Passive cooling for air-conditioning energy savings with new radiative low-cost coatings

Energy and Buildings 42 (2010) 945–954

Contents lists available at ScienceDirect

Energy and Buildings journal homepage: www.elsevier.com/locate/enbuild

Passive cooling for air-conditioning energy savings with new radiative low-cost coatings Marc Muselli * University of Corsica – UMR CNRS 6134, Vignola, Route des Sanguinaires, F-20000 Ajaccio, France

A R T I C L E I N F O

A B S T R A C T

Article history: Received 21 November 2009 Received in revised form 3 January 2010 Accepted 4 January 2010

Passive cooling is considered as an alternative technology to avoid unwanted heat gains, to reduce urban heat islands and to generate cooling potential for buildings (limiting air-conditioning energy). According to materials and surface treatments, the roof can represent to be a major heat gain source from opaque elements of the building fabric, heating up the outer surface and increasing heat flow by conduction. This paper presents low-cost new radiative materials (1 2 = /m2) allowing to limit heat gains during diurnal cycle for hot seasons. To evaluate the relevance of these new substrates, their reflective UV–VIS–IR behavior are studied and compared to classical roofed materials available in industrial and developing countries. A 48 m2 experimental roof having different surfaces (plate steel sheets, fiber cement, terra cotta tiles and corrugated sheets) allows to determine the temperature ratio d between uncoated and coated materials. Up to 34% surface temperature gains are obtained for white coated CS, 25% for FC and 18% for TCT and PSS. According to uncoated materials for a surface temperature T0 = 60 8C, simulations showed that the low-cost white opaque reflective roofs (50 m2) presented in this study would reduce cooling energy consumption by 26–49%. ß 2010 Elsevier B.V. All rights reserved.

Keywords: Passive cooling Air-conditioning energy savings Roof radiative coatings Thermal comfort

1. Introduction Interest on reducing emission of greenhouse gases, caused by fossil fuels to power the heating and cooling requirements of the buildings has stimulated the interest towards adoption of passive heating and cooling techniques for buildings already constructed. Architectures of ancient times have focused on various passive techniques to restrict heat flow to-and-from a building. However in modern times due to the availability of electrical/mechanical power to run active heating/cooling systems, focus on the use of these techniques had been forgotten. The energy crisis, along with the concern about the emission of greenhouse gases has again brought these techniques in the limelight. Economic benefits of the use of these techniques are also being seen, especially in third world countries facing energy crisis. Passive cooling can be defined as the removal/restriction of heat from/to the environment of building by using the natural processes of rejecting heat in the ambient atmosphere by convection, evaporation, and radiation or to the adjacent earth by conduction and convection. New buildings can be designed and oriented in such a way that windows, doors, indoor spaces etc. are located and oriented to

* Tel.: +33 4 95 52 41 41; fax: +33 4 95 52 41 42. E-mail address: [email protected]. 0378-7788/$ – see front matter ß 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.enbuild.2010.01.006

take maximum advantage of the local climate. For existing buildings, new relevant material can be studied to improve the thermal behavior of the building. In hot regions (for example developing countries) where inside building temperature can be high, decreasing inhabitants comfort, roofs are constructed with simple materials as fiber cement, galvanized plate steels, terra cotta tiles or corrugated steels. These buildings need cooling and with the power and economic crisis and high energy cost the only option remaining is the adoption of low-cost passive cooling techniques. The state-of-art of solar passive cooling techniques has been given in Givoni [1] presenting some of the known techniques for passive cooling, insulated roof and wall [2–4], roof pond [5,6], earth-air tunnel [7,8], sky-term cooling [9] and ventilation [10]. The discussion regarding such techniques has been widely addressed in a number of experimental and numerical studies [5,6,8,11,12]. Several studies have been carried out regarding the cooling potential of the application of reflective coatings on buildings. Bansal et al. [13] have studied the effect of external surface color on the thermal behavior of a building. White colored coatings performed better than aluminum-pigmented coatings. Although different types of coatings are characterized by a high solar reflectance, aluminum-pigmented coatings are less desirable because they tend to remain hotter due to their low infrared emittance. The differences in the thermal behavior even among coatings of the same type and color, are due mainly to the

946

M. Muselli / Energy and Buildings 42 (2010) 945–954

differences in their spectral reflectance which mainly affects their performance during the day and their emissivity that is the predominant factor affecting the thermal performance of the various coatings during the night. Roof albedo modifications by simple painted coatings (white [14–16] or silver colors [16]) allows to obtain air-conditioning energy savings such as in Arizona [16] where white roofed coatings (0.9 albedo, 0.98 emissivity compared to dark brown roofing) allow to reduce daily total air-conditioning energy and peak hourly demand for houses by 28 and 18% respectively. In Florida, Parker and Barkaszi [17] have examined the impact of whitened reflective roof coatings on air-conditioning energy use in a series of tests on occupied homes. Measured air conditioner electrical savings in the buildings averaged 19%, ranging from a low of 2% to a high of 43%. Utility peak coincident peak savings averaged 22% with a similar range of values. In California, Akbari et al. [18] have monitored 6 buildings to determine energy saving using reflective roofed coatings and present very interesting absolute results: on 3 potential sites (Sacramento, San Marcos and Reedley), the estimated savings in average air-conditioning energy use can reach 81 Wh/m2/day. Thus, according to climate zones, installing a cool roof can save about 4.5–7.4 kWh/m2/year of conditioned roof area and estimates of average peak demand savings for hours noon–5 p.m. range from 3.9 to 6.6 W/m2. Other works concern the conception of software simulating the thermal behavior of buildings using reflective coatings on its structure under different climates in Iran [19], Brazil [20], Australia [21], where the authors show efficient light reflective coatings technology allowing to decrease by up to 90% the airconditioning energy savings and reduce up to 30% the total heat gain compared with dark surfaces. The Heat Island Group [22] of the Lawrence Berkeley National Laboratory supports the development of cooler roofing and pavement materials as well as urban planting programs (see also [23,24]). The Heat Island Group also develops guideline standards to mitigate the heat island effect through regional and local building design codes. More recently, Santamouris et al. [25] focus on recent progress for passive cooling techniques like the cool reflective coatings to improve outdoor and indoor conditions of low-income households in warm areas of the planet, ground cooling using earth to air heat exchangers, and discusses the potential of new ventilation techniques and systems for improving indoor comfort and air quality. At last, Uemoto et al. [26] present thermal performances of cool colored (white) paints on surface temperature materials concluding that these new substrates enhance thermal comfort inside buildings.

In this paper, we present new low-cost painted coatings for passive cooling techniques with a special focus on air-conditioning energy savings in buildings. These new materials propose a double functionality: passive cooling for buildings reducing air-conditioning energy (summer heat gain and urban heat island) and during the night, dew collection as an alternative drinking water source for arid regions [27–29]. 2. Passive cooling Radiative cooling easily corresponds as the energy balance emitted and received from a surface. This surface receives the solar radiation for short wavelengths (UV, visible, Near-IR) and the atmospheric radiation of long wavelengths (Mid-IR and Far-IR). A variable fraction of the received radiation is absorbed while the other part is reflected (Fig. 1). A given body can emit a maximum of energy corresponding to a black surface with the same temperature (288.1 K dark grey, 303.1 K clear grey in Fig. 1). If the emitted energy is superior to the absorbed energy, passive radiative cooling appears. The global energy received on the ground (diurnal cycle) represents 1290 W m2 in standard atmosphere for a given temperature of 288.1 K. The first component is the solar radiation from UV (2.5%), visible (41.4%) and Near-IR (33.2%) of the whole radiation. For diurnal cycle with a 15 8C ambient temperature, the sky IR emission (Mid and Far components) represents 22.9% of the received energy (i.e. 295.6 W m2). For nocturnal periods, this radiation corresponds to 100% of the received energy. In Fig. 1, the essential deficit of emissivity is observed in the band untitled ‘‘sky window’’ corresponding to 10 mm. At ambient temperature Ta, the emitted spectrum of a black body has his maximum of energy in this band of wavelengths. Thus, a material can dissipate great quantities of energy by radiative transfer between the hot source (the radiative material at Ta) to the cold source (at temperature Tsky < Ta). At Ta = 15 8C, a black body at the same temperature can emit 388.3 W m2. This surface disposed faced to a clear sky will dissipate a power of about 93 W m2 by IR radiation, allowing to cool it without energy adding (for a temperature surface of 25 8C and Ta = 15 8C, the dissipated energy for the black body is evaluated at 150 W m2). This phenomenon corresponds to radiative passive cooling for diurnal cycle. 3. New radiative materials and experimental setup From previous condensing foil materials designed by Nilsson [30,31] (0.39 mm thick foil made with a small % of TiO2 and BaSO4 microspheres embedded in a matrix of low-density polyethylene

Fig. 1. Solar radiation (ASTM G173-03, fine line). IR spectrum received to the Earth’s surface (thick line). Black body emission at 288.1 K on the sky window (grey black surface). Black body emission at 303.1 K (grey clear surface). The received powers (spectrum integration) of each component are expressed in W m2.

M. Muselli / Energy and Buildings 42 (2010) 945–954

947

Table 1 Optical properties for the new films and coatings designed. Materials LDPE films

Thickness (mm) for optical tests

e (4–20 mm) black body 288 K e* (8–13 mm) black body 288 K Solar reflectance (0.280–2.0 mm) Visible reflectance (0.38–0.78 mm)

Painted coatings (2 layers)

Opaque white

Diffuse white

Uncolored

Opaque white

Uncolored

200

215

155

90

94

0.90 0.99 0.84

0.91 0.99 0.78

0.90 0.99 /

0.93 0.99 0.85

0.92 0.99 0.63

0.90

0.86

/

0.90

0.62

(LDPE)), new radiative passive materials have been formulated allowing to obtain a double functionality: passive cooling for diurnal cycle (4 to 5 8C in houses improving inhabitant comfort) and dew production as an alternative source of potable water in nocturnal cycle (up to 0.7 L/(m2 night)) [29]. New patented materials [32] improve the selective optical properties: - Maximum IR emissivity (near to black body) on the whole spectral Mid-IR band (3–50 mm) obtaining a sufficient radiator emission in the ‘‘sky window’’ (8–13 mm) for nocturnal cooling; - Maximum reflectance of solar spectrum (UV/VIS/Near-IR, between 0.2 and 4 mm) where diurnal passive cooling is suited. These new formulations have been obtained by incorporation of radiative mineral components largely distributed in industrial processes. These commercial charges are incorporated in polymer matrix for film substrate (LDPE foil, 200 mm) or in liquid paint basis for a better application on roofs (90–94 mm). To prevent aging by photodegradation in thermoplastic films, the radiative minerals are combined with industrial anti-UV addictives like ‘‘hindered Amine Stabilizers’’ 0.6% in mass of TINUVIN 783 (Ciba Specialty Chemicals) needing an indirect contact with food products. For painted formulations, a commercial polymer basis for outdoor uses (containing anti-UV) is necessary and sufficient. At last in LDPE foil, anti-fog addictives (like ATMER 7340, Ciba Specialty Chemicals, 7% in mass) are incorporated to obtain an hydrophilic property for water gravity collect during night. All incorporated mineral components have permissions for a food use verified by the FDA (USA Food and Drug Administration). A specific experimental protocol has been established:

radiative than the transmission is close to 0, specially in the ‘‘sky window’’. The relative improvements achieved by the designed materials can be estimated from the literature using comparative high reflective cool optical properties. The Heat Island Group (http:// heatisland.lbl.gov/) proposes a large cool roofing material database and especially for white coatings (35 samples). Considering comparable thicknesses, white coatings presented in the data base (76 and 127 mm) present solar reflectance in the range 0.6–0.68 (0.85 for this study, 90 mm thick) and IR emittance equal to 0.91 (0.93 in this study). On the same data base, 18 samples of pigmented white coatings are presented with relevant optical properties (solar reflectance = 0.79  0.057, IR emittance = 0.90  0.012) but the coatings are applied in thicknesses considerably greater than typical white

- The charge dispersion linked to mechanical properties of the thermoplastic foil has been studied at the ICMCB (Institut de Chimie de la Matie`re Condense´e de Bordeaux, France) with Castaing m-sensor; - LDPE films are pressed on a hot press TECNI HISPANA (55 bars, 150 8C, 400 mm  500 mm); - Total solar reflectance is obtained from diffuse spectral reflectance measures using an UV/VIS/Near-IR spectrometer Cary 5000, Hg lamp with a PtFe integration sphere (Ø110 mm); - Transmission IR spectrums have been recorded on a IR Fourier transform spectrometer BRUKER Equinox 55 on the range 7500– 370 cm1 (1.3–27 mm). Table 1 resumes optical properties in thermal IR and solar spectrums. Emittances e (4–20 mm) are obtained from transmission spectrum measures (specular and diffuse) taking into account the diffuse and specular reflection. The IR emissivity e* (8–13 mm) is only determined from the specular transmission on the wavelength range. Fig. 2 shows specular transmission spectrums for designed materials (charged LDPE films and coatings). Materials are much

Fig. 2. Specular transmission spectrums on the sky window for designed charged LDPE films and painted coatings.

M. Muselli / Energy and Buildings 42 (2010) 945–954

948

Fig. 3. Experimental 1 m2 table for UV–VIS–Near-IR and Mid-Far-IR radiations installed in Ajaccio (Corsica Island, France). Description of the CNR1 Kipp and Zonen radiometer. Ta and RH are measured from the laboratory meteorological station and Tc represents the temperature surface under test.

paints, ranging up to about 1 mm. Moreover, Synnefa et al. [33] compare 14 types of reflective coatings selected from the international market. Measured mean solar reflectance and IR emittance values not exceed 0.8 and 0.93 respectively. At last, Uemoto et al. [26] test, in very recent works, conventional and cool white paints presenting reflectance equal to 65.5 and 77.9% respectively (>85% for the white coatings designed in this paper). Thus, the main objective is to transform an inert and nonutilized surface (building roofs) in a heat dissipater to decrease solar radiation unwanted heat gain during the day and to produce drinking water from air atmospheric vapor during night. White opaque, diffusing and uncolored materials have been designed in order to respect the roof substrate initial color. The experimental site is located in the Ajaccio Gulf (Corsica island, France; latitude: 418550 N; longitude: 88480 E), 400 m from the sea at 70 m elevation. The wind regime is characterized by a nocturnal wind with a NE dominant direction (1.8 m/s average) but with two directions (N-W/S-W) for the diurnal dominant wind, characteristic of a Mediterranean island climate. To analyze the radiation balance (solar and IR), a net radiometer CNR1 (Kipp and Zonen) has been mounted on a horizontal 1 m2 table at 1 m from the ground (Fig. 3). The material under test is insulated from the ground by a 3 cm polystyrene foam plate to prevent convective and/or radiative heating. The distance CNR1 device/table is an adjustable parameter. In case of absence of particular value, the distance is taken equal to 30 cm after several test processes. The following physical parameters are continuously recorded: measured solar radiation (W m2) with both CM3 pyranometers (spectral range: 305–2800 nm) incoming from the surface under study CM3 and from the sky CM3+ (08 global solar radiation), measured Far-IR radiation (W m2) with both pyrgeometers (spectral range: 5000–50000 nm) CG3 from the surface and CG3+ from the sky, Tc temperature of the radiator (K), Pt-100 TCNR1 temperature of the radiometer body (K) allowing to correct the CG3 signal to compute the exactly Far-IR radiation generated by the object that is faced by the CG3 using: ECG3 ¼

V 4 þ s  TCNR1 C

(1)

where V (0–25 mV) and C (mV W1 m2) are the voltage signal and the calibration factor of the CG3 pyrgeometer and s the Stefan– Boltzmann constant. At last, a heater prevents dew formation on radiometer glass domes during nocturnal periods. When heater is on, the largest expected deviation between real sensor temperature and Pt-100 is low, inducing a worst case error for the CG3 inferior at 10 W m2. All measured sensors are connected to a Campbell Scientific data logger (CR10X model) recorded with a time-step of 1 min (local time, winter: UT + 1, summer: UT + 2).

The physical properties of the materials under study are determined using classical parameters built from the measured data described previously:

avis ¼

ECM3 ECM3þ

(2)

where avis is the visible albedo (dimensionless) with suppressed values when solar elevation <108 above the horizon (to prevent mask shading). The Net Solar Radiation (NSR, W m2) corresponds to the absorbed energy by the surface under test: NSR ¼ ECM3þ  ECM3 > 0

(3)

The same equation is built for the Net Far Infrared Radiation (NFIR, W m2) where ECM3þ and ECM3 are computed from Eq. (1) and represents the part that contributes to heating or cooling the studied surface: NFIR ¼ ECG3þ  ECG3 < 0

(4) +



Considering that CG3 and CG3 sensors respectively faced the sky and the surface assumed behave like blackbodies, effective ‘‘sky temperature’’ and ‘‘surface temperature’’ can be computed from the classical Stefan–Boltzmann law. At last, the Net Radiation (NR, W m2) can be calculated using the individual sensor measurements results: NR ¼ NSR þ NFIR

(5)

4. Results 4.1. Passive cooling simulations Before testing new materials under CNR1 infrared conditions, we want to compare specific roof materials largely distributed in developing countries on a simulation point of view. Fig. 4 shows the three radiative contributions (incident, reflected and emitted) included in the passive radiative cooling balance. Simulated values of IR dissipated energy from radiator (hot source) to the sky (cold source) are given for two roof materials (new galvanized and aluminum sheets, thickness 1 m) and for the patented new material described in Section 3 (white paint with high emissivity, 90 mm). Materials are considered with a solar contribution of 1 kW m2 and at a temperature 20 8C superior to the mean temperature observed in summer in Corsica (Mediterranean climate 27 8C). Note that the radiative power emitted by our new painted formulation increases by a factor of 4 in relation to aluminum sheet and 23 compared to galvanized sheet.

M. Muselli / Energy and Buildings 42 (2010) 945–954

949

Table 2 Maximum simulated temperatures on different roof materials using TMG parameter. Comparison between standard and new coatings defined in this paper. Description

Commercial typical materials ‘‘Generic black’’ asphalt shingle Brick red paint, 2 layers Brut cement, fiber cement Galvanized steel Commercial white paint, 0.203 mm Commercial white paint, 0.508 mm

Fig. 4. Passive air-conditioning phenomenon with 1 kW m2 solar incident radiation. Indicative values for IR radiation between the hot source (roof radiator) and cold source (sky) are described: galvanized plate steel, aluminum plate steel and white high emissivity coating developed in this work.

The radiative dissipation is better with the increase of the surface temperature. In practical architectural system, the objective is more to limit the heating of roof surfaces rather than to cool theses surfaces. In an urban context, the construction material heating and asphalt on roads induce an increase of the meteorological temperature measured on large areas in the range 1–3 8C. The research team ‘‘Heat Island Group’’ of the Lawrence Berkeley National Laboratory estimates that 5–10% of the Los Angeles electrical consumption in dry and hot season corresponds to active cooling to limit temperature in houses [34]. This team proposed a new parameter TMG (Temperature Maximum Gain) allowing to estimate material behavior exposed to the sun using optical properties (visible reflectance, IR emissivity) of the studied material: ð1  Rvis ÞI0 ¼ ðhc þ hr ÞTMG þ hr ðT s Þ

(6)

where Rvis is the reflectance of solar spectrum, I0 is the incident solar flux considered equal to 1000 W m2, hr = 6.1 ei (W m2 8C1), the radiative heat transfer coefficient; ei emissivity on the thermal spectrum for the studied element (material under study emat or sky es taken classically for Tsky = Ta  10 8C), hc = 12.4 W m2 8C1 convective heat transfer coefficient. The given approximated TMG (8C) parameter is given by [34]: TMG ¼

ð1  Rvis Þ1000  6:1ðes Þ 12:4 þ 6:1ðemat Þ

New materials (this work) White opaque Foil, 0.200 mm White diffusing Foil, 0.215 mm White opaque paint, 0.090 mm Uncolored paint, 0.094 mm

Reflectance solar spectrum 0 < Rvis < 1 (0.280–2.0 mm)

IR emissivity 0 < ei <1 (4–20 mm)

TMG (˚C)

0.05 0.16 0.25 0.61 0.80

0.91 0.91 0.90 0.04 0.91

52.6 46.5 41.6 30.4 10.6

0.85

0.91

8.1

0.84 0.78 0.85 0.63

0.9 0.91 0.93 0.92

8.3 12.0 8.0 20.2

properties for houses cooling, with the better TMG values corresponding to a lower heating effect on the roof. White opaque foil (TMG = 8.3 8C) and paint (TMG = 8 8C) present comparative performances in relation to commercial white paint (TMG = 8.1 8C) but with a lower thickness (90 and 200 mm for paint and foil against 508 mm for commercial product). The Solar Reflectance Index (SRI) is a measure of the roof’s ability to reject solar heat, as shown by a small temperature rise [34]. It is defined so that a standard black (reflectance 0.05, emittance 0.90) is 0% and a standard white (reflectance 0.80, emittance 0.90) is 100%. For example, the standard black has a temperature rise of 50 8C in full sun, and the standard white has a temperature rise of 8.1 8C. Once the maximum temperature rise of a given material has been computed, the SRI can be computed by interpolating between the values for white and black. Materials with the highest SRI values are the coolest choices for roofing. Due to the way SRI is defined, particularly hot materials can even take slightly negative values, and particularly cool materials can even exceed 100%. For a specific coating, SRI index is plotted versus visible albedo avis (%) and corresponds to

(7)

Table 2 compares TMG’s values for typical roof materials with our new formulations (foil and paints) corresponding to temperature rise: T max ¼ T a þ TMGð CÞ

(8)

and Ta = 15 8C, Tsky = 5 8C and es = 0.868. The solar reflectance values of new designed materials measured on the wavelength spectrum 0.28–2 mm represent relevant improvement (0.63 < Rvis < 0.85) compared to uncoated commercial roof material as galvanized steel or fiber cement plate with Rvis = 0.61 and 0.25, respectively. A pertinent behavior is noted on the IR spectrum (4–20 mm) with IR emittance in the range 0.90–0.93 corresponding to a low improvement in relation to traditional roof materials presented in Table 2. Thus, the new designed coatings can be applied on original existing structures to transform it in a radiator in order to limit heat gain into buildings. Moreover, the materials defined in this study present relevant

Fig. 5. Solar Reflectance Index (SRI, %) as defined by Berdahl and Bretz [34]. Performances of materials developed in this work (circled) on the basis of solar reflectivity (albedo) values.

950

M. Muselli / Energy and Buildings 42 (2010) 945–954

a temperature rise value. Fig. 5 shows the effectiveness of our materials contribution with both formulations (LDPE foils and paints) in comparison with traditional commercial roofed structures or coatings defined by the Heat Island Group (black asphalt shingle, red brick paint, concrete fibber cement, galvanized steel and conventional white coatings). New designed white opaque foil (200 mm) and paint (90 mm) respectively correspond to SRI = 105.6% and SRI = 106.7% in relation to white TiO2 (100%) and black (0%) references. The relevance of passive radiative cooling with adapted roof material or substrate in green house concept is demonstrated in Fig. 6 where we have plotted the integrated emitted power for blackbodies faced to sky in standard conditions (Ta = 15 8C, Tsky = 5 8C and relative humidity RH = 80%). For a material surface temperature Tc, multiple applications in real conditions are possible with radiative cooling: - for Tc = Ta  5 8C: dew water condensation (continuous line), - for Tc = Ta + 15 8C: passive cooling (dotted line). For this last case, in the wavelength band 5–9 mm, up to 0.1 kW m2 can be dissipated from the surface to the sky. 4.2. Emissivity In this section, the efficiency of new materials (LDPE foil, 0.2 mm thick, fabrication cost: 1 s m2) are compared to other classical materials to evaluate their relevance for passive radiative cooling on roofs: 1 m2 PtFe sheet (polytetrafluoroethylene, 1 mm, 78 s m2), black PE foil (polyethylene, 0.5 mm, 0.3 s m2),

Fig. 6. IR radiative powers dissipated by a black body exposed to a sky radiation with Ts = 5 8C versus wavelengths and its temperature Tc.

aluminum plate sheet (1 mm, 49 s m2). Experiments are elaborated on the 1 m2 radiometer table described in Section 3. Considering identical meteorological conditions (Tsky < Ta clear sky conditions verified by the SOLIS model [35], Fig. 7a–d plots the VIS–albedo avis and NSR, NFIR and NR balances for the 4 materials under tests for 1 day experiment (PtFe: June 01, 2003; black PE: May 29, 2003; aluminum: June 8, 2003; foil: June 11, 2003). The case of cloudy sky conditions will not be considered because

Fig. 7. UV–VIS–Near-IR albedo (a) and NSR (b), NFIR (c), NR (d) energy balance corresponding to 4 different materials: foil: thick black curve, PtFe: thick grey curve, PE: fine black curve and aluminum: fine grey curve (1 day record with 1 min time-step).

M. Muselli / Energy and Buildings 42 (2010) 945–954

Tsky ! Ta limiting radiative transfer energy between the surface and the sky. Diurnal avis measures present fluctuations taking into account the sun position in the sky and atmosphere perturbations (water vapor in the air for example). Concerning the VIS–albedo, avis(PtFe) > avis(Foil) > avis(Alum)  avis(PE). However, new designed materials present some advantages in relation to the fabrication cost (PtFe about 80 times more expensive than LDPE foil). Based on the identical parameters, same results are obtained with NSR balance with NSR(PtFe) < NSR(Foil)  NSR(Alum) NSR(PE) showing that for the CM3 spectral range, the LDPE foil behaves like about a white body (Fig. 7b). For NFIR balance, NFIR(PE) NFIR(Foil) < NFIR(PtFe) < NFIR(Alum). Thus, in the CG3 spectral range (5–50 mm), LDPE foil presents advantages with a selective behavior according to the wavelengths range (Fig. 7c). Lastly, Fig. 7d allows to conclude that NR(LDPE Foil)  NR (PtFe) < NR(Alum) NR(PE) concluding the great advantage of this material in relation to their fabrication cost. Performances of low-cost designed films are comparable to the PtFe reference. Thus, this discussion is not taking into account both other materials (Alum and PE). On the visible albedo parameter, both materials are comparable with avis(Foil)/avis(PtFe) = 0.908  0.028 i.e. corresponding to a 9% positive gap for PtFe. The comparison on the NSR energy balance shows the competiveness of LDPE foils with ECM3 ðPtFeÞ  ECM3 ðFoilÞ 54:98  20:197 W m2 . For the NR balance, the comparison shows very close behavior on the whole wavelength range (0.3–50 mm) with NR(Foil)  NR(PtFe) = 18.329  16.214 W m2. These results allow to conclude the great relevance of LDPE films compared to PtFe reference in spite of a thickness 5 times lower and a cost divided by 80. 4.3. Experimental instrumented roof (48 m2) In order to validate the designed substrates, an instrumented 48 m2 experimental roof (Fig. 8) composed by 5 different roofed materials has been implemented in Ajaccio (Corsica Island, France) corresponding to the most popular roof structures used in industrial and developing countries: galvanized plate steel sheets (PSS, 12.9 m2), red tiles tinted fiber cement (FC, 7.9 m2), terra cotta tiles (TCT, 11.4 m2), mono-Si photovoltaic modules (4.8 m2, not used in this study) and galvanized corrugated sheets (CS, 11 m2). The roof is 308 tilted from horizontal representing a classical roof inclination in the Mediterranean climate and oriented 2208 for limiting typical climatic risks as storms of wind (W-SW dominant direction). For thermal insulation, a 10 cm thick rock wool layer is

951

deposed under roof materials and on a wooden carpentry. Below this structure, a 1 m layer of air exchanges energy with it. Each roof material has been coated with its specific high emissivity paint defined in this work: - PSS: white opaque deposition, 1 layer, 90 mm, - FC: two surfaced treatments: 50% of the surface with white opaque paint (2.1 in Fig. 8, 1 layer 90 mm) and 50% of the surface with specific uncolored paint for this material (2.2 in Fig. 8, 1 layer, 94 mm), - TCT: same process like FC (3.1 and 3.2 in Fig. 8), - CS: same process like PSS (No. 5 in Fig. 8). The whole architecture allows to modify the configuration by adding other materials presenting different surface coatings. Thus, the passive cooling capacity of each roof material (PSS, FC, TCT and CS) has been tested in 3 situations using type K surfaced thermocouples connected to a 4 input data logger HANNA Instrument (the fourth input measures the ambient temperature Ta at the top of the roof, 2 m from the ground). In Fig. 9, a same material is compared without paint T0 (8C), with on the one hand, white opaque coat Twht (8C) and on the other hand uncolored paint Tunc (8C). The uncolored treatment has been performed for the nocturnal cycle allowing to obtain potable water from air by radiative cooling on porous material like TCT [27–29]. All daytime experiments are performed between July 6 and July 17, 2007 with a 1 min time-step for FC (1147 data), PSS (1337 data), CS (1777 values) and TCT (1425 values). Linear regression Ti = diT0 determines the surface heating or cooling effects (Table 3) using white or uncolored coatings (CS > FC > PSS  TCT). During the daytime period, the white colored coatings have the ability to reduce the surface temperature of the concrete material on which they were applied. The best temperature gradients are obtained for the white opaque coating with temperature decreases of 25% for FC and even 34% for CS. For both TCT and PSS, decreases are about 18%. For the uncolored coating, the gains only are negative for low thickness steel materials (PSS and CS respectively 5 and 9%) but unfavorable for TCT (+1%) and FC (+5%). Similar results have been observed in Synnefa et al. [33] where white coatings on tiles in the summer period (August) allows to obtain temperature gradient up to 15 K in comparison with uncoated materials. In Tucson (Arizona), 20–30 K temperature gradients are measured between white or silver with dark coatings [16]. According to Akbari et al. [36], for a steel concrete roof with 3 cm thick, the temperature of interior surface can reduce 20 8C when its exterior surface is painted white. On a corrugated roof in Australia [21], the relevance of white paint has been demonstrated with a decrease of 20 K during daytime period. In these conditions, the room temperature, relative to ambient, had been lowered by 1.3 K leading to a noticeable improvement in human comfort and reducing the desire for air-conditioning. A ‘‘temperature gain’’ parameter was constructed (Fig. 10, UT + 2):

DT i ¼ ðT i  T a Þ=ðT 0  T a Þ with T i ¼ T wht or T unc in  C

(9)

with

DTi 2 ] 1; +1 [! heating process with the coating i; DTi 2 ] 1;1 [! cooling process with the coating i; DTi  1 ! no thermal effects.

2

Fig. 8. 48 m experimental air-conditioning roof platform with 5 different materials. PSS: 1; FC: 2 (2.1 and 2.2 for white and uncolored coatings); TCT: 3 (3.1 and 3.2 for white and uncolored coatings); PV modules: 4 and CS: 5.

Both steel materials PSS and CS (due to lower thickness) allow air-conditioning process on roofs for uncolored coating DTunc < 1. For TCT and FC, this surface treatment leads to an increase of the support temperature corresponding to heating effects DTunc > 1.

M. Muselli / Energy and Buildings 42 (2010) 945–954

952

Fig. 9. Linear correlation between T0, Twht and Tunc during 1 day (1 min time-step). (a) TCT, (b) CS, (c) FC, (d) PSS.

For white opaque coating on all materials, DTwht < 1 allow cooling processes corresponding to reduce air-conditioning energy for buildings with the following classification: CS > PSS  FC  TCT. These results show interesting perspectives for developing countries where typical roof materials are composed by CS and/ or PSS. These countries presenting a market with a lower technological potential, will use a simple white opaque coating on roof to reduce the external surface temperature and so the heat gain in houses, leading to increase the human comfort. The amount of heat energy (W m2) that gets into the body of one building is determined by: Q ¼ U r  ðT roof  T in Þ

(10)

Table 3 d-Values for 4 materials with white opaque and uncolored coatings. The parenthesis values represent linear regression R2 parameters. Roof materials

FC

PSS

CS

TCT

d = Twht/T0 d = Tunc/T0

0.750 (0.901) 1.053 (0.985)

0.822 (0.959) 0.953 (0.961)

0.664 (0.861) 0.911 (0.976)

0.816 (0.957) 1.011 (0.979)

where Ur represents thermal transmission coefficient (W m2 K1) with Ur = 1/R (R thermal resistance of the given material expressed in m2 K W1) considering Tin = 292.15 K (temperature comfort). Considering Troof = Twht or T0, we have calculated the energy gain (Fig. 11) for each roof white coated material in relation to non-coated equivalent surface, thus:

DQ ¼ U r  ðT wht  T 0 Þ ¼ U r  T 0  ðdwht  1Þ

(11)

with dwht < 1 (Table 3), so DQ < 0. Considering a roof with an external surface resistance Rse = 0.04 m2 K W1, with 10 cm thick rock wool insulation (Rrw = 2.5 m2 K W1), i.e. Rtotal = 2.54 m2 K W1, for T0 = 40 8C and white coating on CS, DQ = 41.4 W m2 (UCS = 0.394 W m2 K1). On FC, TCT and PSS, energy savings respectively reach 30.7, 22.4 and 21.9 W m2 (UFC = 0.3916 W m2 K1, UTCT = 0.3893 W m2 K1, UPSS = 0.394 W m2 K1). Considering a reversible air-conditioning system (with COP = 2), for a 50 m2 roof surface, the white coating allows to avoid 1035 W of electric energy for the compressor during hot season (Fig. 11). For this eventual coated roof, considering COP = 2.5 and COP = 3,

M. Muselli / Energy and Buildings 42 (2010) 945–954

953

Fig. 10. Daily temperature gain DT for white opaque (black continuous line) and uncolored (grey dotted line) paints on roof structures. The grey area corresponds to heating process. (a) FC, July 6, 2007, (b) CS, July 7, 2007, (c) PSS, July 9, 2007, (d) TCT, July 13, 2007.

air-conditioning energy savings (electricity) are respectively 829 and 690 W. Considering T0 = 60 8C (temperature rise, Fig. 9) and COP = 2, white opaque coatings allow to save air-conditioning energy consumption by 26% (TCT) to 49% (CS). 5. Conclusion

Fig. 11. Heating flux (W m2) avoided in the building using white coatings designed in this paper. Not consumed electrical energy (W) by air conditioner corresponding to a 50 m2 simulated roof.

In this paper, the relevance of new radiative passive patented materials was estimated to reduce heat gain from solar direct radiation through buildings roofs available in industrial (terra cotta tiles, TCT) or developing countries (fiber cement FC, corrugated sheets CS, plate steel sheets PSS). Calculated from specular and diffuse transmission spectrums, these designed substrates present competitive solar reflectance values (Rvis  0.63 diffuse to 0.85 white opaque) and IR emittances (ei  0.92). These low-cost materials (1–2 s m2) present the advantage to be distributed in LDPE films (0.200 mm) or painted coatings (90 mm, 2 layers) for application on roofs directly. Using a Kipp and Zonen CNR1 radiometer, LDPE films have shown relevant Solar Reflectance Indexes (SRIs) and IR emissivity properties compared to referenced coatings tested by the National Berkeley Laboratory or typical reference substrates as PtFe presenting great optical properties in visible and IR spectrums. During daytime cycles, white opaque coatings allow

954

M. Muselli / Energy and Buildings 42 (2010) 945–954

to decrease appreciably the surface temperature of traditional roof materials (in the range 18–34% respectively for TCT/PSS and CS, 25% for FC). Simulations from a 48 m2 experimental roof prototype have shown that white painted materials allow to reduce (and so to save) air-conditioning electrical energy by 26 to 49% according to the roof composition cover. Acknowledgments This work is partially financially supported by the Collectivite´ Territoriale de Corse (CTC) and the ADEC agency (Agence de De´veloppement Economique de la Corse). We gratefully thank D. Beysens and O. Clus for their contributions in materials formulations and S. Gregorj for an English critical reading of the manuscript. References [1] B. Givoni, Performance applicability of passive and low-energy cooling systems, Energy and Building 17 (1991) 177–199. [2] H. Asan, Effect of wall’s insulation thickness and position on time lag and decrement factor, Renewable Energy 28 (2008) 299–305. [3] A. Kumar, S. Ashutosh, M.S. Sodha, Optimum distribution of insulation over various components of an air-conditioned building, Building and Environment 24 (2) (1989) 169–178. [4] A. Shariah, B. Tashtoush, A. Rousan, Cooling and heating loads in residential building in Jordan, Renewable Energy 26 (1997) 137–143. [5] M.S. Sodha, A.K. Khatry, M.A.S. Malik, Reduction of heat flux through a roof by water film, Solar Energy 20 (1978) 189–191. [6] M.S. Sodha, N.K. Bansal, P.K. Bansal, A. Kumar, M.A.S. Malik, Solar Passive Building Science and Design, Pergamon, New York, 1986. [7] G. Mihalakakou, M. Santamouris, D. Asimakopoloulos, I. Tselepidakai, Parametric prediction of the buried pipes cooling potential for passive cooling application, Solar Energy 55 (3) (1995) 167–173. [8] M.S. Sodha, S.N. Garg, R.L. Sharma, Solar passive building in Delhi: numerical simulation and validation, International Journal of Energy Research 13 (1989) 339–361. [9] C.R. Prasad, G.S. Dutta, S.R.C. Sathyanarayan, .V.K. Rao, Studies on sky-therm cooling, International Proceedings of the Indian Academy of Science C2 (3) (1979) 339–344. [10] I.F. Hamdy, M.A. Firky, Passive solar ventilation, Renewable Energy 14 (1/4) (1998) 381–386. [11] J.A. Duffie, W.A. Beckman, Solar Engineering of Thermal Process, John Wiley, New York, 1991. [12] A. Srivastava, J.K. Nayak, G.N. Tiwari, M.S. Sodha, Design and thermal performance of passive cooled building for semi-arid climate of India, Energy and Building 6 (1984) 3–13. [13] N.K. Bansal, S.N. Garg, S.S. Kothari, Effect of exterior surface colour on the thermal performance of buildings, Building and Environment 27 (1) (1992) 31–37. [14] P. Berdahl, Building energy efficiency and fire safety aspects of reflective coatings, Energy and Buildings 22 (3) (1995) 187–191. [15] H. Akbari, S. Bretz, D.M. Kurn, J. Hanford, Peak power and cooling energy savings of high-albedo roofs, Energy and Buildings 25 (2) (1997) 117–126.

[16] J.R. Simpson, E.G. McPherson, The effects of roof albedo modification on cooling loads of scale model residences in Tucson, Arizona, Energy and Buildings 25 (2) (1997) 127–137. [17] D.S. Parker, S.F. Barkaszi Jr., Roof solar reflectance and cooling energy use: field research results from Florida, Energy and Buildings 25 (2) (1997) 105–115. [18] H. Akbari, R. Levinson, L. Rainer, Monitoring the energy-use effects of cool roofs on California commercial buildings, Energy and Buildings 37 (10) (2005) 1007–1016. [19] M.S. Hatamipour, H. Mahiyar, M. Taheri, Evaluation of existing cooling systems for reducing cooling power consumption, Energy and Buildings 39 (1) (2007) 105–112. [20] J.T. Oliveira, A. Hagishima, J. Tanimoto, Estimation of passive cooling efficiency for environmental design in Brazil, Energy and Buildings 41 (8) (2009) 809– 813. [21] H. Suehrcke, E.L. Peterson, N. Selby, Effect of roof solar reflectance on the building heat gain in a hot climate, Energy and Buildings 40 (12) (2008) 2224–2235. [22] H. Akbari, R. Levinson, Evolution of cool roof standards in the United States, Advances in Building Energy Research 2 (2008) 1–32. [23] R. Tadeu, A. Prado, F. Lourenc¸o, Ferreira, Measurement of albedo and analysis of its influence the surface temperature of building roof materials, Energy and Buildings 37 (4) (2005) 295–300. [24] A.H. Rosenfeld, H. Akbari, S. Bretz, B.L. Fishman, D.M. Kurn, D. Sailor, H. Taha, Mitigation of urban heat islands: materials, utility programs, updates, Energy and Buildings 22 (3) (1995) 255–265. [25] M. Santamouris, K. Pavlou, A. Synnefa, K. Niachou, D. Kolokotsa, Recent progress on passive cooling techniques: Advanced technological developments to improve survivability levels in low-income households, Energy and Buildings 39 (7) (2007) 859–866. [26] K.L. Uemoto, N.M.N. Sato, V.M. John, Estimating thermal performance of cool colored paints, Energy and Buildings 42 (1) (2010) 17–22. [27] O. Clus, J. Ouazzani, M. Muselli, V. Nikolayev, G. Sharan, D. Beysens, Comparison of various radiation-cooled dew condensers by computational fluid dynamic, Desalination 249 (2009) 707–712. , doi:10.1016/j.desal.2009.01.033. [28] D. Beysens, M. Muselli, I. Milimouk, C. Ohayon, S. Berkowicz, E. Soyeux, M. Mileta, P. Ortega, Application of passive radiative cooling for dew condensation, Energy 31 (13) (2006) 2303–2315. [29] M. Muselli, D. Beysens, I. Milimouk, A comparative study of two large radiative dew water condensers, Journal of Arid Environment 64 (2006) 54–76. [30] T. Nilsson, Optical scattering properties of pigmented foils for radiative cooling and water condensation: theory and experiment, Thesis, Department of Physics, University of Technology, Goteborg, Sweden, 1994. [31] T. Nilsson, Initial experiments on dew collection in Sweden and Tanzania, Solar Energy Materials and Solar Cells 40 (1996) 23–32. [32] O. Clus, Condenseurs radiatifs de la vapeur d’eau atmosphe´rique (rose´e) comme source alternative d’eau douce, Ph.D. Thesis, University of Corsica, 2007. [33] A. Synnefa, M. Santamouris, I. Livada, A study of the thermal performance of reflective coatings for the urban environment, Solar Energy 80 (8) (2006) 968– 981. [34] P. Berbahl, S.E. Bretz, Preliminary survey of the solar reflectance of cool roofing materials, Energy and Buildings 25 (1997) 149–158. [35] R.W. Mueller, K.F. Dagestad, P. Ineichen, M. Schroedter-Homscheidt, S. Cros, D. Dumortier, R. Kuhlemann, J.A. Olseth, G. Piernavieja, C. Reise, L.F D. Wald, et al., Heinemann. Rethinking satellite-based solar irradiance modelling: The SOLIS clear-sky module, Remote Sensing of Environment 91 (2004) 160–174. [36] H. Akbari, S. Konopacki, M. Pomerantz, Cooling energy savings potential of reflective roofs for residential and commercial buildings in the United States, Energy 24 (5) (1999) 391–407.