Content uploaded by Brian Ford
Author content
All content in this area was uploaded by Brian Ford on Dec 27, 2019
Content may be subject to copyright.
This article was downloaded by: [Brian Ford]
On: 25 June 2012, At: 04:34
Publisher: Routledge
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,
37-41 Mortimer Street, London W1T 3JH, UK
Building Research & Information
Publication details, including instructions for authors and subscription information:
http://www.tandfonline.com/loi/rbri20
Passive downdraught evaporative cooling: performance
in a prototype house
Brian Ford a , Robin Wilson a , Mark Gillott a , Omar Ibraheem a , Jose Salmeron b &
Francisco Jose Sanchez c
a Department of Architecture & Built Environment, University of Nottingham, Nottingham,
NG7 2RD, UK
b School of Engineering, University of Seville, 41092, Seville, Spain
c Department of Machines and Thermal Engines, University of Cadiz, Calle de Chile 1,
11002, Cadiz, Spain
Available online: 25 May 2012
To cite this article: Brian Ford, Robin Wilson, Mark Gillott, Omar Ibraheem, Jose Salmeron & Francisco Jose Sanchez (2012):
Passive downdraught evaporative cooling: performance in a prototype house, Building Research & Information, 40:3, 290-304
To link to this article: http://dx.doi.org/10.1080/09613218.2012.669908
PLEASE SCROLL DOWN FOR ARTICLE
Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions
This article may be used for research, teaching, and private study purposes. Any substantial or systematic
reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to
anyone is expressly forbidden.
The publisher does not give any warranty express or implied or make any representation that the contents
will be complete or accurate or up to date. The accuracy of any instructions, formulae, and drug doses should
be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims,
proceedings, demand, or costs or damages whatsoever or howsoever caused arising directly or indirectly in
connection with or arising out of the use of this material.
RESEARCH PAPER
Passive downdraught evaporative cooling:
performance in a prototype house
Brian Ford
1
, Robin Wilson
1
, Mark Gillott
1
, Omar Ibraheem
1
, Jose Salmeron
2
and
Francisco Jose Sanchez
3
1
Department of Architecture & Built Environment,University of Nottingham,Nottingham NG7 2RD, UK
E-mail: Brian.Ford@nottingham.ac.uk
2
School of Engineering, University of Seville,41092 Seville, Spain
E-mail: jms@tmt.us.es
3
Department of Machines and Thermal Engines,University of Cadiz,Calle de Chile 1,11002 Cadiz, Spain
E-mail: francisco.£or@uca.es
The design and performance of a passive downdraught evaporative cooling (PDEC) system are reported. This is the first
application of this technique for housing in Europe. The system was integrated within a house designed and built by
University of Nottingham students for the 2010 Solar Decathlon Europe event in Madrid, Spain. Testing of the thermal
performance of the house occurred for a limited period during June 2010. Performance data have demonstrated that
the combination of a high-performance building envelope and PDEC provided comfortable conditions inside the house
over a period of increasingly warm weather. Estimates of the cooling achieved, based on the measured results, were
compared with the energy required by the system to derive an indicative range of coefficient of performance (CoP)
values under varying ambient conditions. The results suggest that PDEC can deliver significant energy savings and
achieve comfortable thermal conditions without the need for additional mechanical cooling. This technique may
therefore have wider relevance to housing in central and southern Spain, and other hot, dry regions of the world.
Keywords: alternative technology,building performance, cooling, downdraught, evaporative, housing,innovation, passive
Il est rendu compte de la conception et des performances d’un syste
`me de refroidissement passif e
´vaporatif a
`courant d’air
descendant (PDEC). Il s’agit de la premie
`re application de cette technique dans le domaine du logement en Europe. Le
syste
`me a e
´te
´inte
´gre
´a
`une maison conc¸ue et construite par des e
´tudiants de l’Universite
´de Nottingham pour la
manifestation Solar Decathlon Europe 2010 qui s’est tenue a
`Madrid, en Espagne. Les donne
´es de performance ont
de
´montre
´que la combinaison d’une enveloppe de ba
ˆtiment haute performance et d’un syste
`me PDEC ont assure
´des
conditions agre
´ables a
`l’inte
´rieur de la maison sur une pe
´riode ou
`le temps est devenu de plus en plus chaud. Les
estimations quant au refroidissement obtenu, sur la base des re
´sultats mesure
´s, ont e
´te
´compare
´es a
`l’e
´nergie exige
´e
par le syste
`me de fac¸on a
`de
´terminer une plage indicative des valeurs du coefficient de performance (CoP) dans
diffe
´rentes conditions ambiantes. Les re
´sultats sugge
`rent que le PDEC peut assurer d’importantes e
´conomies d’e
´nergie
et cre
´er des conditions thermiques agre
´ables sans qu’il y ait besoin d’un refroidissement me
´canique supple
´mentaire.
Cette technique peut par conse
´quent pre
´senter un inte
´re
ˆt plus large pour le logement en Espagne centrale et du sud,
ainsi que dans les autres re
´gions chaudes et se
`ches du monde.
Mots cle´s: technologie alternative, performance des ba
ˆtiments, refroidissement, courant d’air descendant, e
´vaporatif,
logement, innovation, passif
Introduction
The purpose of this paper is to report on the summertime
performance of a prototype passively cooled house in
Madrid, Spain, as part of ongoing research at the Univer-
sity of Nottingham, UK, into the application of passive
and hybrid downdraught cooling (PHDC) in buildings
BUILDING RESEARCH &INFORMATION (2012) 40(3), 290 – 304
Building Research & Information ISSN 0961-3218 print ⁄ISSN 1466-4321 online #2012 Taylor & Francis
http: ⁄ ⁄www.tandfonline.com ⁄journals
http://dx.doi.org/10.1080/09613218.2012.669908
Downloaded by [Brian Ford] at 04:34 25 June 2012
in different parts of the world. The intention was to
establish whether thermal comfort could be achieved
and what the energy balance (coefficient of performance
– CoP) would be, as part of a search for alternatives to
conventional air-conditioning.
The demand for air-conditioning for housing is increas-
ing in Southern Europe, and is set to double by 2020
(Pout and Hitchin, 2009). This will impact on demands
for electrical energy and the use of environmentally
harmful refrigerants. There is therefore an urgent need
to seek energy-efficient, environmentally benign and
cost-competitive alternatives to conventional air-con-
ditioning. PHDC is one such technique. Previous research
(Bowman et al., 2000; Ford et al., 2010) has found down-
draught cooling to be an energy-efficient and cost-effec-
tive alternative to conventional air-conditioning for
new and existing non-domestic buildings. Much of the
evidence for this is based on post-occupancy performance
evaluation of non-domestic buildings (Thomas and
Baird, 2006; Salmeron et al., 2009). The evaluation of
potential benefits in the residential sector has so far
been limited by the lack of applications in practice. The
research presented here seeks to address this gap.
In the context of rapidly rising demand for convention-
al air-conditioning in housing in Europe, PHDC rep-
resents a potentially significant alternative to meet
cooling demand, while also reducing carbon emissions.
PHDC relies on the effect of gravity on a body of (rela-
tively) cold air, to create a downdraught, and thus cir-
culate air from the source of cooling to the occupied
zone within the building. The source of the cool air
can be either ‘passive’ or ‘active’.
.Passive downdraught cooling
Apassive downdraught can be achieved through the
evaporation of water within an air stream. This is
referred to as passive downdraught evaporative
cooling (PDEC). PDEC is only appropriate in hot,
dry conditions (wet bulb temperature .248C
(Givoni, 1994). In Southern Europe previous
research (Bowman et al., 2000) has demonstrated
that PDEC can meet 25–85% of the cooling load
of non-domestic buildings (equivalent to 15 – 60
kWh/m
2
/year). In residential buildings PDEC can
potentially provide thermal comfort without the
need for additional mechanical assistance
(Schiano-Phan, 2010) (although this will vary in
relation to the external wet bulb temperature).
.Active downdraught cooling
In warm and humid conditions active downdraught
cooling can be achieved by using chilled water
cooling coils or panels. Although this relies on mech-
anical cooling, it avoids the need for fans, which can
represent an energy saving of 25– 35% of the electri-
cal load in a non-domestic building. This approach
also avoids the need for either bulky fan-coil units
or air-handling units and related ductwork. While
there is a risk of condensation on the coils, the
on-coil water temperature can be set to be slightly
above the dew-point if possible. Also, any conden-
sate can be collected in a drip tray.
.Hybrid downdraught cooling
Ahybrid downdraught cooling system combines
both ‘passive’ and ‘active’ downdraught cooling
techniques (Ford and Diaz, 2003). Such a system
can function in both hot and dry conditions
(using PDEC) and warm and humid conditions
(using chilled water cooling coils).
Conventional cooling systems use fans to drive the
airflow to the occupied zones of a building. Down-
draught cooling systems differ in that the primary
mechanism for the distribution of cool air is either by
buoyancy or by wind-assisted natural ventilation.
This means that both energy costs and capital costs
can be significantly reduced (Ford et al., 2010).
A wide range of building types are amenable to the
integration of downdraught cooling systems, including
commercial and industrial, residential, educational,
health, and large-volume buildings such as transport
interchanges. Downdraught cooling has been applied
to buildings in many parts of the world and assess-
ments have been made of its applicability to existing
commercial and residential buildings in Europe, India
and China (Ford et al., 2010).
A study of the application of PDEC to existing commer-
cial buildings in Southern Europe revealed that PDEC is
applicable to potentially 70% of this stock, (saving 3.7 –
5.7 million tonnes of carbon per year) (Ford and Cairns,
2002; Ford and Moura, 2003). Analysis of applicability
to the housing stock in Spain, Portugal, Italy and Greece
has shown that the residential market for PDEC is even
larger, with potential annual savings of 19 – 38 million
tonnes of carbon per year (Ford and Moura, 2003).
However, prior to this project, residential applications
had not been demonstrated in Europe.
Typically the capital cost of a PDEC/hybrid cooling
system for non-domestic buildings is comparable in cost
with comfort cooling, whereas full air-conditioning is sig-
nificantly more expensive. Expressed as a percentage of
new-build costs, PDEC is in the order of 4–12% com-
pared with 19–23% for centralised air-conditioning
installations. Detailed life cycle cost analyses for Europe,
China and India indicate that passive and hybrid down-
draught cooling has significant life cycle cost benefits, typi-
cally providing a return on investment within a few years
(two-five years in India; five to 15 years in Europe) (Ford
et al., 2010; Ford and Moura, 2003).
A recent global review of research and practice in the
application of PHDC (Ford et al., 2010) has shown
Passive downdraught evaporative cooling
291
Downloaded by [Brian Ford] at 04:34 25 June 2012
that it is technically and economically viable in many
parts of the world, but to date has largely been
applied to non-domestic buildings. However, the same
study has shown that the potential market for applying
these techniques to residential buildings is even larger
than for the non-domestic sector. This paper reviews
the first residential application of PDEC in Europe.
The Nottingham Solar Decathlon entry
Participation in the 2010 Solar Decathlon (SDE) compe-
tition provided an opportunity to explore the contri-
bution PDEC can make to providing summer-time
occupant comfort in a high-performance house as part
of a broader passive environmental strategy. The Notting-
ham HOUSE (Home Optimising the Use of Solar Energy)
for the SDE 2010 competition, designed and built by stu-
dents from the University of Nottingham, seeks to achieve
the UK’s Code for Sustainable Homes Level 6 (Zero
Carbon), in an affordable modular terrace (Figure 1).
Although the SDE took place in Madrid, the intention
was to re-erect the house on its return to Nottingham,
and therefore for the house to be designed to meet the
demands of both climates. The high-performance envel-
ope (U-value of walls and roof ≤0.17 W/m
2
/8C),
triple-glazed windows, lobbied entrance and shading to
glazed openings all help to minimize external heat gains.
In addition, the specification of A-rated appliances mini-
mizes internal heat gains. The house was erected and
tested in Madrid in June as part of the SDE competition
and exhibition (Figure 2). The final PDEC system installed
in the house was evaluated for the week of the compe-
tition in Madrid (19–26 June 2010).
The house utilizes PDEC to maintain comfort tempera-
tures during the summer season in Madrid (Figure 3).
Nozzles positioned at the top of the double-height
space generate a mist of water that evaporates in
warm external air drawn through the roof light
(Figure 4). Evaporation of the water cools the air, gen-
erating a plume that drops into the dining area and
then divides, part flowing through the living room
and exiting via a window on the south wall. The
remainder flows through the kitchen, absorbing heat
from any appliances that are operating and exiting
via a window in the north wall. Operation of the
system, including automatic actuators attached to the
windows and roof-light, was designed to respond to
measurements of external and internal temperature
and humidity via a control system. Unfortunately, it
was not possible to complete the installation of the
roof-light actuators and wiring, or the PDEC control
system, in time for the start of the competition on the
18 June. The PDEC system was therefore controlled
manually, with the occupants increasing or reducing
the rate of cooling according to their perception of
how warm they felt. Ventilation was therefore entirely
driven by natural forces (buoyancy and wind), and was
not augmented by mechanical extract from the kitchen.
The system makes use of a novel nozzle design (by Inge-
niatrics-Frialia, Seville, Spain) that combines the motive
energy from a pressurized water circuit with a com-
pressed air flow (Figure 5). This provides the desired
level of atomization of the water flow and avoids the cre-
ation of drips whenthe nozzles start and stop.The system
requires energy to drive a pump and small air compres-
sors, to run an ultraviolet water treatment cell, and to
operate the window controls. Peak power requirement
is approximately 700 W, but this varies with the
number of nozzles in operation. There are eight nozzles
and peak water consumption is 8 litres/h if run continu-
ously. To run the system for an average of 5 h/day would
therefore require 40 litres of water and 3.5 kWh of elec-
tricity. The system is therefore not entirely passive in
operation.
Figure1 Ground £oor (left) and ¢rst £oor (right) plan of Nottingham H.O.U.S.E.
For d et al.
292
Downloaded by [Brian Ford] at 04:34 25 June 2012
Seasonal strategies
The building was designed to operate under three dis-
tinct seasonal modes, which are based on an analysis
of the climate data for Madrid.
Summer
During the hot, dry summers in Madrid, maximum
daily temperatures can frequently be above 358C.
The protection of glazed openings from the high
direct solar radiation is essential. Night-time
temperatures can drop below 208C, providing
an opportunity for convective cooling to be
exploited.
During the day, PDEC may potentially achieve thermal
comfort most of the time, provided solar heat gains can
be minimized and there is sufficient thermal capaci-
tance to stabilize temperatures internally. Plotting
data for Madrid on a psychrometric chart (Figure 6)
reveals that during the summer period thermal
comfort may be achievable by evaporative cooling
alone, even with dry bulb air temperatures above
358C. This is because the external air is so dry –
Figure2 External view of Nottingham HOUSE, Madrid 2010. Photo by Denis Doyle
Figure3 Section showing the summer daytime air £ow path
Passive downdraught evaporative cooling
293
Downloaded by [Brian Ford] at 04:34 25 June 2012
typically below 30% relative humidity in the after-
noon. At night, convective cooling will be achieved
by promoting buoyancy-driven natural ventilation,
pre-cooling the interior before sunrise the next day
(Figure 7).
The shading of exposed perimeter openings is provided
by a woven mesh, keeping high altitude summer sun
out while allowing low altitude winter sun to pene-
trate. Solar gains are also minimized by the high-per-
formance envelope (U-value of roof ¼0.1 W/m
2
K,
U-value of walls ¼0.17 W/m
2
K). Infiltration heat
gains are minimized by reasonably ‘air-tight’
construction and by providing a lobby ‘airlock’
entrance. Solar gains are further reduced by using the
photovoltaic panels to shade the roof, and therefore
reduce roof surface temperatures.
It was postulated that this combination of strategies
would remove the need for conventional cooling.
The results of subsequent monitoring are discussed
below.
Figure 5 Atomizing nozzle design by Ingeniatrics-Frialia of
Spain
Figure6 Psychrometric chart showing annual hourly average climate data for Madrid, Spain, with Givoni’s (1994) boundaries for comfort
and direct evaporative cooling (DEC)
Figure 4 Interior view of the light well showing the four nozzles
in operation. Photo by Denis Doyle
For d et al.
294
Downloaded by [Brian Ford] at 04:34 25 June 2012
Autumn and spring
During autumn and spring, mild external temperatures
(mean external temperatures of 15–168C in May and
October) will potentially allow the internal air temperature
to be brought within the comfort zone by controlling the
rate of natural ventilation. Occupant control of window
openings should enable the regulation of natural venti-
lation to remove any heat gains during the day, while set
to ‘trickle’ ventilation at night to provide minimum fresh
air. On warm days, evaporative cooling may be required
occasionally, although due to the short monitoring
period this has not been evaluated as part of this study.
Winter
Winters can be cool, with temperatures rarely dropping
below freezing, and generally rising above 108C during
the day. The high-performance envelope will minimize
fabric heat losses, and a combination of fabric air tight-
ness and the provision of a lobby ‘airlock’ will minimize
infiltration heat losses. Most of the glazed openings
(53%) are oriented to the south to capitalize on the sig-
nificant potential for beneficial passive solar gains.
The combination of passive solar gain and internal heat
gains coupled with a high-performance building envel-
ope are likely to reduce the residual heating requirement
to well below 15 kWh/m
2
(the ‘Passivhaus’ standard for
heating energy). Top-up heating for this residual
heating load will be provided by a ground source heat
pump arranged to preheat ventilation supply air.
Testing and monitoring methodology.
In order to establish the feasibility of applying PDEC to the
SDE Nottingham HOUSE, a prototype was constructed
and tested during the summer of 2009 at the University
of Seville in collaboration with AICIA (Asociacio
´npara
la Investigacio
´n y Cooperacio
´n Industrial de Andalucı´a)
and Ingeniatrics-Frialia (a company specializing in the pro-
duction of misting nozzles for evaporative cooling). The
objective was to characterize the performance of the pro-
totype and to establish its applicability in the climate of
Madrid. Additionally, in order to establish the air-flow
characteristics of the house, computational fluid dynamics
(CFD) analysis was also undertaken by AICIA. For details
of the prototype, the test results and CFD analysis carried
out by AICIA, see Alvarez et al. (2012).
The final-version house (as built) has eight misting
nozzles located at a high level within the central light-
well and ventilation shaft (Figure 8). These misting
nozzles generate evaporatively cooled air into the
central double-height space, which in turn promotes
the air flow to first-floor bedrooms and to the ground
Figure 8 Nozzle array below the roof-light transom. Photo by
Denis Doyle
Figure7 Section showing the summer night-time air £ow path
Passive downdraught evaporative cooling
295
Downloaded by [Brian Ford] at 04:34 25 June 2012
floor, from where it is exhausted via window vents at the
perimeter. The nozzles can be switched on or off in
pairs, enabling the system to respond to changes in
temperature and humidity. The system is made from
widely available components, and demonstrates an
affordable alternative to conventional air-conditioning.
In order to assist with the operation of the PDEC system,
and to gather data relating both to its performance and to
the performance of the house as a whole, five synchro-
nized ‘Tinytag’ data loggers instrumented with dry bulb
and relative humidity sensors were used. These were set
to record every 60 s and were located inside in the living
room, kitchen and bedrooms. The external sensors were
located beneath the photovoltaic (PV) modules on the
roof to shade them from direct solar radiation. The
back of the PV modules have a silvered surface creating
a ventilated low-emissivity environment. The results
from the Tinytag data loggers are shown in Figures 9 –12.
Preliminary results
Monitored data
The installation of the PDEC pipework and nozzle
array, compressor, hydraulic and electrical cabinets
was completed and tested in the Nottingham HOUSE
in Madrid on 16 June 2010. As noted previously, the
PDEC system was controlled manually, while generally
being switched off during public visits to the house
(11.00–14.00 and 17.00 – 21.00 hours) in line with
the rules of the SDE competition.
Logging of dry-bulb air temperature and relative
humidity in three locations in the house and externally
(in the shade), ran from 14.00 hours on 19 June to
17.00 hours on 28 June. During this period the
weather became warmer than it had during the con-
struction phase of the competition. The external daily
maximum shade air temperature rose from 308Con
20 June to 448C on 24 June.
Figure9 Internal(living room) and external dry bulb and wet bulb temperatures, 22^ 24 June 2010, for the Nottingham HOUSE,Madrid
Figure10 Internal and external dry bulb temperature and relative humidity, 23 June 2010, for the Nottingham HOUSE, Madrid
For d et al.
296
Downloaded by [Brian Ford] at 04:34 25 June 2012
During this whole period, in the absence of an auto-
matic control system, a decision was made to keep
window and roof-light openings open 24 h/day to
promote night-time convective cooling and to allow
ventilation during the day.
In spite of the fact that control of the PDEC system was
therefore very simple, the record of logged temperature
and relative humidity demonstrates that comfortable
conditions were obtained for much of the time
(Figure 9). The data also indicated that there was an
opportunity for the PDEC system to reduce tempera-
tures further at other times if the SDE competition
rules had permitted it to be operated for longer
periods during the day.
The high-performance building fabric, the triple-glazed
windows and the care taken to minimize solar heat
gains all have a major impact on reducing the risk of
overheating. It may also be the case that the use of
high-density ‘Gyproc Rigidur’ panels internally con-
tributes to stabilizing internal temperatures in what is
essentially a ‘lightweight’ building. Rigidur (a
gypsum fibreboard incorporating recycled paper) has
a density of 1200 kg/m
3
,or22kg/m
2
for 18 mm
panels.
It is interesting to note that there is an approximate 4-
hour time-lag between the peak external air tempera-
ture (about 14.00 hours) and the peak internal air
temperature (about 18.00 hours), which reflects the
thermal characteristics of the building envelope.
However, it is also evident that without the PDEC
system internal air temperatures would have risen
well outside internal comfort conditions.
From the recorded data it can be observed that when
the PDEC system was activated, it promoted a rapid
Figure11 External and internal dr y-bulb temperatures and di¡erence in absolute humidity, 22 ^24 June 2010, for t he Nottingham HOUSE,
Madrid
Figure12 Measured dry bulb and absolute humidity dur ing the operating period of the PDEC system on 22 June 2010
Passive downdraught evaporative cooling
297
Downloaded by [Brian Ford] at 04:34 25 June 2012
reduction in temperature within the house. This can be
seen particularly well on 22 June, when PDEC was
operational after about 14.30 hours for approximately
15-minute intervals until about 21.00 hours, maintain-
ing the living room temperature at or below 268C when
the external temp was 38 – 408C.
For the next two days the external air temperature rose
well above 408C and the impact of switching the PDEC
system on and off became even more pronounced. On
23 June (Figure 10) the PDEC system was not switched
on until 13.00 hours, by which time the internal temp-
erature had risen to 27.58C. With intermittent oper-
ation of the nozzles between 13.00 and 15.00 hours
(the nozzles were on for about 50% of the time) the
internal temperature was stabilized, and after 15.00
hours when nozzles were left on continuously, the
living room temperature dropped below 268C, while
measured external shade air temperature was about
408C. When the system was switched off again for
the evening public visit (17.00 – 21.00 hours), the
living room temperature can be seen to rise again.,
until a peak of 288C is reached at 20.00 hours. Had
the PDEC system been run consistently, and if the
house had not had to cope with large numbers of visi-
tors, it is likely it would have held the temperature
below 268C all day.
It can be concluded that in a typical house the evapora-
tive system could run until 20.00 hours keeping the
indoor temperature below 268C. The ideal switch-off
hour is thus when indoor temperature is equal to
outdoor temperature.
It is important to note that during these extremely hot,
dry days, the internal relative humidity did not rise
above 65% during the day, and much of the time it
was below 50% indicating potential for further eva-
porative cooling (Figure 10).
It is also interesting to note that while it was antici-
pated that the PDEC system would primarily serve
the ground floor, the temperature and relative humid-
ity in the main bedroom followed the pattern in the
living room fairly closely, and there was not the level
of stratification that had been anticipated. This may
be because cool air was being caught by the balcony
and passed through the open door of the main
bedroom, before exhausting via the open bedroom
window.
Operation of the PDEC system took place within the
context of the SDE competition and was dictated by
two differing requirements. Part of the competition
was judged on the ability of the house to maintain a
time-averaged internal temperature between 238C
and 258C, for which PDEC was the sole ‘active’ strat-
egy used. The other requirement was the energy
balance of the house, which was judged on surplus
PV generated electricity, and which was impacted
negatively by operation of the PDEC system.
It can be seen that the house provided significant pro-
tection from the extreme daytime conditions prevailing
outside, most of this through passive means alone. The
combination of solar protection (both against direct
gains through glazing and indirect gains through the
fabric), the heavily insulated envelope, the presence
of thermally heavyweight internal finishes and
opening the building up to utilize night cooling
served to limit temperature rises within the space.
This also meant that the need to use the PDEC
system to control excessive temperature rises was
limited and the period of operation was commensu-
rately small. Figure 12 shows the data collected
during the operation period on 22 June. Rises in the
internal absolute humidity of the ground floor living
area can be observed when the system was switched
on and depressions in the internal dry bulb temperature
can be seen to occur in response.
The combination of internal and external temperature
and humidity data, in conjunction with the rate of
water flow to the PDEC misting nozzles, was used in
an attempt to assess the cooling performance by deter-
mining flow rates and supply temperatures from the
downdraught stack. During the periods of operation,
occupancy of the living areas was very low and rep-
resented the only significant source of moisture
released into the space beyond the supply air from
the PDEC system. A sketch of the assumed psychro-
metric process is shown in Figure 13.
This implies that the supply flow rate was sufficiently
high to dictate the internal absolute humidity,
meaning its condition could be determined by
moving up the external air wet bulb line until the
internal absolute humidity was reached. The mass
flow rate, m, of air from the system could then be esti-
mated from the mass flow rate of water from the
misters, q, and the change in absolute humidity,
Figure13 Schematic of t he assumed psychrometric process for
PDEC operation
For d et al.
298
Downloaded by [Brian Ford] at 04:34 25 June 2012
DAH, between the external and supply air flows as:
m=q/DAH
Having determined the supply air mass flow rate, the
effective cooling (equal to the thermal gains to the
space) could then be determined using the resulting
temperature difference between the room air and the
supply air as:
Q=m×c×DT
The results for one of the sampling periods recorded
during operation of the system are shown in
Figure 14 and are representative of all of the data
collected.
These show that the method described above cannot be
used to explore the behaviour of the system suggested
by the monitored data. The room air condition lies
below rather than above the wet bulb line for the exter-
nal air. There are a number of possible explanations for
the observed behaviour, the two most likely of which
are listed below:
.The mass flow rate of the supply air was low and
the room condition represents mixing of the
supply and room air. The measured absolute
humidity would therefore lie somewhere between
that of the supply air and that of the room air
prior to activation of the PDEC system.
.The dry bulb temperature of the external air was
significantly overestimated. The sensors were
positioned beneath the PV modules on the roof
of the building and so were protected from direct
solar radiation for the majority of the day. Small
peaks in the external temperature can be seen in
Figure 11 for a number of hours after sunrise.
These correlate with periods of direct solar
exposure between 06.30 and 07.30 hours caused
by low-angle sun penetrating the region beneath
the PV modules. It is unlikely that heat radiated
from the rear of the PV modules could have had
a similar effect during the remaining hours of the
day, since the back of the panels are silvered
providing a low-emissivity well-ventilated
environment.
If it is assumed the process illustrated in Figure 13 actu-
ally occurred, the overestimate in the external air dry
bulb temperature would have to be in the order of
158C (assuming the absolute humidity was measured
accurately), yielding a value close to the internal
room temperature. What is more likely is that the
mixing condition described above took place and the
proceeding analysis is performed on this basis.
The approach adopted is to use a building simulation
model to estimate the heat gains to the Nottingham
HOUSE on a notional day, and from this estimate
the possible operating condition of the PDEC
system.
Simulation results
Building energy simulation was used to inform the
construction and operation of the house and evidence
of this process was also demanded as part of the SDE
competition. A simple two-zone model used during
Figure14 Measured external and internal temperature and humidity points for 14.30^15.00 on 22 June 2010
Passive downdraught evaporative cooling
299
Downloaded by [Brian Ford] at 04:34 25 June 2012
the design phase was also used to explore the possible
behaviour of the PDEC system both for the con-
ditions anticipated during the SDE competition
week and the cooling season as a whole. The model
was generated within the TAS thermal simulation
software
1
andusedtoestimatethedirectsolar
gains, gains resulting from conduction through the
envelope and ventilation gains driven by wind and
buoyancy forces.
The external dry bulb temperatures were obtained
from the data logger located on the roof, which may
represent an overestimate of the actual conditions.
No solar data were collected during the competition
period: instead, the solar availability was inferred
from the PV output. This was done by identifying the
PV output from a known clear day during the compe-
tition and matching it to a clear day in June within the
Madrid typical metrological year data file (sourced
from the Energy Plus website
2
). Data for the remaining
days were obtained by scaling the weather data by the
ratio of the measured PV outputs. Finally, in the
absence of measured wind data, results for an
average day in June were used. Of all these assump-
tions, that for the wind data probably had the greatest
impact on the simulation for which wind and stack-
driven mechanisms were used to estimate the venti-
lation rate for the model.
Figure 15 shows the measured external and internal
dry bulb temperatures and a comparison with the pre-
dicted internal conditions. The rolling mean was taken
over the duration of the competition period.
The agreement is reasonable given the assumptions
made in determining the input data and suggests that
the model provides a reasonable picture of the
building’s behaviour and may be affected by a
number of assumptions made in compiling the input
climatic data.
Having explored the behaviour of the building energy
model in comparison with the measured behaviour of
the house, the simulations were rerun using the
Madrid TMY file to yield a set of cooling loads for a
notional summer’s day that could then be used to
explore the behaviour of the PDEC system. Figure 16
represents a day (day 186 in the simulation, i.e. a day
in July) when the cooling loads are modest.
The cooling loads were used in conjunction with the
performance specification for PDEC to explore
operation.
Data in Table 1 relate to the total amount of cooling
produced by the system . Not all of this is available
to cool the building as the external air has to be pre-
cooled to the room design temperature before any
benefits start to be realized. The operational CoP will
therefore be lower than these figures and is explored
below.
Figure 17 shows the useful cooling power delivered by
PDEC system (which is equal to the space cooling from
the building simulation model) in comparison with the
cooling available. For the day selected from the simu-
lation, the cooling load is capable of being met by the
lowest setting on PDEC, i.e. the evaporation produced
by two nozzles.
The residual power is assumed to pre-cool the supply
air from the external air dry bulb temperature to
278C: the room design temperature. The CoP based
on the space cooling demand is shown in Figure 18.
Figure15 Measured and predicted dry bulb temperatures, 22 ^ 24 June 2010, for t he Nottingham HOUSE,Madrid
For d et al.
300
Downloaded by [Brian Ford] at 04:34 25 June 2012
Figure17 Cooling power
Figure16 Predicted t emperatures with and without co oling plus cooling load
Ta b l e 1 Performance characteristics of the Not tingham HOUSE PDEC system
Water £ow
rate (l/s)
Operating equipment Electrical operating
power (W)
Maximum
cooling (W)
Maximum coe⁄cient of
performance (CoP)
2 Two nozzles, UV lamp, water pump and one
compressor
212 1256 5.9
4 Four nozzles, UV lamp, water pump, one
actuator and one compressor
2 27 2511 11.1
6 Six nozzles,UV lamp, water pump, two
actuators and two compressors
483 3767 7.8
8 Eight nozzles,UV lamp, water pump, three
actuators and two compressors
498 502 2 10.1
Note: UV ¼ultraviolet.
Passive downdraught evaporative cooling
301
Downloaded by [Brian Ford] at 04:34 25 June 2012
The results suggest that a reasonably high CoP may be
achieved when the system operation is well matched to
the space cooling demand. When the cooling demand is
small, as at the start and end of the day, the majority of
the work that the PDEC system does is in pre-cooling
the supply air and therefore the CoP drops. As the
results represent the PDEC system operating at its
lowest setting, it is not possible to reduce its output
to match demand better.
The supply air flow rate and state required to maintain
the design dry bulb temperature of 278C may be found
from a psychrometric analysis of the system. If it is
assumed that buoyancy provides the motive force to
drive the air flow, it is also possible to use the stack
equation, information about the building geometry
and the predicted supply, and room design temperature
to predict the air flow rate that would be achieved.
These data are shown in Figure 19.
This suggests that when a high proportion of the cooling
delivered by the two nozzles is used to absorb heat from
inside the building, the air change rate required to main-
tain the comfort conditions approaches a minimum.
Implicit in this is that less pre-cooling of the supply air
is taking place. At the same time, flows generated by
buoyancy approach a maximum and the potential for
the required flow rate actually being achieved is at its
greatest. In the case of the Nottingham HOUSE operat-
ing under the conditions assumed in the simulation, suf-
ficient flow would be generated to ensure the comfort
temperature was met.
Figure18 Coe⁄cient of per formance (CoP) based on space cooling
Figure19 Predicted air change rates for t he system
For d et al.
302
Downloaded by [Brian Ford] at 04:34 25 June 2012
When the system is delivering significantly more
cooling than required for maintaining the comfort
temperature, the air flow rate grows. The system
undertakes more pre-cooling, hence the flow rate
increases. At the same time, the temperature difference
between the supply air and the comfort condition
decreases and the buoyancy driven flow drops. In this
example, it drops to the point where it would not be
able to sustain the required flow rate.
An idea of the sensitivity of the model to uncertainty in
the input data is also indicated on Figure 19. The upper
and lower bounds to the curve enclose a range of air
change rates predicted based on changes made to the
predicted space cooling. Increasing the estimated
cooling by 20% reduces the required air change rate
and increases the buoyant flow. Decreasing the esti-
mated cooling by 20% has the opposite effect.
Occupants’ response
During the course of the competition week students
were tasked with guiding visitors around the house,
and responding to their questions. Although limited
to a maximum of six people at any one time, the stu-
dents on duty were rotated and so over the week
about 30 students, staff and observers spent extended
periods of time in the house and experienced the
PDEC system operating.
Anecdotal comments were broadly favourable:
Overall I was really impressed by the system. It’s
difficult to say how it could be improved as it
wasn’t operating fully and the manual control
and filling of the tank did hinder its performance.
However, when it was switched on, its effects
were instant, even from parts of the house I
didn’t expect to feel it, and it seemed to do an
excellent job at lowering the temperature of the
entire house.
Nearly everyone agreed that the most unsatisfactory
aspect of the PDEC system was the noise from the
nozzles. There was a very noticeable ‘hissing’ induced
by the compressed air. One of the students commented:
If the nozzles could be far quieter and less of an
eyesore then I think it could be more successful.
Noise from the nozzles definitely needs to be
addressed, as at its present level it would probably
inhibit take-up of the system.
Conclusions
Measured data from the Nottingham HOUSE during
the competition week of the Solar Decathlon in
Madrid (19–27 June 2010) demonstrate that the com-
bination of a high-performance building envelope,
night-time convective cooling and PDEC can deliver
comfortable conditions throughout the house, even
when maximum external temperatures exceed 358C.
The data also indicated that there was significant
potential for further evaporative cooling, and that
this is most likely to be achieved by automatic
control of the misting system.
Three main factors limit the period of potential over-
heating to the late afternoon and early evening when
PDEC is operated:
.Very low U-values (≤0.17 W/m
2
K) and effective
shading to all openings minimize external heat
gains.
.Night-time convective cooling contributes to low
internal start temperatures each morning (20 –
228C).
.The observed time lag between peak external and
internal temperatures of 3 – 6 h. This may be due
in part to the use of high-density gypsum fibre-
board (‘Rigidur’) panels internally.
The operation of the nozzles is very effective in limiting
any further rise in internal air temperature, which is
kept within acceptable limits throughout the house,
except for 1– 2 h at the hottest time of day. At times
the internal temperature was 168C below external
dry bulb air temperature. Such a significant tempera-
ture depression can only be achieved by evaporative
cooling when the external wet bulb temperature is
extremely low. There is no significant stratification
within the house, which may be ascribed to the effect
of the internal ‘balcony’ catching and directing the
cooled air into the bedrooms.
The requirements of the competition limited the
overall height of the house, which meant that a
supply air ‘tower’ could not be included. However,
results from the week of the competition suggest
that the absence of an inlet tower is no impediment
to the successful integration of a PDEC system
within a two-storey house. In spite of this very
favourable result, the addition of an inlet tower
could improve control over the supply air, and
potentially increase air volume flow rate, and thus
effective cooling. It might also simplify access for
maintenance.
While manual control is clearly a viable option, this
requires close occupant involvement and it is likely
that performance would have improved had auto-
mated control of nozzles and window actuators been
operational. This would require the monitoring of
both internal and external temperatures and relative
Passive downdraught evaporative cooling
303
Downloaded by [Brian Ford] at 04:34 25 June 2012
humidities as well as wind speed and direction. The
latter would assist in avoiding stack reversal when
wind pressures are unfavourable.
Noise from the nozzles was regarded as unsatisfactory
by many of the student occupants. This is one of a
number of product development issues that need to
be addressed. There is also potential to improve the
CoP of the system through the use of variable speed
compressors or an air receiver to avoid the stepped
response of the present design. This may be further
refined by controlling the water supply down to the
level of individual nozzles (rather than pairs of
nozzles as in the present design).
While the data set is limited, these results suggest that
PDEC can provide a viable alternative to conventional
air-conditioning for housing in central and southern
Spain, and in other hot, dry regions of the world.
Although the system components are widely available
and competitive with conventional cooling systems,
take-up in the residential sector will require further
product development if it is to be successful. Take up
for residential applications is also likely to be slow
until more systems have been installed.
The team is seeking the application of this approach to
a demonstration house (new build or existing) with an
occupant family to determine occupant reaction to
PDEC over a cooling season.
Acknowledgements
This work was carried out with the support of Saint-
Gobain (main sponsors of the Nottingham HOUSE)
and Ingeniatrics Tecnologias (Miguel Asensio). The
prototype was constructed and tested by members of
AICIA, University of Seville (directed by Professor Ser-
vando Alvarez). The Nottingham HOUSE was
designed and built by students and staff at the Depart-
ment of Architecture & Built Environment, University
of Nottingham.
References
Bowman, N., Lomas, K. and Cook, M.J. (2000) Passive down-
draught evaporative cooling 1. Concept and precedents.
Indoor and Built Environment,9(5), 284–290.
Ford, B. and Cairns, K. (2002) Market assessment of passive
downdraught evaporative cooling in non-domestic buildings
in Southern Europe, in Proceedings of the 3rd European
Conference on Energy Performance & Indoor Climate in
Buildings, EPIC, Lyon, France, 2002, Vol. 2, pp. 505 – 510.
Ford, B. and Diaz, C. (2003) Passive downdraft cooling: hybrid
cooling in the Malta Stock Exchange, in Rethinking Devel-
opment: Are We Producing a People Oriented Habitat?, Pro-
ceedings of the PLEA Conference, Santiago, Chile, 2003,
PLEA, Brussels.
Ford, B. and Moura, R. (2003) Market Assessment of Passive
Downdraught Evaporative Cooling in Non-domestic Build-
ings in Southern Europe. Final Report, ALTENER II Project
on ‘Solar Passive Heating and Cooling’, European Commis-
sion-DG Research (available at: http://www.phdc.eu/
uploads/media/ALTENER_1_Final_report_extract.pdf).
Ford, B., Schiano-Phan, R. and Francis, E. (2010) The Architec-
ture & Engineering of Downdraught Cooling: A Design
Sourcebook, PHDC Press, London.
Givoni, B. (1994) Passive and Low Energy Cooling of Buildings,
Wiley, New York, NY.
Pout, C. and Hitchin, E.R. (2009) Future environmental impacts
of room air-conditioners in Europe. Building Research &
Information,37(4), 358–368.
Salmeron, J.M., Alvarez, S., Sanchez, J., Ford, B. and Gillot, M.
(2012) Analysis of a PHDC (passive and hybrid downdraft
cooling) experimental facility in Seville and applicability to
the Madrid climate. International Journal of Ventilation,
10(4), 391–404.
Salmeron, J.M., Sanchez, J., Ford, B., van Steenberghe, T. and
Alvarez, S. (2009) Passive and hybrid downdraught cooling
in buildings and software for design. REHVA Journal,
46(6), 34–39.
Schiano-Phan, R. (2010) Environmental retrofit: building inte-
grated cooling in housing. Architectural Research Quarterly,
14(1), 139–151.
Thomas, L.E. and Baird, G. (2006) Post-occupancy evaluation of
passive downdraft evaporative cooling and air-conditioned
buildings at Torrent Research Centre, Ahmedabad, India,
in Proceedings of the 40th Annual Conference of the Archi-
tectural Science Association ANZAScA, pp. 97– 104.
Endnotes
1
TAS 9.2.1 Dynamic Thermal Analysis Software (Environmental
Design Solutions Ltd; available at: http://www.edsl.net/).
2
EnergyPlus Energy Simulation Software and Weather Data
(available at: http://apps.eere.energy.gov/buildings/energyplus).
For d et al.
304
Downloaded by [Brian Ford] at 04:34 25 June 2012
