Article pubs.acs.org/est
Indoor Particle Levels in Small- and Medium-Sized Commercial Buildings in California Xiangmei (May) Wu,† Michael G. Apte,‡ and Deborah H. Bennett*,† †
Department of Public Health Sciences, University of California, Davis, California, United States Lawrence Berkeley National Laboratory, Berkeley, California, United States
‡
S Supporting Information *
ABSTRACT: This study monitored indoor and outdoor particle concentrations in 37 small and medium commercial buildings (SMCBs) in California with three buildings sampled on two occasions, resulting in 40 sampling days. Sampled buildings included offices, retail establishments, restaurants, dental offices, and hair salons, among others. Continuous measurements were made for both ultrafine and fine particulate matter as well as black carbon inside and outside of the building. Integrated PM2.5, PM2.5−10, and PM10 samples were also collected inside and outside the building. The majority of the buildings had indoor/outdoor (I/O) particle concentration ratios less than 1.0, indicating that contributions from indoor sources are less than removal of outdoor particles. However, some of the buildings had I/O ratios greater than 1, indicating significant indoor particle sources. This was particularly true of restaurants, hair salons, and dental offices. The infiltration factor was estimated from a regression analysis of indoor and outdoor concentrations for each particle size fraction, finding lower values for ultrafine and coarse particles than for submicrometer particles, as expected. The I/O ratio of black carbon was used as a relative measure of the infiltration factor of particles among buildings, with a geometric mean of 0.62. The contribution of indoor sources to indoor particle levels was estimated for each building.
■
secondary reactions with terpenes used in cleaning products.15 Indoor PM levels also result from outdoor particles infiltrating into the building, which has been explored in residential buildings,16 but less so in commercial buildings.17−20 We conducted a field study to monitor building ventilation and indoor air quality in a semirandomized sample of 37 SMCBs in California. The objectives of this paper are to (1) present the results for particulate matter levels inside of buildings as well as indoor/outdoor concentration (I/O) ratios for SMCBs serving a variety of functions; (2) to estimate the I/O ratio for particles of outdoor origin and analyze the relationships among building ventilation, filtration, and indoor PM levels; and (3) examine the contribution of indoor sources to indoor particle levels in SMCBs.
INTRODUCTION Multiple epidemiology studies have found exposures to PM2.5 to cause respiratory and cardiovascular health effects.1,2 More recent attention has also focused on the potential health effects of ultrafine particles.3,4 The U.S. population spends approximately 87% of their time indoors, and residential PM concentrations have been well studied. However, 21% of indoor time is spent in nonresidential environments, primarily in commercial buildings,5 therefore, particle concentrations in commercial buildings are of particular interest. Ninety-six percent of existing commercial buildings are small and medium commercial buildings (SMCBs), defined here as being less than 50 000 ft2 (4600 m2), functioning as small offices, restaurants, retail establishments, and health care facilities, among other uses.6 One study collected indoor PM concentrations in 100 randomly selected large office buildings (BASE study).7 Indoor PM concentrations have also been measured in 38 offices and schools of various sizes.8 There have been some studies measuring PM concentrations in nonsmoking restaurants, which have significant indoor particles generated by cooking,9−11 but few data are available in other building types. Potential indoor sources in commercial buildings without cooking include particle resuspension due to occupant movement, which have been found to increase PM in residential settings.12,13 In addition, use of laser printers can also generate ultrafine particles,14 as can ozone-initiated © 2012 American Chemical Society
■
MATERIALS AND METHODS Building Sample. Small and medium commercial buildings (SMCBs), as defined for this study, are any low-rise buildings (less than four stories, less than 50 000 ft2) that are served by rooftop heating, ventilation, and air conditioning units. The building sample is detailed in the Supporting Information (SI). Briefly, the sample included 37 different buildings, with 3 Received: Revised: Accepted: Published: 12355
June 11, 2012 October 1, 2012 October 8, 2012 October 8, 2012 dx.doi.org/10.1021/es302140h | Environ. Sci. Technol. 2012, 46, 12355−12363
Environmental Science & Technology
Article
accuracy for concentrations at 500 000 particles/cm3 is stated by the manufacturer to be within ±10%. Continuous size-fractionated fine particle number concentrations for particles greater than 0.3, 0.5, 0.7, 1.0, 2.0, and 5.0 μm were measured each minute using laser-based optical particle counters, specifically, Met One 237B instruments (Grants Pass, OR). Optical particle counters (OPC) will undercount particles when concentrations are high, due to a tendency for multiple particles passing through the laser beam to be counted as a single particle. For the Met One counters, this undercounting due to coincidence loss (based on information from the manufacturer) is less than 10% at a total particle count of 140 particles/cm3. Approximate mass concentrations were determined, assuming the particle diameter is the geometric midpoint of a size fraction and a particle density of 1 g/cm3 (see SI for details). Note that density is not well quantified and that this value is low but used for simplicity. Integrated PM2.5, PM2.5−10 ,and PM10 mass concentrations were measured beginning with the sixth building surveyed. Samples were collected using two colocated 30 L/min Harvard Cascading Impactors at each sample location, with one sampler collecting PM10 onto a Teflon filter and the second collecting PM2.5 onto a Teflon filter and PM2.5−10 on polyurethane foam (PUF)21 (see SI for details). Black carbon measurements were collected every 5 min using aethalometers (models AE22, AE31, and AE42, MAGEE Scientific, Berkeley, CA), by optical transmission through a quartz fiber tape. Three aethalometers were used, with one located on the roof (representing the concentrations brought in through the HVAC inlet), one located outside the building at ground level (representing concentrations brought in through the building shell), and one located indoors. Building HVAC Characterization. The vast majority of buildings in the study had packaged rooftop HVAC systems. In 62% of the buildings, outdoor air was brought into the building through an intake in the HVAC unit, while in 38%, the HVAC unit did not have an intake to provide outside air. Total building ventilation due to mechanical systems and naturally driven air flows, a, was measured using a Tracer Decay Method.22,23 SF6 was released by a field technician into each room/area to obtain a uniform SF6 concentration. Measurements of the tracer were taken continuously in 1−3 locations spread spatially throughout the building (depending on size). The initial SF6 concentrations of the various locations within the building were generally within 20% of each other. The slopes from multiple sample locations were averaged, with a median percent difference between the highest and lowest location of 16%. The average air exchange rate across the buildings was 1.6 ± 1.7 h−1 (25th percentile: 0.7 h−1; median 1.0 h−1; 75th% percentile: 1.9 h−1).24 Additionally, the rate of outdoor air supplied through the HVAC system, called “HVAC ventilation” (aHVAC), was measured using a calibrated variable speed fan with integral air flow meter, primarily a Duct Blaster (The Energy Conservatory, Minneapolis, Minnesota) connected in series with the outdoor air intake.24 We calculated “additional ventilation” (ashell) based on the difference between the total ventilation and aHVAC, likely caused by a variety of reasons, including pressure differences resulting from mechanical flows as well as natural phenomena, such as temperature differences or wind on the building shell. It was noted if the building doors and windows were kept open the majority of the sampling day.
buildings sampled on two separate occasions, resulting in 40 sampling days, treated as independent observations. The selection of buildings was semirandomized, providing a minimum coverage of the vast variety of SMCBs across space, age, size, and building use. Buildings were almost evenly enrolled across five regions of California; north-coastal, northinland, south-coastal, south-inland, and central-inland, nominally representing various climate regions. Twenty-five buildings were built before 2000 and 12 were built after 2000. There were 24 small buildings (1000−12 000 ft2, two sampled twice), 7 medium buildings (12 000−25 000 ft2, one sampled twice), and 6 medium/large buildings (25 000−50 000 ft2). The buildings varied in their use, including 7 retail establishments, 5 restaurants (one sampled twice), 9 nonmedical offices (one sampled twice), 2 each of gas station convenience stores (one sampled twice), hair salons, healthcare facilities, grocery stores, dental offices, and fitness gyms, along with 4 buildings with other functions, including one daycare facility. Aerosol Sample Collection. Continuous measurements were made for ultrafine particles, size-fractionated fine particulate matter, and black carbon, and time-integrated measurements were made for PM2.5, PM2.5−10, and PM10. Size fractionated and time-integrated particle concentrations were measured simultaneously at one or two indoor locations (Cin,1, Cin,2), depending on building size, and at one outdoor location (Cout); ultrafine particles were measured at one indoor and one outdoor location per building (Cin and Cout); black carbon was measured at two outdoor locations and one indoor location per building (Croof, Cout, Cin) (Figure 1). For the indoor location(s),
Figure 1. Illustration of sampling locations and related parameters. Abbreviations are as follows: Cin, indoor concentration (μg/m3); Cout, outdoor concentration (μg/m3); Croof, rooftop concentration (μg/m3); a, total air exchange rate (1/h), made up of aHVAC, the portion being brought in through the intake on the HVAC unit, and ashell, the portion being brought in through the building shell.; PHVAC, penetration fraction through HVAC supply; P, penetration fraction through the building shell (unitless); λ, rate of air recycled through the HVAC system and thus the filter (1/h); f, fraction of particles removed while being recirculated in the HVAC system (unitless); V, building volume (m3); S, indoor source rate (μg/h); k, deposition rate (1/h).
sampling devices were placed in the main occupied area as centrally located as possible without disrupting business operations, approximately 1.2 m from the floor. All samples were collected for 6−8 h during normal working hours. Ultrafine particle concentrations were measured each minute using portable condensation particle counters (CPC) (model 3781, TSI, Shoreview, MN). These counters can detect particles from 6 nm to over 500 nm in diameter. It has fast response to changes in concentration (providing a 95% change in
