Potential for Free-Cooling by Ventilation

Solar Energy 80 (2006) 402–413 www.elsevier.com/locate/solener

Potential for free-cooling by ventilation

Cristian Ghiaus a,b,*,1, Francis Allard a

a Laboratoire d’Etude des Phenomenes de Transfert Appliques au Batiment (LEPTAB), Universite´ de La Rochelle, Av. M. Cre´peau, 17000 La Rochelle, France b Technical University of Civil Engineering, Bd. P. Protopopescu, Bucharest, Romania

Received 24 November 2004; received in revised form 31 May 2005; accepted 31 May 2005 Available online 29 September 2005

Abstract

Natural ventilation is one of the most effective techniques for cooling. Its potential for cooling may be assessed by using a method based on the indoor–outdoor temperature difference of the free-running building, the adaptive comfort criteria and the outdoor temperature. It is demonstrated that the free-running temperature may be used instead of the balance temperature in energy estimation methods. The indoor–outdoor temperature difference of the free-running building becomes a characteristic of the thermal behavior of the building which is decoupled from comfort range and outdoor temperature. A measure related to the energy saved and the applicability of free-cooling is given by the probabilistic distribution of the degree-hours as a function of the outdoor temperature and time. Weather data for this method are available in public domain from satellite investigation. The method can be applied when buildings similar to existing ones are con- structed in a new location, when existing buildings are retrofitted or when completely new buildings are designed. The method may be used to interpret the results of building simulation software or of the field measurements. 2005 Elsevier Ltd. All rights reserved.

Keywords: Natural ventilation; Energy; Building design; Free-running temperature

1. Introduction Augenbroe, 2002). Building simulation programs are based on first principles and require as inputs The most important decisions that affect the ther- the geometry of the building, the comfort criteria mal performance of buildings are taken in the initial and even the specification of the HVAC technology. stages of design (Holm, 1993; de Wilde et al., 2002). These prerequisites make simulation more adapted Design evaluation may be supported by the results for evaluations in the final stages of design rather of building simulation, by simplified guidelines or than a support for decisions when the building is by expert advice based on experience (deWit and sketched (Clarke, 1985; Shaviv, 1998; Al-Homoud, 2000; Hong et al., 2000; Olsen and Chen, 2003; * Corresponding author. Address: Laboratoire dEtude des Clarke et al., 2004). In the initial design, the archi- Phenomenes de Transfert Appliques au Batiment (LEPTAB), tects have many important issues on which the pro- Universite´ de La Rochelle, Av. M. Cre´peau, 17000 La Rochelle, ject is evaluated and investing more in thermal France. Tel.: +33 5 4645 7259; fax: +33 5 4645 8241. E-mail address: [email protected] (C. Ghiaus). design would make them less competitive. The ther- 1 ISES Member. mal analysis is done usually by engineers after the

Nomenclature

dhc degree-hour for cooling (K h) Qc energy needed for cooling (J) dhfr degree-hour for free-cooling (K h) Tb balance temperature (K) DHc frequency distribution of degree-hour for Tbin temperature interval for a bin (K) cooling Tcl lower limit of comfort temperature (K) DHfc frequency distribution of degree-hour for Tcu upper limit of comfort temperature (K) free-cooling Tdiff temperature difference between indoor DHmc frequency distribution of degree-hour for and outdoor in free-running (K) mechanical cooling Tfr free-running temperature (K) f relative density To outdoor temperature (K) Ktot total cooling loss coefficient of the building (W/K) Greek letters N total number of samples dc condition for cooling () Pdf probability distribution dfr condition for free-cooling () Pr probability density dmc condition for mechanical cooling () qc energy rate needed for cooling (W) qgain total heat gains from sun and internal sources (W) building is designed in order to estimate the heating the thermal mass (Zalba et al., 2003; Khudhair and cooling loads. Energy consumption can be esti- and Farid, 2004). Putting the problem in this way mated and checked if it is within accepted limits, allows us to assess the potential of architectural ele- complying thus with minimal requirements (Mihala- ments, e.g. sun-space for heating and cooling (Argi- kakou et al., 2002; CSTB, 2003). But these calcula- riou et al., 1994; Mihalakakou, 2002), green roofs tions have practically no influence on the building (Del Barrio, 1998) and pond roofs (Kishore, 1988; design (Ellis and Mathews, 2001). Since the thermal Erell and Etzion, 2000), or methods, such as passive calculations are not done in the initial stages of de- cooling (Belarbi and Allard, 2001). sign, it is essential that the architects are able to take The potential of free-cooling represents a mea- decisions that affect the thermal design. They can sure of the capability of ventilation to ensure indoor use simplified models (Ellis and Mathews, 2001), a comfort without using mechanical cooling systems. combination of simplified models and expert knowl- A way to measure this ability is by using a variant edge (Yezioro and Shaviv, 1996; Shaviv, 1999), or of the bin method in which the annual energy con- just rules of thumb based on acquired experience. sumption is evaluated for different temperature Although these results are not optimal, the heuristic intervals and time periods. The energy consumption rules indicate design solutions for building orienta- is estimated for several values of the outdoor tem- tion, geometry, and thermal mass (Shaviv et al., perature, called bins, and then the total is obtained 1996; Shaviv, 1999). by multiplying by the number of hours in the tem- Another approach would be to consider the perature interval (ASHRAE, 2001a,b). The bin building as a system that should have the potential method employs the balance temperature for calcu- to ensure indoor comfort. The indoor temperature lating the degree-hours in a bin. Balance tempera- is the result of the balance of energy fluxes and accu- ture concept supposes that the indoor temperature mulation. Then, from a thermal point of view, the is constant which implies the use of a heating or aim of the architectural design is to provide the po- air conditioning system. In buildings that use natural tential to implement the control of the energy fluxes ventilation, the indoor temperature varies making through the envelope and to allow energy storage in the concept of balance temperature unsuitable. the building structure. Arguably, the main gains are In order to estimate the relative influence of the solar and internal (Givoni, 1991; Littlefair, 1998), building design, of the comfort criteria and of the the main losses are by advection and conduction climate, it is important to have a method that (Florides et al., 2002), and the energy storage is in decouples these three characteristics. 404 C. Ghiaus, F. Allard / Solar Energy 80 (2006) 402–413

2. Thermal comfort zone Fig. 1 shows a comparison of the comfort zones for the European climatic conditions. The comfort Thermal comfort standards specify the range of range in real HVAC controlled buildings (Fountain conditions or comfort zones where 80% of sedentary et al., 1996) is compared to the ASHRAE comfort or slightly active people feel the environment ther- zone (ASHRAE, 2001a,b) and the standard for nat- mally acceptable. ASHRAE Standard 55 indicates ural ventilation (Brager and de Dear, 2000). The summer and winter comfort zones that correspond approximate parameters of the comfort zones are to typical clothing level of 0.5 and 0.9 clo, respec- given in Table 1. According to the comfort standard tively. ASHRAE comfort range is 3 C, with a sea- for natural ventilation, the indoor mean tempera- sonal shift of 3 C(ASHRAE, 2001a,b). Field ture in summer is considered to be 25 C, corre- studies showed that indoor temperature in fully sponding to the mean monthly temperature of HVAC controlled buildings is maintained in nar- 22 C(Fig. 2). This comfort standard may be used rower ranges, with the mean temperature of 23 C, for thermal adaptive control of buildings (McCart- standard deviation of 1–1.5 C and seasonal shift ney and Nicol, 2002). of 0.5–1 C(Fountain et al., 1996). But because people adapt themselves to the environment, ther- 3. Balance temperature and the free-running mal comfort in naturally ventilated buildings has temperature larger seasonal ranges than assumed by ISO 7730 and ASHRAE 55 Standards (de Dear et al., 1997; A common way to compare the energy consump- Brager and de Dear, 1998; Nicol and Humphreys, tion of buildings is by applying the bin method 2002). (ASHRAE, 2001a,b). The estimation of the degree

30 c d 28 C) ° 26 b

24 a

20 (a) A/C 18 (b) ASHRAE

Indoor comfort temperature ( 16 (c) 90% accept. in NV (d) 80% accept. in NV 14 0 51015 20 25 30 3540 Mean monthly temperature (°C)

Fig. 1. Comfort range for air conditioning and for natural ventilation: (a) air conditioning; (b) ASHRAE comfort range; (c) natural ventilation, 90% acceptability limits; (d) natural ventilation, 80% acceptability limits (Table 1).

Table 1 Comfort zones Mean Winter mean Summer mean Range Seasonal shift (C) (C) (C) (C) (C) (a) Air conditioned buildings 23.0 22.5 23.5 1.5 1.0 (b) ASHRAE comfort zone 23.5 22.1 24.9 3.5 2.7 (c) Standard for natural ventilation, 90% acceptability limits 23.9 19.5 25.0 max 28.3 5.0 8.8 (d) Standard for natural ventilation, 80% acceptability limits 23.9 19.5 25.0 max 28.3 7.0 8.9 C. Ghiaus, F. Allard / Solar Energy 80 (2006) 402–413 405

Fig. 2. Mean monthly temperature in Europe during January and July (data from IIASA, 2001).

X X hours in the bin method is done by using the con- Qc ¼ Ktotði; jÞ½T oði; jÞT bði; jÞdcDtðiÞ; cept of balance temperature. The balance tempera- i j ture for cooling, Tb, is the outdoor temperature ð6Þ for which the building having a speciﬁed indoor where summation indexes i and j refer to time inter- temperature, Tcu, is in thermal balance with the out- doors. For this temperature, the heat gains (solar, val and bins of the outdoor temperature, respec- internal) equal the heat losses (ASHRAE, 2001a,b): tively. If the time interval, Dt(i), is one hour, the factor qgain ¼ KtotðT b T cuÞ; ð1Þ dhc ½T oði; jÞT bði; jÞdcDtðiÞð7Þ where q —total heat gains [W]; K —total gain tot is termed degree-hour for cooling, dh . The expres- cooling loss coeﬃcient of the building [W/K]; c sion (7) has the disadvantage of using the concept T —upper limit of comfort temperature [K]; T — cu b of balance temperature which implies that the in- balance temperature [K]. door temperature is controlled at a constant value. The balance point temperature is then We can demonstrate that the degree-hours as q T ¼ T þ gain . ð2Þ used in the bin method can be expressed as a func- b cu K tot tion of the free-running temperature, Tfr (Ghiaus, The energy rate needed for cooling is 2003). The free-running temperature is the indoor temperature of the building when no HVAC system q ¼ KtotðT o T bÞ; if T o > T b; c ð3Þ is used; by ‘‘system’’ we mean heating, air condi- 6 0; if T o T b; tioning and ventilation for cooling. Practically, this condition means that the building is as tight as dur- where To is the outdoor temperature. The energy needed for cooling is ing the heating season. The airtightness may be Z changed by ventilation and it is used for cooling. tfin From the thermal balance Qc ¼ KtotðT o T bÞdc dt; ð4Þ tinit KtotðT fr T oÞqgain ¼ 0; ð8Þ where dc is the condition for cooling. it results the free-running temperature 1; if T o > T b; q d ¼ ð5Þ gain c 6 T fr ¼ T o þ . ð9Þ 0; if T o T b. Ktot

The total cooling loss coeﬃcient of the building is a By replacing Tb in Eq. (7) by the expression (2) and function of outdoor temperature and time. The dis- by using Eq. (9), we obtain an equivalent expression crete equivalent of the integral (4) is for degree-hour for cooling: 406 C. Ghiaus, F. Allard / Solar Energy 80 (2006) 402–413 dhc ¼ðT fr T clÞdc; ð10Þ running temperature, Tfr, may be higher or lower than the outdoor temperature, To; free cooling is where the condition for cooling is feasible when T > T (Fig. 3(a)). fr o By summing the degree-hours given by Eqs. (11), 1; if T fr > T cu; dc ¼ ð11Þ (13) and (15) in bins of outdoor temperature, we ob- 0; if not. tain the degree-hour distribution as a function of the The cooling load may be balanced by free-cool- outdoor temperature for cooling: ing or by mechanical cooling. If the outdoor tem- X DHcðT oÞDHcðjÞ¼ ½T frði; jÞT clði; jÞdc perature, To, is lower than the upper limit of the i comfort range, Tcu, then free-cooling is possible. The condition for free-cooling is ð16Þ for free cooling, 1; if T fr > T cu and T o < T cu; dfr ¼ ð12Þ X 0; if not, DHfcðT oÞDHfcðjÞ¼ ½T frði; jÞT cuði; jÞdfc i resulting the degree-hour for free-cooling ð17Þ dhfr ¼ðT fr T cuÞdfr. ð13Þ and for mechanical cooling, X If cooling is needed, i.e. Tfr > Tcu, but the outdoor DHmcðT oÞDHmcðjÞ¼ ½T frði; jÞT cuði; jÞdmc; temperature, To, is higher than the upper limit of i the comfort temperature, Tcu, then mechanical cooling is required. The condition for mechanical cooling ð18Þ is where Tfr(To), Tcl(To), and Tcu(To) represent the 1; if T fr > T cu and T o P T cu; free-running temperature and the lower and the dmc ¼ ð14Þ 0; if not, upper limits of the comfort temperature thatP correspond to the bin j centered around T ; ½ is o T fr resulting the degree-hour for mechanical cooling the sum for all the values in the bin centered around To. dhmc ¼ðT fr T cuÞdmc. ð15Þ The integral of degree-hour distributions for The conditions expressed by the Eqs. (11), (12) cooling: and (14) are shown in Fig. 3. The comfort range is X X delimited by the lower and the upper comfort limits, DHc ¼ ½T frði; jÞT clði; jÞdc ð19Þ Tcl and Tcu as shown in Fig. 1 and Table 1. The free- j i

Fig. 3. Ranges for heating, free-cooling and mechanical cooling when the free-running temperature is (a) higher and (b) lower than the outdoor temperature. C. Ghiaus, F. Allard / Solar Energy 80 (2006) 402–413 407 for free cooling the degree-hours for mechanical cooling, DH , and X X mc free-cooling, DH , respectively. At 18 h, about 20% DH ¼ ½T ði; jÞT ði; jÞd ð20Þ fc fc fr cu fc of the energy needed for cooling may be saved by j i using ventilation. and for mechanical cooling X X 4. Obtaining the degree-hour distribution from the DH ¼ ½T ði; jÞT ði; jÞd ð21Þ mc fr cu mc probability distribution j i are mathematically equivalent with the classical def- Degree-hours may be calculated when the time inition of degree-hours. variation of the outdoor temperature is known. Fig. 4 shows an example of the degree-hour dis- However, this approach has two disadvantages: tribution for 12 h and for 18 h obtained from build- the data are not easily available and, if accessible, ing simulation. The panels in the ﬁrst row show the they should be available for more years (typically comfort zone, the outdoor temperature and the free- 5–20) in order to be statistically signiﬁcant. An running temperature. The panels in the second row alternative to using time series is to use the prob- show the frequency distribution of the outdoor tem- ability distribution. The probability distribution is perature in bins of 1 C at 12 and 18 h for a year. obtained on measurements achieved during long The integral of this distribution is the sum of occur- periods of time, more than 5 years (IIASA, 2001). rences over a year, i.e. 365. The last row shows the degree-hours obtained with Eqs. (17) and (18).We 4.1. Number of days in a month with a given may notice that free cooling is not possible at 12 h temperature because the free-running temperature is lower than the outdoor temperature. The integrals of degree- A histogram shows the distribution of data val- hour distribution DHmc(To)andDHfc(To) represent ues. It bins the values of a variable in equally spaced

Time 12h Time 18h 40 To 40 To C]

° Tfr Tfr

Tc Tc 20 20 Tcl Tcl

Temperature [ Temperature 0 0 (a) 0 10 20 30 40 0 10 20 30 40

40 40

20 20 of occurrences]

[No

o 0 0 T 0 10 20 30 40 0 10 20 30 40 (b)

150 150

100 Free cooling DHmc(To) 100 DHfc(To) DHmc(To)

)[K h] )[K Mechanical cooling o

T ( 50 50

DH 0 0 0 10 20 30 40 0 10 20 30 40 (c) Outdoor Temperature [°C] Outdoor Temperature [°C]

Fig. 4. Degree hour distribution for cooling: (a) Free-running temperature, (b) outdoor temperature distribution (c) degree-hour for free cooling and mechanical cooling. 408 C. Ghiaus, F. Allard / Solar Energy 80 (2006) 402–413 containers and returns the number of elements in grated over an interval. For example, the probability each container. The number of values in an interval of the temperature to be between 9.5 and 10.5 Cis (bin) divided by the total number of values repre- the integral of the appropriate Pdf from 9.5 to sents the relative frequency of that variable for a 10.5 C(Fig. 6(b)). bin. The probability density is the limit of relative For a discrete distribution having bins of Tbin, density, when the number of values of the variable the probability of the temperature being between is inﬁnity (Fig. 5): T T /2 and T + T /2, T 2 {T ,T + T , bin bin min min bin f Tmin +2Tbin,...,Tmax}is Pr ¼ lim . ð22Þ N!1 N PrðT 2½T T bin=2; T þ T bin=2Þ ¼ T bin P df ðT Þ.

The probability density function, Pdf, has a diﬀerent ð23Þ meaning depending on whether the distribution is For a continuous distribution, the probability of the discrete or continuous (Fig. 6). For discrete distri- temperature being between T T /2 and T + butions, the probability density function is the prob- bin Tbin/2, T 2 R,is ability of observing a particular outcome. Let us Z consider, for example, that the temperature is mea- T þtbin=2 PrðT 2½T T =2; T þ T =2Þ ¼ P ðT ÞdT . sured in discrete values of 1 C. In the case of the bin bin df T tbin=2 discrete probability density function (Pdf) repre- ð24Þ sented in Fig. 6(a), the probability that the temperature is 10 C is given by the value of the Pdf at 10. Let us consider a random variable, T, that has N Unlike discrete distributions, the Pdf of a continu- values. The probable frequency of variable T in the ous distribution at a value is not the probability of bin [T Tbin/2, T + Tbin/2] is observing that value. For continuous distributions, f ðT 2½T T =2; T þ T =2Þ the probability of observing any particular value is bin bin ¼ N PrðT 2½T T =2; T þ T =2Þ. ð25Þ zero. To obtain probabilities, the Pdf must be inte- bin bin

0.2 0.2 0.2

0.15 0.15 0.15

0.1 0.1 0.1

0.05 0.05 0.05 Relative frequency, f/N frequency, Relative Relative frequency, f/N frequency, Relative 0 0 P(T) density, Probability 0 0 10 20 30 0 10 20 30 0 10 20 30 (a)Temperature, [oC] (b)Temperature, [oC] (c) Temperature, [oC]

Fig. 5. Relative distribution frequency and probability density for a random process with m = 15, s = 3. (a) Relative frequency, N = 31; (b) Relative frequency, N =20· 31; (c) probability density function. )

( (T)

df df

P P 0.1 0.1

0.05 0.05

Probability density, Probability 0 density, Probability 0 0 10 20 30 0 10 20 30 Temperature [°C] (b) Temperature [°C]

Fig. 6. Probability density functions (a) discrete and (b) continuous. C. Ghiaus, F. Allard / Solar Energy 80 (2006) 402–413 409

For a discrete distribution, Eq. (25) becomes T diff ¼ T fr T o. ð31Þ

Tdiff is a property of the building that is almost inde- f ðT 2½T T bin=2; T þ T bin=2Þ ¼ N T bin P df ðT Þ; pendent of the outdoor temperature. A concept sim- ð26Þ ilar to the temperature difference in free-running is and for a continuous distribution it becomes the temperature difference ratio used in assessing the potential of cooling by using the thermal mass f ðT 2½T T bin=2; T þ T bin=2Þ (La Roche and Milne, 2004). Z T þT bin=2 The free-running temperature can be expressed ¼ N P df ðT ÞdT . ð27Þ as the sum of two values that are independent: the T T bin=2 temperature difference in free-running and the outdoor temperature. By introducing the relation (31) If the random variable T represents the daily mean in Eqs. (28)–(30), we obtain the degree-hour distri- temperature in a month, then N is the total number bution for cooling: of days in a month, N 2 {28,30,31} and f(T 2 [T T /2, T + T /2]) gives the number of bin bin DHcðT oÞ¼N T bin P df ðT oÞðT diff þ T o T cuÞdc; days in a month that have the temperature ð32Þ T 2 [T Tbin/2, T + Tbin/2]. Having the probable number of days in a month with a given mean tem- for free-cooling, perature, we can calculate the number of degree- hours for that month. DHfcðT oÞ¼N T bin P df ðT oÞðT diff þ T o T cuÞdfc; ð33Þ 4.2. Degree hours for cooling and for mechanical cooling, The degree-hours are a function of outdoor tem- DHmcðT oÞ¼N T bin P df ðT oÞðT diff þ T o T cuÞdmc. perature. For a given outdoor temperature, To, the degree-hour distribution is obtained as the product ð34Þ of The degree-hour distributions expressed by Eqs. 1. the number of days having the temperature To, (32)–(34) depend on the characteristics of the build- number expressed by NÆTbin ÆPdf(To), ing, Tdiff, of the climate, To, and of the comfort cri- 2. and the temperature difference (Tfr Tcu), teria, Tcu. 3. when there is a need for cooling, condition given by Eq. (11): 5. Obtaining the temperature difference in free-running DHcðT oÞ¼N T bin P df ðT oÞðT fr T cuÞdc. ð28Þ Similarly, the degree hour distribution for free- The temperature difference in free-running may cooling is be an educated guess or it may be calculated by simulation or measured during building operation. Typ- DHfcðT oÞ¼N T bin P df ðT oÞðT fr T cuÞdfc; ical values of the daily mean temperature difference ð29Þ in free-running can be obtained from the experience acquired during the heating season, when Tdiff has and for mechanical cooling is values of 3–10 C. But since the building inertia has an important influence on the daily variation DHmcðT oÞ¼N T bin P df ðT oÞðT fr T cuÞdmc; of Tdiff, the use of daily mean can be misleading. ð30Þ URBVENT project created a database of daily vari- where the conditions for cooling, dc, free-cooling, ations of Tdiff based on results of building simula- dfc, and mechanical cooling, dmc, are given by Eqs. tion. Three types of buildings (Table 2) were (11), (12) and (14), respectively. The free-running placed in different conditions (Table 3). The free- temperature in Eqs. (28)–(30) depends on the build- running temperature of the three types of buildings ing characteristics and on the outdoor temperature. located in Rome, Munich and Moscow was obtained In order to obtain a characteristic of the building for the orientation and occupancy given in Table 3. that is independent of the climate, we define the Then, the mean of the temperature difference temperature difference in free-running as between the indoor and the outdoor temperature 410 C. Ghiaus, F. Allard / Solar Energy 80 (2006) 402–413

Table 2 Types of buildings Surf. (m2) Vol. (m3) Occupants no. Heat gains (W/m2) ACH Sensible Latent House 20 50 4 7 5 0.5 Glass façade office 300 750 30 7 5 0.5 Brick wall office 75 200 10 7 5 0.5

Table 3 the need for cooling is lower as compared to the case Conditions for URBVENT database of buildings in which the comfort standard for air conditioning Type of building Location Orientation Occupancy would be used. Let us take the example of a zone House North North Full time in an office building having the dimensions of 2 Glass facade office Center East Working time 10 m long · 7.5m wide · 2.5 m height (75 m of Brick wall office South South floor), occupied by 10 persons and internal sources West of 30 W/m2. The results of the degree-hours for Eur- ope are given in Fig. 8. The first row shows the results for the case of the comfort range found in in free-running regime was calculated for each of the real air conditioned buildings for which the comfort hours: 0, 6, 12, and 18 h and for each month. An range is indicated in Fig. 1(a). The second row example of the result for the ‘‘glass façade office’’ shows the results for the ASHRAE comfort range with south orientation located in Rome is given in as presented in Fig. 1(b). The third and the fourth Fig. 7. rows show the results for the comfort range found in buildings with natural ventilation, as indicated 6. Estimation of potential for energy savings in Fig. 1(c) and (d). Comparison of the rows of Fig. 8 reveals that the The energy savings for cooling may be estimated distribution is almost the same but the energy con- by comparing the integrals of degree-hour distribu- sumption in air conditioned buildings would be al- tion. The energy consumption for cooling depends most halved if ASHRAE comfort range were used. on the adopted standard for comfort. Since the The reduction would be even more important if the comfort range for natural ventilation is much larger, standard for natural ventilation were used. The pat-

Jan 7

Feb 6.5

Mar 6 C] ° Apr 5.5 May 5 Jun 4.5

Month Jul 4 Aug 3.5

Sep Temperature difference [ 3 Oct 2.5 Nov

Dec 2 0h 6h 12h 18h Hour

Fig. 7. Example of temperature diﬀerence of a free-running building. C. Ghiaus, F. Allard / Solar Energy 80 (2006) 402–413 411

Fig. 8. Energy consumption and savings for an oﬃce building as a function of comfort standards. Case of comfort range of fully HVAC buildings: (a) degree-hours for cooling; (b) percentage of free-cooling. Case of ASHRAE comfort range: (c) degree-hours for cooling; (d) percentage of free-cooling. Case of natural ventilation with 90% acceptance: (e) degree-hours for cooling; (f) percentage of free-cooling. Case of natural ventilation with 80% acceptance: (g) degree-hours for cooling; (h) percentage of free-cooling. tern of energy savings is also similar. For all comfort gions located northern Danube and Loire but the ranges, free-cooling may save more than 50% for re- need for cooling is much lower in these regions. 412 C. Ghiaus, F. Allard / Solar Energy 80 (2006) 402–413

7. Conclusions de Dear, R., Brager, G., Cooper, D., 1997. Developing an adaptive model of thermal comfort and preference, final The free-running temperature may be used in- report ASHRAE RP-884, American Society of Heating, Refrigerating and Air Conditioning Engineers, Inc., and stead of the base temperature to define the degree- Macquarie Research, Ltd. hours in the bin method. The two definitions are Del Barrio, E.P., 1998. Analysis of the green roofs cooling equivalent, but the use of the free-running tempera- potential in buildings. Energy and Buildings 27, 179–193. ture has the advantage that it can be applied when de Wilde, P., Augenbroe, G., van der Voorden, M., 2002. Design the indoor temperature varies, which is the case analysis integration: supporting the selection of energy saving building components. Building and Environment 37, 807–816. for non-air conditioned buildings in summer. deWit, S., Augenbroe, G., 2002. Analysis of uncertainty in The method proposed has the advantage that the building design evaluations and its implication. Energy and three important aspects that influence the results of Buildings 34, 951–958. the building simulation, i.e. building type, comfort Ellis, M.W., Mathews, E.H., 2001. A new simplified thermal range and local climate, are decoupled. The building design tool for architects. Building and Environment 36, 1009–1021. is characterized by the temperature difference in Erell, E., Etzion, Y., 2000. Radiative cooling of buildings with free-running; the comfort is characterized by the flat-plate solar collectors. Building and Environment 35, 297– temperature mean, range and seasonal shift; the 305. local climate is characterized by the time series of Florides, G.A., Tassou, S.A., Kalogirou, S.A., Wrobel, L.C., outdoor temperature or by the probabilistic distri- 2002. Review of solar and low energy cooling technologies for buildings. Renewable and Sustainable Energy Reviews 6, bution of outdoor temperature for different hours 557–572. of the day (0, 6, 12 and 18 h) of each month. Fountain, M., Brager, G., deDear, R., 1996. Expectations of Based on this method, we can obtain quick esti- indoor climate control. Energy and Buildings 24, 179–182. mations of energy need for cooling and of the po- Ghiaus, C., 2003. Free-running building temperature and HVAC tential of energy savings for cooling by using climatic suitability. Energy and Buildings 35, 405–411. Givoni, B., 1991. Characteristics, design implications, and ventilation. In Europe, ventilation has the potential applicability of passive solar heating systems for buildings. to be used for cooling especially in north. However, Solar Energy 47, 425–435. the need for cooling in these regions is also much Holm, D., 1993. Building thermal analyses: what the industry smaller. Free-cooling alone can cover only a small needs: the architects perspective. Building and Environment part of the need for cooling in southern Europe. 28, 405–407. Hong, T., Chou, S.K., Bong, T.Y., 2000. Building simulation: an overview of developments and information sources. Building References and Environment 35, 347–361. IIASA, 2001. LUC GIS Database. Laxenburg, Austria, www. Al-Homoud, M.S., 2000. Computer-aided building energy anal- iiasa.ac.at. ysis techniques. Building and Environment 36, 421–433. Khudhair, A.M., Farid, M.M., 2004. A review on energy Argiriou, A., Santamouris, M., Assimakopoulos, D.N., 1994. conservation in building applications with thermal storage Assessment of the radiative cooling potential of a collector by latent heat using phase change materials. Energy Conver- using hourly weather data. Energy 19, 879–888. sion and Management 45, 263–275. ASHRAE, 2001. Energy estimating and modeling methods. In: Kishore, V.V.N., 1988. Assessment of natural cooling potential ASHRAE Fundamentals. for buildings in different climatic conditions. Building and ASHRAE, 2001. Thermal comfort. In: ASHRAE Fundamentals. Environment 23, 215–223. Belarbi, R., Allard, F., 2001. Development of feasibility La Roche, P., Milne, M., 2004. Effects of window size and approaches for studying the behavior of passive cooling thermal mass on building comfort using an intelligent systems in buildings. Renewable Energy 22, 507–524. ventilation controller. Solar Energy 77, 421–434. Brager, G., de Dear, R., 1998. Thermal adaptation in the built Littlefair, P., 1998. Passive solar urban design: ensuring the environment: a literature review. Energy and Buildings 27, penetration of solar energy into the city. Renewable and 83–96. Sustainable Energy Reviews 2, 303–326. Brager, G., de Dear, R., 2000. A standard for natural ventilation. McCartney, K., Nicol, F., 2002. Developing an adaptive ASHRAE Journal, 21–28. control algorithm for Europe. Energy and Buildings 34, Clarke, J.A., 1985. Energy simulation in building design. Adam 623–635. Hilger, Bristol. Mihalakakou, G., 2002. On the use of sunspace for space Clarke, J.A., Conner, S., Fujii, G., Geros, V., Johannesson, G., heating/cooling in Europe. Renewable Energy 26, 415–429. Johnstone, C.M., Karatasou, S., Kim, J., Santamouris, M., Mihalakakou, G., Santamouris, M., Tsangrassoulis, A., 2002. On Strachan, P.A., 2004. The role of simulation in support of the energy consumption in residential buildings. Energy and internet-based energy services. Energy and Buildings 36, 837– Buildings 34, 727–736. 846. Nicol, F., Humphreys, M.A., 2002. Adaptive thermal comfort CSTB, 2003. RT 2000 Re´glementation thermique 2000, CSTB, and sustainable thermal standards for buildings. Energy and www.rt2000.net. Buildings 34, 563–572. C. Ghiaus, F. Allard / Solar Energy 80 (2006) 402–413 413

Olsen, E.L., Chen, Q.Y., 2003. Energy consumption and comfort architectural design (CAAD) systems for passive and low analysis for diﬀerent low-energy cooling systems in a mild energy architecture. Energy and Buildings 23, 257–269. climate. Energy and Buildings 35, 560–571. Yezioro, A., Shaviv, E., 1996. A knowledge based CAAD system Shaviv, E., 1998. Computer aided energy conscious building for determining thermal comfort design strategies. In: Pro- design. Renewable Energy 15, 343–348. ceedings of WREC. Shaviv, E., 1999. Design tools for bio-climatic and passive solar Zalba, B., Marin, J.M., Cabeza, L.F., Mehling, H., 2003. Review buildings. Solar Energy 67, 189–204. on thermal energy storage with phase change: materials, heat Shaviv, E., Yezioro, A., Capeluto, I.G., Peleg, U.J., Kalay, Y.E., transfer analysis and applications. Applied Thermal Engi- 1996. Simulation and knowledge based computer aided neering 23, 251–283.