Science of the Total Environment 634 (2018) 1119–1125

Contents lists available at ScienceDirect

Science of the Total Environment

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

A new structure of permeable pavement for mitigating urban heat island

Yong Liu, Tian Li ⁎, Hangyu Peng
State Key Lab of Pollution Control and Resource Reuse, Tongji University, Shanghai 200092, China

H I G H L I G H T S G R A P H I C A L A B S T R A C T

• The capillary column enhances evapora-

tion of the new pavement.
• The new pavement is cooler 9.4 °C in
maximum than conventional pavement.
• The new pavement provides a longer
cooling duration.
• Figuline pavers present a better cooling
effect than fine concrete pavers.

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

Article history: The urban heat island (UHI) effect has been a great threat to human habitation, and how to mitigate this problem
Received 11 January 2018 has been a global concern over decades. This paper addresses the cooling effect of a novel permeable pavement
Received in revised form 23 March 2018 called evaporation-enhancing permeable pavement, which has capillary columns in aggregate and a liner at the
Accepted 4 April 2018
bottom. To explore the efficiency of mitigating the UHI, bench-scale permeable pavement units with capillary col-
Available online 11 April 2018
umns were developed and compared with conventional permeable pavement. Criteria of capillary capacities of
Editor: P. Kassomenos the column, evaporation rates, and surface temperature of the pavements were monitored under simulated rain-
fall and Shanghai local weather conditions. Results show the capillary column was important in increasing evap-
Keywords: oration by lifting water from the bottom to the surface, and the evaporation-enhancing permeable pavement was
Urban heat island cooler than a conventional permeable pavement by as much as 9.4 °C during the experimental period. Moreover,
Permeable pavement the cooling effect of the former pavement could persist more than seven days under the condition of no further
Capillary columns rainfall. Statistical analysis result reveals that evaporation-enhancing permeable pavement can mitigate the UHI
Water-holding paver effect significantly more than a conventional permeable pavement.
Evaporation cooling effect
© 2018 Elsevier B.V. All rights reserved.

1. Introduction phenomenon of climate change and is a major factor in observed tem-

The urban heat island (UHI) effect has become a global issue and a
great threat to human habitation and the urban environment as urban-
ization advances. This has been conceptualized and studied thoroughly
since 1982 (Oke, 1982). The UHI effect, characterized by higher temper-
atures in urban areas than those in rural areas, is a most documented
⁎ Corresponding author.
E-mail address: tianli@tongji.edu.cn (T. Li).
perature increments (Santamouris, 2007; Nakayama and Fujita, 2010;
Founda, 2011). The effect occurs because of the changing nature of a
city and consequent reduction in vegetation and evapotranspiration.
This results in a heat balance between natural ground and building ma-
terials different than before (Arnfield, 2003; Stone et al., 2010). The UHI
increases air temperature in urban areas with temporal and regional
variability, and the value of this excess has been accurately shown to
be 5–15 °C in recent research (Santamouris, 2013a). Up to now, quanti-
ties of researches on UHI including carbon footprint and energy

https://doi.org/10.1016/j.scitotenv.2018.04.041
0048-9697/© 2018 Elsevier B.V. All rights reserved.
1120 Y. Liu et al. / Science of the Total Environment 634 (2018) 1119–1125
consumption have been conducted (Rossi et al., 2016), and more studies
on effective countermeasures still need to be explored.
Cool pavement is one method to mitigate the UHI effect by changing
city surfaces and has been well researched in recent years (Qin, 2015).
Existing methods of cool pavement mainly include increasing solar re-
flectance, enhancing evaporation, and reducing sensible heat transfer
to the urban atmosphere. These methods are respectively associated
with reflective, evaporative, and heat-harnessing pavements (Qin,
2015). To keep pavement cool, reflective pavements usually resurface
the pavements with light-colored materials or mix them with light-
colored aggregates and whitish cementitious material when con-
structed to increase pavement albedos (Levinson and Akbari, 2002;
Levinson et al., 2010), besides, retroreflective materials were also pro-
posed recently (Morini et al., 2017). Heat-harnessing pavements usually
extract absorbed heat as renewable energy to reduce surface tempera-
ture (Mallick et al., 2011; Wu and Yu, 2012). Evaporative pavements
typically use water-holding pavers to suppress surface temperatures,
and mainly include permeable and water-retentive pavements (Qin,
2015).
Typically, permeable pavement cooling durations are only 1–2 days
following rainfall (Li et al., 2013). Sometimes water is sprayed on the
pavement to stay cool (Yamagata et al., 2008), which has substantial
cost and is inconvenient. In the present study, a new permeable pave-
ment called evaporation-enhancing permeable pavement was devel-
oped as a potential method of UHI mitigation. This pavement is
designed to manage stormwater in a region where the groundwater
table is high. Therefore, a liner is necessary for the permeable pavement
to prevent infiltration from polluting groundwater. In the new pave-
ment, capillary columns are installed in aggregate to lift runoff captured
by the liner to the surface, which can promote evaporation and cool the
pavement for a longer period. Moreover, cooling effects of two kinds of
pavers (figuline and fine concrete) were investigated, which are usually
applied to sidewalks and light parking lots respectively, considering
compressive strength.
Fig. 1. Sections of conventional and evaporation-enhancing permeable pavement units.
2. Materials and methods
2.1. Evaporation-enhancing permeable pavement
The structures of common and evaporation-enhancing permeable
pavements are shown in Fig. 1. A common permeable pavement con-
sists of a permeable paver, laying course, geotextile, and aggregate in se-
quence. In the new permeable pavement, there are capillary columns
inbuilt in aggregate, and all of its materials are hydrophilic, including
hydrophilic geotextile, fine sand laying course, and water-holding
pavers. These promote the lifting of water by capillary columns to
reaching the surface of pavers. In the present study, several bench-
scale permeable pavement units were made with PVC with dimensions
210 mm × 210 mm × 400 mm (length × width × height). These were
then loaded with a 30-cm graded aggregate with nominal maximum
size 19 mm, geotextile, 2-cm laying course, and pavers of height 6 cm
in turn. There was a transparent U-pipe placed at the bottom to monitor
water level in each unit. At height 15 cm from the bottom, another pipe
was placed to overflow captured runoff.
2.2. Capillary column and capillary capacity measurement
The capillary columns were made of figuline with dimensions
300 mm × 100 mm × 50 mm (length × width × height), from a brick-
yard in Yixing, Jiangsu Province, China. Measured compressive strength
of the capillary column was 20 MPa. To evaluate the capillary effect of
the column, water absorption capacity and capillary height variation
were measured. The water absorption capacity was evaluated by
weighing the mass of water absorbed by the columns at a given time,
and capillary height was determined by the maximum saturated height
that water could reach. First, the columns were dried in a drying oven
for 24 h at a temperature of 105 °C. After cooling to room temperature,
the columns were weighed and then put in a tank with a constant 10-
mm water depth. Finally, the masses of columns and column capillary
 Y. Liu et al. / Science of the Total Environment 634 (2018) 1119–1125
height were recorded every 10 min during the first hour and at a one-
hour interval subsequently (Karasawa et al., 2006).
2.3. Experimental design and method
The bench-scale experiment was conducted on the roof of a building
in Tongji University, Shanghai, China. There were eight permeable pave-
ment units of the same dimension and one free surface unit (Fig. 2). As
Table 1 shows, F1, F2 and F3 were parallel units of figuline paver with
capillary columns, for investigating reproducibility of the cooling effect
of evaporation-enhancing permeable pavements. The capillary column
ratios (CR = capillary column section / unit section) were 1/4. F0 was
the control unit for F1, F2 and F3 without capillary column. C1, C2 and
C3 were another set of parallel units with fine concrete paver, for eval-
uating the effect of paver character on evaporation and cooling effi-
ciency. For this set, the same capillary columns were used, and C0 was
the control unit for C1, C2 and C3, without a capillary column. Character-
istics of the two types of pavers are shown in Table 2. Pore diameter was
measured by mercury intrusion porosimetry. Maximum capillary height
was the saturated height of the paver in 24 h. A free surface unit (B0)
was established to evaluate maximum evaporation rates of days and
nights.
At the beginning of the experiment, the units were placed on the
building roof. Then, 2500 mL of tap water (equivalent to 56 mm of
rain) was sprayed on each unit with a sprinkler. Masses of all units be-
fore and after the simulated rainfall were weighed using an electronic
scale with accuracy 1 g. On subsequent days, masses of each unit were
weighed at 8:00 and 20:00 every day, and surface temperatures of
each unit were measured from 8:00 to 20:00 at a one-hour interval
every day. The experiment was conducted for seven days and repeated
three times during March–May 2017. Data analyzed were from 20:00
on 17 April through 20:00 on 24 April 2017.
2.4. Weather data logging
Weather data were logged with a portable meteorograph
(NHQXZ-W-609, manufactured by Wuhan Zhongke Nenghui Tech-
nology Development Co., Ltd.) that was mounted on the building
roof. Monitored data included air temperature, wind speed, and rel-
ative humidity.
Fig. 2. The bench scale permeable pavement units.
1121
Table 1
The experimental permeable pavement units and a free surface unit.
Pavement type CR
1/4 0 B0
Figuline pavers F1, F2, F3 F0 (Free surface unit)
Concrete pavers C1, C2, C3 C0
2.5. Evaporation rate and latent heat flux calculation
Mass losses during daytime and nighttime for each unit were used to
calculate evaporation rate via Eq. (1). Daytime evaporation rates of the
units were obtained by unit mass losses between 8:00 and 20:00, and
nighttime rates were from 20:00 through 8:00 the following day.
M t2 −Mt1
ER ¼ ; ð1Þ
Aðt 2 −t 1 Þ
where t1 is the time of the beginning (hr), t2 is the time of ending (hr),
and Mt1 and Mt2 are masses of the unit in the two periods (kg), A is the
area of the unit (m2) and ER is the evaporation rate (kg/m2/h).
2.6. Surface temperature measurement and data analysis
Surface temperatures of the units were measured by infrared ther-
mometer at a one-hour interval during daytime. Before the temperature
measurement, emissivities of the figuline block and concrete block were
tested, because the emitted infrared radiation of an object is a function
of its emissivity. Both pavers were measured to be 0.95 using hot plate
techniques (Doulos et al., 2004). During the measurement, there could
be non-uniformities of surface temperature. The infrared thermometer
can calculate average surface temperature automatically. ANOVA and
Tukey HSD tests were used when data had normal distributions with
homoscedasticity, and the Jonckheere-Terpstra test was used when
data had unknown distributions.
3. Results and discussion
3.1. Capillary capacities of figuline columns
Capillary capacity of the figuline column was measured before the
experiment. Monitored criteria including water absorption and capillary
1122 Y. Liu et al. / Science of the Total Environment 634 (2018) 1119–1125
Table 2
Characteristic of two kinds of pavers.
Paver type Size (mm) Pore diameter (μm) Water absorption capacity (g/cm3)
Figuline paver 200 × 100 × 50 10 0.14
Concrete paver 200 × 100 × 60 0.05 0.07
height varied over time. The results in Fig. 3 show good performance and
reproducibility among the three samples. The amount of water absorbed
by columns increased rapidly at the beginning and became saturated
after 6 h. The average maximum adsorption quantity for each column
was 208 g, which was 0.14 g/cm3. Capillary height of the columns
showed a pattern similar to the water absorption curve. The three sam-
ples all reached the top of the columns (30 cm) within 5 h. This perfor-
mance was better than that of the capillary columns in the road-runoff
infiltration facility (Furumai, 2016). Slopes of the two sets of curves de-
creased with time. This revealed that water flux in capillary columns de-
creased as capillary height increased, with 2.28 mL/cm2 h in the first
hour and 0.7 mL/cm2 h when water reached the top. These findings
show that the figuline column can effectively suck water from the bot-
tom to the surface and promote evaporation of the pavements.
3.2. Weather data and evaporation rate
Weather during the experimental period was sunny or cloudy. Air
temperature and relative humidity are presented in Fig. 4. Air tempera-
ture during the period fluctuated because of cloudy skies on the fourth
and fifth days. Evaporation rates of the units and free surface unit during
daytime and nighttime are shown in Fig. 5. F1, F2 and F3 had similar pat-
terns (Fig. 5a). Mean evaporation rates of the three parallel units during
daytime and nighttime were 4.15 and 0.44 mm/d, respectively, with co-
efficients of variation (C·V) 0.89% and 1.13%. This demonstrates favor-
able reproducibility for the three parallel units.
As shown in Fig. 5b, B0 had the highest evaporation rates compared
with the other units because of sufficient available water at the surface,
with maxima 7.6 mm/d during daytime and 1.2 mm/d during night-
time. Evaporation of F1 fluctuated with B0 but the percentage of F1/B0
decreased with time; its value was 82.9% on the first day and 38.1% on
the last day. Evaporation of F0 mainly occurred on the first two days
and there was little on subsequent days. This is consistent with results
of other research (Li et al., 2013; Nemirovsky et al., 2013). Descriptive
statistics show that mean evaporation rates of F1 and F0 during the
first two days reached 6.47 and 4.81 mm/d during daytime, respec-
tively, whereas those rates during the next five days were 3.23 and
0.48 mm/d. These results indicate that capillary columns did increase
evaporation of the permeable pavers.
Similar results were found for fine concrete paver units. C1, C2 and
C3 also had good parallel performance (Fig. 5a) with mean 3.59 mm/d
Fig. 3. Capillary capacities of three figuline column samples.
Maximum capillary height (cm) Compressive strength (Mpa)
20 20
7.5 30
during daytime and 0.42 mm/d during nighttime, with corresponding
C·V 2.67% and 8.56%. However, differences were found when compar-
ing the two types of paver units. Figuline paver unit F1 had greater evap-
oration than concrete paver unit C1 beginning the first day, but the
difference declined gradually. The evaporation of F1 was even lower
than C1 from the sixth day (Fig. 5b). The water level of figuline paver
units began to be lower than concrete paver units beginning the first
day, and this difference amplified on subsequent days (Fig. 5c). That
probably accounted for the difference of water absorption capacity be-
tween the figuline and concrete paver. That is, the figuline paver had a
greater water absorption capacity and could sustain more evaporation
than the concrete paver, which caused water remaining in the figuline
unit decrease faster than that in the concrete unit. The reason why evap-
oration conditions of the two types of units changed over time might be
the amplified water level difference between them. This is because
water flux on top of the capillary columns decreased when distance to
the water level increased (Fig. 3). This caused water to reach the surface
and evaporation of F1 lessened in subsequent days.
After seven days, there were still 3.0 and 5.6 cm water depths in F1
and C1, respectively, while water levels in F0 and C0 nearly retained
the initial height of the first day. The monitoring program partially ter-
minated because the mean rainfall interval during summer and autumn
in Shanghai was less than seven days.
3.3. Surface temperature
(1) Effect of capillary columns
Surface temperature variations of the model units are shown in
Fig. 6a. Variations of F1, F2 and F3 were consistent during the experi-
ment (C·V = 0.41%). There was no significant difference among them
using Tukey HSD (p N 0.05), demonstrating that the evaporation-
enhancing permeable pavement also functioned consistently from the
perspective of surface temperature variation.
As Fig. 6 shows, the temperature fluctuation of F0 had almost the
same pattern as F1 on the first day. This was mainly because water evap-
orated during the period was from the pavers themselves, which was
retained from the simulated rainfall. There was a difference of temper-
ature between F0 and F1 beginning on the second day, when the tem-
perature of F0 began to be higher than that of F1; this persisted on
subsequent days. This was because the capillary columns provided
water for pavers of F1, which caused F1 to have more latent heat loss
 Y. Liu et al. / Science of the Total Environment 634 (2018) 1119–1125 1123

Fig. 4. Weather data during experimental period.

than F0. Pavers of F0 became dry after the second day and temperature
increased rapidly, while pavers of F1 remained wet and the temperature
was much lower than F0 beginning the second day. This means that the
cooling effect of evaporation-enhancing permeable pavement was
much stronger and of longer duration than conventional permeable
pavement. Moreover, the surface temperature of F1 was lower by a
maximum 9.4 °C relative to F0 at noon on the second day.
During daytime of 18 and 19 April, mean temperatures of F1 and F0
were 22.9 and 26.2 °C, respectively (Fig. 7). There was no significant
difference between them from the ANOVA test (p N 0.05), revealing
that the capillary columns were not important during the first two
days. However, the ANOVA test showed a significant difference between
F1 and F0 during 20–24 April (p b 0.05), with mean temperatures 26.3
and 30.9 °C, respectively. This indicates that the evaporation-enhancing
permeable pavement could reduce surface temperature significantly. In
such a case, less sensible heat is released to the atmosphere, therefore,
the UHI could be effectively mitigated (Takebayashi and Moriyama,
2012).

Fig. 5. Evaporation rate and water level variations of the units.

1124 Y. Liu et al. / Science of the Total Environment 634 (2018) 1119–1125
Fig. 6. Surface temperature variations of the two kind paver units.
(2) Effect of paver type
Temperature variations of the three concrete paver units in parallel
over seven days had a coefficient of variation of 1.1%, there was no sig-
nificant difference among them from the Jonckheere-Terpstra test (p N
0.05). Nevertheless, there were differences between the two types of
paver units. Temperature variations of C0 and C1 were very similar at
the beginning, whereas the value of F0 exceeded F1 from the second
day. In addition, the surface temperature of C1 was a maximum 6.7 °C
lower than that of C0 on the fifth day, whereas F1 was a maximum 9.4
°C cooler than F0 on the second day. The reason accounting for these
was differences in water absorption and capillary capacities of the two
types of pavers. Because the capillary capacity of the concrete paver
was less than that of the figuline paver (Table 2), evaporation of C1
less than that of F1, which resulted in different latent heat losses and
cooling effects of two types of paver units.
According to the previous research (Santamouris, 2013b), surface
temperature and albedo of the paver had a negative relationship. In
this study, F0 had a higher temperature than C0 when the two types
of pavers were completely dry, with respective maxima of 47.2 °C and
43.2 °C on the last day. This suggests that the figuline paver has a
lower albedo than the concrete block. However, mean temperature
and the temperature range box of F1 were lower than C1 during
both 18–19 and 20–24 April (Fig. 7). This indicates that the figuline
Fig. 7. Boxplots of surface temperature range of the units.
paver has a greater cooling effect than the concrete paver in the
evaporation-enhancing permeable pavement. Statistical analysis
with the Jonckheere-Terpstra test shows a significant difference be-
tween the two types of pavers (p b 0.05), which proves that the
paver type produces a difference in effect on UHI mitigation.
There are a number of problems to be solved before the application
of evaporation-enhancing permeable pavement. What kind of material
that capillary column adopted plays a most important part. And a paver
with effective water absorption capacity is also necessary, because it can
help the water lifted by capillary columns reach surface of the pave-
ment. Evaporation-enhancing permeable pavement may be a good
choice for low impact development in a region with a high groundwater
table.
4. Conclusions
In this study, the cooling effects of new permeable pavements with
capillary columns and conventional permeable pavements were inves-
tigated. Capillary capacities of the columns, evaporation rates, and sur-
face temperatures of the pavement units were measured and analyzed
for UHI mitigation. The main findings are as follows.
(1) The figuline columns have good capillary capacities, which can
lift water to a height of 30 cm within 5 h. Water absorption ca-
pacity of the column was 0.14 g/cm3 on average.
(2) Evaporation and latent heat loss of evaporation-enhancing per-
meable pavements could persist more than seven days during
the experiment, but only during the first two days for the con-
ventional permeable pavement.
(3) Evaporation-enhancing permeable pavement could contribute
substantially to UHI mitigation, and was a maximum 9.4 °C
cooler than conventional permeable pavement.
(4) The cooling effect of the figuline paver was greater than that of
the fine concrete paver because of a greater water absorption
capacity.
Acknowledgements
This research was supported by the Science and Technology Com-
mission of Shanghai Municipality (No. 16DZ1202106) and National Wa-
ter Pollution Control and Management Technology Major Project (No.
2017ZX07207001).
References
Arnfield, A.J., 2003. Two decades of urban climate research: a review of turbulence, ex-
changes of energy and water, and the urban heat island. Int. J. Climatol. 23 (1), 1–26.
Doulos, L., Santamouris, M., Livada, I., 2004. Passive cooling of outdoor urban spaces. The
role of materials. Sol. Energy 77 (2), 231–249.
Founda, D., 2011. Evolution of the air temperature in Athens and evidence of climatic
change: a review. Adv. Build. Energy Res. 5, 7–41.
Furumai, H., 2016. Evaluation of Rainwater Harvesting and Use Potential Considering Cli-
mate Change in Arakawa Watershed. 2016. NOVATECH, pp. 1–4.
Karasawa, A., Toriiminami, K., Ezumi, N., Kamaya, K., 2006. Evaluation of performance of
water retentive concrete block pavements. Proceedings of the 8th international con-
ference on concrete block pavement, pp. 1–8 (San Francisco, CA2006).
Levinson, R., Akbari, H., 2002. Effects of composition and exposure on the solar reflectance
of portland cement concrete. Cem. Concr. Res. 32 (11), 1679–1698.
Levinson, R., Akbari, H., Berdahl, P., et al., 2010. A novel technique for the production of
cool colored concrete tile and asphalt shingle roofing products. Sol. Energy Mater.
Sol. Cells 94 (6), 946–954.
Li, H., Harvey, J.T., Holland, T.J., Kayhanian, M., 2013. The use of reflective and permeable
pavements as a potential practice for heat island mitigation and stormwater manage-
ment. Environ. Res. Lett. 8 (1), 015023.
Mallick, R., Carelli, J., Albano, L., Bhowmick, S., Veeraragavan, A., 2011. Evaluation of the
potential of harvesting heat energy from asphalt pavements. Int. J. Sustain. Eng. 4,
164–171.
Morini, E., Castellani, B., Presciutti, A., et al., 2017. Experimental analysis of the effect of ge-
ometry and façade materials on urban district's equivalent albedo. Sustainability 9,
1245.
Nakayama, T., Fujita, T., 2010. Cooling effect of water-holding pavements made of new
materials on water and heat budgets in urban areas. Landsc. Urban Plan. 96 (2),
57–67.
 Y. Liu et al. / Science of the Total Environment 634 (2018) 1119–1125
Nemirovsky, E.M., Welker, A.L., Lee, R., 2013. Quantifying Evaporation from Pervious Con-
crete Systems: Methodology and Hydrologic Perspective. vol. 139(4) pp. 271–277.
Oke, T.R., 1982. The energetic basis of the urban heat island. Q. J. R. Meteorol. Soc. 108
(455), 1–24.
Qin, Y., 2015. A review on the development of cool pavements to mitigate urban heat is-
land effect. Renew. Sust. Energ. Rev. 52, 445–459.
Rossi, F., Bonamente, E., Nicolini, A., et al., 2016. A carbon footprint and energy consump-
tion assessment methodology for UHI-affected lighting systems in built areas. Energy
Build. 114, 96–103.
Santamouris, M., 2007. Heat island research in Europe: the state of the art. Adv. Build. En-
ergy Res. 1, 123–150.
Santamouris, M., 2013a. Energy and Climate in the Urban Built Environment. Routledge,
Abingdon-on-Thames, UK.
1125
Santamouris, M., 2013b. Using cool pavements as a mitigation strategy to fight urban heat
island—a review of the actual developments. Renew. Sust. Energ. Rev. 26, 224–240.
Stone, B., Hess, J.J., Frumkin, H., 2010. Urban form and extreme heat events: are sprawling
cities more vulnerable to climate change than compact cities. Environ. Health
Perspect. 118 (10), 1425–1428.
Takebayashi, H., Moriyama, M., 2012. Study on surface heat budget of various pavements
for urban heat island mitigation. Adv. Mater. Sci. Eng. 2012, 1–11.
Wu, G., Yu, X., 2012. System Design to Harvest Thermal Energy Across Pavement Struc-
ture. Energytech, 2012 IEEE, pp. 1–4.
Yamagata, H., Nasu, M., Yoshizama, M., Miyamota, A., M, M., 2008. Heat island mitigation
using water retentive pavement sprinkled with reclaimed wastewater. Water Sci.
Technol. 57, 763–771.