Assessing the relationship between outdoor air pollution and indoor air quality in naturally ventilated classrooms: A case study from Chile Christopher Cáceres-Araya1, Hector Altamirano-Medina1 and Clive Shrubsole1 1
UCL Institute for Environmental Design and Engineering, University College London
Corresponding email: [email protected]
, [email protected]
Keywords: IAQ, Naturally ventilated schools, Particulate matter (PM) SUMMARY This paper presents the findings of a research study on the influence of outdoor pollution on the indoor air quality (IAQ) in naturally ventilated classrooms. The study was carried out in a school located in Rancagua (Chile), a city where annual average concentrations of particulate matter (PM10 and PM2.5) have been found to exceed several times the maximum levels as recommended by the World Health Organization (WHO). Classrooms were monitored and occupants surveyed to gain an understanding of the conditions affecting IAQ. In addition simulations using CONTAM, a validated software package, were carried out to predict the movement of pollutants within classrooms. The simulations considered two scenarios of outdoor PM10 and PM2.5, five ventilation rates (ACH) with no indoor sources of particulate matter (PM) and/or re-suspension effects. Results show that indoor concentrations of PM represent at least 40% of the outdoor levels even when all doors and windows are closed and ventilation is provided purely by infiltration (0.5 ACH). Ventilations rates above 6 ACH resulted in high indoor concentrations of both PM10 and PM2.5 (67 µg/m3 and 45 µg/m3 respectively), which are similar to concentrations found in smokers’ apartments in Germany (Fromme et al. 2005). This study suggests that natural ventilation may not be a good strategy for educational buildings located in areas with high levels of pollution, since penetration of pollutants was found to be high even in buildings with low infiltration rates, leading to negative health impacts. 1
Naturally ventilated buildings have demonstrated their capability to provide healthy and comfortable environments, with a reduced environmental impact by using lower amounts of energy compared to buildings with HVAC systems (Allard, 1998). Natural ventilation techniques seem to be a suitable option for many types of buildings, especially for those located in temperate climates. However, the design of naturally ventilated buildings is complex and needs to consider factors that are not always controllable (e.g. local weather, air pollution and noise) and which may in turn affect the ventilation strategy hence the indoor air quality (IAQ) and occupants’ wellbeing (Mumovic & Santamouris 2009).
The challenge of providing good IAQ has been addressed through the development of standards and recommendations such as ASHRAE 62 ‘Ventilation for acceptable IAQ’ and CIBSE AM10 ‘Natural ventilation in Non-Domestic Buildings’. They provide guidance for estimating the minimum requirements of air exchange according to the activities performed by occupants in each type of building but also highlight the need of a thorough analysis of external environments. However, this last is not always considered when the aforementioned standards are applied. A good example is the published design guide for learning spaces developed by the Chilean government in 2000. The guide, used in the design of more than 2,000 new schools, encourages the practice of natural ventilation regardless the serious problems of air pollution existing in most of the locations where the schools have been built. Exposure to high levels of PM have been shown to be extremely harmful to the health and educational performance of children (Vela & Miller, 2013). This paper presents the findings of a study that looked at the IAQ of a naturally ventilated classroom located in Rancagua, which is one of the most polluted cities in Chile (MMA Chile, 2013). The study intended to quantify the influence of outdoor levels of PM10 and PM2.5 on classrooms’ IAQ as well as exploring the possible adverse effects on students. Furthermore, the study discusses whether the use of natural ventilation is a suitable strategy for areas with high levels of air pollution. 2
2.1 Study design Environmental parameters of two naturally ventilated classrooms were monitored for approximately a month in winter and during a period of time that coincides with the highest levels of air pollution recorded in the city where the studied school is located. Internal and external environmental parameters were recorded including indoor/outdoor dry bulb temperatures (ºC), relative humidity (RH,%), winds direction and speed (m/s), outdoor concentrations of PM10, PM2.5 and indoor levels of CO2. A survey was carried out to confirm the physical features of the classrooms. In addition, a standardised occupant survey was carried out among the students to measure their perception of IAQ during the monitoring period. The survey was designed by adapting questionnaires previously tested by CBES Berkeley (2014) and Engvall et al. (2003). Furthermore, a series of simulations using the software package CONTAM, a validated multi-zone airflow and contaminant transport model developed by The National Institute of Standards and Technology of the United States of America (NIST) (Shrubsole et al. 2012), allowed to simulate the influence of external pollution in the indoor depositions of PM10 and PM2.5 according to two scenarios of outdoor pollution and five ventilation rates. Resultant PM concentrations were compared with the results of previous studies, to estimate the impact of the outdoor pollution on classrooms IAQ and the possible effects on occupants. 2.2 Case study The school building studied is located in the urban area of Rancagua (Chile). The city has a warm-summer Mediterranean climate as defined by the Köppen climate classification. In the last years the average level of PM10 and PM2.5 have exceeded the levels recommended by the WHO (20 µg/m3 and 10 µg/m3 respectively) and the
Chilean norm (50 µg/m3 and 20 µg/m3 respectively). In the period 2009-2011 the monthly PM10 concentration was above 50 µg/m3 the whole year with a maximum average of 100 µg/m3 in June and July. The main cause is the massive use of biomass for heating in winter along with a high-pressure weather system that provides poor conditions for natural ventilation at urban scale. Moreover, it should be noted that the Chilean school term coincides with the period of highest level of pollution in Rancagua (Figure 1Figure 1).
Figure 1: Rancagua’s average outdoor concentrations of PM10 (2009-2011) The classrooms studied are located on the ground floor (GF) and first floor (1F) of a two-storey building (reinforced concrete construction without insulation and simple pane windows) which faces two courtyards located on the east and west side of the building. Both classrooms are naturally ventilated by cross ventilation and there are no heating or cooling systems (free-running). Each room have a floor area of 56 m2 (145 m3 volume) and capacity for 36 students and 1 teacher. Male and female students aged between 16 and 19 years old uses these classrooms every weekday from 07:50 to 14:20 hours. Figure 2Figure 2 provides an overview of the classrooms’ layout.
Figure 2: Classrooms’ floor plan and indoor view 2.3 Modelling methods A base model was built in order to reproduce the environmental conditions of the classroom that showed the worst IAQ during the period surveyed. Building envelope features and pollutants were modelled following published methods (Ozkaynak et al.
1996; Shrubsole et al. 2012; Jones et al. 2013). Models did not consider indoor generation of PM10 and PM2.5 or re-suspension of PM. In addition, models included a constant indoor CO2 generation of 30 mg/s per student (Cooper et al. 1987). This rate of CO2 generation was used to replicate the total levels of CO2 produced by 36 occupants using the classroom according to the typical pattern of occupancy during weekdays. The infiltration rate for the simulated classroom was calculated using a modified CO2 decay equation (1) proposed by Chatzidiakou et al. (2014) where A=ventilation rate (h–1); C(t), internal CO2 concentration at time t (ppm); C(ex), external CO2 concentration (ppm); and t= time (h). Thereby a minimum ventilation rate of 0.5 ACH was calculated for the base model. A dataset obtained from Meteonorm 7.0 (Meteotest 2014) was employed to run the simulations under transient weather conditions.
(1) Simulations were organised according to two scenarios of external pollution levels. Each group of simulations included the testing of the base model with five levels of ventilation rates; a minimum of 0.5 ACH (purely by infiltration) and a maximum of 15 ACH (Chilean standard recommendation). The first group of simulations (G1) considered 67 µg/m3 and 45 µg/m3 outdoor concentrations of PM10 and PM2.5 respectively, as these were the average outdoor concentrations registered during the monitoring stage. A constant indoor temperature of 13 ºC was set in the simulation, which corresponds to the average indoor temperature logged during occupied hours. The second group of simulations (G2) was run under a hypothetical pollution scenario, which complies with the acceptable outdoor concentrations of PM10 (20 µg/m3) and PM2.5 (10 µg/m3) stated by WHO (2005). In addition, the indoor temperature was set at a constant 21 ºC (estimated with Humphreys et al. 2010 method), which is more likely to produce thermal comfort considering Rancagua’s outdoor temperatures. Table 1 provides a summary of all simulations performed.
Table 1: Summary of simulations Outdoor Conditions Model 1 Model 2 Model 3 Model 4 Model 5 3 Avg. PM10 67 µg/m 0.5 ACH1 3 ACH 6 ACH2 9 ACH3 15 ACH4 Avg. PM2.5 45 µg/m3 Indoor Temp 13ºC (constant) Avg. outdoor CO2 600 ppm 36 students, 30 mg/s of CO2 generated per student Weather Transient No indoor sources of PM10 / PM2.5
Group 2 (G2)
Outdoor Conditions Model 6 Model 7 Model 8 Model 9 Model 10 Avg. PM10 20 µg/m3 0.5 ACH1 3 ACH 6 ACH2 9 ACH3 15 ACH4 Avg. PM2.5 10 µg/m3 Indoor Temp 21ºC (constant) Avg. outdoor CO2 400 ppm 36 students, 30 mg/s of CO2 generated per student Weather Transient No indoor sources of PM10 / PM2.5 1 Air changes provided purely by infiltration. 2 Recommended air changes according to ASHRAE standard 3 Recommended air changes according to CIBSE standard 4Recommended air changes according Chilean standard
RESULTS AND DISCUSSION
3.1 Simulation results for Indoor PM10 and PM2.5 concentrations Table 2 presents the predicted PM10 concentrations for the first set of models (G1). Results show that the highest mean value of indoor pollution was 64.3 µg/m3 and it corresponded with Model 5 that has the highest ventilation rate (15 ACH). Conversely, the lowest mean indoor concentration of PM10 was 28 µg/m3 for Model 1, which has a ventilation rate of 0.5 ACH. The highest increase in indoor PM 10 concentrations was observed in Model 2 where indoor PM10 levels rose 77% compared to Model 1. Similarly, Table 3 presents the predicted PM10 concentrations for G2 models. Overall, absolute values for indoor levels of PM10 in G2 are by far lower than results in G1. In this case, the lowest mean indoor PM10 level was 9 µg/m3 for Model 6 and the highest mean value was 17.5 µg/m3 for Model 9 and 10. Table 2: Indoor PM10 concentration (µg/m3) results for G1 simulations Model 1 Model 2 Model 3 Model 4 Indoor PM10 statistics (G1 = outdoor 67 µg/m3) 0.5 ACH 3 ACH 6 ACH 9 ACH Mean 28.0 49.7 55.7 63.0 Standard deviation 8.7 10.2 8.8 4.2 Minimum 11.2 17.5 18.8 38.9 Maximum 57.8 65.2 66.1 66.8
Model 5 15 ACH 64.3 3.4 39.8 66.9
Table 3: Indoor PM10 concentration (µg/m3) results for G2 simulations Model 6 Model 7 Model 8 Model 9 Indoor PM10 statistics 3 (G2 = outdoor 20 µg/m ) 0.5 ACH 3 ACH 6 ACH 9ACH Mean 9.0 15.0 16.7 17.5 Standard deviation 2.4 2.9 2.5 2.2 Minimum 3.2 5.8 6.6 6.8 Maximum 17.3 19.5 19.7 19.8
Model 10 15 ACH 17.5 2.2 6.8 19.8
Table 4 and Table 5 summarise the results of indoor concentrations of PM2.5 for the two groups of simulations. As can be predicted, the highest indoor PM2.5 concentrations were observed in the simulations run under the most polluted environment (G1). Detailed results for G1 simulations show that Model 5 (15 ACH) obtained a mean indoor PM2.5 level of 42.4 µg/m3 in contrast to 24.2 µg/m3 achieved in Model 1 (0.5 ACH). PM2.5 results for G2 simulations also show that the model with the highest airflow achieved the high concentration of pollutant. In this case, Models 9 and 10 predict 9.2 µg/m3 as mean indoor level of PM2.5. Conversely, Model 6 (0.5 ACH) obtained the lowest indoor concentration of PM2.5 for G2 simulations with a mean value 5.7 µg/m3. Table 4: Indoor PM2.5 concentration (µg/m3) results for G1 simulations Model 1 Model 2 Model 3 Model 4 Indoor PM2.5 statistics (G1 = outdoor 45 µg/m3) 0.5 ACH 3 ACH 6 ACH 9 ACH Mean 24.2 36.9 39.9 41.2 Standard deviation 5.5 5.4 4.4 3.8 Minimum 12.6 18.3 19.7 20.4 Maximum 41.1 44.3 44.6 44.8
Model 5 15 ACH 42.4 3.1 21.1 44.9
Table 5: Indoor PM2.5 concentration (µg/m3) results for G2 simulations Model 6 Model 7 Model 8 Model 9 Indoor PM2.5 statistics (G2 = outdoor 10 µg/m3) 0.5 ACH 3 ACH 6 ACH 9 ACH Mean 5.7 8.3 8.9 9.2 Standard deviation 1.1 1.1 0.9 0.8 Minimum 2.8 4.4 4.9 5.0 Maximum 9.1 9.8 9.9 9.9
Model 10 15 ACH 9.2 0.8 5.0 9.9
3.2 Influence of atmospheric pollution on IAQ Simulation results for a classroom with 0.5 ACH (air change purely by infiltration) shown that the average indoor level of PM10 is 42% of the outdoor PM10 concentration. In the case of PM2.5, the indoor concentration is 52% of outdoor PM2.5. It can be argued that they are in agreement with the estimations of WHO (2005), which established that for a home with an air infiltration of 0.75 ACH and no indoor sources of pollution, average values of indoor PM can be 65% of the outdoor concentration. In the case of ventilation rates above 6 ACH, simulation results show that indoor PM levels tend to be equal to outdoor concentrations for both sizes of PM. These results coincide with findings from six studies reviewed by Chen & Zhao (2011). However, it should be noted that the transfer of particles is determined by many factors that include particle size, external wind speeds, building shape and geometry of cracks in the building envelope, among others (Mumovic & Santamouris 2009; Jeng et al. 2006, 2007). Indeed, other factors that were not considered in this study, for example wall roughness, can have a significant influence on pollutant penetration (Liu & Nazaroff 2003). Considering the evidence presented above, it is possible to state that atmospheric pollution has a large influence on IAQ of naturally ventilated classrooms in Rancagua, because at the ventilation rates recommended by both CIBSE and ASHRAE standards (above 6 ACH), indoor concentrations of PM10 and PM2.5 will be almost equal to outdoor levels. Therefore, if the outdoor atmosphere is highly polluted, the indoor environment will also show high levels of air pollution. 3.3 Adverse effects on students for indoor pollutants' concentrations The predicted indoor pollutants' concentrations could be comparable with those found in a monitoring study carried out in a German nursery school by Fromme et al. 2005. The study showed that the average indoor concentration of RPM (PM10 and PM2.5 are not differentiated and are referred in the study as ‘respirable particulate matter’ (RPM)) was 52.6 µg/m3, with a maximum of 128 µg/m3. The authors qualified such levels of indoor pollution as ‘relative high’ concentrations. They established that those values were similar to the levels found in smoker’s apartments. Comparing Fromme’s recorded levesl with some results of this study, for example Model 3 (G1, 6 ACH), it is possible to say that predicted concentration of indoor PM10 and PM2.5 would be ‘relative high’ as well, because the values seen were 55.7 ± 8.7 µg/m3 and 39.9 ± 4.4 µg/m3 respectively. In addition, it should be noted that these values correspond to a model that provides an air exchange rate of 6 ACH and therefore in order to meet 9 ACH (CIBSE standard) or 15 ACH (Chilean standard) of air renovation, the indoor air concentration of PM10 and PM2.5 will increase even more.
With regards to adverse effects on occupants, previous studies have correlated high indoor levels of PM10 and PM2.5 with a wide range of short term health problems, such as mucosal symptoms (Walinder et al. 1997; Chatzidiakou et al. 2014), chronic bronchitis, chronic pulmonary disease, emphysema (Bruce et al. 2002), and in the long-term, lung cancer and cardiovascular disease (WHO 2005). In addition, Janssen et al. (2003) linked indoor concentration of PM2.5 equal to 20.5 ± 2.2 µg/m3 with conjunctivitis, hay fever, current itchy rash and sensitisation to outdoor allergens. This concentration is similar to indoor PM2.5 predicted by Model 1 in G1 (24.2 ± 5.5 µg/m3), which only considered the transfer of pollutants purely by infiltration. In addition, Sanhueza et al. (2009) related PM2.5 increments of 50 µg/m3 with an increase of 9% in the lower respiratory disease in children between 3 and 9 years old in Temuco, another Chilean city with serious air pollution problems. In 2013, Vela & Miller analysed the results of 3,880 Chilean schools in the national test for learning outcome assessment (SIMCE) between 1997 and 2012. They found that 10 units increase of PM10 is linked with a reduction between 2.3% and 4.9% of a standard deviation in students’ performance for reading, math and social / natural science tests when PM10 concentrations exceed 50 µg/m3. These results agree with findings of previous studies on air pollution and academic performance by Lavy et al. (2012) and Zweig et al. (2012). 4
This research aimed to assess the impact of a polluted outdoor environment on the classroom’s IAQ as well as to discuss the suitability of using natural ventilation strategies to provide fresh air. Simulations predicted high indoor PM10 and PM2.5 for all models run under Rancagua’s average PM during monitoring period (PM10 67 µg/m3 and PM2.5 45 µg/m3), especially for those with high ventilation rates. In the studied classrooms, regardless of outdoor PM10 and PM2.5 levels and with the increase of ventilation rates, the indoor concentrations of PM tend to be equal to outdoor concentrations. Therefore, if the external environment is highly polluted and classrooms are naturally ventilated by cross ventilation, the indoor PM concentration will be high even at ventilation rates below 6 ACH (ASHRAE standard) and 9 ACH (CIBSE standard). This study suggests that the most basic strategies for natural ventilation should not be used in environments with high levels of pollution because the penetration of pollutants is high even in buildings with low infiltration rates, leading to adverse effects on occupants’ health and performance. In particular, Chilean authorities should strengthen building regulations to assure the development of schools with healthy indoor environments, especially when they are located in polluted areas. In addition, current levels of external air pollution should be significantly reduced according to international standards.
REFERENCES Allard, F., 1998. Natural ventilation in Buildings. A design Handbook American Society of Heating, Refrigeration and Air-Conditioning Engineers (ASHRAE) (2010): Standard 62.1: Ventilation for Acceptable Indoor Air Quality, Atlanta, GA, American Society of Heating Refrigerating and, Air-Conditioning Engineers, Atlanta, USA Bruce, N., Perez-Padilla, R. & Albalak, R., 2002. The health effects of indoor air pollution exposure in developing countries. CBES Berkeley 2014, Ocuppant IEQ survey. University of California Berkeley Chatzidiakou, L. et al., 2014. A Victorian school and a low carbon designed school: Comparison of indoor air quality, energy performance, and student health. Chen, C. & Zhao, B., 2011. Review of relationship between indoor and outdoor particles: I/O ratio, infiltration factor and penetration factor. Atmospheric Environment, 45(2), pp.275–288 CIBSE AM10, 2005. Natural ventilation in Non-Domestic Buildings, London, UK. Cooper, D. et al., 1987. Coupling of Ventilation and CO2 Production during Exercise in Children. Engvall, K., Norrby, C. & Sandstedt, E., 2003. The Stockholm Indoor Environment Questionnaire : a sociologically based tool for the assessment of indoor environment and health in dwellings. Indoor Air, 14, pp.24–33. Fromme, H. et al., 2005. Elemental carbon and respirable particulate matter in the indoor air of apartments and nursery schools and ambient air in Berlin. Humphreys, M.A., Rijal, H.B. & Nicol, J.F., 2010. Examining and developing the adaptive relation between climate and thermal comfort indoors. , (April), pp.9–11 Janssen, N. a. H. et al., 2003. The Relationship between Air Pollution from Heavy Traffic and Allergic Sensitization, Bronchial Hyperresponsiveness, and Respiratory Symptoms in Dutch Schoolchildren. Jeng, C.J., Kindzierski, W.B., Smith, D.W., 2006. Particle penetration through rectangular-shaped cracks. J. Environ. Eng. Sci. 5, S111eS119. Jeng, C.-J., Kindzierski, W.B., Smith, D.W., 2007. Particle penetration through inclined and Lshaped cracks. J. Environ. Eng. 133, 331e339. Jones, B. et al., 2013. The Effect of Party Wall Permeability on Estimatios of Infiltration from Air Leakage. International Journal of Ventilation. Jones, B., 2013. Tutorial and Design Exercises for Airflow Studies. , pp.1–43. Lavy, V., Ebenstein, A., and Roth, S. 2012. The Impact of Air Pollution on Cognitive Performance and Human Capital Formation. Liu, D. & Nazaroff, W.W., 2003. Particle Penetration Through Building Cracks. Aerosol Science and Technology, 37(7), pp.565–573. Meteotest., 2014. Meteonorm software (Vesion 7.3.1). Available at http://meteonorm.com/en/downloads MMA Chile, 2013. Primer Reporte Reporte del Estado Estado del del Medio Ambiente Ambiente Mumovic, D. & Santamouris, M., 2009.The Handbook of Sustainable Building Design & Engineering. Routledge, UK Ozkaynak, H., Xue, J., Spenglar, J.D., Wallace, L.A., Pellizzari, E.D. and Jenkins, P., (1996). Personal exposure to particles and metals: results from a particle TEAM study in Riverside CA. Sanhueza, P. a. et al., 2009. Particulate Air Pollution and Health Effects for Cardiovascular and Respiratory Causes in Temuco, Chile: A Wood-Smoke-Polluted Urban Area. Journal of the Air & Waste Management Association, 59(12), pp.1481–1488. Shrubsole, C., Ridley, I., Biddulph, P., Milner, J., Vardoulakis, S., Ucci, M. & Davies, M. (2012). Indoor PM2.5 exposure in London’s domestic stock: modelling current and future exposures following energy efficient refurbishment. Atmospheric Environment, 62, 336-343. Vela, M., & Miller, S. (2013). The Effects of Air Pollution on Educational Outcomes : Evidence from Chile Educational Outcomes. Walinder, R. et al., 1997. Nasal Mucosal Swelling in Relation to Low Air Exchange Rate in Schools. , pp.198–205. WHO, 2005. Global Update Air Quality Guidelines. Zweig, J., Ham, J., and Avol, E. 2012. Air Pollution and Academic Performance: Evidence from California Schools