INDOOR AIR QUALITY AND THE ADEQUACY OF CLEANING IN 25 FINNISH SCHOOLS T Pasanen 1*, T Keskikuru 1, J Ridell 2 and P Pasanen 1 1 Department of Environmental Science, University of Kuopio, Finland 2 Jyväskylä Environment Office, Finland ABSTRACT Indoor air quality and the adequacy of cleaning were studied in 25 schools in a Finnish city. Methods included a questionnaire for teachers (Örebro MM-40-FIN) and measurements of ventilation rate, CO 2, temperature, particulate matter (PM) and surface dust in 56 classrooms. In most classrooms, ventilation rates did not fulfill the rate required by the Finnish building code. CO 2 -concentrations were lowest in classrooms with mechanical ventilation and highest in classrooms with natural ventilation. Ventilation rates correlated with PM in all particle size groups except the smallest one. Total PM-levels exceeded the recommended values, and almost all classrooms were too warm. No correlation was found between PM and surface dust. According to the questionnaire (n=424 teachers), main problems in the working environment were dust, stuffy or dry air and noise. There was a statistically significant correlation between perceived indoor air quality and symptoms. The prevalence of symptoms did not explain dissatisfaction with work. INDEX TERMS Dust, Schools, Perceived air quality, Ventilation rate, Cleaning quality INTRODUCTION The prevalence of indoor air problems, including moisture and mold problems, found in Finnish schools has increased in recent years. According to recent studies, the most common problems are draft in the winter, insufficient ventilation, and visible mold or the smell of mold. The consequences of low ventilation rates include elevated levels of carbon dioxide (CO 2 ) and other air contaminants in classrooms. Poor indoor air causes various symptoms, such as fatigue, headache, stuffy nose, and eye and skin problems. Efficient ventilation and adequate cleaning can reduce the level of contaminants. It is important to check the ventilation system regularly and to repair it when needed. If the ventilation system is not adjusted properly and it is not used correctly, actual ventilation rates can be significantly lower than the intended rates. As group sizes in classrooms have been increasing, the number of students in classrooms may be higher than what the ventilation system is designed for. The purpose of this study was to determine the indoor air quality and adequacy of cleaning in schools. This survey was needed because previous studies have indicated that the indoor air quality in schools is often poor. Over the past years, resources for school cleaning have been reduced, which has raised concerns about the effects of diminished cleaning on indoor air quality. The methods included a questionnaire for teachers and measurements of ventilation rate, CO 2, particulate matter (PM), amount of surface dust, and temperature in the classrooms. * Contact author email: tjpasane@email.unc.edu 87
The relationship between teachers symptoms, perceived air quality and measured factors was analyzed statistically. METHODS All schools that were located within the city boundaries were included in the study. Out of the 25 schools, 17 were elementary schools, six were junior high schools and three were high schools. Depending on the size of the school, one-to-three classrooms were randomly chosen for measurements. The measurements were done in typical classrooms, not in special classrooms, for example those used for teaching arts or sports. Sixteen schools were ventilated with mechanical supply and exhaust, seven were ventilated with mechanical exhaust and three had natural ventilation. The following factors were measured in the classrooms: ventilation rate, CO 2, PM, amount of surface dust, room temperature, temperature of supply air, and relative humidity (RH). Ventilation rates were measured according to the European Standard (EN 12599, 2000). CO 2, room temperature and RH were measured with a manual measuring instrument about 5-10 minutes before breaks. PM was measured at the same time with a laser particle counter (Climet Instruments CI-500), and the results were converted to mass concentration. The amount of surface dust was measured with an optical instrument (BM-Dustdetector), which is based on the transmission of a laser beam through a sampling tape. The proceedings of the measurements have been presented in more detail previously (Pasanen, 2001). Measurements were carried out in February-March 2001 except for ventilation rates, which were measured in March-April 2001. Perceived air quality and teachers symptoms during the past three months were investigated with a questionnaire (Örebro MM-40-FIN). The questionnaire was delivered in January 2001. The relationship between teachers symptoms, perceived air quality and measured indoor air factors was analyzed statistically. The objective was to find out the relationship between teachers symptoms and the perceived air quality. In addition, the relationship between symptoms and confounding factors, such as gender, age, length of employment, smoking, and dissatisfaction with work was also investigated. Finally, the relationship between different indoor air factors was studied. To study the prevalence of symptoms, they were categorized in the same way as in the Örebro-questionnaire. The first category included general symptoms, such as fatigue, headache, dizziness and difficulties concentrating, the second category included irritation symptoms related to the eyes and the upper respiratory tract, and the third category included skin symptoms. Teachers responses about the prevalence of symptoms were coded by giving the answer never, 0 points, yes, sometimes, 1 point and yes, every week, 2 points. Overall scores for each symptom group and for all symptoms together were calculated from these results. The normality of the overall scores was tested with Kolmogorov-Smirnov-test. The responses about perceived air quality were managed the same way. The difference between the overall scores of the symptom categories for 2-classed factors was tested with Mann-Whitney-test. Examples of 2-classed factors include gender, smoking status, and satisfaction with work. When the factors had more than two classes Kruskal- Wallis-test was used. The effect of different confounding factors on satisfaction with work, such as existing allergy, gender, age, smoking, length of employment, and the overall scores of symptoms, was tested using logistic regression. 88
The relationship between measured factors as well as the relationship between symptoms and perceived air quality was investigated by determining their correlation coefficient (Pearson s and Spearman s correlation coefficients). For example, the correlation between PM in different particle sizes and the amount of surface dust on different surface types was investigated. The relationship between ventilation rate and temperature, CO 2 -level, PM 10 and total PM was studied. In addition, the relationship between room temperature and the temperature of supply air was determined. RESULTS AND DISCUSSION Ventilation rates and CO 2 -concentration Almost without exception the ventilation rates failed to fulfill the rate of fresh air required by the Finnish building code (D2, 1987) for classrooms. The mean ventilation rate in the classrooms ventilated with mechanical exhaust was 1.4 l/s, person or 0.5 l/s, m 2 and the required rate (6 l/s, person) was not fulfilled in any of these classrooms. In classrooms ventilated with mechanical exhaust and supply the ventilation rates were higher, the mean ventilation rate being 3.3 l/s, person or 1.3 l/s, m 2. Low ventilations rates resulted in high CO 2 -concentrations. In classrooms ventilated with mechanical exhaust and supply the CO 2 - concentrations (mean 1190 ppm, n=40) were clearly lower than in classrooms that were ventilated with natural ventilation (mean 2020 ppm, n=4) or mechanical exhaust only (mean 1850 ppm, n=12). The limit concentration value of 1500 ppm for indoor air (Finnish building code) was exceeded only in 10% of classrooms ventilated with mechanical exhaust and supply, whereas the limit value was exceeded in two thirds of classrooms ventilated with mechanical exhaust. There was a strong negative correlation between CO 2 -levels and ventilation rates (r=-0.759, p=0.000). Temperature and relative humidity Almost all classrooms were too warm (mean 23,3 ºC) in relation to comfortable room temperature. Only in 10 classrooms the temperature was within the limits of S2-class for winter (20-22 ºC) of the Classification of Indoor Climate (Classification of Indoor Climate 2000). In statistical analysis no significant correlation between room temperature and ventilation rate was found. Relative humidity in the classrooms was low (mean 23.1 %), which is typical during heating season. Relative humidity was lowest in classrooms ventilated with mechanical exhaust and supply (mean 20.4 %), which resulted from higher ventilation rates in these classrooms. The mean temperature of supply air was 20.3 ºC (16.3-22.7 ºC). There are no requirements for the temperature of supply air, but supply air should not cause a draft or increase the temperature burden in the room. In many classrooms where room temperature was too high, the temperature of supply air was high as well. However, a statistically significant correlation between room temperature and the temperature of supply air was not found. Particulate matter and the amount of surface dust In classrooms ventilated with mechanical exhaust and supply, PM 10 was lower (mean 120 µg/m 3 ) than in classrooms ventilated with mechanical exhaust (mean 210 µg/m 3 ) or natural ventilation (mean 280 µg/m 3 ). The highest concentrations were measured in elementary schools or in classrooms where students moved around often during class hours. Respectively, the lowest results were obtained in junior high and high schools, where students 89
stayed most of the time at their desks. Except for a few cases, PM 10 did not vary significantly within schools. Statistical analysis showed no correlation for ventilation rate and the smallest particle fraction (0.3-0.5 µm). This was expected because fine particles do not settle on surfaces but instead drift in the air. Significant negative correlation was found between ventilation rates (l/s, person) and all other fractions of PM. There was a clear negative correlation between ventilation rates and PM 10 (r=-0.447, p=0.001) as well as ventilation rates and total PM (r=- 0.423, p=0.002). However, none of the fractions of PM correlated with the amount of dust on different surface types. The amount of surface dust was measured on three kinds of surfaces: on floors, on easily accessible surfaces (surfaces between floor and 180 cm) and hard-to-access surfaces (higher than 180 cm). The present results show that the floors of the schools were very clean. Except for three schools, all schools were in class 5, the cleanest class (less than 2.5 % of dust when measured with BM-Dusdetector) of INSTA 800 cleanliness classification (Nilsen et al., 2000). The remaining three schools were in class 4 (less than 5.0 % of dust). This was expected since the floors of the schools are wiped daily. The easily accessible surfaces varied more in their dust content than the floors. Eight schools were in class 5 (< 1.0 % of dust), three schools were in class 4 (<1.5 % of dust), five schools were in class 3 (<2.5 % of dust), seven schools in class 2 (<5.0 % of dust) and one school was in class 1 (>5.0 % of dust). The amount of dust on easily accessible surfaces varied considerably within the same school. Even though the floors of the schools are cleaned daily, easily accessible surfaces may have been cleaned on different days. Usually they are cleaned once a week, or when the surfaces seem dusty. High surfaces, such as the tops of cupboards are usually cleaned only once a year. This is seen in the results: most of the schools (76%) were in class 1 (>8 % of dust). In many classrooms these surfaces were covered by a thick, visible layer of dust. The effect of dust settled on high, hard-to-access surfaces on indoor air quality has not been studied sufficiently. The amount of dust on hard-to-access surfaces did not correlate with the amount of dust on floors or on easily accessible surfaces, but there was a weak correlation between the amount of dust on floors and on easily accessible surfaces (r=0.312, p=0.024). The amount of surface dust on any type of surface did not correlate with PM in any size range. The surface dust results produced with the reliable optical method are of special interest because they reveal severe problems in the quality of school cleaning. Symptoms According to the questionnaire (n=424, reply percentage 69%), most significant problems in the working environment were dust and dirt, stuffy or dry air and noise. Most common symptoms related to working environment were irritation of nose, hoarseness and dryness of throat, skin symptoms on hands and fatigue. In elementary schools and junior high schools, more than 20% of teachers reported these symptoms. Women had more symptoms than men in all symptom categories except irritation symptoms. Other background factors, such as age, length of employment, or smoking did not correlate significantly with the prevalence of symptoms. Dissatisfaction with work did not correlate with the prevalence of symptoms either. However, the difference between men and women still existed (i.e., among teachers who were dissatisfied with their work, women had more symptoms than men). Symptoms correlated moderately with perceived air quality. When using logistic regression to find the reasons for teachers being dissatisfied with work, the main factors turned out to be previous 90
or current asthma and the length of employment, although their overall significance was small. The prevalence of symptoms did not explain dissatisfaction with work. These findings indicate that other reasons, such as psychological factors, affect work dissatisfaction even more than environmental factors. Most of the teachers chose insufficient ventilation and/or crowded classrooms as the reasons for poor air quality. Many teachers commented that air quality depends on the number of students in the classroom and the time of the day, so that the air quality is worst in the afternoons. Teachers also mentioned that when the air quality is poor, children are more restless and tired. When asked about what should be done to improve air quality, most teachers chose increasing the ventilation, improving cleaning and adjusting the heating system. The symptoms of building users and perceived air quality have previously been studied in various working environments, but an extensive study has not been made in schools in Finland. It should be noticed that the participants of the questionnaire were teachers, and children can be more sensitive and thus more susceptible to adverse health effects. CONCLUSIONS The results of this study show that improvements in many areas of indoor climate of the schools are urgently needed. Ventilation rates and room temperatures were not adjusted according to the number of people in the room. The observed high PM-levels can be decreased by improving ventilation, cleaning the classrooms efficiently, and reducing dust sources indoors. Efficiency of cleaning, especially in areas that are typically cleaned only rarely, and the effect of excessive dust on these surfaces on indoor air quality should be assessed. This study revealed new information about teachers symptoms and the perceived air quality, as well as the relationships between different indoor air factors. In this study, there was a statistically significant correlation between symptoms and perceived air quality. ACKNOWLEGEMENTS The authors would like to thank all the people in the City of Jyväskylä who contributed to this project. We would like to thank especially the teachers and the students for their patience when the measurements were carried out. REFERENCES Classification of Indoor Climate 2000. Target Values, Design Guidance and Product Requirements. FiSIAQ publication 5 E. Espoo, Finland 2001. D2, 1987. Finnish building code for indoor air and ventilation requirements. Ministry of Environment. EN 12599, 2000. European standard. Ventilation for Buildings Test procedures and measuring methods for handing over installed ventilation and air conditioning systems. March 2000. Nilsen SK, Schjønning AL, Dahl I, et al. 2000. INSTA 800. New internordic standard for measuring cleaning quality. Proceedings of Buildings 2000, Vol 4, 375-378. Pasanen, T, 2000 [text in Finnish]. Sisäilman laatu ja pintapölyn määrä Jyväskylän kouluissa. Kuopion yliopiston ympäristötieteiden monistesarja 8/2001. 91