Available via license: CC BY 4.0
Content may be subject to copyright.
Cooling without Air Conditioning: Membrane-Assisted RadiantCooling without Air Conditioning: Membrane-Assisted Radiant
Cooling for Expanding Thermal Comfort Zones GloballyCooling for Expanding Thermal Comfort Zones Globally
This paper was downloaded from TechRxiv (https://www.techrxiv.org).
LICENSE
CC BY 4.0
SUBMISSION DATE / POSTED DATE
26-03-2020 / 29-03-2020
CITATION
Teitelbaum, Eric; Rysanek, Adam; Pantelic, Jovan; Aviv, Dorit; Ruefenacht, Lea; Teitelbaum, Megan; et al.
(2020): Cooling without Air Conditioning: Membrane-Assisted Radiant Cooling for Expanding Thermal Comfort
Zones Globally. TechRxiv. Preprint. https://doi.org/10.36227/techrxiv.12034971.v1
DOI
10.36227/techrxiv.12034971.v1
Cooling without Air Conditioning:
Membrane-Assisted Radiant Cooling for
Expanding Thermal Comfort Zones Globally
Eric Teitelbauma,b,c,2, Kian Wee Chenc, Dorit Avivb,d, Kipp Bradfordb, Lea Ruefenachta, Denon Shepparde, Megan
Teitelbaumf, Forrest Meggersb,c, Jovan Pantelicf,g, and Adam Rysaneke
a
Singapore-ETH Centre, ETH Zurich, Singapore, 318602, Singapore;
b
School of Architecture, Princeton University, Princeton, NJ 08544, USA;
c
Andlinger Center for Energy
and the Environment, Princeton University, Princeton, NJ 08544, USA;
d
Weitzman School of Design, University of Pennsylvania, Philadelphia, PA, USA;
e
School of Architecture
and Landscape Architecture, University of British Columbia, Vancouver, CAN;
f
Berkeley Education Alliance for Research in Singapore, 138602, Singapore;
g
Center for the Built
Environment, University of California, Berkeley, CA, USA
This manuscript was compiled on March 26, 2020
We present results of a world-first radiant cooling system that made the hot and humid tropical climate of Singapore feel relatively cool
and comfortable. Thermal radiation exchange between occupants and surfaces in the built environment can augment thermal comfort.
Even in air-conditioned spaces, radiation exchanged between occupants and their surroundings accounts for approximately 50% of their
perceived comfort(1). The lack of widespread commercial adoption of radiant cooling technologies for indoor air conditioning is due to two
widely-held views: (1) the low temperature required for radiant cooling in hot and humid environments will form condensation and (2) cold
surfaces will still cool adjacent air via convection, limiting overall radiant cooling effectiveness. This work directly challenges these views
and dispenses with them. We constructed a demonstrative outdoor radiant cooling pavilion in Singapore that used an infrared-transparent
low density polyethylene membrane to provide radiant cooling at temperatures up to 20 ◦C below the dew point. Surrounding the radiant
cooling surfaces by an air-gap and infrared-transparent membrane permits radiation exchange to occur between the human body and cold
surfaces whilst avoiding condensation on any exposed material as well as significant convective heat transfer losses. Test subjects who
experienced the pavilion (n=37) reported a ‘cool’ to ‘neutral’ thermal sensation 81% of the time, despite experiencing 29.6 ±0.9 ◦C air at 66.5
±5 %RH and with low air movement of 0.26 ±0.18 m s−1. Comfort was achieved with a coincident mean radiant temperature of 23.9 ±0.8
◦C, requiring a chilled water supply temperature of 17.0 ±1.8 ◦C. The pavilion operated successfully without any observed condensation on
exposed surfaces despite an observed dewpoint temperature of 23.7 ±0.7 ◦C. The coldest conditions observed without condensation used
a chilled water supply temperature 12.7 ◦C below the dew point, which resulted in a mean radiant temperature 3.6 ◦C below the dew point of
23.7 ◦C.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Radiant Cooling |Thermal Comfort |Energy Efficiency |Photonics
F
or the first time in known records, a radiant cooling system
1
that makes people comfortable in the hot-humid tropical
2
outdoors, and yet does not condense water, has been created.
3
The cooling panel operates below dew-point temperatures, but
4
is insulated from humid air by a membrane transparent to
5
longwave radiation. It successfully makes people feel comfort-
6
able in conditions exceeding 30
◦
C and 65% relative humidity
7
without modifying the air temperature or humidity circulating
8
around human bodies. By relying instead on thermal radiation,
9
the system created and investigated in this paper made people
10
feel cold outdoors in tropical Singapore, reporting thermal
11
comfort sensations of “cool" as assessed by a thermal comfort
12
survey, despite the unconditioned outdoor air temperature
13
and humidity.14
While thermal radiation has been studied for over a century
15
in the context of thermal comfort (
2
–
5
), a database of build-
16
ings spanning 23 countries containing 81,846 complete sets of
17
objective indoor climatic observations (
6
) does not contain a
18
single data point with a mean radiant temperature more than
19
4
◦
C below the air temperature, for air temperatures above
20
28
◦
C. This fact, in conjunction with further literature review
21
(
3
,
7
,
8
) leads the authors to believe such an environment has
22
never been designed or studied. For reference, mean radiant
23
temperature is a proxy for the view factor-weighted average
24
temperature of the surroundings. 25
In 1963, Morse proposed a method for radiant cooling in the
26
tropics, using a membrane-assisted approach to convectively
27
isolate chilled surfaces from the surrounding air (
7
). The
28
membrane is transparent to thermal radiation in the 5-50
29
micron range where humans emit, allowing for radiant cooling
30
to occur between the chilled surface and a person through the
31
membrane. 32
While this idea has been proposed, a full scale system has
33
never been built testing whether the uniqueness of conditions
34
will actually provide comfort for people(
6
). The conditions of
35
high air temperature and low mean radiant temperature do
36
not occur naturally anywhere, as chilled surfaces act as heat
37
exchangers, cooling the air. Using the thermally transparent
38
membrane as a convection shield, we eliminate this mechanism
39
of heat transfer. Further, we transformed the initial 1963
40
concept with modern analytical techniques to improve the
41
system’s performance in the tropics, eliminating the need for
42
components such as an internal heater and originally proposed
43
by Morse to avoid condensation on the outer surface of the
44
membrane (
9
). Promising results from this initial study (
9
)
45
were scaled up to a full scale demonstrator, in which a thermal
46
comfort study was conducted to monitor occupants’ responses
47
to the low radiant temperature environment with high outdoor
48
Fig. 1.
Schematic of a Cold Tube radiant cooling panel (left) and radiant heat transfer
through the infrared-transparent membrane (right).
air temperatures for the first time (6).49
Typically, building occupants associate comfort with air
50
temperature and relative humidity, and in traditional build-
51
ings, only air temperature is required for a comfort setpoint
52
(
8
). To demonstrate that our system provides comfort while
53
operating outside the conventional comfort modes, we con-
54
ducted a thermal comfort study, surveying participants to
55
gauge the perception of the new thermal environment.56
Figure 1schematically illustrates how the system functions,
57
allowing radiation to pass, but not air and humidity, thereby
58
reducing convection and eliminating condensation. Chilled
59
water is circulated in a dense capillary mat internally in the
60
panels. These cold surfaces extract heat independent of the
61
air temperature, but it is previously impossible to remove heat
62
from people radiatively without also cooling the air.63
Such a radiative cooling system is notable since a carbon-
64
constrained world is an air conditioning-constrained world, an
65
unavoidable fact as global air conditioning demand is expected
66
to reach 50 exajoules (EJ) by the end of the century, eclipsing
67
global heating demand around 2070 (
10
). Already in the
68
United States, air conditioning is responsible for nearly 9%
69
of all primary energy demand (
11
) and is one of the primary
70
CO2emission sectors.71
Air conditioning is an attractive choice for comfort systems
72
as the refrigeration cycle both dehumidifies and cools air,
73
an important function since much of the ventilation load in
74
the United States and tropics is dehumidification, known
75
as the latent load (
12
). However, dehumidification requires
76
subcooling the air, an energetically and exergetically intensive
77
Fig. 2. The completed Cold Tube.
process (
13
), and the two processes cannot be decoupled with
78
conventional vapor compression techniques. Using radiant
79
systems for cooling and desiccants for dehumidification is an
80
efficient combination (14). 81
With the recent excitement surrounding tunable nanopho-
82
tonic materials for passive daytime and radiative cooling (
15
–
83
17
), this study helps advance the understanding for the po-
84
tential of direct occupant radiant cooling. Utilizing these
85
materials for comfort can increase the utility of outdoor space,
86
manage thermal comfort of walking people, and rapidly pro-
87
vide cooling comfort to people outdoors, perhaps at bus stops,
88
all without wasting cooling energy to the air. 89
Results 90
The completed pavilion, known further as the Cold Tube, is
91
shown in figure 2. Three vertical panels are shown on the
92
image in the left, and in the interior image on the right both
93
vertical and horizontal ceiling panels are shown. The opti-
94
cally clear membrane is also transparent to infrared radiation,
95
with a hemispherical transmissivity of 0.824 at 300
K
. The
96
blue capillary mats inside the panels circulated chilled water
97
produced by a heat pump. The capillaries were in thermal
98
contact with a thin metal sheet painted white (emissivity 0.95
99
at 300
K
). Sensible heat in the air prevents condensation on
100
the membrane surface, maintaining temperatures above the
101
dew point for chilled water up to 20
◦
C below the dew point
102
supplied to the capillary mats, allowing comfortable conditions
103
with exclusively radiant cooling, no air conditioning. 104
The coldest mean radiant temperature produced in the
105
Cold Tube was 19.9
◦
C with a coincident air temperature of
106
29.3
◦
C and supply water temperature of 10.8
◦
C, producing
107
no condensation despite a dew point of 23.5
◦
C. Not only was
108
the chilled water supply temperature 12.7
◦
C below the dew
109
point, but the resulting mean radiant temperature was 3.6
110
◦
C below the dew point. Such conditions have never been
111
achieved (6) in the built environment.112
55 individuals participated in a subjective thermal comfort
113
study in the Cold Tube carried out from January 8 through
114
January 27 in 2019. 37 of the test subjects experienced the
115
Cold Tube operating, and the remaining 18 were a control
116
group experiencing the Cold Tube when turned off (and thus
117
providing shade only). All test subjects were first asked to
118
sit in a shaded outdoor space adjacent to the Cold Tube for
119
a period of 15 minutes in order to achieve thermal neutrality
120
with outdoor conditions.121
Figure 3shows histograms of cumulative data for thermal
122
responses on a 7 point scale, ranging from -3 (cold) to 3
123
(hot) with 0 as neutral. After reaching thermal neutrality
124
in the shade, which was confirmed verbally by participants,
125
participants were surveyed three more times: 1) after walking
126
seven minutes to the Cold Tube, 2) after sitting in the Cold
127
Tube for one minute, and 3) after sitting in the Cold Tube
128
for 10 minutes. Data from both the operational and non-
129
operational Cold Tube participants are displayed side by side
130
in the histograms. Statistics about the distributions, as well
131
as p-values assessing the likelihood the responses from both
132
the Cold Tube on and off groups are related based on a t-test.
133
Data in figure 3shows that when the Cold Tube is on, there
134
is never a ‘Hot’ population in the Cold Tube, and after pro-
135
longed sitting in the pavilion, ‘Slightly Warm’ is the warmest
136
vote. While 46% of Cold Tube on responses were warm after
137
only 1 minutes in the Cold Tube, which is greater than the
138
initial state population, this number fell to 27% after being in
139
the Cold Tube for 10 minutes. More importantly, the mean
140
vote drops below 0, implying the mean of the perception is
141
cool. Such a result is without precedent for conditions where
142
air velocities are below 0.4
m s−1
and air temperature exceeds
143
30
◦
C. The t-test provides a p-value less than 0.02, implying a
144
98% confidence interval that both survey groups were report-
145
ing feeling different thermal sensations. Much higher p-values
146
were observed between the populations of Initial State and
147
Walking responses. Similarly, the p-value of the Cold Tube off
148
group compared to the Initial State groups together is 0.74,
149
compared to 0.002 with the Cold Tube on compared with the
150
Initial State population. This implies that the Cold Tube,
151
when turned off, was perceived to provide a similar degree of
152
comfort as sitting under any shaded outdoor structure, but
153
sitting inside the Cold Tube when it was on was absolutely
154
not perceived as similar to a shading-only scenario.155
Data from both Cold Tube on and off groups were inter-
156
preted in the adaptive comfort framework, plotted in figure
157
4a. Using the operative temperature calculated in equation 1,
158
the outdoor air temperature was used as the x-axis and data
159
is shaded based on the satisfaction response. When the Cold
160
Tube was operational, 21% of participants were dissatisfied,
161
which is nearly an allowable design criteria within the adap-
162
tive comfort framework (80% satisfaction interval), however
163
when the Cold Tube was off, 73% of participants were dissat-
164
isfied. There is a clear segmentation between the on and off
165
groups, and shows that this type of system has potential for
166
augmenting comfort in naturally ventilated spaces without air
167
conditioning.168
The same data is transformed in figure 4b, plotting the raw
169
mean radiant temperature data against the air temperature
170
for each survey point. Again, there is a clear separation of
171
32101234
0.0
0.2
0.4
0.6
0.8
1.0
p
< 0.39
On group: 32% > 0; = 0.41; = 0.72
Off group: 47% > 0; = 0.60; = 0.74
(a) Initial State
32101234
0.0
0.2
0.4
0.6
0.8
1.0
p
< 0.25
On group: 97% > 0; = 1.73; = 0.80
Off group:: 100% > 0; = 2.0; = 0.65
(b) After Walking
32101234
0.0
0.2
0.4
0.6
0.8
1.0
p
< 0.02
On group: 46% > 0; = 0.30; = 1.02
Off group: 80% > 0; = 1.07; = 1.03
(c) 1 Minute in Cold Tube
32101234
0.0
0.2
0.4
0.6
0.8
1.0
p
< 0.02
On group: 27% > 0; = -0.05; = 0.81
Off Group: 53% > 0; = 0.53; = 0.74
(d) 10 Minutes in Cold Tube
Fig. 3.
The thermal sensation votes reported by occupants are compared between the
Cold Tube on and off groups. The histograms show the thermal perception response
data from the survey participants. A vote of -3 is very cold, 0 is neutral, and +3 is
Very Warm. The subplots are responses during the initial conditioning period (a),
after 7 minutes of walking (b), after spending 1 minute in the Cold Tube (c), and 10
minutes in the Cold Tube (d). Responses with the Cold Tube on are solid gray bars,
and responses with the Cold Tube off is the solid black line. Included are confidence
intervals that the off population is different from the experimental population from a
t-test, the measured mean vote,
µ
, the standard deviation among responses,
σ
, and
the percentage of responses above 0 (warm votes). Within 1 minute of entering the
Cold Tube, occupants report feeling cool, and after 10 minutes the mean vote shifts
cool, going below 0.
20.0 22.5 25.0 27.5 30.0 32.5 35.0
Air Temperature C
20
22
24
26
28
30
32
34
Operative Temperature C
Very Satisfied
Satisfied
Somewhat Satisfied
Neutral
Somewhat Dissatisfied
Dissatisfied
Very Dissatisfied
(a) Operative Temperature
20.0 22.5 25.0 27.5 30.0 32.5 35.0
Air Temperature C
20
22
24
26
28
30
32
34
Mean Radiant Temperature C
Very Satisfied
Satisfied
Somewhat Satisfied
Neutral
Somewhat Dissatisfied
Dissatisfied
Very Dissatisfied
(b) Mean Radiant Temperature
Fig. 4.
(a) Adaptive comfort window for air speed of 0.3 m/s appended with data from
the thermal comfort survey responses. (b) The mean radiant temperature plotted
against air temperature for each survey response. The color of the data is assigned
based on occupant satisfaction votes. Each point is placed at the coincident operative
temperature. Clusters emerge with the Cold Tube on and off, with clear differences in
the response profiles for nearly the same range of air temperatures.
Cold Tube on and off clusters.172
Physiological Measurements.
Skin heat flux and temperature
173
measurements are plotted against system measurements in
174
figure 5b. Figure 5a shows an image of an author standing
175
in front (50 cm away) of a radiant cooling panel in the Cold
176
Tube taken using a thermal and visible light camera. The
177
color gradient shows the driving force for radiant heat transfer
178
from a person’s skin to the cooling panel. As expected, the
179
net heat flux from a person’s skin to the radiant cooling panel
180
scales proportionally to the supply water temperature. The
181
maximum value occurred when the water temperature was 13
182
◦
C, which corresponded to 156.8
W m−2
. With this 13
◦
C
183
Off High Temp. Cooling Cold
15
20
25
30
35
Temperature, C
89.8 W m
2
131.0 W m
2
156.8 W m
2
Skin
Air
Panel
Water
Fig. 5.
Heat flux measured from occupants’ wrists at three water temperature ranges,
showing the full temperature profile in the system from air to water and the associated
heat flux.
water supply, there was not a significant decrease in the air
184
temperature, from 31 to 30 ◦C. The large increase in radiant 185
heat flux occurred due to the radiant losses to the chilled
186
water. 187
Comparing the incremental increase in heat flux as water
188
temperature decreases allows one to extrapolate that if the
189
water temperature was the skin temperature, i.e. no radiant
190
heat exchange, allows us to extrapolate that 52.5
W m−2
were
191
due to convection for each dataset, and the remaining
W m−2192
were therefore attributed to radiation. For the cold 13
◦
C
193
water case, this means that 104.3
W m−2
were due to radiant
194
heat transfer. This further allows us to back-calculate a
TMRT 195
of 15.7
◦
C on the hemisphere of the body facing the panel.
196
This is consistent with the panel temperature measurement
197
produced with the radiometer. 198
More importantly, this physiological data offers an expla-
199
nation for the thermal comfort survey responses. As thermal
200
comfort requires metabolic heat to be lost, the increase in
201
heat flux from a person to the panel as the water temperature
202
decreases despite a nearly constant (close to skin temperature)
203
air temperature confirms that heat is being lost primarily to
204
the panels via radiation. 205
Condensation Prevention.
A primary research objective was
206
to observe chilled water supply temperatures that would be
207
allowable without condensation observed on any surface of the
208
radiant cooling panel. Such an environment has never been
209
constructed before. The membrane surface temperature is
210
difficult to directly measure since sensors placed on the infrared-
211
transparent material locally differed from their surroundings
212
due to radiant cooling. Instead, we slowly lowered the water
213
temperature at a rate of 4
◦
C per hour and watched for
214
signs of condensation. When condensation occurred, the air
215
01234567
Tair Tdp
,
C
0
2
4
6
8
10
12
14
16
Tdp Twater
,
C
Tdp Twater
= 2.0(
Tair Tdp
)
Fig. 6.
Chilling water slowly until the onset of condensation is observed allows the
air temperature minus the dew point temperature to be plotted against the dew point
minus water temperature to understand how cold water can be chilled for supply to
the Cold Tube.
temperature and supply water temperature were recorded. A
216
plot of this data is shown in figure 6a. The data is plotted
217
as the difference in the air temperature,
Tair
, and dew point,
218
Tdp
on the x-axis, and the y-axis is the difference in
Tdp
219
and the water temperature,
Twater
. This representation of
220
the data is done to reparametrize the data in terms of the
221
maximal convective heating provided from the air as dictated
222
by
Tair −Tdp
before the membrane goes below
tdp
. This control
223
logic is elegant, as it implies that as more heat in the air is
224
available for membrane heating, more cooling can be provided
225
through cooler chilled water without energy penalties since
226
the chilled membrane is convectively isolated from the warmer
227
air.228
Discussion229
The Cold Tube was an exciting step forward for exploring
230
novel modes of providing thermal comfort. As previously
231
discussed, the temperature range produced in the Cold Tube
232
has never been observed in the built environment (
6
), however
233
the findings presented in figure 4appear to be consistent
234
with the adaptive comfort framework (
18
). More specifically,
235
the environment produced in the Cold Tube is predicted to
236
be comfortable not only with a heat balance described in
237
the Methods section, but with the existing adaptive comfort
238
framework. Typically in the adaptive framework, the required
239
operative temperatures for comfort would be produced with
240
air or air and radiant systems, not a radiant system alone as
241
achieved in the Cold Tube. The Cold Tube is therefore a first
242
step in validating the adaptive comfort region with radiant
243
heat transfer only, implying that separation of comfort and
244
ventilation air is a plausible method of climate conditioning
245
for the tropics.246
Such a requirement is particularly important when large
247
air exchange rates are required to maintain ventilation rates
248
in spaces such as auditoriums, laboratories, classrooms, and
249
shared office spaces. If fresh air can be supplied at an arbi-
250
trary rate with little or no energy or comfort penalty, this
251
fundamentally changes the climate conditioning paradigm.
252
Further, as preliminarily demonstrated with the data from
253
the Cold Tube, strict dehumidification is also not necessary,
254
which could reduce large dehumidification loads across humid
255
climate regions worldwide (
19
). Using higher temperature
256
hydronic radiant cooling has also been demonstrated to reduce
257
the energy consumption of climate conditioning, as higher
258
2 4 6 8 10 12 14
Wavelength,
m
0.0
0.2
0.4
0.6
0.8
1.0
Value
300
K
= 0.824
300
K
= 0.052
300
K
= 0.124
Fig. 7. FTIR spectra of the LDPE infrared-transparent membrane material.
temperatures of 17-20
◦
C can be used instead of the more
259
traditional 4-8 ◦C used by conventional air systems (14). 260
Conclusions.
For the first time, a system was designed to
261
achieve 10 K of separation between the mean radiant temper-
262
ature and the air temperature, producing no condensation as
263
the supply temperatures and mean radiant temperatures were
264
well below the dewpoint, up to 20 K and 3.5 K, respectively.
265
The Cold Tube is an exciting step forward for demonstrating
266
(1) that radiation and convection can be separated for comfort
267
conditioning (2) to rely on radiation alone to produce com-
268
fortable conditions based on existing metrics. The thermal
269
comfort study conducted in Singapore in January 2019 is a
270
strong preliminary investigation into the applicability of such
271
a membrane assisted radiant cooling technology applied at
272
scale to reduce comfort-related energy demand worldwide. 273
Materials and Methods 274
Cold Tube Design, Construction, and Evaluation.
The Cold Tube was
275
constructed at the United World College, Southeast Asia (UWC-
276
SEA), Dover campus, in Singapore from August to October 2018.
277
The pavilion is enclosed by ten 1.2m x 2.1m (4’ x 8’) panels; two
278
horizontal panels at the top and eight vertical panels, with north
279
and south facing entrances. The surface of the panels are cooled
280
down below the dew point by chilled water from custom variable
281
speed chillers to provide radiant cooling. It is separated from the
282
hot and humid environment to avoid condensation by infrared trans-
283
parent membranes that are 82.4% transparent to thermal blackbody
284
radiation. A schematic of heat transfer about a single vertical panel
285
is shown in figure 1and the FTIR spectra of the 50 micron thick
286
LDPE infrared-transparent material is shown in figure 7.287
The supply and return temperatures of representative panels
288
were measured with high-precision thermistors (10K Precision Epoxy
289
Thermistor - 3950 NTC; +/- 1%). Net radiant heat transfer between
290
occupants and surfaces within a 150
◦
field of view was measured
291
with a pyrgeometer (Apogee, SL-510-SS; 0.12 mV per
W m−2
; 1%
292
measurement repeatability) and pyranometer (Apogee SP-510; 0.057
293
mV per
W m−2
; 1% measurement repeatability), which were manu-
294
ally directed in the direction of heat flux sensing. Skin temperature
295
and heat flux were measured with a skin temperature and heat
296
flux sensor (gSKIN
®
BodyTEMP Patch; +/- 0.3
◦
C). Air tempera-
297
ture and globe temperature were measured inside the pavilion with
298
Pt-100 thermistors (
±
0
.
1
◦
C). The panel temperature was mea-
299
sured with a non-contacting infrared temperature sensor (Melexis
300
®
MLX90614; +/- 0.3
◦
C), sealed inside the radiant panel facing the
301
chilled capillary mats. In addition, an air temperature sensor, rela-
302
tive humidity sensor, and air speed sensor from the ThermCondSys
303
5500 measurement system were placed at the location of the occu-
304
pant. The air temperature sensor was a Pt-100 thermistor (
±
0
.
1
◦
305
C). The air temperature sensor was shielded from radiation with
306
a highly reflective silver cone. The air speed sensor is a spherical
307
omnidirectional air speed sensor with temperature compensation,
308
vacuum covered with an aluminum coating that increases resistance
309
to contamination and decreases the effect of thermal radiation on
310
the accuracy of the measurement (
±
0.02
m s−1
). The relative
311
humidity sensor has a
±
2% accuracy. Measurements were taken at
312
10 second intervals, which were further smoothed by the minute for
313
analysis in this paper. Smoothed measurements for air speed,
vair
,
314
air temperature,
ta
, and mean radiant temperature,
tr
, were used
315
to compute the operative temperature, to, using equation 1(20).316
to=tr+ (ta×√10vair)
1 + √10vair
[1]317
Heat flux measurements from the gSKIN sensor were net heat
318
flux, meaning both convection and radiation fluxes were measured
319
simultaneously. Heat flux measurements were taken with three
320
supply water conditions, warm at 26
◦
C, ‘LowEx’ (short for low
321
exergy (
13
)) at 17
◦
C, and cold at 13
◦
C. If the air temperature
322
is consistent during these measurements, these three data points
323
allow for the regression of heat flux to be made back for water
324
temperature. This regression can be used to find the condition of no
325
radiant heat flux when
TMRT
=
Tskin
. This extrapolated heat flux
326
with no radiant heat flux would represent the convective heat flux,
327
Qconv
that occurs at
Tair
. This was treated as a constant value,
328
and allowed correction of the net heat flux,
Qnet
for the radiant
329
heat flux, Qrad as in equation 2.330
Qrad =Qnet −Qconv [2]331
Further, once a value of
Qrad
was calculated, knowing the skin
332
temperature,
Tskin
[K], the mean radiant temperature in the hemi-
333
sphere of the gSKIN sensor’s exposure,
TMRT ,hemi
[
◦
C], could be
334
back-calculated as shown in equation 3. In this equation
ε
is set
335
to 0.95 and
σ
is the Stephan-Boltzmann constant, 5
.
67
∗
10
−8
[
W336
m−2K−4
]. This value was compared to the measured values with
337
the pyrgeometer and pyranometer.338
TMRT ,hemi =4
rQrad
εσ −T4
skin [3]339
Mean Radiant Temperature Simulation.
Weather data collected at the
340
site was used to determine the required setpoint for comfort in the
341
constructed pavilion using a heat balance approach to expanding the
342
psychrometric comfort zone (
21
,
22
). The measured air temperature,
343
relative humidity, and average air speed of 0.3
m s−1
were used
344
in conjunction with the metabolic rate of a resting person, 1.2
345
met or 69.8
W m−2
and a skin wettedness of 0.06 for dry skin.
346
The color gradient in figure 8covered by the air temperature and
347
humidity data points shows the range of required mean radiant
348
temperature that the system must produce, in order for occupants
349
to feel comfortable, roughly between 23
◦
C and 25
◦
C depending
350
on the precise environmental condition. The white line traversing
351
the chart through the environmental data points shows the set of
352
points where the required mean radiant temperature for comfort is
353
the dew point temperature. Points above this line require a mean
354
radiant temperature lower than the dew point for occupants to
355
feel comfortable. This analysis demonstrates the need for a panel
356
construction separating the surface from the humid air to prevent
357
condensation.358
To achieve these required mean radiant temperatures, a geomet-
359
ric simulation was conducted to spatially map the mean radiant
360
temperature in the Cold Tube. To do this, first a grid of 750 points
361
is created on a plane at a fixed height of 1
m
above the floor. At each
362
location on this grid 1,280 geodesically distributed rays emanate.
363
They intersect the surfaces around them, with assigned known
364
surface temperatures, and the the temperature value at each inter-
365
section is averaged and recorded as the mean radiant temperature
366
at each point on the grid. A color gradient is then created based
367
on the MRT values. Further discussion of this simulation method
368
0 5 10 15 20 25 30 35 40
Air Temperature C
0
5
10
15
20
25
Humidity Ratio (g water/kg dry air)
MRT C
10%
20%
30%
40%
50%
60%
70%
100%
22.4
23.2
24.0
24.8
25.6
26.4
27.2
28.0
28.8
29.6
30.4
31.2
Metabolic Rate: 1.2 (met); Skin Wettedness: 0.06;
vair
: 0.3 m/s
20.3
22.0
23.7
25.4
27.1
28.8
30.5
32.2
Fig. 8.
Expanded Psychrometrics heat balance to determine the mean radiant
temperature required to produce comfort.
Fig. 9.
A simulated map of the mean radiant temperature distribution at a 1m height
in the Cold Tube with a supply water temperature of 18 ◦C.
from our previous work can be found in (
23
). The result from this
369
simulation is shown in figure 9. This simulation was conducted with
370
a supply water temperature of 18
◦
C water to the panels, with every
371
other temperature set to 31
◦
C. The simulation indicates that the
372
required range of mean radiant temperatures required for comfort
373
shown in figure 8can be met in the Cold Tube. The mapping of
374
MRT within the Cold Tube space allows for an understanding of the
375
effect of view factor on the perceived temperature as an occupant
376
walks through the space. 377
Thermal Comfort Study.
The primary goal of the thermal comfort
378
study was to assess whether individuals felt cooler in the Cold
379
Tube than just in shade, and whether the cooling provided by the
380
infrared transparent panels maintained to avoid condensation and
381
air conditioning was sufficient to cool occupants at short (1 minute)
382
and longer (10 minute) time intervals. These time intervals are
383
indicative of transient comfort or thermal delight, and steady state
384
thermal comfort. 385
Thermal delight refers to the instantaneous perception of comfort
386
when one has quickly transitioned from an uncomfortable environ-
387
ment to an environment more amenable to providing thermal com-
388
fort. An example is the experience of entering an air-conditioning
389
lobby after walking in a hot outdoor environment for a prolonged
390
duration. Those individuals who feel pleasure when a rush of cold air
391
blows over their hot and sweaty bodies are said to be experiencing
392
“thermal delight”. 393
Thermal comfort is the condition of the mind that expresses
394
satisfaction with one’s thermal environment. It is assessed empiri-
395
cally by subjective evaluation, often through the administration of
396
surveys. International standardization organizations, such as the
397
American Society for Heating, Refrigeration, and Air-Conditioning
398
Engineers (ASHRAE), nevertheless publish mathematical models
399
for estimating perceived thermal comfort of typical humans. Such
400
models are based on the estimated characteristics of clothing levels,
401
metabolic rates of occupants in an environment, and the estimated
402
air temperature, mean radiant temperature, humidity, and wind
403
speed of the environment. Measured data on these parameters are
404
often collected during survey-based studies of thermal comfort in
405
order to compare model predictions of thermal comfort to actual
406
responses.407
For the study, participants were escorted by a study administra-
408
tor to the experimental site on the United World College Southeast
409
Asia (UWCSEA) Dover campus. Once participants arrived at the
410
first location, the study commenced using the following procedure.411
Permission for the study was obtained from the Institutional Review
412
Board at the University of California, Berkeley who approved the
413
study (CPHS Protocol No. 20180-12-11636).414
1.
Each participant reached a state of thermal neutrality by
415
sitting 10-15 minutes in a shaded area exposed to elevated air
416
movement. Each participant was given control over the use of
417
a fan to make sure that thermal neutrality would be reached
418
in sufficient time.419
•
After 10 minutes, the participants would evaluate their
420
thermal comfort, and decide if an additional 5 minutes
421
beneath the fan would be required. After reaching the
422
thermal neutrality state, 15 minutes maximum under the
423
fan, the participant would be given a thermal comfort
424
survey for the first of four times. The entire thermal
425
comfort survey can be found in Supplemental Materials.
426
•
During this time, participants were asked to complete
427
a survey asking about their air conditioning and fan
428
preferences at home. This is an important step to under-
429
standing how closely our sample resembles the general
430
population. We asked participants what type of cooling431
they use at home and how often they use it.432
•
The participants clothing level was then be recorded by
433
the survey administrator.434
2.
The participant was asked to spend 7 minutes walking through
435
the shaded, covered and uncovered (sun-exposed) outdoor envi-
436
ronment on a predetermined path. After the walk participants
437
were surveyed about the thermal comfort right at that moment.
438
This is the second time they are filling out the thermal comfort
439
survey.440
3.
Next, the participant was asked to step into the pavilion.
441
Participants were subsequently surveyed after 1 min and after
442
10 minutes sitting in the pavilion, the third and fourth time
443
they will complete the survey, respectively.444
•
The objective of the third survey (1 min after entering
445
the pavilion) is to evaluate whether there is the effect of
446
thermal delight or significant feeling of heat relief due to
447
rapid heat release.448
•
The objective of the fourth survey (10 minutes after
449
entering the pavilion) is to understand how participants
450
respond to the pavilion’s environment with respect to
451
overall thermal comfort.452
4.
Finally, participants were asked to qualitatively compare the
453
pavilion environment to the first environment beneath the
454
fan. Participants were also asked to provide feedback about
455
what types of environments they would most like to see this
456
technology installed around Singapore.457
This experimental sequence was used to facilitate two different
458
experiments using the Cold Tube pavilion. These are:459
1.
Evaluation of thermal comfort of people in the active pavilion
460
- This study served as the benchmark information for the
461
pavilion. The pavilion was supplied with 10-15
◦
C water to
462
the radiant cooling panels, which created a perceived mean
463
radiant temperature between 22-24
◦
C. The air temperature
464
would be outdoor conditions of 28-32
◦
C and 60-80 % RH. 39
465
participants were recruited for this study, yet only 37 survey
466
responses were analyzed due to ambient weather condition
467
changes.468
2.
Control for comfort caused by the shade provided by the
469
pavilion - The pavilion will provide cooling to individuals
470
by providing shade only, with the active cooling turned off.
471
During the experiment, chilled water will not be supplied to
472
the pavilion, therefore this study is important to understand
473
the contribution of shading to cooling and to demonstrate the
474
additional benefit to the cooling that the active cooling of the
475
water supplies to occupants. 18 participants were recruited for
476
this study, yet only 16 survey responses were analyzed due to
477
ambient weather condition changes and data loss. 478
ACKNOWLEDGMENTS.
This project was funded by the Na-
479
tional Research Foundation IntraCREATE Grant No. NRF2016-
480
ITC001-005 [Pantelic and Rysanek]. The authors would also like
481
to personally thank Simon Thomas and the United World Col-
482
lege Southeast Asia (UWCSEA) Facilities staff for all of their help
483
facilitating the construction of the Cold Tube demonstrator. 484
1. DA McIntyre, I Griffiths, Radiant temperature and thermal comfort. Vol. CIB Commission 485
W45, (1972). 486
2. C Yagoglou, Report of committee to consider the report of the new york state commission on 487
ventilation. Am. Soc. Heat. Vent. Eng.30, 254–256 (1924). 488
3. T Bedford, C Warner, The globe thermometer in studies of heating and ventilation. The J. 489
Hyg.34, 458–473 (1934). 490
4. A Standard 55, Thermal environmental conditions for human occupancy. Am. Soc. Heating, 491
Refrig. Air conditioning Eng. (2017). 492
5. PO Fanger, Thermal comfort. Analysis and applications in environmental engineering. 493
(Copenhagen: Danish Technical Press.), (1970). 494
6. VF Licina, et al., Development of the ashrae global thermal comfort database ii. Build. Environ.495
142, 502 – 512 (2018). 496
7. R Morse, Radiant cooling. Archit. Sci. Rev.6, 50–53 (1963). 497
8. T Cheung, S Schiavon, T Parkinson, P Li, G Brager, Analysis of the accuracy on pmv–ppd 498
model using the ashrae global thermal comfort database ii. Build. Environ.153, 205–217 499
(2019). 500
9. E Teitelbaum, et al., Revisiting radiant cooling: condensation-free heat rejection using 501
infrared-transparent enclosures of chilled panels. Archit. Sci. Rev., 1–8 (2019). 502
10. M Isaac, DP van Vuuren, Modeling global residential sector energy demand for heating and 503
air conditioning in the context of climate change. Energy Policy 37, 507–521 (2009). 504
11. EI Agency, Residential energy consumption survey, 2015 recs survey data. Tables HC6 8505
(2015). 506
12. LG Harriman III, D Plager, D Kosar, Dehumidification and cooling loads from ventilation air. 507
ASHRAE journal 39, 6 (1997). 508
13. F Meggers, V Ritter, P Goffin, M Baetschmann, H Leibundgut, Low exergy building systems 509
implementation. Energy 41, 48–55 (2012). 510
14. A Schlueter, A Rysanek, F Meggers, 3for2: Realizing spatial, mater ial, and energy savings 511
through integrated design. CTBUH J., 40–45 (2016). 512
15. AP Raman, MA Anoma, L Zhu, E Rephaeli, S Fan, Passive radiative cooling below ambient 513
air temperature under direct sunlight. Nature 515, 540 (2014). 514
16. Y Zhai, et al., Scalable-manufactured randomized glass-polymer hybrid metamaterial for day- 515
time radiative cooling. Science 355, 1062–1066 (2017). 516
17. T Li, et al., A radiative cooling structural material. Science 364, 760–763 (2019). 517
18. F Nicol, Temperature and adaptive comfort in heated, cooled and free-running dwellings. 518
Build. Res. & Inf.45, 730–744 (2017). 519
19. B Wellig, B Kegel, M Meier, E Basler, A Partner, "verdopplung der jahresarbeitszahl von 520
klimakälteanlagen durch ausnützung eines kleinen temperaturhubs“. Ernst Basler+ Partner, 521
Zürich, iA Forschungsprogramm UAW, Bundesamt für Energie, Bern (2006). 522
20. ISO, 7726, ergonomics of the thermal environment, instruments for measuring physical quan- 523
tities. Geneva: Int. Standard Organ. (1998). 524
21. E Teitelbaum, F Meggers, Expanded psychrometric landscapes for radiant cooling and natu- 525
ral ventilation system design and optimization. Energy Procedia 122, 1129–1134 (2017). 526
22. E Teitelbaum, P Jayathissa,C Miller, F Meggers, Design with comfort: Expanding the psychro- 527
metric chart with radiation and convection dimensions. Energy Build.209, 109591 (2020). 528
23. TKFM Dorit Aviv, Eric Teitelbaum, Generation and simulation of indoor thermal gradients. 529
Proc. IBPSA Build. Simul. Conf. In Press.Rome (2019). 530