1、NA-04-2-5 Thermal Comfort in the Climatic Conditions of Southern Italy Ida Fato Francesco Martellotta, Ph.D. Cecilia Chiancarella ABSTRACT This paper presents the results of four thermal comfort surveys conducted in naturally ventilated and air-conditioned buildings located in Bari, in southern Ital
2、y, during winter and summer seasons. The buildings were of dierent types, includ- ing ofices, lecture rooms, and library reading rooms. The sample of subjects consisted mostly of students. A total of 20 rooms were analyzed by measuring indoor climatic parame- ters. During the measurements, the subje
3、cts located near the probes were asked top11 in a questionnaire to rate the thermal environment ut that moment. A total of 1840 valid question- naires were collected during the four suwqs. Clothing insu- lation levels were 0.45 CIO in summer and 0.90 clo in wintev. Metabolic rate was assumed equal t
4、o 1.2 met. Thermal neutral- ity, according to the ASHRAE seven-point scale, occurred at 24.4“C and 26.3“C in summer and at 20.7“C and 20.6“C in winter, respectively in naturally ventilated (Nv) and air-condi- tioned (AC) buildings. Preferred temperature, based on the Mclntyre preference scale, was c
5、ooler in summer (24.2“C and 25.6“C in NV and AC buildings, respectively) and warmer in winter (23.1“C and 21.2“C in NV and AC buildings, respec- tively). Thermal acceptability was investigated by means of all the available scales, showing that the indirect estimation of acceptability (based on the t
6、hree central categories of the ASHRAE scale) onlyprovidespartial information about occu- pants conditions. Finally, the whole set of data was analyzed in order to propose an adaptive algorithm for this climaticzone. The resulting equation fl, = 17.82 + 0.315 Td, where T, is the comfort temperature a
7、nd T, is the running mean outdoor air temperature, proved to be in good agreement with other studies. INTRODUCTION Many thermal comfort surveys have been carried out in different parts ofthe world, and a database including a selected sample of these is currently available on-line (De Dear 1998). The
8、 spread of field studies is mostly due to the development of laboratory-grade field instruments and to an enhanced set of protocols, which make comfort data obtained in the field as rigorous as their climate chamber counterparts. in addition, chamber research tended to simplifi comfort models, negle
9、ct- ing many factors (such as expectations and adaptation) that are now fully recognized as a part of the problem. Most of the research is aimed at investigating how subjects living in different climatic zones react to their thermal environment. Some research has been explicitly commis- sioned by AS
10、HRAE in order to have a set of state-of-the-art, fully comparable field experiments across different climatic zones, including: temperate (Schiller et al. 1988), hot-humid (de Dear and Fountain 1994), cold (Donnini et al. 1996), and hot-arid (Cena and de Dear 1999). However, most of the research is
11、represented by independent studies carried out in thermally critical climatic zones, such as tropical and subtrop- ical (Busch 1990; Chan et al. 1998; Kwok 1998; Nicol et al. 1999; Wong and Khoo 2002; Kwok and Chungyoon 2003). Another important subject of investigation is how well Fangers comfort eq
12、uation, on which both IS0 7730 (1994) and current revisions to ASHRAE Standard 55 (1992) are based, works in real environments. In fact, several papers have pointed out significant discrepancies between rational indices and actual comfort ratings in both naturally ventilated build- ings and air-cond
13、itioned ones (de Dear and Aulicems 1985; Humphreys and Nicol 2002). The origin of the discrepancies may be found in biases in all the contributing variables I. Fato is an associate professor, E Martellotta is an assistant professor, and C. Chiancarella is a Ph.D. student in the Dipartimento di Fisic
14、a Tecnica, Politecnico di Bari, Italy. 578 Q2004 ASHRAE. (Humphreys and Nicol 2002) or, as suggested by Fanger and Toftum (2002), in different “expectations” of the people. A further argument ofresearch is the observation that, due to its inaccuracy, the PMV model is no better at predicting the comf
15、ort vote than simpler equations based on regression anal- yses using temperature alone as the independent variable. According to Nicol and Humphreys, this effect could be the result of a feedback between the comfort of the subjects and their behavior so that they “adapted” to the climatic conditions
16、 in which the field study was conducted (Nicol and Humphreys 2002). Important evidence that supports this adaptive approach is the good performance of the outdoor temperature as a predictor of the comfort temperature. In fact, variables such as clothing, posture, use of building controls, and meta-
17、bolic rate all depend on outdoor temperature. The feedback between climate, and these actions explain why only the latter need be considered in real situations in real buildings. The present paper provides a comprehensive analysis of the results of four transverse field studies carried out in Bari (
18、in southern Italy) and partly presented elsewhere (Conte and Fato 2000; Fato and Chiancarella 2003). The collected data have been used to investigate the effects of indoor climate on thermal perception, the relation between different scales, and the possibility of developing an adaptive algorithm to
19、 predict comfort temperature. THE THERMAL COMFORT SURVEYS Climatic Environment The thermal comfort surveys were carried out in Bari, in Apulia, in southern Italy. The city is located on the Adnatic Sea at a latitude of 41 O 8 N, in a large and homogeneous rain- fall region including the entire easte
20、rn Mediterranean (Tuni- sia, Sicily, Greece, Asia Minor, and the Black Sea region, the Levantine and Egyptian coasts as far as Cyrenaica) character- ized by rainfall from September to March (Littmann 2000; Petrarca et al. 1999). The rainfall is much lower than that observed in the northern part of I
21、taly (more influenced by the continental climate) and is mostly due to air mass movements caused by southern winds. During the spring, there are frequent day-to-day thermal variations due to the different origins of the air mass (from the Balkan peninsula and north- ern Europe or from the African re
22、gion). During the summer, the rainfall is negligible and the temperatures are quite stable with peaks due to southern winds. During the winter, the weather is quite unstable, with frequent alternation of cloudy and rainy days with sunny, but equally cold, days. According to the meteorological data b
23、ased on ten-year statistics (Petrarca et al. 1999), the range of daily mean temperatures in the winter season varies from 8.7”C to 10.9”C, with the lowest temperature observed in January. In summer, the mean temperatures vary from 21.4”C to 24.3OC, with the highest temperature observed in August (wi
24、th maximum temperatures up to 36C). The relative humidity varies from 60% to 75% in winter and from 50% to 70% in summer. The thermal comfort surveys were carried out in summer 1995 (S95), winter 1996 (W96), summer 1999 (S99), and winter 2000 (WOO). Sample of Buildings Five different university buil
25、dings were investigated. Four of them were located within the same campus in the city center, and the other, a research center, was located in a subur- ban area. The first building (identified with the code “A”) was built in the sixties and, in agreement with the “modem” architec- tural trend, has a
26、 steel frame finished with concrete panels and single-glazed strip windows. The building is not air condi- tioned, but there is a centralized heating system. The windows are openable with a shading system. The building includes both office and public rooms, among which two libraries, a reading room,
27、 a computer room, and two lecture rooms were investigated. The second building (identified with the code “B”) is an addition to building A built in the late nineties. The frame is in reinforced concrete with brick walls and double-glazed strip windows. This new building is fully air conditioned with
28、 a centralized system and fan coil units to allow local control of the thermal environment. The windows are openable with movable curtains. Two reading rooms were surveyed in this building. The third building (identified with the code “C”) was built in the eighties and is a concrete structure with b
29、rick walls and single-glazed windows. There is no air conditioning, and only a centralized heating system. Four lecture rooms were inves- tigated. The fourth building (identified with the code “D”) was built in the sixties and is a concrete structure with brick walls and strip windows with movable c
30、urtains. The building is air conditioned with fan coil units. The building includes offices, classrooms, and a library, but only the latter was analyzed. The fifth building (identified with the code “E”) was built in the late nineties and has a reinforced concrete frame finished with a curtain-wall.
31、 The building is fully air conditioned with fan coil units. In this building, the survey was carried out in a library reading room. A summary of the different buildings taken into account during the four surveys on which this study is based is in Table 1, where the number of rooms per building and t
32、he total number of interviewed subjects are reported. Instruments and Physical Measurements The physical quantities were measured by means of an automatic data logger with four sensors complying with the IS0 7726 Standard (1998). The air temperature was measured using a platinum (Pt 100) thermistor
33、with an accuracy within 0.2”C and carefully shielded in order to prevent the influence of radiant sources. The globe temperature was measured using a globe thermometer with a 15 cm copper sphere painted black and a platinum thermistor. Its accuracy was within OSOC, and its response time was ten minu
34、tes (at 90%). Relative humidity ASHRAE Transactions: Symposia 579 Table 1. Summary of Field Surveys Building code Number of rooms Sample size s95 W96 s99 woo A C A C B B D E 4 3 6 1 2 2 1 1 248 175 972 62 250 63 44 26 was given by an aspirated psychrometer that measured w- bulb and wet-bulb temperat
35、ure using two platinum thermistors with a response time of ten minutes. Finally, the air velocity was measured with an omni-directional, fully temperature- compensated, hot wire anemometer. The data logger recorded the minimum, maximum, mean, and standard deviation of globe temperature, air temperat
36、ure, relative humidity, and air velocity, as well as calculated oper- ative temperature and mean radiant temperature. However, since the data logger calculated the operative temperature as the simple arithmetic mean of air temperature and mean radi- ant temperature, it was later recalculated accordi
37、ng to the IS0 7730 (1994) definition, taking into account air velocity varia- tions. Due to the nature of the rooms surveyed whereas in the hot season, cool becomes the desirable state. According to de Dear and Brager (1 998), this effect should be taken into account because “preferred temper- ature
38、” could be more appropriate than “neutral temperature” as a descriptor of subjective thermal conditions. Thermal Acceptability Thermal acceptability is a quite controversial aspect of thermal comfort because, among other things, it can be defined with reference to different scales. The traditional a
39、nd most commonly used method is based on the IS0 7730 (1994) definition of “satisfaction” and equates acceptability with the three central categories of a seven-point thermal sensation scale (corresponding to “slightly cool,” “neutral,” and “slightly warm”). In this way, the percentage of subjects r
40、ating the environment thermally unacceptable should be ideally comparable with the calculated PPD. A second possible defi- nition of acceptability is based on the McIntyre preference scale and assumes that only the subjects who want “no change” are satisfied with the thermal environment. A further a
41、cceptability rating may be obtained using a comfort scale. IS0 1055 1 Standard (1 995) suggests to assume as satisfied all ASHRAE Transactions: Symposia 585 Table 6. Cross-Tabulation of ASHRAE Scale vc. Mclntyre Preference Scale, Direct Acceptability Scale, and Comfort Scale; Naturally Ventilated Bu
42、ildings, Winter Season Wants cooler (%) No change (Yo) I Wants warmer (%) I Thermal Sensation Scale I -3 -2 -1 O +1 +2 +3 Yo ROW 0.0 3.2 2.2 3.0 35.8 51.2 87.5 10.4 0.0 0.0 7.4 67.8 42.0 39.5 12.5 34.8 100.0 96.8 90.4 29.2 22.2 9.3 0.0 54.8 Acceptable (%) Unacceptable (/O) Very uncomfortable (YO) Si
43、ightly uncomfortable (YO) I Comfortable (%) Uncomfortable (%) 10.8 33.3 49.1 59.1 74.1 72.1 37.5 54.6 89.2 66.7 50.9 40.9 25.9 27.9 62.5 45.4 62.2 10.7 0.6 0.3 0.0 0.0 0.0 3.5 29.7 61.3 6.5 0.0 0.6 11.6 0.0 9.2 8.1 25.8 85.5 11.2 50.6 46.5 87.5 43.9 0.0 2.2 7.4 88.5 48.8 41.9 12.5 43.4 I % column I
44、3.6 9.0 31.3 35.5 15.7 4.2 0.8 I 100.0 Total responses 37 93 324 367 162 43 8 1034 the subjects rating the thermal environment as “comfortable” (i.e., the top category of a four-point scale), even though it has been proposed (Brager et al. 1993) to extend the acceptability to “slightly uncomfortable
45、.” Finally, a direct assessment of the acceptability is possible when the questionnaire includes a personal acceptability question (Le., “Taking into account your personal preference, would you accept this environ- ment?”), so that all the subjects responding “yes” are consid- ered satisfied. The fo
46、ur acceptability ratings are expected to differ from each other because they are based on scales designed to judge different psychological aspects of the ther- mal comfort. In order to investigate these aspects, three differ- ent procedures were followed. First, the percentages of the responses to t
47、he different scales were cross-tabulated, then the values of the percentage of dissatisfied (PD) calculated using the four different methods were plotted as a function of the operative temperature, and finally the averages of the PDs were plotted as a function of binned PPD values. Table 6 reports t
48、he cross-tabulation ofthe responses to the different scales for naturally ventilated buildings during the winter season. Subjects that proved to be satisfied according to the TS scale were 82.5% of the whole sample. However, the comparison with the preference scale shows that more than 30% of the “n
49、eutral” subjects had a preference (mostly for warmer temperature), while the percentages of dissatisfied in the “slightly cool” and “slightly warm” group were, respec- tively, 92.6% and 58%. For the “slightly cool” group, 90.4% of the preferences were for a warmer environment, while in the “slightly wan” group, 22.2% wanted warmer and the other 35.8% wanted cooler. The direct acceptability question shows that only 59% of the neutral subjects rated the thermal envi- ronment as acceptable, while higher percentages of accept- ability (up to 74%) were observed on the warm side of the sc
