Download - Effect of ageing on solar spectral reflectance of roofing membranes: Natural exposure in Roma and Milano and the impact on the energy needs of commercial buildings

Paolini et al. (2014). Energy and Buildings - http://dx.doi.org/10.1016/j.enbuild.2014.08.008

1/26

Effect of ageing on solar spectral reflectance of roofing membranes:

natural exposure in Roma and Milano and the impact on the energy

needs of commercial buildings

Riccardo Paolini1,2*, Michele Zinzi3, Tiziana Poli1,

Emiliano Carnielo4, Andrea Giovanni Mainini1

Published on Energy and Buildings

http://dx.doi.org/10.1016/j.enbuild.2014.08.008

© 2014. This manuscript version is made available under the CC-BY-NC-ND 4.0 license

http://creativecommons.org/licenses/by-nc-nd/4.0/

Disclaimer

This document was prepared as an account of work sponsored by the Italian Revenue Agency and the Italian Ministry of Economic Development. While this document is believed to contain correct information, neither the Italian Government nor any agency thereof, nor the Research Institutions to which the authors are affiliated, nor any of their employees, makes any warranty, express or implied, or assumes any legal responsibility for the accuracy, completeness, or usefulness of any information, apparatus, product, or process disclosed, or represents that its use would not infringe privately owned rights. Reference herein to any specific commercial product process, or service by its trade name, trademark, manufacturer, or otherwise, does not necessarily constitute or imply its endorsement, recommendation, or favouring by the Italian Government or any agency thereof, or the Research Institutions to which the authors are affiliated. The views and opinions of authors expressed herein do not necessarily state or reflect those of the Italian Government or any agency thereof, or their Research Institutions.

1 Politecnico di Milano, Department of Architecture, Built environment and Construction engineering 2 Heat Island Group, Environmental Energy Technologies Division, Lawrence Berkeley National Laboratory, Berkeley, CA, USA

* Corresponding author: [email protected] – Via Ponzio, 31 – 20133 Milano, Italia – Tel. +390223996015 3 ENEA – UTEE-ERT Italian National Agency for New Technologies, Energy and Sustainable Economic Development 4 Università degli Studi Roma Tre


2/26

Abstract

Highly reflective roofs, widely known as cool roofs, can reduce peak surface temperatures

and the energy required to cool buildings, mitigate urban microclimates, and offset CO2.

However, weathering, soiling, and biological growth affect their solar reflectance. In this

study, the solar spectral reflectances of 12 roofing membranes were measured before the

exposure and after 3, 6, 12, 18, and 24 months of natural ageing in Roma and Milano, Italy.

The membranes with an initial solar reflectance greater than 0.80, for example, decreased in

reflectance by 0.14 in Roma and 0.22 in Milano after 2 years. Then, for a typical highly

insulated commercial building, the annual cooling savings were calculated to be reduced by

4.1-7.1 MJ m-2 y-1 per 0.1 loss in reflectance. When the buildings are non-insulated, the

savings reduction is 58-71 MJ m-2 y-1 in Milano and 70-84 MJ m-2 y-1 in Roma. Ageing yielded

a reduction of the cooling savings that could be achieved with a new white membrane of 14-

23% in Roma and of 20-34% in Milano. Moreover, in Milano, an aged, white, highly

insulated roof, which has a solar reflectance of 0.56, may reach a surface temperature 16°C

higher than a new roof, which has a solar reflectance of 0.80.

Keywords: solar reflectance; ageing; soiling; cool roof; building energy need; UV-Vis-NIR;

natural exposure; building envelope.

1. Introduction

The microclimate in urban areas is very different from that in non-urban adjacent areas [1–3]

and typically leads to relevant differences in the energy needs of urban buildings [4–7] and

the peak electricity demand [8]. For instance, for office buildings within the urban area of


3/26

Milano, the heating energy need is estimated to be from 30% to 60% lower than the needs

outside the city, while the cooling need is from 15% to 70% higher [9,10].

To mitigate urban climates and to reduce the cooling energy required, highly reflective

roofing materials have been widely suggested and recommended [11–18]. However,

especially in an urban environment where air pollution is significantly higher than in rural

areas, the surfaces of buildings are subject to weathering and to the deposition of soot and

other particulate matter, which cause a change in their reflectance [19–26].

Even if data are available, and unfortunately, the available data are seldom spectral,

they are provided by short exposure research programmes or concern only North America. In

the U.S. especially, extensive data about more than 2500 roofing products are made available

by the Cool Roofing Rating Council (CRRC) for three exposure sites: one in a temperate sub-

urban environment in Ohio, which has moderate air pollution [27]; one in a hot-dry extra-

urban climate in Arizona; and one in a hot and humid extra-urban climate in Florida. The

CRRC reports the clean and aged average solar reflectance ( Sρ ) and thermal emittance (ε )

measured at the three sites after three years of natural exposure. At the CRRC's sites, all of the

exposed products, excluding those with an initial solar reflectance ( S0ρ ) lower than 0.20,

present losses increasing with S0ρ [22]. For instance, for products with S0ρ greater than 0.80,

the three-site average loss of Sρ after three years is equal to 0.16, with a maximum of 0.24 in

Florida and a minimum of 0.08 in Arizona. With regard to the thermal emittance, in the

CRRC's database, most variations for products with an initial ε greater than 0.85 are within ±

0.05 after 3 years, while if the initial ε is lower than 0.50, there is an average increase of

approximately 0.07.

The ageing of high-albedo roofing may naturally lead to a decrease in the energy

savings achievable using cool materials (i.e., up to 20% compared to the first year for

detached houses in Sacramento, CA) as assessed by Bretz and Akbari [28]. Even if techniques


4/26

to restore the initial solar reflectance exist [29,30], cleaning does not seem economically or

environmentally sustainable when only trying to achieve energy savings.

To complete the available information with data measured in urban environments,

which is where most buildings with high energy needs are typically located, a selection of 12

roofing membranes, including synthetic, factory-applied coatings on synthetic membranes,

field-applied coatings on modified bitumen, and modified bitumen with roofing granules,

were exposed and analysed in Milano and Roma, Italy. Their solar spectral reflectance was

measured when new and after 3, 6, 12, 18, and 24 months of exposure; the impact of solar

reflectance variation on the surface temperatures of insulated or non-insulated roofs was then

measured and was used to calculate the resulting energy needs for heating and cooling of a

typical commercial building in Milano and Roma.

2. Experiment

2.1 Selected materials

Twelve roofing membrane products available on the market with S0ρ values ranging from

0.26 to 0.85 (Figure 1) and having varying surface roughnesses were selected. Some

membranes were somewhat glossy, while others were matte; for each roofing material class

(e.g., modified bitumen), products offering different spectral features (i.e., cool and non-cool

coloured) were chosen. Those selected comprise a wide set of features of non-black

waterproofing materials made of modified bitumen, PVC and polyolefin, with different

spectral reflectances, surface roughnesses and open porosity.


5/26

Selected samples

m01 (ρs0 = 0.26) m02 (ρs0 = 0.84) m03 (ρs0 = 0.76) m04 (ρs0 = 0.82) Grey flexible polyolefin (matte

and with anti-slip surface) White factory-applied coating (glossy) on flexible polyolefin

White flexible polyolefin (matte and anti-slip)

White thermoplastic polyolefin (glossy)

m05 (ρs0 = 0.46) m06 (ρs0 = 0.85) m07 (ρs0 = 0.59) m08 (ρs0 = 0.73)

Grey PVC membrane (glossy) White PVC membrane (matte) Cool beige thermoplastic

polyolefin (matte) Modified-bitumen with white field-applied coating (matte)

m09 (ρs0 = 0.72) m10 (ρs0 = 0.39) m11 (ρs0 = 0.28) m12 (ρs0 = 0.23)

Modified-bitumen with white field-applied coating (glossy)

Modified-bitumen with cool coloured field-applied coating

Modified-bitumen with photoactive roofing granules

Modified-bitumen with standard roofing granules

Figure 1: Selected roofing membranes, and initial solar reflectance ( S0ρ ).

2.2 Natural exposure procedure

The selected roofing membranes were exposed to the natural elements at two urban sites: in

Roma (41° 55' 57" N, 12° 27' 54" E; 35 m above mean sea level) and in Milano (45° 28' 48"


6/26

N, 9° 13' 46" E; 123 m above mean sea level), offering different climates and pollutant

concentrations. In both cases, the exposures occurred approximately halfway between the city

centres and the peripheries on two non-shaded roofs and were distant from the primary

sources of pollution. The roofing membranes were exposed at a low slope (i.e., 1.5%

according to Italian code practices [31] and Swiss standards [32]). In Milano, additional

specimens were also exposed facing south with a slope of 45°.

The samples were measured when new and after 3, 6, 12, 18, and 24 months of natural

exposure, which began on April 18th, 2012. In addition, in Milano, one year later on May 3rd,

2013, a second low-sloped exposure of membranes from m02 to m10 (Figure 1) was started to

assess the variability caused by different ageing conditions occurring in short-term

programmes. At each time point, the samples were retrieved, measured in the laboratory, and

re-exposed; each sample remained unexposed for approximately one week when the

measurements were taken. Three samples of 10 cm x 10 cm in size per product were exposed

for each site and slope condition; they were fastened to metal frames according to ISO 2810

[33] 80 cm above the roof (Figure 2). To observe the speed of the reflectance loss, in the

second year, 39 additional specimens of membrane m06, which was a white PVC single-ply,

were exposed, retrieving three coupons and measuring them each week during the first two

months and then every two weeks for the following two months.

2.3 Reflectance measurement method

The spectral reflectance was measured with two identical Perkin Elmer Lambda 950

spectrophotometers, one in Roma and one in Milano; the latter was used after the first year.

Both machines were equipped with a 150 mm Spectralon-coated integrating sphere, a

photomultiplier tube, and lead sulphide detectors.


7/26

Figure 2: Exposure facility in Milano. A weather station on the same roof monitors solar global radiation, air temperature, relative humidity and pressure, wind velocity and direction, and precipitation.

Reflectance measurements were carried out and compared to a Spectralon calibrated reference

in the 300-2500 nm wavelength range with a spectral resolution of 5 nm. The centre point of

each sample lit by the measurement beam was used for the analysis; thus, soiling edge effects

were excluded. The slit aperture was set to 2 nm in the visible range and in servo mode in the

near-infrared range. The servo mode allowed the instrument to automatically change the slit

aperture in order to optimize the energy input as a function of wavelength. Broad band values

were calculated from the spectral data according to ASTM E 903 [34] using the global solar

horizontal irradiance distribution given for air mass 1 at 5 nm intervals, as described by

Levinson et al. [35]. The visible band was considered to range from 380 to 780 nm, according

to ISO 9050 [36]. For each product and exposure condition (i.e., site, orientation, and slope)


8/26

for the three specimens, we computed the average spectral curve and then the integrated

values.

3. Building energy simulations

3.1 Simulation tool

The evolution over time of solar reflectance already provides an indication of the possible

variation in the surface energy balance of the building envelope. However, dynamic heat and

moisture transport numerical simulations may provide a deeper insight into the impact of

reflectance changes on the hygrothermal performance of the building envelope and on the

building energy needs for heating and cooling. The software model WUFI Plus 2.5.3 [37] was

used, which was validated within the context of IEA Annex 41 [38]. This model resolves the

enthalpy balance with the finite control volumes method, coupling the heat transfer with the

liquid and vapour moisture transport in porous media and accounting for both latent heat

transformations and the influence of moisture content and temperature on the thermal and

moisture transport properties of building materials [39].

As in the work of Levinson et al. [40], the building simulations included the

temperature dependency of thermal conductivity (λ). For expanded polystyrene (EPS), λ =

0.034 W m-1 K-1 at -20°C and λ = 0.054 W m-1 K-1 at +80°C from WUFI database were used,

in agreement with the provisional formulas for EPS given in ISO 10456 [41]. Simulations

were performed with a time step of 15 minutes.

3.2 Case study

As a case study, a typical one-storey commercial building located in Milano or Roma was

considered, which was modelled as a single zone measuring 50 m x 40 m x 6.5 m (i.e., inner


9/26

dimensions) and east-west oriented with ribbon windows on the south and north façades,

precast walls, and a flat precast roof with unprotected membrane (Table 1).

Table 1: Building envelope components surface area, and thermal transmittance (U). The thermal conductivity (λ) of expanded polystyrene (EPS) is set to 0.04 W m-1 K-1, while for

reinforced concrete we considered λ = 1.6 W m-1 K-1 and volumic mass equal to 2300 kg m-3. Windows are described also by frame incidence, and by their total solar energy transmittance for normal incidence (g⊥).

Building envelope

component

Orientation and area

(m2) Highly insulated case No insulation

Roof (flat)

2000

U = 0.19 W m-2 K-1 Double tee precast reinforced

concrete slab (flange 0.05 m thick), 0.05 m of concrete screed, vapour

barrier (water vapour diffusion equivalent thickness sd = 100 m), 0.20

m of EPS, and roofing membrane

U = 4.7 W m-2 K-1 Double tee precast reinforced

concrete slab (flange 0.05 m thick), 0.05 m of concrete screed, and

roofing membrane

Floor 2000

U = 0.19 W m-2 K-1 0.15 m of reinforced concrete slab over the ground (no direct contact),

0.20 m of EPS, and 0.08 m of concrete (screed and cement

pavement finishing).

U = 3.8 W m-2 K-1 Reinforced concrete slab (0.15 m,

with cement pavement finishing) over the ground (no direct contact).

Walls (precast concrete)

N: 281 S: 269 E: 248 W: 260

(tot: 1058)

U = 0.27 W m-2 K-1 Sandwich panel made of two skins of

0.01 m thick high density fibre-reinforced concrete each, and 0.14 m

of EPS

ρs = 0.50

U = 1.4 W m-2 K-1 Panel made of two skins of 0.05 m thick reinforced concrete each, and

0.02 m of low density EPS as lightening (we considered only 2 cm of EPS to take into account thermal bridges between panels, and those

due to the edges and the ribs of each panel)

ρs = 0.50

Ribbon windows (5 m from ground)

S: 44 N: 44

Doors E: 12 S: 12

Double-glazing with low emissivity

coating (ε = 0.20) U = 1.6 W m-2 K-1

g⊥ = 0.60 Frame incidence = 30%

Uncoated single-glazing U = 5.0 W m-2 K-1

g⊥ = 0.81 Frame incidence = 30%

The building typology and envelope technology were representative of industrial and

commercial sectors in Italy. Internal partitions and furniture provided low thermal inertia,

which was offered by the precast building envelope and by 75 m3 of bottled liquid


10/26

merchandise, which was modelled as an internal partition 0.10 m thick with water’s thermal

properties within a polyethylene envelope. The exterior climate was defined by hourly typical

weather data from Meteonorm, and the inner set point conditions and inner loads are detailed

in Table 2 and Table 3.

Table 2: Inner heat loads (internal gains), moisture and CO2 production for the department store (0.20 persons m-2 and 10 W m-2 of electrical loads) and for the wholesale store (low occupancy) during opening hours (08:00-22:00).

Building Heat

(W m-2) Moisture (g h-1 m-2)

CO2 (g h-1 m-2)

Human activity (Met)

Department store 34.2 11.8 7.30 1.16 Wholesale store 10.0 0.00 0.00 0.00

Table 3: Design conditions (set-point values for indoor air temperature and relative humidity). Hour Tmin (°C) Tmax (°C) RHmin (%) RHmax (%)

0 12 35 10 80 8 20 26 40 60 22 12 35 10 80

3.3 Variables

As variables, the building envelope insulation (i.e., high or no thermal insulation) and the

inner loads and the presence of occupants (i.e., average-high loads for a department store or

low loads for a wholesale store) were considered. During the business’ open hours from 08:00

to 20:00, mechanical ventilation with 65% heat recovery efficiency provided an air change

rate of 0.4 air changes per hour (ACH) for the wholesale store or 1.2 ACH for the department

store according to the specifications given in EN 15251 [42], with a constant air change rate

by infiltration equal to 0.05 ACH. The building with a new white membrane was modelled as

clean ( Sρ = 0.80) or as aged with aSρ computed based on regression of the experimental data

after 24 months of natural exposure in Milano and Roma. For comparison, the buildings with

an aged black roofing membrane, whose Sρ was assumed to be 0.10 due to the lack of

experimental data, was also modelled. For the purpose of this study, the solar reflectance


11/26

decay was accounted for only in the case of the roofing membrane, while for the glazing and

opaque façades were considered constant values.

4. Results and discussion

4.1 Experimental results

Both in Milano and in Roma after 3 months of exposure (i.e., from mid-April to mid-July

2012), the three membranes with S0ρ values greater than 0.80 lost an average of 0.13 in solar

reflectance (Figure 3). At 6 months of exposure (i.e., mid- to late October 2012), remarkable

variations were not measured. After the initial drop during the first 3 months, the greatest

losses in reflectance occurred during the first winter (i.e., from 6 to 12 months of exposure

time, namely, from October 2012 to April 2013), which is when building heating systems are

operating. The second winter did not yield further loss of the same magnitude, and stability

was achieved for most of the exposed products. After 24 months of natural ageing,

membranes withS0ρ values greater than 0.80 were found to have lost an average of 0.14 in

Roma and 0.22 in Milano, while membranes with S0ρ values in the range of 0.60 to 0.80 lost

an average of 0.15 in Roma and 0.24 in Milano.

Low-reflectivity membranes (S0ρ = 0.20-0.30) showed modest absolute variations (i.e.,

within 0.02-0.05) with a Sρ of modified bitumen with roofing granules (i.e., m11 and m12)

that fluctuated over time, sometimes exceeding the initial value; this occurred because of soot

deposits on the modified bitumen substrate (Sρ ~ 0.05) underneath the granules, as discussed

in detail by Berdahl et al. [43]. In some cases (e.g., for m02 and m04), fluctuations were also

shown for the glossy single-ply membranes. Differences between the ageing conditions in

Roma and Milano emerged during the wintertime because it is in average colder in Milano

(i.e., 2404 heating degree days with 20°C as set-point temperature) than in Roma (i.e., 1415


12/26

heating degree days) with higher pollutant concentrations in Milano due to the increased

building heating (e.g., over one year, the 95th percentile of PM10 concentration was equal to

98.7 49 µg m-3 in Milano and to 49 µg m-3 in Roma [44,45]). These differences were also due

to the weak air circulation in Milano, where the average wind velocity was equal to 1.5 m s-1,

while the average wind velocity was 3.3 m s-1 in Roma. For membranes with S0ρ values

greater than 0.60, the difference between Roma and Milano was nearly 0.08.

Sol

ar r

efle

ctan

ce v

aria

tion

0 3 6 12 18 24-0.30

-0.25

-0.20

-0.15

-0.10

-0.05

0.00

0.05

0.10Roma horMilano horMilano 45° south

m01ρs0 = 0.26

Sol

ar r

efle

ctan

ce v

aria

tion

0 3 6 12 18 24-0.30

-0.25

-0.20

-0.15

-0.10

-0.05

0.00

0.05

0.10Roma horMilano horMilano 45° southMilano hor 2 expnd

m02ρs0 = 0.84

Sol

ar r

efle

ctan

ce v

aria

tion

0 3 6 12 18 24-0.30

-0.25

-0.20

-0.15

-0.10

-0.05

0.00

0.05


m03ρs0 = 0.76

Sol

ar r

efle

ctan

ce v

aria

tion

0 3 6 12 18 24-0.30

-0.25

-0.20

-0.15

-0.10

-0.05

0.00

0.05


m04ρs0 = 0.82

Sol

ar r

efle

ctan

ce v

aria

tion

0 3 6 12 18 24-0.30

-0.25

-0.20

-0.15

-0.10

-0.05

0.00

0.05


m05ρs0 = 0.46

Sol

ar r

efle

ctan

ce v

aria

tion

0 3 6 12 18 24-0.30

-0.25

-0.20

-0.15

-0.10

-0.05

0.00

0.05


m06ρs0 = 0.85

Sol

ar r

efle

ctan

ce v

aria

tion

0 3 6 12 18 24-0.30

-0.25

-0.20

-0.15

-0.10

-0.05

0.00

0.05


m07ρs0 = 0.59

Sol

ar r

efle

ctan

ce v

aria

tion

0 3 6 12 18 24-0.30

-0.25

-0.20

-0.15

-0.10

-0.05

0.00

0.05


m08ρs0 = 0.73

Sol

ar r

efle

ctan

ce v

aria

tion

0 3 6 12 18 24-0.30

-0.25

-0.20

-0.15

-0.10

-0.05

0.00

0.05


m09ρs0 = 0.72

Sol

ar r

efle

ctan

ce v

aria

tion

0 3 6 12 18 24-0.30

-0.25

-0.20

-0.15

-0.10

-0.05

0.00

0.05


m10ρs0 = 0.39

Sol

ar r

efle

ctan

ce v

aria

tion

0 3 6 12 18 24-0.30

-0.25

-0.20

-0.15

-0.10

-0.05

0.00

0.05


m11ρs0 = 0.28

Sol

ar r

efle

ctan

ce v

aria

tion

0 3 6 12 18 24-0.30

-0.25

-0.20

-0.15

-0.10

-0.05

0.00

0.05


m12ρs0 = 0.23

Exposure time (months) Exposure time (months) Exposure time (months)

Figure 3: Solar reflectance variation (aged - initial) with exposure time (measurements before the exposure, and after 3, 6, 12, 18, and 24 months).


13/26

Excluding the case of m09 (i.e., a field-applied coating with uneven thickness), all of

the membranes showed a standard deviation of less than 0.035 (median < 0.008) after 24

months of exposure in Milano and less than 0.017 in Roma (median < 0.010); the clean

membranes showed a maximum standard deviation of 0.017 (median < 0.006).

The membranes exposed and tilted by 45° in Milano showed intermediate values

between the low-sloped specimens in both Roma and Milano, with the exception of m09,

which was a modified bitumen with a glossy field-applied coating that retains much of its

initial reflectance due to its high slope. After 12 months, the second low-sloped exposure in

Milano, which included membranes with S0ρ values greater than 0.40, showed trends similar

to those of the first exposure, with absolute losses lower than 0.04 on average. During the

spring of 2013, when the second exposure began, the climate was exceptionally rainy (i.e.,

374 mm in 2012 from mid-April to mid-July and 467 mm in the same period in 2013). This

led to less soil deposition on the membranes during the first three months. After 24 months of

exposure, the Sρ values for all of the exposed membrane reached a stable value within 0.01 of

each other; this resulted in a linear relationship between the initial and aged reflectances

(Figure 4).

For m06, which was a matte white PVC membrane withS0ρ = 0.84, the loss of

reflectance was measured to be 0.08 after the first few days of exposure, especially in the

visible spectrum (Figure 5). In certain cases, reflectance decreased even after a sequence of

rainy events (i.e., between day 7 and 28 of exposure, or after day 86); these events probably

cleaned the air somewhat but also carried airborne particulate matter onto the specimens. In

other circumstances (i.e., after day 35), the reflectance increased after showers, likely caused

by the cleaning effects of the rain.


14/26

Focusing on the spectral data, the portions of the solar spectrum where weathering and

soiling have the largest impact and the factors that primarily affect the dirt pickup and the

variations in reflectance can be determined (Figure 6).

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

Initial solar reflectance

Age

d so

lar

refle

ctan

ce

ρRM 24 hor = 0.78 * ρs0+ 0.02; R = 0.972

ρMI 24 hor = 0.64 * ρs0 + 0.05; R = 0.922

ρMI 24 tilt45 = 0.67 * ρs0 + 0.05; R = 0.872

Figure 4: Linear regression of solar reflectance after 2 years of natural exposure in Roma (low sloped) and in Milano (low sloped and 45° tilted south oriented) as a function of initial solar reflectance.

02468

m s

−1 Wind velocity

0

10

20

30

mm

h−1 Precipitation

01020304050

µ g

m−3

PM2.5 PM10

(d.u

.)

0 7 21 35 49 71 86 101 116-0.20

-0.10

0.00∆ρs ∆ρv ∆ρn

Exposure time (days)

Figure 5: Solar, visible, and near infrared reflectance variation (aged - initial) during the first four months of natural ageing for a white PVC membrane (m06) low sloped exposure in Milano, measured each week for the first two months, and each two weeks in the following two months. The decreasing trend of reflectance is plotted together with the average daily concentration of fine particulate matter (PM2.5 and PM10), and hourly values of precipitation and wind velocity (measured on the same roof of the exposure). Air pollution data are provided by ARPA Lombardia (regional environmental protection agency), for the station 'Pascal', about 250 m far from the exposure site [45].


15/26

500 1000 1500 2000 2500

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

Wavelength (nm)

Ref

lect

ance

T0 s = 0.84 u = 0.07 v = 0.91 n = 0.83T3 s = 0.70 u = 0.08 v = 0.73 n = 0.73T6 s = 0.69 u = 0.07 v = 0.72 n = 0.72T12 s = 0.61 u = 0.09 v = 0.62 n = 0.66T18 s = 0.59 u = 0.08 v = 0.59 n = 0.64T24 s = 0.62 u = 0.07 v = 0.63 n = 0.67

a) m02 - MILANO

uv vis nir

500 1000 1500 2000 2500

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

Wavelength (nm)

Ref

lect

ance


b) m02 - ROMA

uv vis nir

500 1000 1500 2000 2500

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

Wavelength (nm)

Ref

lect

ance

T0 s= 0.59 u= 0.07 v= 0.54 n= 0.73T3 s= 0.53 u= 0.07 v= 0.47 n= 0.67T6 s= 0.50 u= 0.07 v= 0.45 n= 0.64T12 s= 0.42 u= 0.07 v= 0.36 n= 0.54T18 s= 0.42 u= 0.07 v= 0.36 n= 0.54T24 s= 0.42 u= 0.06 v= 0.36 n= 0.54

c) m07 - MILANO

uv vis nir

500 1000 1500 2000 2500

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

Wavelength (nm)

Ref

lect

ance

T0 s= 0.59 u= 0.07 v= 0.54 n= 0.73T3 s= 0.53 u= 0.07 v= 0.47 n= 0.66T6 s= 0.52 u= 0.07 v= 0.47 n= 0.66T12 s= 0.48 u= 0.07 v= 0.43 n= 0.61T18 s= 0.45 u= 0.08 v= 0.39 n= 0.58T24 s= 0.45 u= 0.08 v= 0.39 n= 0.57

d) m07 - ROMA

uv vis nir

500 1000 1500 2000 2500

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

Wavelength (nm)

Ref

lect

ance


e) m10 - MILANO

uv vis nir

500 1000 1500 2000 2500

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

Wavelength (nm)

Ref

lect

ance


f) m10 - ROMA

uv vis nir

Figure 6: Spectral and computed solar (s), UV (u), visible (v), and near infrared (n) reflectance after 3, 6, 12, 18 and 24 months of low sloped exposure for m02 (white factory applied coating on TPO) in Milano (a) and Roma (b), and for m07 (a cool beige TPO) in Milano (c) and Roma (d), and for m10 (a cool grey field applied coating on modified bitumen) in Milano (e) and Roma (f).


16/26

For the white membranes m02 and m08, or for the membranes with a similar spectrum,

the abatement of reflectance is evident mainly in the visible spectrum and in the first part of

the NIR. The shape of the spectra is altered, especially between 420 and 600 nm, as a result of

UV irradiation, water and early physical degradation. Especially for the field-applied coatings,

some yellowing was noticed after only a few months of exposure, and the non-linearity in the

reduction of reflectance in this portion of the spectrum was also a symptom of physical

degradation (Figure 7). After approximately 600-800 nm, the spectrum was shifted down,

retaining its original shape; then, between 2200 and 2500 nm and between 300 and 400 nm,

there was almost no variation. This had little significance for the surface energy balance

because in the UV band, all of the exposed membranes had reflectances near 0.07, and the

portion of the spectrum up to 400 nm contributed to the broadband solar irradiance by

approximately 7%, while after 2200 nm, there was only a small fraction of the available solar

power in peak clear sky conditions.

Cool-coloured membranes such as m10, which was a field-applied coating on modified

bitumen, suffered a small absolute loss (near 0.05) because the absorption coefficient of soot

decreases with wavelength [19,46], which was less than the other non-cool membranes with

similar S0ρ values (e.g., m05). The spectrum of membrane m10 was almost unchanged in the

visible region (Figure 6e,f), while for m07, which was a cool beige polyolefin membrane and

had higher reflectance in the first part of the NIR region (780-1400 nm), the average relative

loss was higher (Figure 6c,d).

Although the absolute values of the reflectance of samples exposed at the two sites

were different, the shapes of the spectra of the aged membranes were almost the same in both

Roma and Milano, suggesting that in metropolitan areas, the intensity of the deposition is

different although the basic ingredients of soiling are the same (e.g., products of combustion

from vehicles' engines and heating plants, etc.). Instead, there was a difference between the


17/26

shape of the spectra of the horizontally exposed samples and the spectra of those tilted by 45°

facing south. For the latter, the effect of weathering and soiling between 420 and 600 nm was

shown to be slightly less pronounced than for the low sloped specimens (Figure 8).

Figure 7: Coefficient of determination (R2) of a linear regression (soiled spectral reflectance as a function of clean reflectance) of all the UV-Vis-NIR measurements performed each 5 nm for all the samples. Within the chart, an example of visual comparison between a new and an aged white field applied coating, showing yellowing of the surface.

300 350 400 450 500 550 600 650 700 750 800

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

Wavelength (nm)

Ref

lect

ance

T0 T12 Roma horT12 Milano hor (2nd exp)T12 Milano 45° southT12 Milano hor

uv vis nir

Figure 8: Spectral reflectance between 300 and 800 nm of a white factory applied coating onto a single-ply TPO membrane (m02): comparison between time zero, and values after 12 months of natural exposure in Roma (low sloped), and in Milano (low sloped, first and second exposure), and 45° tilted facing south.


18/26

4.2 Simulations results

For a department store with a high inner heat load and high levels of insulation in the roof and

pavement in Milano and in Roma, the impact of membrane ageing on the annual cooling load

is not significant (11-15 MJ m-2 y-1) but already reduces the cooling savings achievable with a

new white membrane by one-fifth in Roma and by one-third in Milano, while the difference

for heating energy needs is negligible (Table 4).

Obviously, the impact of reflectance variation is maximized for non-insulated

buildings. For highly insulated buildings, the annual cooling savings are shown to be reduced

by 4.1-7.1 MJ m-2 y-1 per 0.1 loss in reflectance, while non-insulated buildings could save 58-

71 kWh m-2 y-1 in Milano and 70-84 kWh m-2 y-1 in Roma. With poor or no insulation, ageing

of the membrane might require cooling the building, which otherwise might not be necessary.

Ageing of the roofing membrane obviously also impacts the peak conditions. Focusing

on the highly insulated case with low inner loads in Milano and in Roma, the increase in solar

reflectance yields a higher peak power demand for cooling by 2-3 kW.

Table 4: Annual building energy need (useful energy) for heating and cooling for a typical commercial building whether located in Milano or in Roma, with new or aged white membrane, or with a black aged membrane.

Milano Roma Inner loads

Insulation Energy need

(kWh m-2)

New white

ρs = 0.80

Aged white

ρs = 0.56

Aged black

ρs = 0.10

New white

ρs = 0.80

Aged white

ρs = 0.64

Aged black

ρs = 0.10

Heating 1.92 1.81 1.57 0.59 0.59 0.58 High insulation

Cooling 64.0 68.0 76.1 90.6 93.8 104

Heating 246 213 170 122 102 59.7 High

No insulation

Cooling 14.8 49.7 150 32.0 59.8 192

Heating 20.0 18.9 17.3 4.00 3.63 2.60 High insulation

Cooling 9.68 12.4 17.7 15.2 17.2 24.4

Heating 296 255 203 171 144 87.1 Low

No insulation

Cooling 2.00 23.7 111 10.2 28.9 143


19/26

While the impact on energy needs from the increase in solar reflectance depends on

the building inner loads and the building envelope design, the impact on the surface

temperatures of the roofing membrane itself is extremely relevant, especially for high-

insulation setups. For instance, in Milano, the computed exterior surface temperatures of a

new white membrane (Sρ = 0.80) over 20 cm of thermal insulation is between -10°C and

+36°C, while the same membrane when aged (Sρ = 0.56) may reach +52°C with frequent

differences exceeding 10°C during the hot season and peak differences of 19°C (Figure 9).

Month

Tem

pera

ture

diff

eren

ce (

°C)

J F M A M J J A S O N D

0

2

4

6

8

10

12

14

16

18

20

22 ∆ Tmax (aged-new) insulateda)

Month

Tem

pera

ture

diff

eren

ce (

°C)

J F M A M J J A S O N D

0

2

4

6

8

10

12

14

16

18

20

22 ∆ Tmax (aged-new) insulatedb)

Figure 9: Difference between maximum daily temperatures of an aged white membrane (ρs =

0.56) and a white new membrane (ρs = 0.80) over 20 cm of expanded polystyrene (a) or over a non-insulated roof (b).

When the roof is not insulated, the surface temperature of the new membrane would

range from -2°C to +31°C, while if the membrane is soiled, the temperature would exceed

+40°C. These peak temperatures differ because the maximum temperatures are reached at

different times due to the different relative importance of air temperature and solar radiation

and the variation in the thermal conductivity of building materials with temperature.


20/26

On a typical peak summer day (i.e., maximum irradiance of 860 W m-2 and air

temperature of 30°C), a highly insulated soiled roof may be 16°C hotter than the same roof

when the membrane is new; even without insulation, the peak difference between the soiled

and clean membrane is nearly 10°C (Figure 10a).

2 4 6 8 10 12 14 16 18 20 22 24

15

20

25

30

35

40

45

50

55

Local standard time

Tem

pera

ture

(°C

)

Exterior surface - 19 JULAged - insulatedAged - non-insulatedNew - insulatedNew - non-insulated

a)

2 4 6 8 10 12 14 16 18 20 22 24

-10

-5

0

5

10

15

20

25

30

Local standard time

Tem

pera

ture

(°C

)

Exterior surface - 16 JANAged membrane - non-insulatedNew membrane - non-insulatedAged membrane - insulatedNew membrane - insulated

b)

2 4 6 8 10 12 14 16 18 20 22 24

16

18

20

22

24

26

28

30

32

34

36

Local standard time

Tem

pera

ture

(°C

)

Interior surface - 19 JUL

Aged - non-insulated

New - non-insulated

Aged - insulated

New - insulated

c)

2 4 6 8 10 12 14 16 18 20 22 24

4

6

8

10

12

14

16

18

20

22

24

Local standard time

Tem

pera

ture

(°C

)

Interior surface - 16 JAN

Aged - insulated

New - insulated

Aged - non-insulated

New - non-insulated

d)

Figure 10: Exterior surface temperature of a white soiled membrane and of the same membrane when clean (a) for 19th July and (b) for 16th January; difference between the interior (ceiling) surface temperature of a roof with a white soiled membrane and with the same membrane when clean (c) for 19th July and (d) for 16th January.


21/26

On a cold but sunny day in Milano (i.e., maximum irradiance of 230 W m-2 and air

temperature of 3.5°C), the peak differences between the soiled and clean membranes over an

insulated or non-insulated roof may be, respectively, 5.7°C and 2.6°C (Figure 10b). If the roof

is highly insulated, the influence of the variation in solar reflectance on the ceiling surface

temperature is negligible, while in the case of poor or no insulation, a soiled roof may be even

5.6°C hotter on the interior than a clean roof in summer and nearly 1.4°C hotter in winter

(Figure 10c,d).

6. Conclusions

Cool roofs are effective at mitigating heat islands and reducing the cooling energy needs of

buildings, but weathering and soiling may strongly affect their solar reflectance, especially in

polluted urban environments.

Twelve commercially available roofing membranes with initial solar reflectances

ranging from 0.26 and 0.84 were selected. Samples of each were exposed to the natural

rooftop environments in two major European urban areas: Roma and Milano. Their solar

spectral reflectances were measured before exposure and after 3, 6, 12, 18, and 24 months of

exposure. The results showed that ageing, which was mainly caused by weathering and

soiling but not mould growth, strongly affects the solar reflectance of roofing membranes,

even within the first few months. The most reflective membranes (S0ρ > 0.80) lost 0.14 in

Roma and 0.22 in Milano on average after 2 years of natural ageing. Cool-coloured materials

were subject to lower losses due to soot than standard materials with the same solar

reflectance but lower NIR reflectances because the absorbance of soot particles decreases

with increasing wavelength and because the most relevant impact of weathering is between

420 and 600 nm.


22/26

The loss of reflectance impacts the energy needs of buildings and the surface

temperature of the roofing membrane. To quantify this aspect, representative commercial

buildings in Roma and Milano with a new white flat roof ( Sρ = 0.80) and with an aged roof

( Sρ equal to 0.64 for Roma and 0.56 for Milano, as computed from the regression of

experimental data to take into account the average loss) were modelled. For highly insulated

buildings, the annual cooling savings were shown to be reduced by 4.1-7.1 MJ m-2 y-1 per 0.1

loss in reflectance, while for non-insulated buildings, the loss of cool roof savings due to

ageing was shown to be 58-71 MJ m-2 y-1 in Milano and 70-84 MJ m-2 y-1 in Roma. In all

considered cases, ageing yielded a reduction of the cooling savings that could be achieved

with a new white membrane of approximately 14-23% for the building in Roma and of 20-

34% in Milano.

Reflectance loss also impacted the peak cooling power demand, with a difference of 2-

3 kW in the considered case study. Moreover, an aged white ( Sρ = 0.56) highly insulated roof

in Milano under peak summer conditions may reach a surface temperature 16°C higher than a

new roof ( Sρ = 0.80).

The impact of ageing on the optical-radiative performance of building envelope

materials is thus relevant and must be considered in the modelling and design of sustainable

and durable near-zero energy buildings. Thus, there is a need to develop a new generation of

cool roofing materials that should be easy to clean and able to retain their initial solar

reflectance for as long as possible.

Acknowledgments

This work was funded by the Italian Ministry for Economic Development with the projects

“Valutazione delle prestazioni di cool materials esposti all'ambiente urbano” and “Sviluppo di


23/26

materiali e tecnologie per la riduzione degli effetti della radiazione solare”; and by Politecnico

di Milano & Agenzia delle Entrate (Italian Revenue Agency) with the project “Cinque per

mille junior - Rivestimenti fluorurati avanzati per superfici edilizie ad alte prestazioni”. The

authors wish to thank Florian Antretter and Daniel Zirkelbach (Fraunhofer Institut für

Bauphysik) for precious suggestions about hygrothermal simulations; and Hugo Destaillats,

Ronnen Levinson, and Mohamad Sleiman (Lawrence Berkeley National Laboratory) for

valuable exchanges about the analysis of UV-Vis-NIR spectral data. The authors gratefully

acknowledge anonymous reviewers, who provided relevant and helpful suggestions to

improve the presentation of results.

References

[1] P. Bacci, M. Maugeri, The urban heat island of Milan, Nuovo Cim. C. 15 (1992) 417–424. http://dx.doi.org/10.1007/BF02511742.

[2] A.M. Rizwan, L.Y.C. Dennis, C. Liu, A review on the generation, determination and mitigation of Urban Heat Island, J. Environ. Sci. 20 (2008) 120–128. http://dx.doi.org/10.1016/S1001-0742(08)60019-4.

[3] C.S.B.S.B. Grimmond, M. Roth, T.R.R. Oke, Y.C.C. Au, M. Best, R. Betts, et al., Climate and more sustainable cities: Climate information for improved planning and management of cities (Producers/Capabilities Perspective), in: Procedia Environ. Sci., 2010: pp. 247–274. http://dx.doi.org/10.1016/j.proenv.2010.09.016.

[4] S. Hassid, M. Santamouris, N. Papanikolaou, A. Linardi, N. Klitsikas, C. Georgakis, et al., Effect of the Athens heat island on air conditioning load, Energy Build. 32 (2000) 131–141. http://dx.doi.org/10.1016/S0378-7788(99)00045-6.

[5] M. Santamouris, N. Papanikolaou, I. Livada, I. Koronakis, C. Georgakis, a Argiriou, et al., On the impact of urban climate on the energy consumption of buildings, Sol. Energy. 70 (2001) 201–216. http://dx.doi.org/10.1016/S0038-092X(00)00095-5.

[6] M. Kolokotroni, I. Giannitsaris, R. Watkins, The effect of the London urban heat island on building summer cooling demand and night ventilation strategies, Sol. Energy. 80 (2006) 383–392. http://dx.doi.org/10.1016/j.solener.2005.03.010.

[7] M. Kolokotroni, X. Ren, M. Davies, A. Mavrogianni, London’s urban heat island: Impact on current and future energy consumption in office buildings, Energy Build. 47 (2012) 302–311. http://dx.doi.org/10.1016/j.enbuild.2011.12.019.

[8] H. Akbari, M. Pomerantz, H. Taha, Cool surfaces and shade trees to reduce energy use and improve air quality in urban areas, Sol. Energy. 70 (2001) 295–310. http://dx.doi.org/10.1016/S0038-092X(00)00089-X.


24/26

[9] T. Poli, L. Pietro Gattoni, R. Paolini, G. Pansa, M. Favaron, S. Pilati, The Influence of the Urban Heat Island over Building Energy Demand: The Case of Milan, in: Seventh Int. Conf. Urban Clim., Yokohama, Japan, 2009: pp. 1–4. http://www.ide.titech.ac.jp/~icuc7/extended_abstracts/pdf/384633-1-090521000651-002.pdf.

[10] T. Toppi, P. Zangheri, R. Paolini, Simplified Method for Modification of Weather Data File for Energy Simulations within Urban Areas, in: R. Paolini (Ed.), 8th Work. Meet. CIB W108 - Clim. Chang. Built Environ., Maggioli editore, Milan, Italy, 2009: pp. 65–70. http://www.irbnet.de/daten/iconda/CIB16454.pdf.

[11] A.H. Rosenfeld, H. Akbari, J.J. Romm, M. Pomerantz, Cool communities: strategies for heat island mitigation and smog reduction, Energy Build. 28 (1998) 51–62. http://dx.doi.org/10.1016/S0378-7788(97)00063-7.

[12] R. Levinson, H. Akbari, S. Konopacki, S. Bretz, Inclusion of cool roofs in nonresidential Title 24 prescriptive requirements, Energy Policy. 33 (2005) 151–170. http://dx.doi.org/10.1016/S0301-4215(03)00206-4.

[13] A. Synnefa, M. Santamouris, H. Akbari, Estimating the effect of using cool coatings on energy loads and thermal comfort in residential buildings in various climatic conditions, Energy Build. 39 (2007) 1167–1174. http://dx.doi.org/10.1016/j.enbuild.2007.01.004.

[14] M. Zinzi, Cool materials and cool roofs: Potentialities in Mediterranean buildings, Adv. Build. Energy Res. 4 (2010) 201–266. http://dx.doi.org/10.3763/aber.2009.0407.

[15] M. Zinzi, S. Agnoli, Cool and green roofs. An energy and comfort comparison between passive cooling and mitigation urban heat island techniques for residential buildings in the Mediterranean region, Energy Build. 55 (2012) 66–76. http://dx.doi.org/10.1016/j.enbuild.2011.09.024.

[16] C. Romeo, M. Zinzi, Impact of a cool roof application on the energy and comfort performance in an existing non-residential building. A Sicilian case study, Energy Build. 67 (2013) 647–657. http://dx.doi.org/10.1016/j.enbuild.2011.07.023.

[17] A.L. Pisello, F. Cotana, The thermal effect of an innovative cool roof on residential buildings in Italy: Results from two years of continuous monitoring, Energy Build. 69 (2014) 154–164. http://dx.doi.org/10.1016/j.enbuild.2013.10.031.

[18] P.J. Rosado, D. Faulkner, D.P. Sullivan, R. Levinson, Measured temperature reductions and energy savings from a cool tile roof on a central California home, Energy Build. (2014). http://dx.doi.org/10.1016/j.enbuild.2014.04.024.

[19] P. Berdahl, H. Akbari, L.S. Rose, Aging of reflective roofs: soot deposition., Appl. Opt. 41 (2002) 2355–60. http://www.ncbi.nlm.nih.gov/pubmed/12003230.

[20] P. Berdahl, H. Akbari, R. Levinson, W. a. Miller, Weathering of roofing materials – An overview, Constr. Build. Mater. 22 (2008) 423–433. http://dx.doi.org/10.1016/j.conbuildmat.2006.10.015.

[21] M. Ichinose, T. Inoue, Y. Sakamoto, Long-term performance of high-reflectivity exterior panels, Build. Environ. 44 (2009) 1601–1608. http://dx.doi.org/10.1016/j.buildenv.2008.10.003.

[22] M. Sleiman, G. Ban-Weiss, H.E. Gilbert, D. François, P. Berdahl, T.W. Kirchstetter, et al., Soiling of building envelope surfaces and its effect on solar reflectance—Part I: Analysis of roofing product databases, Sol. Energy Mater. Sol. Cells. 95 (2011) 3385–3399. http://dx.doi.org/10.1016/j.solmat.2011.08.002.

[23] M.V. Diamanti, R. Paolini, M. Zinzi, M. Ormellese, M. Fiori, M.P. Pedeferri, Self-cleaning ability and cooling effect of TiO2-containing mortars, in: NSTI Nanotechnol.


25/26

Conf. Expo, NSTI-Nanotech, Washington, DC, USA, 2013: pp. 716–719. http://www.nsti.org/procs/Nanotech2013v3/10/T9.232.

[24] G.M. Revel, M. Martarelli, M.Á. Bengochea, A. Gozalbo, M.J. Orts, A. Gaki, et al., Nanobased coatings with improved NIR reflecting properties for building envelope materials: Development and natural aging effect measurement, Cem. Concr. Compos. 36 (2013) 128–135. http://dx.doi.org/10.1016/j.cemconcomp.2012.10.002.

[25] M. Sleiman, T.W. Kirchstetter, P. Berdahl, H.E. Gilbert, S. Quelen, L. Marlot, et al., Soiling of building envelope surfaces and its effect on solar reflectance – Part II: Development of an accelerated aging method for roofing materials, Sol. Energy Mater. Sol. Cells. 122 (2014) 271–281. http://dx.doi.org/10.1016/j.solmat.2013.11.028.

[26] A.G. Mainini, T. Poli, R. Paolini, M. Zinzi, L. Vercesi, Transparent Multilayer ETFE Panels for Building Envelope: Thermal Transmittance Evaluation and Assessment of Optical and Solar Performance Decay due to Soiling, Energy Procedia. 48 (2014) 1302–1310. http://dx.doi.org/10.1016/j.egypro.2014.02.147.

[27] N. Moise, M. Rogers, J. Rush, P. Shah, Annual Air Quality Monitoring Report 2010, Cleveland, OH, USA, 2010. http://www.clevelandhealth.org/assets/documents/health/communicable_disease/air_quality_report_2010_final_draft.pdf.

[28] S.E. Bretz, H. Akbari, Long-term performance of high-albedo roof coatings, Energy Build. 25 (1997) 159–167. http://dx.doi.org/10.1016/S0378-7788(96)01005-5.

[29] H. Akbari, A.A. Berhe, R. Levinson, S. Graveline, K. Foley, A.H. Delgado, et al., Aging and weathering of cool roofing membranes, Berkeley, CA, USA, 2005. http://escholarship.org/uc/item/3qb8j3k7.

[30] R. Levinson, P. Berdahl, a Asefawberhe, H. Akbari, Effects of soiling and cleaning on the reflectance and solar heat gain of a light-colored roofing membrane, Atmos. Environ. 39 (2005) 7807–7824. http://dx.doi.org/10.1016/j.atmosenv.2005.08.037.

[31] IGLAE, Codice di pratica delle impermeabilizzazioni (Practice code for waterproofing. In Italian), 2012. http://www.iglae.org.

[32] SIA, SIA 271. Waterproofing for buildings, (2007).

[33] ISO, ISO 2810. Paints and varnishes – Natural weathering of coatings - Exposure and assessment, (2004).

[34] ASTM, ASTM E 903-12. Standard Test Method for Solar Absorptance, Reflectance, and Transmittance of Materials Using Integrating Spheres, (2012). http://dx.doi.org/10.1520/E0903-12.

[35] R. Levinson, H. Akbari, P. Berdahl, Measuring solar reflectance-Part I: Defining a metric that accurately predicts solar heat gain, Sol. Energy. 84 (2010) 1717–1744. http://dx.doi.org/10.1016/j.solener.2010.04.018.

[36] ISO, ISO 9050 - Glass in building - Determination of light transmittance, solar direct transmittance, total solar energy transmittance, ultraviolet transmittance and related glazing factors, (2003).

[37] A. Holm, H.M. Kuenzel, K. Sedlbauer, The Hygrothermal Behaviour of Rooms : Combining Thermal Building Simulation and Hygrothermal Envelope Calculation, in: Eighth Int. Build. Perform. Simul. Assoc. Conf., Eindhoven, Netherlands, 2003: pp. 499–506. http://www.ibpsa.org/%5Cproceedings%5CBS2003%5CBS03_0499_506.pdf.


26/26

[38] F. Antretter, F. Sauer, T. Schöpfer, A. Holm, Validation of a hygrothermal whole building simulation software, in: 12th Conf. Int. Build. Perform. Simul. Assoc., Sydney, Australia, 2011: pp. 1694–1701. http://www.ibpsa.org/proceedings/bs2011/p_1554.pdf.

[39] H.M. Künzel, Simultaneous Heat and Moisture Transport in Building Components One- and two-dimensional calculation using simple parameters., Fraunhofer IRB Verlag, Stuttgart, Germany, 1995. http://www.hoki.ibp.fraunhofer.de/ibp/publikationen/dissertationen/hk_dissertation_e.pdf.

[40] R. Levinson, H. Akbari, L.M. Gartland, Impact of the temperature dependency of fiberglass insulation R-value on cooling energy use in buildings, in: ACEEE 1996 Summer Study Energy Effic. Build., 1996. http://www.osti.gov/scitech/biblio/451183.

[41] ISO, ISO 10456. Building materials and products - Hygrothermal properties - Tabulated design values and procedures for determining declared and design thermal values, (2007).

[42] CEN, EN 15251. Indoor environmental input parameters for design and assessment of energy performance of buildings addressing indoor air quality, thermal environment, lighting and acoustics, (2007).

[43] P. Berdahl, H. Akbari, R. Levinson, J. Jacobs, F. Klink, R. Everman, Three-year weathering tests on asphalt shingles: Solar reflectance, Sol. Energy Mater. Sol. Cells. 99 (2012) 277–281. http://dx.doi.org/10.1016/j.solmat.2011.12.010.

[44] ARPA Lazio, Air quality data of ARPA Lazio, Agenzia Regionale per la Protezione Ambientale del Lazio (Environmental Protection Agency of Lazio region, Italy), (2015). www.arpalazio.it.

[45] ARPA Lombardia, Air quality data of ARPA Lombardia, Agenzia Regionale per la Protezione Ambientale della Lombardia (Environmental Protection Agency of Lombardia region, Italy), (2015). http://www2.arpalombardia.it/.

[46] J.D. Lindberg, R.E. Douglass, D.M. Garvey, Carbon and the optical properties of atmospheric dust., Appl. Opt. 32 (1993) 6077–81. http://dx.doi.org/10.1364/AO.32.006077.