. Some recent theless, the interest in water wall systems seems to be declining in
reviews of these technologies can be found in [7–10]. Chan et al. the 1990s and 2000s. Currently only a very few companies still offer
[7] reviewed the passive solar technologies for space heating and water walls for building applications, and none of them has been
cooling according to their working mechanisms. Trombe wall, solar commercially successful [17]. One of the major concerns regarding
chimney, unglazed transpired solar façade, and solar roof are some the development of water wall systems is the possibility of rusting
of the passive solar heating and natural ventilation strategies cov- and leakage [17]. However, with the adoption of rust-control addi-
ered in their review, which also includes passive solar cooling tive, sacrificial anodes and corrosion free materials (e.g., fibreglass),
by direct evaporative cooling and indirect evaporative cooling. no corrosion was found in any of the water wall homes designed by
200 T. Wu, C. Lei / Energy and Buildings 112 (2016) 198–208

San Luis Sustainability Group, which had been in operation for 20 performance will be achieved. They also studied the arrangement
years [17]. Leakage can be avoided either by air-leakage testing or of a water wall with both concrete and insulation panels. The
by adopting modular water wall design. Therefore, water wall has results indicated that for the summer climate in New Delhi, the
a great potential for future building development. water–concrete–insulation configuration with an ascending order
of the thicknesses was the best combination, whereas for the win-
3. Configurations of water wall ter climate, the same combination with a descending order of the
thicknesses was preferable. Another investigation of the optimum
According to the different configurations, water walls can be thickness distribution between the inside and outside insulation
generally grouped into the following four types: water wall with layers of a thermal storage water wall under the constraint of a
an opaque building envelope, water wall with a semi-transparent given total thickness of the insulation layers was presented in [21].
building envelope, water wall with phase change materials, and Periodic solar radiation and atmospheric air temperature using typ-
water wall combined with other passive technologies. In what fol- ical Delhi summer weather data were prescribed on the external
lows, the configuration and the status of research of each type of side of the thermal storage water wall, whereas a constant room
the water walls are described. air temperature was set on the internal side (corresponding to an
air-conditioned room). They found that an equal thickness between
3.1. Water wall with an opaque building envelope the inside and outside insulation layers gave the minimum temper-
ature fluctuation. Moreover, the temperature swing was smaller
Water walls are often incorporated into an opaque building when the entire insulation layer was applied outside rather than
envelope such as metallic plates, opaque PVC pipes, concrete, or inside the thermal storage water wall.
insulation panels, which separate living spaces from the ambient Nayak et al. [22] compared the thermal performance of four
(refer to Fig. 1). This is the basic configuration of the water walls. typical passive heating concepts, namely a Trombe wall, a water
wall, and a solarium under two different configurations, one with
3.1.1. Performance comparisons with other strategies the glazing uncovered and the other with the glazing covered by a
The thermal performance of water walls with opaque building moveable insulation during off sunshine hours. Their calculations
envelopes have been widely compared against the performance of showed that the water wall system resulted in a less temperature
conventional walls. Balcomb and McFarland [18] applied a sim- fluctuation and a higher average heat flux than the Trombe wall
ple empirical method called Solar Load Ratio Method to estimate when night-time insulation was used, and a phase shift of almost
the thermal performance of a Trombe wall and water wall with 12 h was observed for a 0.22 m thick concrete wall next to a 0.10 m
or without night-time insulation and with or without a reflector. thick water wall.
The results showed that the water wall achieved a higher monthly A comparison between a water wall plus concrete and a water
solar heating fraction (i.e., the percentage of the space heating load wall plus insulation panels was studied by Kaushik and Kaul [23].
supplied by the passive solar system) than that of the Trombe wall. They found that the water wall with insulation panels gave an
They further carried out a parametric analysis of the annual energy almost uniformly low level heat flux both during and after sunshine
savings using the Trombe wall and water wall [19]. The parameters hours for a mild winter day of New Delhi. As a consequence, the
considered in their analysis included the R-value of the night-time hourly indoor air temperature with the water wall plus insulation
insulation, the number of glazing, the wall absorptivity and emis- panels was lower than that of the water wall plus concrete at all
sivity, the thermal storage capacity, and the additional building hours. They also analysed the effects of different combinations
mass, etc. They found that the performance of the water wall was of water, concrete, and insulation. The results showed that the
enhanced by decreasing the R-value of the night-time insulation, concrete–water–insulation configuration resulted in less fluctu-
the number of glazing and the wall absorptivity and by increasing ation in the hourly heat flux than that of the insulation–water–
the wall emissivity, the thermal storage capacity, and the additional concrete configuration. However, a concrete–water–concrete
building mass. configuration was more desirable than the other combinations
Sodha et al. [20] compared the thermal performance of a Trombe for space heating, with which the indoor air temperature can be
wall with a water wall plus concrete, a water wall plus insu- maintained in a comfortable range. Subsequently Sodha et al. [24]
lation panels and a simple water layer by calculating the heat compared the effects of two types of thermal storage materials
transfer through these walls. All the walls were subjected to a (i.e., water and concrete) on the thermal performance of a non-air-
periodic solar radiation with atmospheric air on one side while conditioned room. It was found that the concrete storage was less
a constant air temperature on the other side. Their results illus- effective than the water storage in reducing the swing of the room
trated that a water wall plus concrete was more desirable than the air temperature for the same storage mass because of its lower
other configurations for the winter climate in New Delhi, and the specific heat capacity.
more water contained in the water wall plus concrete, the better The thermal performance of a vertical cellular clayey concrete
wall fitted with a water piping system was experimentally com-
pared with a floor heating system by Marmoret et al. [25]. The water
pipe wall system was monitored during a cyclic process which
included circulating hot water for 8 h and then passively releasing
heat for 16 h. The monitored data showed that the water flow rate in
the charging period had only a small influence on the wall temper-
ature, whereas the amount of energy stored in the wall increased as
the water temperature was increased. It was reported that heat was
still released from the water pipe wall to the room after discharging
for 16 h, but was less than that of the floor heating installation for
the same water intake temperature of 42 ◦ C.
A two-dimensional (2D) comparative evaluation between a
water pipe wall heating system and a floor heating system was car-
ried out for a residential building located in Kocaeli, Turkey [26].
Fig. 1. Sketch of a water wall with an opaque building envelope. Steady-state simulations were carried out using the Shear Stress
 T. Wu, C. Lei / Energy and Buildings 112 (2016) 198–208 201

Transport k–ω turbulence model coupled with Surface-to-surface and was large in winter and small in summer. An experiment was
radiation model. The air flow and heat transfer inside the room were carried out to validate the computer model and a fair agreement
numerically investigated with three different constant water tem- between the predicted and experimental results was reported.
peratures, that is, 30 ◦ C, 40 ◦ C, and 50 ◦ C, respectively. The results Din et al. [31] also carried out a transient analysis to investigate
showed that a weak circulation formed in the room with the water the TES effect of a water wall in the northern side of a greenhouse.
wall heating system, which was gentle enough to be unnoticed Various temperatures, namely the plant, the water wall, and the
by occupants, whereas a strong circulation occurred in the room room temperatures for a typical winter day in New Delhi were cal-
with the floor heating system. The water pipe wall heating sys- culated by changing the absorptivity of the blackened surface, the
tem caused a thermal stratification of less than 1 ◦ C/m in the room, thickness of the water wall, and the fraction of the transmitted solar
which did not cause discomfort. For the floor heating system, the energy. It was reported that a north-facing water wall of 27.5-cm
temperature difference between the surface of the floor and the air thickness gave an increase of the plant/room temperature by up to
stream above the floor was about 5 ◦ C. It was then concluded that 4–5 ◦ C at night, whereas a reduction of the plant/room temperature
the water pipe wall heating system had a better thermal perfor- by 3–4 ◦ C was reported during the day.
mance than the floor heating system since better thermal comfort
conditions can be achieved in the former system with a lower water 3.2. Water wall with a semi-transparent building envelope
temperature.
Water wall can also be coupled with a semi-transparent out-
3.1.2. Applications side building façade such as light-permeable plastic or glass, which
The application of a water wall with an opaque building enve- allows part of the sunlight to penetrate from outside into the
lope in a residential house was reported by Sutton and McGregor water wall. The internal side of the water wall can be either semi-
[12]. They built two passive solar heating houses to compare a transparent or opaque (refer to Fig. 2), and depending on the
north-facing concrete wall with a water storage wall in Australia. material of the internal wall, sunlight may penetrate further into
A total of 27 galvanised steel tubes filled with water were installed the room.
in the north side of the room. The monitored data showed that
the house with the water-tube wall saved more than half of the 3.2.1. Performance comparisons with other strategies
annual energy consumption for heating compared to the conven- A series of comparative studies have been carried out to
tional house with the concrete wall and consumed only about 70.8% demonstrate the advantages of water walls coupled with a semi-
of the energy consumed by the concrete wall house. transparent building envelope. Among them, Balcomb et al. [32]
Turner et al. [27] fabricated a water wall comprising 7.6-cm studied the thermal performance of water storage walls under five
diametre plastic tubes embedded into a conventional stud wall. different configurations including: (1) the water storage was placed
The system was operated over a 24-h cycle including 6 h active inside the room with the same temperature as the room; (2) the
charging with hot ambient air during the day and 18 h passive dis- water storage was placed inside the room as an internal wall; (3)
charging. The results showed that the wall temperature was still the water storage was placed at the back of a glass panel; (4) the
2.6 ◦ C higher than the room temperature after 18 h of discharging, water storage was placed at the back of an opaque wall; and (5) the
which reduced the heating load of the house. Similarly, daytime water storage was placed at the back of a transparent insulation
thermal comfort in summer can be achieved by nocturnal water panel. Their results indicated that the water storage placed at the
wall TES actively charged by cool ambient air at night. back of a glass panel and that placed at the back of a transparent
Apart from the residential applications described above, water insulation panel performed appreciably better than the others in
wall with an opaque building envelope in greenhouse has also terms of solar heat gain in winter.
attracted much research attention. Dutt et al. [28] presented a A popular configuration of water walls with a semi-transparent
transient analysis of a winter greenhouse by incorporating vari- building envelope is transwall, which was first designed by Fuchs
ous features such as water storage and movable insulation, etc. and Mcclelland [33]. The transwall had a semi-transparent baffle
It was concluded that the plant temperature would be stabilised inside the water wall. Fuchs and Mcclelland [33] compared the ther-
when a large drum of water was kept inside the greenhouse. Govind mal performance of the transwall with Trombe wall and direct solar
et al. [29] further experimentally and theoretically investigated the gain systems and found that the solar heating fraction in the tran-
effect of TES in a winter greenhouse. A greenhouse with a floor swall passive heating system could be very close to or exceed that
area of 15.4 m2 and a water drum capacity of 3.02 m3 was con- of the Trombe wall.
structed for the purpose of growing early summer vegetables. The The quest for understanding the thermal performance of tran-
results indicated that the thermal energy storage by water offered swall motivated a number of numerical and experimental studies.
a much higher air temperature than the ambient and minimised Nayak [14] compared the thermal performance of two types of
the fluctuations of the air temperature in the greenhouse. south-facing water walls including a water wall with concrete and a
Another TES experiment was carried out by Sørensen [13] in transwall in terms of the heat flux entering a heated space through
a 12 m2 greenhouse in Denmark with four black-painted drums these walls. The results showed that the transwall is more effective
containing 0.8 m3 water. It was reported that a considerable strati-
fication formed in the drums with the bottom temperature 4–10 ◦ C
lower than the top temperature, depending on the season. With the
assistance of the inexpensive thermal energy storage by water, the
air temperature inside the greenhouse was maintained at approx-
imately 5–8 ◦ C above the outside temperature in winter. As a
consequence, the period during which the indoor climate in the
greenhouse was suitable for growing plants could be extended by
at least a month both in spring and in late autumn in Denmark.
A transient analysis using a computer model to investigate the
TES effect of a water mass in a passive greenhouse was reported by
Gupta and Tiwari [30]. They concluded that the temperature fluctu-
ation inside the room decreased with an increase of the water mass Fig. 2. Sketch of a water wall with a semi-transparent building envelope.
202 T. Wu, C. Lei / Energy and Buildings 112 (2016) 198–208

than the concrete water wall in meeting the daytime heating load. performance was evaluated by considering the heat flux entering
However, a concrete water wall is better from the viewpoints of a heated space through the transwall. Their analytical results
reducing temperature swings and the overall day-and-night per- showed that the placement of the trap material between the
formance because it caused a significant phase shift. Using the water columns and the increase of the water column thicknesses
same boundary conditions as those adopted in [14], Nayak [34] both resulted in a reduction of the temperature fluctuation of the
also compared the thermal performance of various types of thermal system. They also found that the thickness of the water column
storage walls, namely Trombe wall, transwall, and water Trombe between the trap and the outside glazing should be smaller than
wall. It was concluded that the transwall was more effective than the thickness of the water column between the trap and the room.
the Trombe wall in meeting the daytime heating load, but the water Following the above investigation, a transient analysis of the
Trombe wall had the best thermal performance with the maximum transwall for a heated room was presented by Upadhya et al. [15],
heat flux released from the wall to the room, less temperature fluc- who calculated the heat flux entering the room through direct and
tuations, and a moderate phase shift, and thus was the preferred indirect gains and investigated the optimum distribution of the
option. water layer thickness in the transwall. The water temperatures and
Nisbet and Kwan [35] studied the energy savings potential of the heat flux entering the room were computed on an hourly basis
a transwall inside a horticultural glasshouse using a computer over a number of days for the typical cold climatic conditions of
model. Two 2.3 m transwalls consisting of clear plastic bags filled Srinagar, India. The results showed that, when the water columns
with water–dye mixture were erected behind external glazing. The inside and outside of the trap material were of an equal thickness,
annual energy savings of this system was reported to be around the water temperature, the indirect gain and the heat flux from the
20%, depending on evapotranspiration, for two sites including the wall to the room were maximised, and the room air temperature
West of Scotland and Southeast England. It was also found that the fluctuation was minimised. They also concluded that in order to
transwall could reduce the energy consumption of the glasshouse achieve the maximum heat gain with less temperature fluctuation,
more effectively at higher latitudes. the thickness of the inside water column should be minimised and
Tiwari et al. [36] compared the net heat gain of various south- the thickness of the trap material should be maximised.
facing walls including a glass wall, a water wall, an active air Nisbet and Mthembu [40] considered a water–dye filled tran-
collector wall, and a transwall. The design parameters such as the swall module and a water–gel filled transwall module in a
thickness of the water wall and transwall and the flow rate of the conventional residential house in the UK. The transwalls were
air collector, etc. were varied under winter conditions for a heated installed behind a double-glaze wall of the house which was well
room. The results showed that the transwall and the water wall insulated. They numerically obtained the optimum dye concentra-
were preferred for residential heating at night due to their large tion and the optimum thickness of the transwalls in terms of annual
thermal capacity, whereas the glass wall and the air collector wall energy savings and experimentally validated the results. The results
were desirable for heating during sunshine hours. Tiwari [37] also showed that the optimum dye concentration of Lissamine Red 3GX
compared the performance of a transwall, a water wall and an was 20 ppm and the optimum water layer thickness was about
isothermal mass for heating in a non-air-conditioned passive solar 150 mm for the weather conditions of western Scotland and south-
house for the harsh cold climate of Srinagar, India. They found that ern France. The energy savings due to the adoption of the two
the room air temperature with the transwall was higher than that different modules located at various sites were also compared. It
with the isothermal mass and the water wall in winter night. was concluded that a transwall module filled with a water–dye
Fernández-González [16] summarised the thermal performance solution released more heat during the evening compared with the
of five different passive solar test-cells (direct gain, Trombe wall, water–gel filled module of a similar size.
water-tube wall, sunspace, and roof pond). For the water-tube Recently, a transparent water storage envelope (TWSE) module,
wall, four translucent water tubes, each holding 0.25 m3 of water, which was very similar to the transwall, was developed by Liu and
were installed in an 11.9 m2 (4.88 m × 2.44 m) test-cell. The results Shen [41,42] for improving the overall energy efficiency of commer-
showed that, despite its large thermal storage capacity (e.g., 2.84 cial buildings. Based on theoretical analyses, they concluded that a
times the thermal storage capacity of the Trombe wall test-cell), the 30–50 mm thick water layer could produce the maximum thermal
water-tube wall test-cell had the second highest maximum swing resistance. Therefore, it was suggested that polycarbonate sheets be
of the diurnal operative temperature, surpassed only by the direct added every 30–50 mm in the TWSE modules to further increase the
gain test-cell. thermal resistance and enhance the overall thermal performance.
Recently, Liu and Feng [38] also compared the performance of Consequently, a 378 mm thick TWSE module was designed which
two passive heating strategies including a Trombe wall and a water- consisted of nine separated water layers with a thickness of 40 mm
tube wall for optimising indoor thermal environment and reducing each, eight polycarbonate sheets with a thickness of 0.25 mm each,
building energy consumption. Semi-transparent water tubes with and the interior and exterior walls of the container made of 8 mm
a total surface area of 57 m2 were installed behind the east and thick high quality resin. The thermal resistance and thermal iner-
west façades of a building so that the room can receive both light tia index of the TWSE module were investigated. It was reported
and heat from the water wall. The average daily heat gain for each that an excellent heat enduring ability in terms of temperature lag
month during the winter was calculated using SolPass software. (15.38 h) was obtained compared with other traditional envelopes
The results indicated that the water-tube wall obtained more than of a similar thickness. They also suggested that the water should be
ten times solar gains than that of the Trombe wall in winter. Their discharged and replaced by air in order to achieve effective thermal
simulation results also showed that these passive solar techniques insulation in winter.
could reduce the heating loads by about 25% when compared to the More recently, Adams et al. [43] conducted a comparative lab-
heating load of the building without adopting these techniques. oratory experiment in order to obtain the optimal thickness of the
water layer in a small-scale water wall. They used a halogen bulb to
3.2.2. Optimization of thermal performance simulate solar irradiation. Three different water layer thicknesses
A number of studies have considered optimising the thermal (i.e. 3”, 6”, and 9”, respectively) were tested by turning on the heat
performance of a water wall with a semi-transparent building source for 5 h, then switching off the heat source and monitoring
envelope. Among them, Sodha et al. [39] investigated the effects the temperature changes of the water and the room. The results
of a trap material inside a transwall and the thickness distribu- showed that the three-inch water wall could not regulate the room
tion of the transwall on the thermal performance. The thermal temperature as the other water walls with larger thicknesses, and
 T. Wu, C. Lei / Energy and Buildings 112 (2016) 198–208 203

water wall to the ambient reduced significantly during night time

and in cold cloudy climates. It was found that such a heat pipe
augmented water wall system provided 52.2% to 107% more solar
thermal energy than the conventional concrete wall and 16.6% to
59.8% more thermal energy than the water wall in three different
locations in the USA.
More recently Albanese et al. [47] developed a computer model
for a passive solar space heating system utilising heat pipes to
transfer latent heat to a storage water tank inside the building.
A bench-scale experimental model was also constructed to ver-
ify the computer model. It was found that the solar heat pipe
system gave a significantly higher solar heating fraction than the
Fig. 3. Sketch of a water wall with PCMs.
other passive technologies, especially in cold and cloudy climates,
and a good agreement was obtained between the computer model
the room temperature variation associated with the six and nine- and the laboratory experiment. In order to further validate the
inch water walls was minimal. above-described computer model and the bench-scale experiment,
a full-scale prototype of the heat pipe system was designed by
3.3. Water wall with phase change materials Robinson et al. [48] in a classroom at the University of Louisville
during the spring heating season of 2010. During this season, the
Although water wall has almost the largest heat capacity among
maximum hourly average heat gain was only 163 W/m2 . Their field
the sensible heat storage systems, its heat capacity is still relatively
results indicated that the thermal storage water tank was heated
small compared to that of the latent heat storage systems (e.g.,
to a sufficiently high temperature to supply heat to the classroom
PCMs). The conjunctive use of water wall with PCMs may take the
even during the coldest days of the season. During a long period
advantages of both the water wall and the PCMs and thus achieve
(4 consecutive days) of low solar isolation, the average hourly heat
better thermal performance. The PCMs may be placed at the exter-
delivery to the classroom remained positive and was never less than
nal side, internal side, or both sides of the water wall (refer to Fig. 3),
16.6 W/m2 . During January to February of 2013, an improved model
resulting in different thermal performance.
of the passive heat pipe system was tested alongside the previous
Tiwari et al. [44] used PCMs in conjunction with a water wall to
prototype by Robinson and Sharp [49]. Significant improvement
form a link wall in a sunspace for a cold climate. The thermal per-
in increasing the heat transfer to the classroom and reducing heat
formance of this composite water/PCM wall was compared with
losses was achieved by adding a copper absorber, thicker insulation,
several other configurations including a sunspace without a link
a rubber adiabatic section, and exposing one condenser directly to
wall, a sunspace with a water wall as a link wall, a sunspace with
the room air.
a PCM wall as a link wall, and a sunspace with movable insulation
panels. They found that the combined PCMs and water wall as a
link wall gave the best thermal performance among all the config- 3.4. Water wall in conjunction with other passive technologies
urations. The lowest temperature fluctuation was observed when
the thickness of the water wall and the PCM wall was equal, and a 3.4.1. With a sunspace
comfortable temperature for the living space was achieved when Sunspace is a direct gain system, whereas water wall is an indi-
the ambient temperature was around 0 ◦ C. rect gain system. When a water wall and a sunspace are adopted
A passive solar heating system using PCMs in heat pipes to col- together (refer to Fig. 5), the solar gains and the thermal per-
lect and transfer solar energy to a short-term TES (refer to Fig. 4) formance of the passive solar heating strategy may be improved
was described by Rice [45]. In this system, solar thermal energy was significantly.
absorbed by an absorber plate and the absorbed heat was trans- Yadav and Tiwari [50] carried out a transient analysis of an
ferred to the PCMs inside the heat pipes by conduction. The liquid opaque water wall acting as the link wall between the sunspace and
form of the PCMs in the evaporator end of the heat pipes thus vapor- the living space in a single-storey residential house. Their results
ised due to the heat from solar energy and the PCM vapour rose to showed that the water wall combined with the sunspace resulted
the condenser end, where it condensed and released its latent heat in significant reduction of the temperature fluctuation in the living
to the water wall. The resultant condensate then flowed back to the space. It was also reported that, for such a system with a 0.05 m
lower end of the heat pipes by gravity, completing the full thermal thick water wall, a quasi-steady state was achieved in the system
cycle. within 3–4 days. Further, it was found that the temperatures of both
Susheela and Sharp [46] also designed a passive solar heat pipe the sunspace and the living space reduced when the thickness of
augmented system using the same concept as the system illustrated the water wall was increased from 0.05 m to 0.10 m.
in Fig. 4 in a south-facing water wall of a building. Their exper- A comparative study of the thermal performance of an opaque
imental and simulation results indicated that the heat loss from water wall, a transwall and an air collector in a sunspace has been

Fig. 4. Schematic of a water wall with heat pipes [46].

204 T. Wu, C. Lei / Energy and Buildings 112 (2016) 198–208

to the building. The TRNSYS software was used to simulate the

hourly indoor temperature and the required heating or cooling load
in Northern China. The results showed that active cooling was not
required in summer as the extreme indoor temperature of 35 ◦ C was
exceeded for only 47 out of 8760 h. In winter, the measured indoor
air temperature remained at 13.7 ◦ C under the condition of no space
heating when the outside air temperature dropped to −0.4 ◦ C. The
minimum indoor temperature was 0.5 ◦ C when the outdoor tem-
perature was at an extreme of −13.1 ◦ C. Therefore, the combined
water wall with natural ventilation system can significantly reduce
the required heating and cooling loads. Compared with massless
Fig. 5. Sketch of a water wall combined with sunspace. walls, the passive system may reduce the maximum indoor tem-
perature by up to 4 ◦ C in summer and increase the minimum indoor
temperature by up to 3 ◦ C in winter.
conducted by Tiwari and Kumar [51]. They numerically compared
the temperatures in the sunspace and the living space obtained
with the different link walls. According to their calculations, the 3.4.3. With evaporative cooling
transwall as a link wall produced the highest room air temperature Evaporative cooling works by adding water vapour into air,
in winter night among the three different configurations consid- resulting in a decrease of the air temperature and an increase of
ered in their investigation. its water content. When the evaporative cooling is coupled with
Tiwari and Singh [52] carried out another comparative study of the water wall, the thermal comfort in terms of both temperature
various passive heating strategies including a glass wall, a semi- and humidity may be improved.
transparent water wall, a sunspace, a wall air heater, and a roof Moustafa and Aripin [57] evaluated the thermal performance
air heater. The hourly and monthly room temperatures in a non- of combined water wall and porous ceramic pipes for evaporative
air-conditioned building were calculated for sunny cold climatic cooling in a hot arid climate using CFD simulation. A three-
conditions in India. It was found that the use of a semi-transparent dimensional (3D) model with inner dimensions of 5 m × 5 m × 3 m
water wall in a sunspace can significantly reduce the fluctuation of (L × W × H) was established using ANSYS Workbench, and ANSYS
the temperature and the peak temperature was shifted towards the Fluent was used for the simulation of heat transfer using a lam-
night time. The minimum temperature fluctuation was achieved inar flow model coupled with the Rosseland radiation model. A
with the configuration of sunspace + water wall + roof air heater. steel container water wall with a dimension of 4 m × 0.4 m × 3 m
Recently, Sánchez-Ostiz et al. [53] constructed four proto- (L × W × H) was placed behind the south façade, and the sur-
types in two containers to explore the thermal performance of face area of the porous ceramic pipe was varied from 4.62 m2 to
an attached sunspace with water for thermal energy storage. The 10.85 m2 . The water temperature was set to 15, 17.5, 20, 22.5, and
results showed that a sunspace with water for TES improved the 25 ◦ C, respectively in summer and to 30, 32.5, 35, 37.5, and 40 ◦ C,
indoor thermal comfort of the adjacent room during both winter respectively in winter, whereas the air inlet temperature was fixed
and summer. The experimental data was also used to validate a at 40 ◦ C in summer and 3 ◦ C in winter with a fixed air flow veloc-
TRNSYS model. A fair agreement between the experimental data ity of 1 m/s. Steady-state simulation was conducted for every hour
and the numerical simulation performed using the TRNSYS model during the 24 h of one typical summer day and one typical winter
was reported. day, representing the extreme weather conditions in Luxor, Egypt.
They found that using this type of water wall, the room tempera-
3.4.2. With a solar chimney ture was 4–10 ◦ C cooler than the outdoor temperature in summer
Solar chimney is a passive strategy for enhancing stack-driven and was 4–15 ◦ C higher than the outdoor temperature in winter.
ventilation by buoyancy [54]. When incorporating a water wall into
the solar chimney (refer to Fig. 6), the night time ventilation of the 3.5. Summary of water wall research
solar chimney is expected to be improved.
Wang et al. [55,56] studied numerically and experimentally the The above review has covered the most common configu-
natural ventilation in a passive solar house by solar chimney inte- rations of water wall systems reported in the literature. Other
grated with a water thermal storage wall. The single-storey passive configurations are possible and are worth investigating. The actual
solar house with a sunspace had a floor area of 700 m2 . A water configuration of a water wall system will depend on the location
thermal storage wall with a total of 41 t of water was retrofitted and climate condition for its application. Table 1 summaries the

Table 1
Locations and climate conditions for which water wall research has been conducted.

Location Climate References

Winter Summer
√
Australia [12]
√
Europe [13,53]
√ √
[40]
√
India [14,15,20,23,28–31,34–37,44,50–52,59–61]
√
[21,24]
√
USA [16,18,19,22,27,32,33,39,46–49]
√
Turkey [26]
√
China [38]
√
[41,42]
√ √
[55,56]
√
Africa [57]
√
Unspecified [25,43]
Fig. 6. Sketch of a water wall with solar chimney (self-ventilation mode).
 T. Wu, C. Lei / Energy and Buildings 112 (2016) 198–208 205

location and climate information for the studies reviewed above. It of solar radiation, whereas the solarium was the most effective in
is clear in Table 1 that only a limited number of locations have been transferring heat into living spaces at a low level of solar radiation.
studied, and thus it is worth considering the application of water
wall systems in other locations. Moreover, it is also clear in this
4.1.2. Transient analysis
table that much more attention has been paid to the winter climate
Transient analysis is more common in water wall applications.
condition than the summer climate condition. Since a water wall
A large body of literature on the transient analysis of water wall
system will operate over all seasons around the year once it is con-
systems is available, which focuses mainly on the temperature
structed, its performance during other seasons including summer
fluctuation and thermal lag (the shift of the peak temperature) in
conditions at various locations needs to be evaluated.
building systems incorporating a water wall [15,28,36,40,50,52,61].
Transient analysis has been commonly used in simulations over a
4. An overview of methodologies for water wall research period of 24 h, that is, one full day and night cycle only. An example
of such analysis can be found in [61], which evaluated the overall
In general, three types of research approaches have been heat transfer coefficients of four passive heating concepts including
adopted in the water wall research, which are analytical approach Trombe wall, Trombe wall with vents, water wall and transwall. It
based on a heat balance model (HBM), experimental modelling was reported in [61] that the water wall yielded a positive heat
and numerical modelling based on computational fluid dynamics flux entering a room of a constant temperature throughout the
(CFD). Each approach has its pros and cons. In what follows, the 24-h period, whereas a negative heat flux was obtained with the
advantages and disadvantages of these research methodologies are transwall. Transient analysis has also been used to simulate water
described. wall systems for longer periods of time. Dutt et al. [28] presented a
transient analysis of a greenhouse with water drums for two con-
secutive days in winter. The results showed that the fluctuations
4.1. Heat balance model
of the plant temperature decreased in the second day. Other tran-
sient analyses over longer periods include the study by Upadhya
The heat balance model uses the energy conservation concept
et al. [15] for 3 days, that of Yadav and Tiwari [50] for 4 days, and
to establish energy balance associated with conductive, convective,
the study of Tiwari et al. [36] for up to 6 days. It is worth noting that
and radiative heat transfer in order to obtain the room air temper-
Nisbet and Mthembu [40] and Tiwari and Singh [52] extended the
ature and the water temperature. This approach usually embodies
transient analyses for a period of a full year.
the following five major assumptions [58]:

• uniform fluid (air or water) temperature throughout a selected 4.1.3. Building energy simulation programs
zone; There is a wide range of building energy simulation (BES) pro-
• uniform surface temperatures; grams which use the heat balance model as the core of their
• grey radiative surfaces; analyses. Examples of BES programs include EnergyPlus and TRN-
• diffuse radiative surfaces; SYS, etc. [62]. With the assistance of these programs, the simulation
• one-dimensional heat conduction through all the surfaces of of a large building over an annual cycle could be completed in
interest. just a few minutes. Therefore, the BES programs are very useful
for the design of energy efficient buildings, especially for mak-
ing critical decisions at the early stage of the design. Wang et al.
It is clear that the temperature stratification in the water col-
[55,56] used TRNSYS to investigate the energy consumption and
umn and the living space is neglected in the heat balance model.
thermal comfort of a passive solar house with a water thermal
As a consequence, the HBM can only capture the main features
storage wall. The simulation of a whole year cycle was performed,
of a system and is suitable for assessing the bulk performance of
and they reported that their building design could reduce 8.6%
the system, rather than providing the details of temperature varia-
of the yearly energy consumption and raise the thermal comfort
tions throughout the system. Nevertheless, this approach is highly
evaluation index by 12.9%. Another simulation using EnergyPlus
effective and is capable of simulating large-scale systems over an
and TRNSYS was reported by Sánchez-Ostiz et al. [53] for a sun-
extended period of time. Therefore, the HBM can be used for energy
space with water storage in both winter and summer conditions.
savings analysis over a daily, monthly, or even yearly cycle.
Their results showed that a sunspace with water storage required
12.9–16.7% less energy for heating than a sunspace without water
4.1.1. Steady state analysis storage.
Steady-state analyses of water wall systems using the heat It is seen from the above that the BES is a powerful tool and has
balance model have been reported in the literature. Using this a great potential for applications to the research of water wall sys-
approach, Bansal and Thomas [59] studied the steady-state ther- tems. However, as for the HBM described above, the BES programs
mal efficiency of four indirect gain passive solar heating systems, cannot resolve the details of the complex heat transfer processes
namely mass wall, Trombe wall, pure water wall, and solarium. occurring both inside the water wall and in the living space. Fur-
The calculated results indicated that the pure water wall and the thermore, the various heat transfer coefficients characterising the
Trombe wall were more efficient than the other systems at high heat transfer between surfaces and fluids are usually determined
incident solar radiation. When it was applied to a building, in using empirical correlations in the BES programs. Different BES pro-
January, of India the pure water wall delivered the largest amount of grams may adopt different empirical correlations, and thus the
heat into the living spaces during the day but also lost the maximum results obtained with different BES programs may vary signifi-
amount of heat to the ambient at night. Accordingly, night-time cantly from one to another even for the same building configuration
insulation was recommended for the pure water wall system in [63]. Mirsadeghi et al. [64] reviewed the various empirical correla-
winter. A similar study was carried out by Bhandari and Bansal tions for determining external convective heat transfer coefficients
[60] to identify the heat gains and losses in various building config- adopted in some popular BES programs and reported deviations
urations including direct gain, mass wall, water wall, Trombe wall, up to ±30% in the yearly cooling energy demand and ±14% in the
and solarium, etc. They concluded that the water wall and Trombe hourly peak cooling energy demand of an isolated, well-insulated
wall were more suitable for applications with a slightly higher level building.
206 T. Wu, C. Lei / Energy and Buildings 112 (2016) 198–208

In summary, the heat balance model can be used to quickly

predict the major features of a building system and provide a use-
ful basis for further analysis or testing. Nevertheless, the results
obtained with this method are dependent on the selected empiri-
cal correlations for determining the heat transfer coefficients and
thus may vary significantly from real conditions.

4.2. Experimental modelling

Experimental modelling is a direct and effective way to inves-

tigate the thermal performance of water wall systems. Most of the
existing experimental studies have been conducted under real cli-
mate conditions, whereas a few studies have been conducted in
laboratory under controlled conditions (e.g., [43]). Experimental Fig. 7. Statistics of the research publications related to water wall for thermal com-
fort and thermal energy storage.
modelling under real climate conditions has many advantages. It
gathers real-time performance data under realistic climate condi-
tions that include all aspects, for example solar irradiation, solar Despite its ability to accurately resolve the full details of the
angle, ambient temperature, wind direction, and speed, etc. In the convective flows in both water and air as well as the tempera-
previous research on water walls, significant attention has been ture stratification in buildings, the CFD approach also has some
paid to the heating performance of water walls in winter, whereas limitations. Firstly, the application of the CFD approach requires
using the water wall as a passive cooling strategy in summer has good knowledge of and experience with both the CFD technique
received relatively less attention. and the physical processes under investigation. Secondly, solving
On the other hand, experimental modelling under realistic the governing equations numerically using a computer inevitably
climate conditions also has many disadvantages. Firstly, the exper- introduces numerical errors, though the errors may be minimised
imental systems are generally inflexible. Once an experimental by proper verification and validation. The data for validating CFD
prototype is established, it is usually difficult or too expensive to models usually comes from experiment. Finally, some problems
modify the prototype for different designs. In contrast, reduced- may be computationally intensive due to the complex geometry
scale models for laboratory experiment may be easily modified for and physical processes involved in the systems. It is not uncom-
different purposes. Secondly, full-scale experimental modelling is mon to take up to several months to simulate a complex problem
very expensive compared with reduced scale lab experiment. Full- using CFD for only a few hours of the flow time.
scale building models require significant land area for installation In summary, CFD simulations are relatively inexpensive and
and involve significant costs for construction. Thirdly, it usually have the ability to simulate both realistic and idealised conditions,
takes a long time to test full-scale building models. If the testing but they require significant knowledge of the physical processes
involves seasonal changes from winter to summer, or vice versa, under consideration and experience with the CFD technique, and
it may take as long as half a year to complete the testing. Further- thus should be applied with care.
more, the full-scale prototype experiment is often constrained by
unpredictable climate conditions. The full-scale building models 4.4. Summary of methodologies for water wall research
may be damaged under extremely bad weather conditions. Lastly
but not the least, unexpected system and recording errors may The application of water wall in buildings has a history of more
occur, though they can be minimised to a low level with careful cal- than half a century. Nevertheless, the research on the thermal
ibration. It is usually impossible to repeat full-scale testing under performance of water wall systems has lagged behind their appli-
realistic climate conditions. cations. Fig. 7 shows a statistics of the publicly available research
In conclusion, experimental modelling is an effective way for papers on water wall systems for TES and thermal performance pur-
studying the thermal performance and thermal comfort of water pose. The statistics is obtained from the literature reviewed in the
wall systems. However, the many limitations of this approach present study, which is fairly comprehensive. It is seen in this figure
described above have prevented it from being widely adopted by that the research on water wall systems did not receive much atten-
researchers. tion before 1980s, and back then the heat balance model was the
only method applied to this research. However, significant research
4.3. CFD modelling attention was paid to water wall systems in the 1980s. During that
period of time experimental modelling was also adopted in water
CFD is a powerful numerical approach that has been widely wall research, though the majority of the studies were still based
adopted to solve a wide range of scientific and engineering prob- on the heat balance model. In the subsequent two decades, the
lems involving fluid flow, heat transfer and species transport, etc. research interest on water wall seemed to decline, and the methods
There are many successful examples of using CFD simulation to adopted in the water wall research were still limited to the heat bal-
investigate passive solar systems (e.g., [65–67]). However, few ance model and experimental modelling. It is worth noting that the
papers have reported CFD modelling of water wall systems. Karabay water wall has attracted significant research interest again over the
et al. [26] studied the thermal performance of a concrete wall with past five years, and for the first time, the CFD approach has been
embedded water pipes filled with constant-temperature water. applied to the water wall research although it is for steady state
The flow was assumed to be turbulent and two-dimensional, and conditions only. The adoption of the CFD modelling approach will
the simulations were performed for steady state conditions only. push the water wall research onto a high level.
Moustafa and Aripin [57] evaluated the thermal performance of a
combined water wall and porous ceramic pipes system for evapo- 5. Conclusions
rative cooling in a 3D model. In their numerical model, the flow was
assumed to be laminar, and the simulations were again performed The water wall research has experienced a prosperous period
for steady state conditions. as well as a depressed time over the last few decades. In recent
 T. Wu, C. Lei / Energy and Buildings 112 (2016) 198–208 207

years, the interest in water walls has resurged due to the concerns Acknowledgement
regarding energy crisis and the global climate change. The current
status of research and development in water wall systems for both The financial support of the Australian Research Council through
residential and commercial buildings and the relevant research the Discovery Projects grant DP130100900 is gratefully acknowl-
methodologies have been reviewed in this paper. edged.
In typical winter applications, the water wall absorbs solar radi-
ation and stores the heat during the day. It then releases the heat References
at night to partially offset the heating load of the buildings. In sum-
[1] Energy Information Administration, International Energy Outlook 2013, US
mer applications, the water wall prevents the living space from
Department of Energy, Washington DC, 2013.
overheating during the day and thus makes indoor environments [2] L. Pérez-Lombard, J. Ortiz, C. Pout, A review on buildings energy consumption
more comfortable. The water wall also has the capacity to moderate information, Energy Build. 40 (3) (2008) 394–398.
[3] V. Basecq, G. Michaux, C. Inard, P. Blondeau, Short-term storage systems of
large temperature swings between the day and night. Therefore, it
thermal energy for buildings: a review, Adv. Build. Energy Res. 7 (1) (2013)
has a great potential for maintaining thermal comfort and cutting 66–119.
energy consumption in buildings. [4] F. Kuznik, D. David, K. Johannes, J.J. Roux, A review on phase change materials
A number of different configurations of water wall systems have integrated in building walls, Renew. Sustain. Energy Rev. 15 (1) (2011)
379–391.
been studied previously, and they can be generally grouped into [5] L.F. Cabeza, A. Castell, C. Barreneche, A. De Gracia, A.I. Fernández, Materials
four types: (1) a water wall with an opaque building envelope; (2) a used as PCM in thermal energy storage in buildings: a review, Renew. Sustain.
water wall with a semi-transparent building envelope; (3) a water Energy Rev. 15 (3) (2011) 1675–1695.
[6] A. Pasupathy, R. Velraj, Effect of double layer phase change material in
wall with PCMs; and (4) a water wall in conjunction with other building roof for year round thermal management, Energy Build. 40 (3) (2008)
passive technologies. Different configurations have different ther- 193–203.
mal performance, and thus are suitable for different applications, [7] H.Y. Chan, B. Riffat, J. Zhu, Review of passive solar heating and cooling
technologies, Renew. Sustain. Energy Rev. 14 (2) (2010) 781–789.
depending on the location and/or the season. [8] S.B. Sadineni, S. Madala, R.F. Boehm, Passive building energy savings: a review
Three types of approaches have been adopted in the water of building envelope components, Renew. Sustain. Energy Rev. 15 (8) (2011)
wall research including analysis using the heat balance model, 3617–3631.
[9] O. Saadatian, K. Sopian, C.H. Lim, N. Asim, M.Y. Sulaiman, Trombe walls: a
experimental modelling, and CFD modelling. The present litera-
review of opportunities and challenges in research and development, Renew.
ture survey shows that the majority of the previous investigations Sustain. Energy Rev. 16 (8) (2012) 6340–6351.
have adopted the heat balance model which can produce data [10] S. Stevanović, Optimization of passive solar design strategies: a review,
Renew. Sustain. Energy Rev. 25 (2013) 177–196.
for a quick assessment of the system performance. However, the
[11] D.A. Bainbridge, A Water Wall Solar Design Manual, 2005 [Online]. Available
heat balance model has many limitations as it cannot resolve from http://www.builditsolar.com/Projects/.
the details of the complex heat transfer processes occurring both SpaceHeating/AWaterWallIntro.pdf.
inside the water wall and in the adjacent space. In contrast, the [12] R.G. Sutton, R.J. McGregor, Solarwall project: two demonstration houses with
passive solar heating in Tasmania, Archit. Sci. Rev. 29 (1) (1986) 2–11.
CFD approach can account for the complex heat transfer processes [13] B. Sørensen, Experiments with energy-storage in a high-latitude greenhouse,
and resolve full details of the convective flows in both water and Sol. Energy 42 (4) (1989) 293–301.
air. However, the application of the CFD approach to water wall [14] J.K. Nayak, Thermal performance of a water wall, Build. Environ. 22 (1) (1987)
83–90.
research is still in its infancy, which is in part due to the fact that [15] M. Upadhya, G.N. Tiwari, S.N. Rai, Optimum distribution of water-wall
the CFD modelling of building systems is usually very resource- thickness in a transwall, Energy Build. 17 (2) (1991) 97–102.
intensive. [16] A. Fernández-González, Analysis of the thermal performance and comfort
conditions produced by five different passive solar heating strategies in the
The present review has revealed a number of contradictory United States midwest, Sol. Energy 81 (5) (2007) 581–593.
claims reported in the literature. For instance, Balcomb et al. [32] [17] D.A. Bainbridge, K. Haggard, P. Cooper, Return of the Water Wall, Solar Today,
concluded that a water wall was better than a double-glass wall Boulder, Colorado, 2007, pp. 38–41 (July/August).
[18] J.D. Balcomb, R.D. McFarland, A simple empirical method for estimating the
in a winter application, whereas Liu and Shen [42] argued that the
performance of a passive solar heated building of the thermal storage wall
water wall should be converted into a double-glass wall because type, Proceedings of the 2nd National Passive Solar Conference, 1978.
water was not a prominent heat insulating material compared with [19] R.D. McFarland, J.D. Balcornb, Effect of design parameter changes on the
performance of thermal storage wall passive systems, Proceedings of the 3rd
air. Another example of contradicting claims is regarding the heat
National Passive Solar Conference, 1979.
flux entering a heated space through transwall. Sodha et al. [39] [20] M.S. Sodha, S.C. Kaushik, J.K. Nayak, Performance of trombe walls and roof
found that the thickness of the water column between the trap pond systems, Appl. Energy 8 (3) (1981) 175–191.
and the outside glazing should be smaller than the width of the [21] M.S. Sodha, S.C. Bhardwaj, S.C. Kaushik, Thermal load leveling of heat flux
through an insulated thermal storage water wall, Energy Res. 5 (2) (1981)
water column between the trap and the room. However, Upadhya 155–163.
et al. [15] reported that the thickness of inside water column of [22] J.K. Nayak, N.K. Bansal, M.S. Sodha, Analysis of passive heating concepts, Sol.
the transwall should be minimised. These controversies suggest Energy 30 (1) (1983) 51–69.
[23] S.C. Kaushik, S. Kaul, Thermal comfort in buildings through a mixed
that, although much attention has been paid to the water wall water-mass thermal storage wall, Build. Environ. 24 (3) (1989)
research, the understanding of the complex flow and heat trans- 199–207.
fer processes in water wall systems is still incomplete. Moreover, [24] M.S. Sodha, J. Kaur, R.L. Sawhney, Effect of storage on thermal performance of
a building, Int. J. Energy Res. 16 (8) (1992) 697–707.
there are still many gaps in the available information about the flow [25] L. Marmoret, P. Glouannec, O. Douzane, A. t’Kint de Roodenbeke, M.
and heat transfer processes in water wall systems. A particular issue Queneudec, Use of a cellular clayey concrete for a wall specially fitted with
which is important but has been commonly ignored in the exist- water pipes, Energy Build. 31 (1) (2000) 80–95.
[26] H. Karabay, M. Arıcı, M. Sandık, A numerical investigation of fluid flow and
ing literature is the temperature stratification in the water column
heat transfer inside a room for floor heating and wall heating systems, Energy
and in the adjacent space. The issue may be resolved by running a Build. 67 (2013) 471–478.
CFD model of the physical system. Furthermore, the existing water [27] R.H. Turner, G. Liu, Y.A. Cengel, C.P. Harris, Thermal storage in the walls of a
solar house, J. Sol. Energy Eng. 116 (4) (1994) 183–193.
wall research is mainly concerned with the thermal performance of
[28] D.K. Dutt, S.N. Rai, G.N. Tiwari, Y.P. Yadav, Transient analysis of a winter
water wall systems under winter climate conditions, whereas their greenhouse, Energy Convers. Manag. 27 (2) (1987) 141–147.
performance in other seasons has received little attention. There- [29] Govind, Rashmi, N.K. Bansal, I.C. Goyal, An experimental and theoretical study
fore, further attention should be paid to the water wall research of a plastic film solar greenhouse, Energy Convers. Manag. 27 (4) (1987)
395–400.
in order to advance our understanding of the water wall systems [30] A. Gupta, G.N. Tiwari, Computer model and its validation for prediction of
and to guide the design of such systems for various locations and storage effect of water mass in a greenhouse: a transient analysis, Energy
climate conditions. Convers. Manag. 43 (18) (2002) 2625–2640.
208 T. Wu, C. Lei / Energy and Buildings 112 (2016) 198–208

[31] M. Din, G.N. Tiwari, M.K. Ghosal, N.S.L. Srivastava, M. Imran Khan, M.S. Sodha, [51] G.N. Tiwari, S. Kumar, Thermal evaluation of solarium-cum-passive solar
Effect of thermal storage on the performance of greenhouse, Int. J. Energy Res. house, Energy Convers. Manag. 32 (4) (1991) 303–310.
27 (1) (2003) 79–92. [52] G.N. Tiwari, A.K. Singh, Comparative studies of different heating techniques of
[32] J.D. Balcomb, J.C. Hedstrom, R.D. Mcfarland, Simulation analysis of passive a non-air conditioned building, Build. Environ. 31 (3) (1996) 215–224.
solar heated buildings—preliminary results, Sol. Energy 19 (3) (1977) [53] A. Sánchez-Ostiz, A. Monge-Barrio, S. Domingo-Irigoyen, P.
277–282. González-Martínez, Design and experimental study of an industrialized
[33] R. Fuchs, J.F. Mcclelland, Passive solar heating of buildings using a transwall sunspace with solar heat storage, Energy Build. 80 (2014) 231–246.
structure, Sol. Energy 23 (2) (1979) 123–128. [54] R. Khanal, C. Lei, Solar chimney—a passive strategy for natural ventilation,
[34] J.K. Nayak, Transwall versus Trombe Wall—relative performance studies, Energy Build. 43 (8) (2011) 1811–1819.
Energy Convers. Manag. 27 (4) (1987) 389–393. [55] W. Wang, Z. Tian, X. Niu, X. Xu, Investigation on a passive solar house
[35] S.K. Nisbet, C.M. Kwan, The application of the transwall to horticultural equipped with water thermal storage wall, Appl. Mech. Mater. 178 (2012)
glasshouses, Sol. Energy 39 (6) (1987) 473–482. 193–196.
[36] G.N. Tiwari, M. Upadhya, S.N. Rai, Relative thermal performances of south [56] W. Wang, Z. Tian, Y. Ding, Investigation on the influencing factors of energy
walls in winter, Energy Build. 17 (4) (1991) 313–320. consumption and thermal comfort for a passive solar house, Energy Build. 64
[37] G.N. Tiwari, Design of a non-airconditioned passive solar house for the cold (2013) 218–223.
climate of Srinagar, Build. Environ. 26 (4) (1991) 371–380. [57] M.A. Moustafa, S. Aripin, CFD evaluation of the pottery water wall in a hot arid
[38] Y. Liu, W. Feng, Integrating passive cooling and solar techniques into the climate of Luxor, Egypt, J. Green Build. 9 (4) (2014) 175–189.
existing building in South China, Adv. Mater. Res. 368 (2012) 3717–3720. [58] C.O. Pedersen, D.E. Fisher, R.J. Liesen, Development of a heat balance
[39] M.S. Sodha, N.K. Bansal, S. Ram, Periodic analysis of a transwall: a passive procedure for calculating cooling loads, ASHRAE Trans. 103 (2) (1997)
heating concept, Appl. Energy 14 (1) (1983) 33–48. 459–468.
[40] S.K. Nisbet, N.S. Mthembu, Transwall modeling using effective conductivities, [59] N.K. Bansal, P.C. Thomas, A simple procedure for selection and sizing of
Sol. Energy 49 (2) (1992) 127–138. indirect passive solar heating systems, Build. Environ. 26 (4) (1991) 381–387.
[41] X. Liu, T. Shen, Conceptual development of transparent water storage [60] M.S. Bhandari, N.K. Bansal, Solar heat gain factors and heat loss coefficients
envelopes, Archit. Sci. Rev. 50 (1) (2007) 18–25. for passive heating concepts, Sol. Energy 53 (2) (1994) 199–208.
[42] X. Liu, T. Shen, The development of transparent water storage envelopes [61] N.K. Bansal, M.S. Sodha, S. Ram, S.P. Singh, Evaluation of complex heat transfer
(TWSE) through theoretical thermal and optical analyses, Archit. Sci. Rev. 51 coefficients for passive heating concepts, Build. Environ. 22 (4) (1987)
(2) (2008) 109–123. 259–268.
[43] S. Adams, M. Becker, D. Krauss, C. Gilman, Not A dry subject: optimizing water [62] S. Oh, Origins of Analysis Methods in Energy Simulation Programs Used for
trombe walls, Proceedings of the Solar 2010 ASES conference, 2010. High Performance Commercial Buildings, A & M University, Texas, 2013
[44] G.N. Tiwari, Y.P. Yadav, S.A. Lawrence, Performance of a solarium: an (Master thesis).
analytical study, Build. Environ. 23 (2) (1988) 145–151. [63] H.S. Rallapalli, A Comparison of EnergyPlus and eQUEST Whole Building
[45] W.J. Rice, Performance of passive and hybrid solar heating systems, Int. J. Energy Simulation Results for a Medium Sized Office Building, Arizona State
Ambient Energy 5 (4) (1984) 171–186. University, Tempe, Arizona, 2010 (Master thesis).
[46] N. Susheela, M.K. Sharp, Heat pipe augmented passive solar system for [64] M. Mirsadeghi, D. Cóstola, B. Blocken, J.L.M. Hensen, Review of external
heating of buildings, J. Energy Eng. 127 (1) (2001) 18–36. convective heat transfer coefficient models in building energy simulation
[47] V. Albanese, S. Robinson, G. Brehob, M.K. Sharp, Simulated and experimental programs: Implementation and uncertainty, Appl. Therm. Eng. 56 (1) (2013)
performance of a heat pipe assisted solar wall, Sol. Energy 86 (5) (2012) 134–151.
1552–1562. [65] R. Khanal, C. Lei, Flow reversal effects on buoyancy induced air flow in a solar
[48] B.S. Robinson, E. Chmielewski, A. Knox-Kelecy, G. Brehob, M.K. Sharp, Heating chimney, Sol. Energy 86 (9) (2012) 2783–2794.
season performance of a full-scale heat pipe assisted solar wall, Sol. Energy 87 [66] S. Hussain, P.H. Oosthuizen, Numerical investigations of buoyancy-driven
(2013) 76–83. natural ventilation in a simple three-storey atrium building and thermal
[49] B.S. Robinson, M.K. Sharp, Heating season performance improvements for a comfort evaluation, Appl. Therm. Eng. 57 (1) (2013) 133–146.
solar heat pipe system, Sol. Energy 110 (2014) 39–49. [67] B.K. Koyunbaba, Z. Yilmaz, K. Ulgen, An approach for energy modeling of a
[50] Y.P. Yadav, G.N. Tiwari, Analytical model of a solarium for cold climate—a new building integrated photovoltaic (BIPV) Trombe wall system, Energy Build. 67
approach, Energy Convers. Manag. 28 (1) (1988) 15–20. (2013) 680–688.