1、4695 (RP-1161) Operable Windows, Personal Control, and Occupant Comfort Gail S. Brager Fellow ASHRAE Gwelen Paliaga Student Member ASHRAE Richard de Dear ABSTRACT Past research (ASHRAERP-884) demonstrated that occu- pants of naturally ventilated buildings are comfortable in a wider range of temperat
2、ures than occupants of buildings with centrally controlled HVAC systems. Howevel; the exact inju- ence ofpersonal control in explaining these differences could only be hypothesized because of the limits of the existingJeld study data that formed the basis of that research. The objective of ASHRAE RP
3、-1161 was to quantitatively investigate how personal control of operable windows in ojce settings inju- ences local thermal conditions and occupant comfort. We conductedafieldstudy in a naturally ventilatedbuilding where occupants had varying degrees of control over the windows. Utilizing continuous
4、 measurement of each subject S worksta- tion microclimate, plus a Web-based survey that subjects took several times a day and was cross-linked to concurrentphys- ical assessments of workstation microclimatic conditions, we collected over 1000 survey responses in each of the two main seasons. The dat
5、a show that occupants with different degrees of personal control had signijkantly diverse thermal responses, even when they experienced the same thermal envi- ronments and clothing and activity levels. Our findings offer further empirical support for the role of shifting expectations in the adaptive
6、 model of thermal comfort. INTRODUCTION Thermal environments in buildings with operable windows are typically more variable than conditions found in fully air-conditioned buildings, but research studies have demonstrated that they are not necessarily less comfortable. In particular, ASHRAE RP-884 (d
7、e Dear and Brager 1998) developed and analyzed a worldwide database from thermal comfort field experiments conducted in buildings that were either naturally ventilated (ie., occupant-controlled operable windows) or had centrally controlled HVAC systems (in which occupants had no control over their e
8、nvironment, simi- lar to the laboratory studies). One of their primary findings was that occupants in the naturally ventilated buildings accepted, and actually preferred, a significantly wider range of temperatures compared to occupants of the HVAC buildings. Furthermore, these comfortable indoor te
9、mperatures were noted to follow the seasonal shifts in outdoor climate and often fell beyond the ASHRAE Standard 55-1992 (ASHRAE 1992) comfort zones. These differences could not be entirely accounted for by conventional thermal comfort theory and the factors that affect a bodys heat balance (i.e., d
10、ry-bulb temper- ature, mean radiant temperature, air speed, humidity, clothing insulation, and metabolic rate). One of the hypotheses advanced for this anomaly was that naturally ventilated build- ings afford their occupants greater degrees of thermal control than air-conditioned buildings, and that
11、 this sense of control leads to a relaxation of expectations and greater tolerance of temperature excursions. Environmental psychologists have long known that human reaction to sensory stimulus is modi- fied when a person has control over that stimulus (Brager and de Dear 1998). A related explanatio
12、n is that people are more accepting of variations that come from a known source having predictable behavior (Bordass et al. 1994), which is often the case in a naturally ventilated building. A greater understanding of the influence of personal control has implications for building design, occupant c
13、omfort, and energy use. If people remain comfortable in a wider range of conditions in naturally ventilated buildings that provide personal control, significant energy can be saved by Gail S. Brager is a professor of architecture and Gwelen Paliaga is a graduate student research assistant at the Cen
14、ter for Environmental Design Research, University of California, Berkeley. Richard de Dear is an associate professor, Division of Environmental and Life Sciences, Macquarie University, Sydney, Australia. 02004 ASHRAE. 17 relaxing thermal comfort standards and allowing more vari- able indoor temperat
15、ures that cycle or drift in response to the natural swings of the outdoor and indoor climate (Milne 1995; Baker and Standevenl996). While the standards do provide some allowances for varying thermal conditions, the limits are fairly limited and are again based on laboratory studies in which subjects
16、 were given minimal or no control over the conditions they were experiencing. These laboratory studies may not necessarily be directly transferable to real buildings. When thinking about naturally ventilated buildings, prob- ably the most important architectural issue is the window. Windows can be u
17、sed for ventilative cooling of the building structure and, more importantly for this paper, the attainment of thermal comfort by moving air through the building. However, our understanding of the effect of air movement on occupant comfort in real buildings is limited. The draft limits in ASHRAE Stan
18、dard 55 are very low, and a literature review by Fountain and hens (1 994) explored a number of studies that indicate that personally controlled air movement is an underutilized cooling method in contemporary design. Specific knowledge about the influence of operable windows and the personal environ
19、mental control they afford on indoor thermal conditions and occupant comfort will give designers much needed information on how to design natu- rally ventilated buildings. ASHRAE RP-884 began this process by developing an adaptive model of comfort that was incorporated into the revised ASHRAE Standa
20、rd 55 (ASHRAE 2004) as an alternative compliance method for naturally ventilated buildings. However, the research was not able to disentangle the precise effect of personal control from all the other potential explanations for peoples acceptance of more variable thermal conditions for two important
21、reasons. First, the empirical basis of ASHRAEs adaptive model project, namely, data fiompast field studies, was mostly based on traditional, single-point-in-time thermal comfort measure- ments. As such, we dont know anything about the thermal conditions people had been exposed to prior to the condit
22、ions that were measured and assessed by questionnaire. Secondly, past field studies in naturally ventilated buildings did not typi- cally ask detailed questions about whether or not each of the subjects actually had the ability to personally control a window, nor how they used that control or percei
23、ved its ther- mal comfort effectiveness. Without that information, we cannot make a direct connection between the effects of personal control and thermal perceptions. Toward these ends, the objective of this project was to design and carry out a field study to quantitatively investigate how personal
24、 control of operable windows in ofice settings influences local thermal conditions and occupant thermal comfort, particularly the acceptability of thermal variability. Figure 1 Berkeley Civic Center, west facade. METHODS Description of Building Following an extensive search, we selected the Berkeley
25、 Civic Center (BCC), located in the San Francisco Bay area, as the building meeting the greatest number of our selection criteria. The five-story, 77,000 fi2, U-shaped building (shown in Figure I) houses city government and administrative offices. There are approximately 230 people working in the bu
26、ilding, It is predominately open plan (approximately 80% of the offices), two workstations deep from the perimeter, with a regular layout and access to windows. There are also private offices primarily in the comers of the building. This is a purely naturally ventilated building (i.e., no air-condit
27、ioning) with varied cooling strategies that include cross-ventilation through operable windows, stack ventilation through dedi- cated ventilation stacks, ceiling fans, and exposed thermal mass. The building perimeter has manually controlled radi- ators for heating but is mostly passive in its coolin
28、g mode except for centrally controlled inlet and outlet vents on the ventilation shafts and fan-driven stack assist when needed. The physical layout presented the opportunity to survey occu- pants with varying levels of direct or indirect personal control based on their proximity to the operable win
29、dows. Subjects on the perimeter (open plan and private offices) have direct access to at least three operable windows-two casements and one hopper. Subjects in the core zones are usually one desk away from the window, are most likely directly affected by it, but have less control over its operation.
30、 Most subjects in the core have easy access to windows but must interact with people on the perimeter to use the windows. 18 ASHRAE Transactions: Research Table 1. Field Experimental Methodology Physical Measurements Web-Based Survey Subject Pool Background Study Detailed Study Continuous ambient in
31、door conditions: tempera- Continuous desktop indoor conditions: dry-bulb ture and humidity in different zones within the temperature, globe temperature, air velocity, plus building nearby meteorological station Background Survey: one time, Repetitive Survey: several timedday, seasonal impressions, 1
32、 O- 15 min. long current impressions, 1-2 min. long Administer survey to entire building occupant Administer survey and monitor workstation poimiation microclimate for 38 subjects Duration and Frequency Experimental Methodology 2-week period in each of 2 seasons (warm subjects took the survey only o
33、nce, at their following the Background Study; subjects took convenience, during that period. 2-week period in each of 2 seasons, immediately the survey repetitively during that period, on average 2-3 times per day Our methodology included automation of physical (workstation microclimatic) and subjec
34、tive (questionnaire) data over different time scales and building occupant groups, plus data collected by objective observations of the worksta- tion and exterior facade characteristics of the building as they relate to our research questions. Table 1 shows the two complementary levels of investigat
35、ion. Additional documentation during the detailed study included faade photographs of operable window and blind position (warm season only) and sketches of workstations showing researchers observations about availability and use of adaptive mechanisms. All procedures were reviewed and approved by UC
36、 Berkeleys Committee for the Protection of Human Subjects (CPHS) to ensure that subjects knew their participation was voluntary and their identities and individual responses would be confidential, and that researchers would obtain informed consent and minimize risks and disruption to the participant
37、s. In addition, we consulted with CPHS and others to ensure that our procedures minimized any bias in our data. Physical Microclimatic Measurements A combination of commercially available equipment was assembled to allow close-to-laboratory-grade continuous monitoring of 38 subjects. We collected bo
38、th continuous desk- top and ambient measurements to allow us to sufficiently char- acterize the spatial and temporal variability experienced by the subjects and calculate PMV as a comparative index: Indoor physical measurements were made at each workstation with a custom designed and fabricated inst
39、rument, the Indoor Comfort Monitor (ICM), that was placed on each subjects desk, as near to them as possible without unnecessary expo- sure to heat sources. The ICMs are designed to collect contin- uous measurements of dry-bulb temperature, globe temperature, and air velocity (from which MRT could b
40、e derivedusing the equation in chapter 14 ofthe ASHRAE Hand- book-Fundamentals (ASHRAE 2001). The ICM is housed in a 2 in. x 6 in. x 8 in. electronic enclosure with three stainless steel tubes supporting the three sensors. The ICM “globe” thermometer (diameter 1.6 in.) is painted matte grey to becom
41、e a radiation absorber, or radiant temperature sensor, mimicking the human bodys emissivity of E = 0.95. In contrast, the dry-bulb temperature sensor is shielded from radiation by a highly reflective aluminized film, but it is subject to free-flowing ventilation through the large openings at the top
42、 and bottom. In the center of the ICM is the heated thermistor anemometer. The commercially available glass bead thermistor anemometers adopted in this study have a 0.01 m/s resolution, are factory calibrated to NIST standards, and do not have a strong directional bias compared to other low- cost an
43、emometers that fit within our research budget. After testing the anemometers in our wind tunnel, we determined that they would comply with ASRHAE Standard 55 if we aligned them within 30 degrees of the dominant direction of air movement (in our case, toward the window). As input for our analysis, wa
44、rm season air speeds were calculated by taking a three-point average of the instantaneous air speed (over 15 minutes). Cool season air speeds were a continuous average of the last five minutes (three-second sample period). A signal conditioner for the anemometer and a data logger are housed inside t
45、he ICMs box. After testing the thermistor sensors in a standardized calibration chamber, the ICMs were individually assembled, calibrated, and quality controlled in our laboratory before going out into the field. This device exceeds ASHRAE Standard 55 specifications for ambient temperature and radi-
46、 ant temperature accuracy, while approximating the guideline for air speed. More details about the design of our instrumen- tation can be found in the final report (Brager et al. 2004). Humidity was monitored with separate data loggers distributed one per cluster of subjects because of its rela- tiv
47、ely homogeneous distribution within the occupied zones of our subjects. Meteorological data were obtained from two local stations that meet standardized measurement guidelines and from our own outdoor temperature sensors and weather station on the roof. After analysis, the data from the UC Berkeley
48、Environmental Health and Safety (EH this results in the line starting at 24.5”C 76.l0F instead of 26C 78.8”F). In both cases we see that the air speeds people were experiencing were much less than the recommended air speed to offset the rise in temperature, which partly explains the finding that inc
49、reased numbers of people are wanting even more air movement than is being provided. Although 80% of the subjects in the warm season were voting in the “comfortable” categories of the ther- mal sensation, the 17% who were voting +2 and +3 were not obtaining high enough levels of air movement from the windows and/or ceiling fans to offset the warmer tempera- tures. But the desire for more air movement is not entirely explained by the need for cooling. Figure 5 shows that the average sensation at an operative temperature of 26C (783F) was less than +1 (slightly warm) in both s
