1、VOLUME 13, NUMBER 2 HVAC accepted December 4, 2006Two independent field intervention experiments were carried out in mechanically ventilatedclassrooms receiving 100% outdoor air. Outdoor air supply rate and filter condition weremanipulated to modify indoor air quality, and the performance of schoolw
2、ork was measured.The conditions were established for one week at a time in a blind crossover design with repeatedmeasures on 10- to 12-year-old children in two classes. Seven exercises exemplifying differentaspects of schoolwork (numerical or language-based) were performed as part of normal lessonsb
3、y pupils who also marked visual analogue scales to indicate their environmental perceptionsand the intensity of any symptoms. The children indicated that the air was fresher but otherwiseperceived little difference when the outdoor air supply rate increased from 3.0 to 8.5 L/s(6.418 cfm) per person,
4、 while the speed at which they performed two numerical and twolanguage-based tasks improved significantly. A significant effect of ventilation rate wasobserved in 70% of all the statistical tests for an effect on work rate, but there were no signifi-cant effects on errors. The effects were probably
5、due to improved air quality in the classroomsas judged by a sensory panel of adults blind to conditions, as perceived by children, and as indi-cated by the reduction in the average CO2concentration from 1300 to 900 ppm, taking this as amarker of reduced bioeffluent concentration. It was not possible
6、 to test the effect of replacing asoiled filter with a new one because very little dust had been retained by the “used” filter andbecause of an incompletely balanced design. The unbalanced design also made it impossible totest for an interaction between filter condition and ventilation rate. These r
7、esults indicate theimportance of improving indoor air quality and ventilation in classrooms.INTRODUCTIONIt is almost incredible that no effects of poor indoor air quality on performance had been doc-umented until about ten years ago, when studies showed that the performance of office work byadults i
8、s negatively affected by mediocre indoor air quality (Wyon and Wargocki 2006). In theearly 1900s, the New York State Commission on Ventilation (NYSCV 1923) performed somequite thorough and realistic climate chamber experiments on poor ventilation in which CO2lev-els were allowed to rise to 4000 ppm,
9、 but they did not find any effects of bioeffluents on the per-formance of office work. Recent research has now provided good evidence that other almostuniversal sources of indoor air pollution (carpet, books, paper, and computers) do have a nega-tive effect on office performance (Wargocki et al. 199
10、9, 2002a; Bak-Bir et al. 2004) and thatPawel Wargocki and David P. Wyon are with the International Centre for Indoor Environment and Energy, Departmentof Mechanical Engineering, Technical University of Denmark, Lyngby, Denmark. Wargocki is also vice-president forresearch of the International Society
11、 for Indoor Air Quality. 2007, American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc. (www.ashrae.org). Published inHVAC Tham 2004),in which call completion time was reduced by up to 9% by poor air quality. The design of thesefield experiments was capable of eliminating the
12、effect of external factors, and their duration(nine weeks of full-time work for all the operators involved, each condition being established fora full week at a time) greatly exceeded that of the laboratory experiments (which involved onlyup to five-hour exposures to each condition). The finding tha
13、t mediocre air quality can reducecall center performance is partially supported by the results of a third field experiment, this oneperformed at much higher outdoor air supply rates (Federspiel et al. 2004). In view of these recent findings regarding the performance of office work by adults, it seem
14、slikely that the performance of schoolwork by children would also be affected by poor indoorair quality in classrooms. Very little is known on this issue, even though poor performance ofschoolwork can have lifelong consequences for a student and for society. A wide-ranging andauthoritative review of
15、 the research in these areas that is relevant to school classrooms (Men-dell and Heath 2005) concluded that there have been no satisfactory studies of how poor airquality in classrooms affects the performance of schoolwork by children. The authors were notpersuaded that a classroom study by Myhrvold
16、 et al. (1996) that found a weak associationbetween CO2levels and simple reaction time, with a correlation coefficient of 0.11 (suggestinga positive effect of increased ventilation on performance), had convincingly eliminated con-founding factors nor that simple reaction time was a useful predictor
17、of the performance ofschoolwork. Apart from this they cite only one other study showing an effect of any aspect ofair quality on children: Pilotto et al. (1997) showed in a cohort study that air pollutants fromgas heaters had a negative effect on attendance at school, which was presumed to be due to
18、 anegative effect on childrens health. The reviewers considered that Rosen and Richardson(1999) did not succeed in proving that the 35%55% reductions in absenteeism that theyrecorded during periods of ionization of the air in a day care center were clearly different fromthe 36% reduction observed in
19、 a period of no ionization, and they pointed out that an associa-tion between poor maintenance of schools and the poor academic achievement of the childrenattending them, as reported by Berner (1993), may not be causal (i.e., both may simply beexamples of the negative consequences of the multiple fa
20、ctors that make up an underachievingschool collection area). The remaining studies they reviewed concerned either hygrothermaleffects on children or indoor environmental effects on adults. The reviewers (in their Table 3)list over 50 studies indicating that one or more indoor environmental quality f
21、actors may havenegative effects on health, and 23 of these concern children. They infer from this that the samefactors may have negative effects on childrens performance of schoolwork but consider thatthis remains to be demonstrated.The current lack of knowledge concerning the effects of poor indoor
22、 air quality on school-work is surprising considering that environmental conditions in schools have been found inmany studies to be both inadequate and frequently much worse than in office buildings. Forexample, measurements in 39 schools in Sweden showed that 77% of schools did not meetbuilding cod
23、e regulations (Smedje and Norbck 2000). The most common defects in schoolsVOLUME 13, NUMBER 2, MARCH 2007 167include insufficient outside air supplied to occupied spaces, water leaks, inadequate exhaust air-flows, poor air distribution or balance, and poor maintenance of heating, ventilating, andair
24、-conditioning (HVAC) systems, as indicated by the analysis of 88 National Institute of Occu-pational Safety and Health (NIOSH) Health Hazard Evaluation Reports for educational facilitiesin the USA where formal complaints had been registered (Angell and Daisey 1997; Daisey et al.2003). Outdoor air su
25、pply rates per person in classrooms are often much lower than they shouldbe, according to current recommendations for classrooms (Daisey et al. 2003; Dijken et al.2005). For example, ANSI/ASHRAE Standard 62.1 (ASHRAE 2004) recommends for class-rooms 5 L/s (10 cfm) per person plus 0.60.9 L/s per m2fl
26、oor (0.120.18 cfm/ft2). This is aboutthe same as is recommended in the Danish Building Regulations (DHBA 1995), which addition-ally waives the requirement for mechanical ventilation if other measures to provide healthyindoor climate are in place (e.g., large rooms, provision of natural ventilation b
27、y means of open-able windows, etc.). Low ventilation rates often lead to carbon dioxide (CO2) levels being wellabove the recommended level of 8001000 ppm (Sowa 2002; Dijken et al. 2005; Boxem et al.2006), implying that the concentration of other pollutants, not only the bioeffluents from chil-dren f
28、or which CO2is a good indicator, will be high and that classroom air quality is conse-quently poor.It is not only the indoor air quality that differs between schools and offices. Classrooms aremore crowded, and they are occupied by children who are much more vulnerable than adults,whose bodies are s
29、till growing, and who in spite of poor classroom conditions must attendschool. Children have far fewer ways of registering complaints and cannot simply absent them-selves or find another school. The work that children are obliged to perform in school is notoptional and it is almost always new to the
30、m. Taking into account these factors, the effects ofpoor indoor air quality on the performance of schoolwork by children are likely to be moremarked than they are on the work that adults perform in their workplaces, which for the mostpart is routine and well-practiced and thus less affected by exter
31、nal factors. The objective of the present study was to determine whether classroom air quality affectsschoolwork and thus to extend knowledge of the effects of poor air quality on performance fromadults in offices to children in schools. Classroom air quality was modified by changing outdoorair supp
32、ly rates and by exchanging used supply air particle filters in the HVAC system with newones. Both interventions are well documented to affect indoor air quality (Seppnen et al. 1999;Wargocki et al. 2002b; Clausen 2004) and can be implemented so that the pupils and teachersremain blind to the interve
33、ntions. METHODSExperimental DesignThis study was designed as a series of field experiments in existing classrooms occupied bychildren performing their normal schoolwork, which is more natural for children than transport-ing them to a laboratory where they might behave abnormally, e.g., by exerting e
34、xtra effort toperform well. Three experiments in which the outdoor air supply rate to classrooms was manip-ulated were performed in the present series, all of them in the same school in Denmark, which issituated in the cool temperate area of Northern Europe. One of these experiments was a 2 2design
35、in which the classroom temperature was also manipulated; this is reported in anotherpaper (Wargocki and Wyon 2007). The two experiments reported here were both crossoverexperiments in pairs of classrooms in which two outdoor air supply rates were imposed in thesame week, one to each adjacent classro
36、om. The ventilation conditions were switched betweenthe classrooms the following week (crossover design). One experiment (Experiment 1V, per-formed in wintertime) was a 2 2 design in which each outdoor air supply rate was re-imposed168 HVAC the valves were reset toabout 21C (69.8F) at the end of eac
37、h week by the experimenters. No changes in the scheduleof normal school activities were made so as to maintain the teaching environment and routinesas normal as possible. The interventions were all improvements to existing conditions and wereapproved by parents, teachers, the School Board, the respo
38、nsible local authority, and the DanishEthics Review Board once this had been satisfactorily explained. Children were not asked fortheir consent so that they would remain unaware that they were taking part in an experiment inwhich air quality was being manipulated.SchoolThe school is in a small and a
39、ffluent suburban town on the east coast of the northern part ofthe island of Zealand, in Denmark, about 30 km north of Copenhagen. Like the majority ofschools in Denmark, it is a public school run by the local authority. It is an elementary school forchildren aged 6 to 16 years. At the time of the e
40、xperiments, the school had 470 pupils and anaverage class size of 23 children. In the last four school years for which records are available(20002004), the average grades obtained by school leavers have been better than the nationalaverage. The school buildings were constructed in the 1950s and are
41、architect-designed withlarge glazed areas. The main construction is of bricks and concrete, and all the buildings havesloping roofs. Mechanical ventilation was installed in 1997 in part of the school. The entireschool is a nonsmoking area.ClassroomsThe classrooms used in the experiments (Figure 1) c
42、onstituted a row of six identical wingsopening off the same straight north-south corridor, all with cathedral-height ceilings and largeglazed south-facing facades with openable windows. Each classroom consisted of an entrancehall with a floor area of 15 m2(161.5 ft2) and a ceiling height of 2.2 m (7
43、.2 ft), which was opento a classroom with a floor area of 50 m2(538.2 ft2) and a ceiling height of 33.65 mTable 1. The Partially Balanced Design of Experiments 1V and 2VWeekExperiment 1V, Winter Experiment 2V, SummerClassroom 1 Classroom 2 Classroom 1 Classroom 21Filter used Filter newVentilation ra
44、te high Ventilation rate low Ventilation rate high Ventilation rate low2Filter new Filter newVentilation rate high Ventilation rate low Ventilation rate low Ventilation rate high3Filter newVentilation rate low Ventilation rate high4Filter usedVentilation rate low Ventilation rate highVOLUME 13, NUMB
45、ER 2, MARCH 2007 169(9.812.0 ft) (the ceiling sloped upward toward the south faade). The total volume of eachclassroom was 187.5 m3(6621.5 ft3). The lower window area was 13 m2(139.9 ft2) and above2.2 m (7.2 ft) it was 7 m2(75.3 ft2). Five narrow windows in the lower area, each with an area of0.4 m2
46、(4.3 ft2), could be opened by the teacher. The classrooms were heated by water-filledradiators with thermostatic valves, which were located under the windows. There was no cool-ing. Supply air entered each classroom through four supply grilles with a total cross-sectionalarea of 0.072 m2(0.775 ft2),
47、 located in the wall under the upper set of windows, and left throughexhaust grilles close to the floor in the west wall adjacent to the corridor (Figure 1). The class-rooms have typical school furniture (wooden tables and plywood shelves) and floors coveredwith linoleum; outdoor clothing is left on
48、 hooks in the corridor, just outside the classrooms.VentilationThe classrooms were supplied with 100% outdoor air, filtered and preheated to a nominal20C (68F) from central air-handling units (AHUs) situated in the basement and controlled by acomputer. There was a cross-flow plate heat exchanger for
49、 heat recovery between separatestreams of supply and exhaust air. Each pair of classrooms in a given experiment was suppliedwith outdoor air from the same AHU. The supply air filters (F7 class bag filters) were either newat the outset of each experiment or had been installed in the system for 12 months. It was foundthat very little dust had been retained in the used filter. One reason for this could be that theschool is in a clean area in a suburban town with no industries, near the sea coast. Another rea-son could be the intermittent operation of the AHUit is run for
