Improving indoor air quality in New Zealand classrooms with a solar ventilation unit: the air particulate part. Mikael Boulic1,*, Bill Trompetter2, Travis Ancelet2, Juan C. Garcia-Ramirez3, Perry Davy2, Robyn Phipps1, Chris Cunningham4, Michael Baker5, Philippa Howden- Chapman5 1
Massey University, Auckland, New Zealand GNS Sciences, Lower Hutt, New Zealand 3 Massey University, Palmerston North, New Zealand 4 Massey University, Wellington, New Zealand 5 University of Otago, Wellington, New Zealand 2
*
Corresponding email: m.boulic@massey.ac.nz
SUMMARY There is a lack of research on particulate matter (PM) concentration within the New Zealand (NZ) schools. A study was conducted in two NZ classrooms over a three-week period. A solar ventilation unit (treatment) was activated in one classroom. PM 10 was monitored inside the classrooms plus outdoors. Hourly-resolved coarse and fine PM was collected onto substrates for elemental composition (ion beam analysis) and source apportionment (positive matrix factorization). PM 10 concentrations increased within both classrooms during school hours when outside level remained stable. PM 10 increase was not related to outdoor conditions, but to children’s activity re-suspending dust. The solar ventilation unit had a positive impact in decreasing the PM 10 concentrations by a factor of 1.5 in the treatment classroom. PM in the classrooms was predominantly from crustal sources (fine and coarse soil dust). There is a need for dust exposure mitigation strategies (carpet cleaning regime, dust reducing carpet) in NZ classrooms. PRACTICAL IMPLICATIONS This study is the starting point for documenting the impact of ventilation on particulate matter within New Zealand classrooms. The results support the need to investigate for dust exposure mitigation strategies and for ventilation rate increase in classrooms. KEYWORDS Intervention study, dust elemental composition, dust source contribution. 1 INTRODUCTION Ninety percent of the New Zealand (NZ) classrooms are naturally ventilated through open window (McIntosh, 2011). Due to the combination of a high density of occupants and a reliance on natural ventilation, it is challenging to provide the classrooms with adequate ventilation and consequently an acceptable indoor air quality (IAQ) during the winter months (Jurelionis and Seduikyte, 2008). The NZ Standard “Ventilation for acceptable IAQ” (NZ Standard, 1990) requires a ventilation rate of eight litres of fresh air per second and per child. The NZ Ministry of Education recommends teachers open windows to reach an acceptable ventilation rate (BRANZ, 2012). A NZ study showed that classroom ventilation rates will meet the 8 l.s-1 recommended value only if four windows are kept open simultaneously (Cutler-Welsh, 2006) however this comes at the expense of allowing heat to escape and thermal discomfort.
Conventional mechanical ventilation systems are capital and energy expensive as well as need maintenance; they are not affordable for most NZ schools (Cutler-Welsh, 2006). Consequently, an alternative and affordable method for increasing classroom ventilation rate in winter is needed. As the school hours (9am to 3pm) are closely aligned with the optimum solar radiation, a solar ventilation unit (solar collector) might be a good solution (no need for energy storage) to ventilate the classroom. So far, PM monitoring has primarily been focused on outdoor locations through ambient air quality monitoring networks. During winter 2010, a monitoring study outside two NZ schools found that PM 10 largely originated from wood burning emissions (Ancelet et al., 2012). Because little is known about PM concentrations inside classrooms, and even less is known about the origin of the PM sources and the effect of ventilation on the PM concentration, a pilot study was undertaken in two primary school classrooms. The aim of this pilot study was to assess the impacts of outdoor PM concentrations, coming in the classroom through the solar ventilation unit, on the indoor environment and investigate sources of PM generated within the classrooms. 2 MATERIALS/METHODS The study was undertaken over a three-week period (from 12 August to 1 September 2013, winter season in the Southern Hemisphere) inside two classrooms and outside in a Palmerston North school, NZ. Palmerston North is located at about 30 km from the Tasman Sea and subjected to marine aerosol. One of the two classrooms was a treatment classroom (solar ventilation unit installed and activated) while the other classroom was a control classroom (solar ventilation unit installed but disabled). The solar collector units were located on the north facing roof (southern hemisphere). The unit is made of a transpired solar panel collector, a fan, ducting and a controller. The warm air is pumped into the inside through the ducting by a photovoltaic powered fan. Both, the control classroom and the treatment classroom were located side by side, so that they were matched for location and construction type (Figure 1).
Figure1. Control classroom and treatment classroom located side by side. A solar ventilation unit (collector) is located on the north facing roof of each classroom. Outside the classrooms, hourly PM 10 concentrations were monitored using a continuous EBAM (Met One Instruments, Inc., OR, USA). The E-BAM calculates particle mass based on
beta-particle attenuation through a filter tape. For monitoring the PM 10 concentrations inside the classrooms, much quieter light-scattering devices were required to minimise operational noise and disruption to each class. A TSI Dusttrak II Aerosol Monitor (Model 8530, TSI Inc., MN, USA) and a Grimm (Model 1.108, GRIMM Technologies Inc., GA, USA) were used to hourly monitor PM 10 in the treatment classroom and in the control classroom respectively. Subsequent co-located monitoring gave good agreement between the three devices. In addition, in the two classrooms and outside, fine (PM 2.5 ) and coarse (PM 10 ) particles were hourly collected on two separate filters using a StreakerTM (PIXE International Corporation, FL, USA). Then, the elemental compositions of the samples were measured using ion beam analysis at the New Zealand Ion Beam Analysis Facility in Gracefield, Lower Hutt. The full suite of analyses included particle-induced X-ray emission (PIXE), particle-induced gammaray emission (PIGE) and Rutherford backscattering (RBS). Positive matrix factorization (PMF) techniques were used to identify the PM sources and their hourly contributions. In the treatment classroom, the temperature (°C) and the velocity (m.s-1) of the incoming air (from the solar ventilation unit) were monitored at 10-min intervals using a hot wire anemometer (Lutron Electronic AM-4214SD, Lutron Electronic Enterprise, Taipei, Taiwan). 3 RESULTS AND DISCUSSION Temperature and velocity of the air coming from the solar ventilation unit Figure 2 shows the three week period averaged incoming air temperature (±95CI) during the school day (9am – 3pm).
Figure 2. Averaged incoming air temperature over a three week period (± 95 CI) Figure 2 shows that during school hours the incoming temperature was always above 18.0 ºC which is the minimum threshold targeted to follow the World Health Organization (WHO) recommended temperature (WHO, 1987). Air temperatures below 18.0 ºC should not be pumped into the classrooms. The average temperature was 30.9 ºC [30.5 ºC - 31.4 ºC]. These results confirmed that the solar ventilation unit contributed also to the classroom heating during winter season. Increasing the ventilation rate (m3.h-1) was the main goal in this study in order to expel the PM to significantly benefit children’s health. Complying with the NZ Standard (NZ Standard,
1990) should provide eight litres of fresh air per second and per child. Assuming 30 children per classroom, around 850 m3 of fresh air should be provided to the classroom per hour. However, this standard value is the total ventilation rate including infiltration and natural/mechanical ventilation. Over this three week period, the air velocity of the incoming air peaked at a maximum of 2.65 m.s-1 which gives a flow rate of 117 m3.h-1. It will take 1.7 hours to change the whole volume of the classroom (assuming a classroom volume of 200 m3). Over the three week period, the average air velocity was 1.81 m.s-1 which gives a flow rate of 80 m3.h-1 (11 times lower than the 850 m3.h-1 recommended value). Insufficient airflow was provided to comply with the required ventilation rate; however this was not surprising given that the unit was set up for ventilation and heating, not just ventilation. Further research is currently being undertaken to reduce the time lag of the air in the solar ventilation unit to increase the flow rate and keep the incoming air temperature around 20 ºC and not at the current daily average of 30.9 ºC. In addition, the number of collector will be increased as a single unit will not be able to provide a flow rate of 850 m3.h-1. Despite not being sufficient to comply with the Standard, this higher ventilation rate in the treatment classroom should impact on the classroom PM concentration. Concentrations of PM10 Figure 3 shows that during school hours, the PM10 concentrations increased within both control and treatment classrooms when outside level remained stable. A higher PM10 concentration was found in control classroom than in treatment classroom. On average, the treatment decreased the PM10 concentrations by a factor of 1.5 (from 36 μg.m-3 to 23 μg.m-3). The PM10 increase originated from the children’s activity re-suspending dust by air movement and was not related to outdoor conditions.
Figure 3. The hourly average mass concentration measurements for the PM 10 samples
Sources of PM Overall, 940 samples (470 coarse filters and 470 fine filters) were collected from each of the three locations over this three week monitoring period. Using PMF techniques, the same three primary source contributors were identified at each of the sampling locations. The sources identified were soil, marine aerosol and combustion. Source compositions are well described in the literature. The soil profile features elements associated with crustal matter (soil), including Al, Si, Ca and Fe. The marine aerosol (sea salt) based on the high concentrations of fine and coarse Cl in the profile, and the presence of other elements found in sea water, including S and K. The combustion source is based on the presence of black carbon (a marker of incomplete combustion) and other indicators like Cu and Zn for motor vehicle exhaust (Thorpe and Harrison, 2008), S for coal combustion and fine K for wood combustion (Khalil and Rasmussen, 2003). Wood combustion for home heating is common during the winter in New Zealand and is often the major contributor to poor air quality in New Zealand (Trompetter et al, 2010; Ancelet et al, 2014). On average, the three sources identified accounted for 83%, 81% and 83% of the measured PM 10 from treatment classroom, control classroom and outdoors respectively. The unaccounted mass percentage is possibly related to organic sources made up of compounds that we could not measure using IBA techniques or limitations associated with particle mass measurements. Table 1 shows the source apportionment and the concentration for each of the three locations. Table1. Sources, apportionments and daily concentrations The three identified PM 10 sources Locations % Treatment Control Outside
41.5 42.9 29.1
Soil Concentration (μg.m-3) 5.3 7.2 3.1
Marine aerosol Concentration % (μg.m-3) 24.9 3.3 21.9 3.6 28.2 3.1
Combustion Concentration % (μg.m-3) 16.6 2.2 16.2 2.6 25.7 2.8
Unaccounted mass percentage % 17 19 17
Table 1 showed that the soil was the main source of dust followed by the marine aerosol and the combustion sources. A higher soil contribution to indoor PM concentrations is sensible, given activity within classrooms that can result in re-suspension of dust that has been tracked in by children’s footwear. Marine aerosol is ubiquitous throughout New Zealand, even at inland locations like Palmerston North over 30 km from the nearest coastline. Both the marine aerosol contribution and the combustion contribution showed similar concentrations at each of the sampling sites. These results confirmed that the increased of PM 10 in both classrooms, compared to outside, was due to crustal sources, most likely related to re-entrained dust from activities within the classrooms. 4 CONCLUSIONS Results showed that significantly higher PM concentrations occurred within both classrooms during school hours (9am to 3pm). It is also apparent that on average 1.5 times higher concentrations occurred in the unventilated control classroom. The source apportionment results showed infiltration of marine and traffic PM, however the increased PM in the classrooms was predominantly from crustal sources originating from the children’s activity re-suspending dust by air movement and was not related to outdoor conditions. The WHO recommends for PM 10 a maximum annual mean of 20 μg.m-3 and 24h mean of 50 μg.m-3; so
there is a need for dust exposure mitigation strategies (carpet cleaning regime, dust reducing carpet, increase ventilation rate) for NZ classrooms. The treatment (solar roof collector unit) had a positive impact on the PM concentrations in the ventilated treatment classroom. However, there is a need to optimize the performance of the solar collector unit for a higher ventilation rate to keep lowering the indoor PM 10 concentration to outdoor concentrations. In addition, the number of collector per classroom will need to be increased as a single unit will not be able to provide a flow rate of 850 m3.h-1 to comply with the NZ Standard. There is a potential for solar technology to increase the ventilation rate and expelling PM whilst increasing the temperature in learning environments. ACKNOWLEDGEMENT This study was funded by the Health Research Council of New Zealand, the New Zealand Lottery Grants Board, the Building Research Association of New Zealand, Massey University, GNS Science Direct Core Funding and University of Otago. The researchers would like to thanks the schools’ communities and also C. Purcell of GNS Science for technical assistance with accelerator operation. 5 REFERENCES Ancelet T., Davy P.K., Mitchell T., Trompetter W.J., Markwitz A., and Weatherburn D.C. 2012. Identification of Particulate Matter Sources on an hourly time scale in a wood burning community. Environ. Sci. Technol., 46:4767- 4774. Ancelet T., Davy P.K., Trompetter W.J., and Markwitz A. 2014. Sources of particulate matter pollution in a small New Zealand city. Atmospheric Pollution Research, 5, 572–580. BRANZ. 2012. Designing quality learning spaces: ventilation & indoor air quality. Wellington, New Zealand: Building Research Association of New Zealand. Cutler-Welsh M. 2006. Thorrington school classroom energy and climate management Final Report, Christchurch, New Zealand: Natural Resources Engineering, Canterbury University. Jurelionis A. and Seduikyte L. 2008. Indoor environmental conditions in Lithuanian schools. In: 7th International Conference on Environmental Engineering, Wien, Austria. Khalil M.A.K. and Rasmussen R.A. 2003. Tracers of wood smoke. Atmos. Environ. 37 (9−10), 1211−1222. McIntosh J. 2011. The indoor air quality in 35 Wellington primary schools. Master thesis, Victoria University of Wellington, Wellington, New Zealand. NZ Standard 4303. 1990. Ventilation for acceptable indoor air quality. Wellington, New Zealand: Standards Association of New Zealand. Thorpe A. and Harrison, R.M. 2008. Sources and properties of non- exhaust particulate matter from road traffic: A review. Sci. Total Environ., 400 (1−3), 270−282. Trompetter W.J., Davy P.K. and Markwitz A. 2010. Influence of environmental conditions on carbonaceous particle concentrations within New Zealand. Journal of Aerosol Science, 41, 134–142. WHO. 1987. Health impact of low indoor temperatures WHO report. Copenhagen, Denmark: World Health Organization.