Environ. Sci. Technol. 2010, 44, 5775–5780
Long-Term Characterization of Indoor and Outdoor Ultrafine Particles at a Commercial Building Y U N G A N G W A N G , † P H I L I P K . H O P K E , * ,† DAVID C. CHALUPA,‡ AND MARK J. UTELL‡ Center for Air Resources Engineering and Science, Clarkson University, Potsdam, New York, and Department of Environmental Medicine, University of Rochester Medical Center, Rochester, New York
Received January 17, 2010. Revised manuscript received May 22, 2010. Accepted June 21, 2010.
The particle number concentrations in the size range of 10-500 nm were measured inside and outside of a commercial building in Rochester, New York from 2005 to 2009. The indoor ventilation conditions were controlled by a heating, ventilation, and air-conditioning (HVAC) system. The overall average indoor and outdoor particle number concentrations were 2166 cm-3 and 5214 cm-3, respectively. Comparison of the indoor and outdoor ultrafine particles (UFP) distributions revealed that indoor sources contributed to the indoor UFP concentrations. The indoor/outdoor (I/O) ratio generally increased with particle size. The I/O ratios in the summer months were higher than those in the winter months. Indoor and outdoor correlations of particle concentrations were slightly weaker in warmer months. These results indicated that changes in the air exchange rates (AER) may have affected the correlation between indoor and outdoor UFP number concentrations. Moreover, indoor activities such as food preparation and cleaning may have contributed to the indoor UFP number concentrations.
Introduction Toxicological and epidemiological studies have shown that ultrafine particles (UFP) are associated with significant health risks (1–3). High number concentrations of UFP near urban freeways and intersections have been reported (4, 5). Outdoor particles can infiltrate and penetrate through the mechanical ventilation system and envelope of the building into indoors, respectively (6). Considering that urban population generally spends most time indoors, people living in urban indoor environments are exposed to significant number concentrations of outdoor UFP. Studies have been conducted to investigate the relationship between indoor and outdoor number concentrations of UFP. In Helsinki, indoor and outdoor particles were measured in and near a building with a mechanical ventilation system. The outdoor number concentrations showed that meteorological conditions had a large effect on the number concentrations and the size distributions. The temporal variations in the indoor particle number concentration (7-600 nm) closely followed the corresponding temporal * Corresponding author e-mail:
[email protected]; phone: 315-268-3861; fax: 315-268-4410. † Clarkson University. ‡ University of Rochester Medical Center. 10.1021/es1001677
2010 American Chemical Society
Published on Web 06/30/2010
variation outdoors (7). Variations in outdoor particles sources were an important factor influencing indoor particle levels (8). Open doors and windows caused the indoor particle number concentrations to have similar temporal patterns as the outdoor values (9). In a study in Los Angeles, the indoor/outdoor (I/O) UFP number concentration ratio depended strongly on particle size near urban freeways (10). The largest I/O ratios (0.6-0.9) were observed for larger particles (70-100 nm), while the smallest I/O ratios (0.1-0.4) were observed for particles of 10-20 nm when air exchange rates (AER) ranged from 0.31 to 1.11 h-1. Higher ventilation rates seemed to increase I/O ratio for particles larger than 90 nm. The I/O ratio of nucleation mode particles did not seem to be as strongly dependent on the ventilation rate as for accumulation mode particles (11). Transport of outdoor particles through the building heating, ventilation, and air-conditioning (HVAC) system was the most important contributor to the indoor concentration in commercial buildings (6). Sources, ambient conditions, building structure and materials, occupant behavior and activities, ventilation, and AER are all major factors affecting indoor particle number concentrations (12, 13). Indoor UFP can be generated by high temperature operations such as combustion or electrical arcing (14). Elevated indoor UFP concentrations have been related to cooking activities, cigarette smoking, candle burning, and vacuum cleaning (8, 15, 16). All of the prior studies of indoor environments with the HVAC systems were conducted over relatively short periods with limited data on UFP number concentrations. In this study, the long-term variations in the indoor and outdoor UFP number concentrations and size distributions, indoor/ outdoor particle number concentration ratios, and indoor and outdoor correlations at a commercial building in urban areas were examined over almost 5 years. The major foci are (1) the indoor-to-outdoor relationships of UFP number concentrations and size distributions, (2) the seasonal variations of indoor and outdoor UFP number concentrations and size-resolved I/O ratios on a monthly basis, and (3) the effects of indoor sources on the indoor particle number concentrations.
Experimental Section Site Description and Measurement Period. The continuous measurements were performed from January 2005 to the end of November 2009 at a commercial building in Rochester, New York. Rochester is the third largest city in New York State with a 2008 population of 206,886. It is an urban area influenced by locally transported UFP from the interstate highways 390 and 590 that are 350 m south of the site. Meteorological parameters were measured along with other pollutants (PM2.5, O3, CO, and SO2) 5.4 km northeastward from the site by New York State Department of Environmental Conservation (NYS DEC). The predominant wind direction was southwest making the site downwind of the highways. The location of this building is shown in Figure S1 in the Supporting Information. The measurements were made in a 2700 m2, two-floor commercial building where part of the space is used as a cardiac rehabilitation center. It is open from 06:00 to 18:00 on Monday through Friday and closed on Saturdays and Sundays. It has a typical HVAC system with an economizer, filtration, and variable-volume makeup air. Mechanical ventilation air ducts are installed throughout the whole building to supply air to each room through diffusers and ducts that remove air through return-air grilles. Ventilation VOL. 44, NO. 15, 2010 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
5775
includes both the exchange of air to the outside as well as circulation of air within the building. A medium-efficiency pleated furnace filter (MERV 8 per ASHRAE 52.2-1999) is used in the HVAC system. The filters are changed approximately every three months to ensure continued good performance. The minimum efficiency during conditioning is 85% for 3-10 µm particles (coarse particles). For the particles measured in this study (10-500 nm), the efficiency data are not currently available. Measurements were made in a break room on the ground floor of the building. A side door is about 5 m to the east of the break room door. The room volume was approximately 50 m3 and had one window. Typical of current commercial buildings, the windows in the building were sealed. The room housed a refrigerator, a toaster, a microwave oven, a coffee maker, and twenty lockers. Indoor activities included regular office work, people exercising on the exercise machines 10 m west of the room where measurements were taken, and bread toasting, coffee making, and microwave oven heating in the break room during the day on weekdays. The floors were cleaned with a vacuum cleaner and wet mopping between 00:00 and 05:00 each weekday. No smoking was allowed inside the building. Particle Measurements. Particle size distributions between 10 and 500 nm were measured both indoor and outdoor simultaneously using a wide-range particle spectrometer (WPS, model M-1000XP, MSP Inc.). The sample flow rate in this unit was 1.0 L/min. Measurements were made through a common switching valve alternately sampling indoor and outdoor air every 3.5 min. One size distribution sample was taken during each 3.5-min interval. Two 1-cm conducting tubes conducted the aerosol to the valve. One was located at ceiling height (approximately 2.4 m) that extended to the outside edge of the break room next to the door. Thus, the break room and the main side door had considerable impact on the indoor measurements. The other extended through a hole in the wall to the outside of the building at the same height of the indoor inlet. The outdoor air sampling inlet was as close to the building envelope as possible to shorten the sampling line in order to minimize the particle losses (17). Tests were performed to ensure that the indoor and outdoor inlet sections yielded the same size distribution when sampling from the same location. Air Exchange Rate Measurements. Air exchange rates (AER) were determined by measuring the rate of decay in CO2 concentrations in an exercise room (10 m to the west of the break room) on the first floor with a GE Telaire 7001 CO2 monitor. The AER is related to the change in CO2 concentration with time by λ)
(
j out Cin(t) - C 1 ln j out (t - t0) Cin(t0) - C
)
(1)
where λ is the AER (h-1), and t and t0 are the end and the beginning of the sampling interval (h), respectively (10). The average AER based on a 9-day measurement in August 2006 was 0.25 h-1. Date Analysis. Particle number concentrations were averages of hourly values to minimize the instrument uncertainties and short-period fluctuations. The Pearson correlation coefficients of indoor and outdoor particle number concentration data were calculated to assess the association between indoor and outdoor levels of particles.
Results and Discussion Indoor and Outdoor Particle Size Distributions and SizeResolved I/O Ratios. The average size distributions of both indoor and outdoor UFP and size-resolved I/O ratios obtained during the five-year measurement program are shown in 5776
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 15, 2010
FIGURE 1. Average size distributions of indoor and outdoor UFP and size-resolved I/O ratios during the measurements. Figure 1. The I/O ratios were then truncated to the 0-1 range and the overall average I/O ratio distributions in three particle size bins are shown in Figure S2 (Supporting Information). There were values greater than 1 when the indoor sources were active. It can be seen in Figure 1 that the overall outdoor UFP number concentrations were much higher than those indoors. Both the outdoor and indoor size distributions exhibited a unimodal shape with the peak number concentrations at 30-40 nm and 50-60 nm, respectively. The lowest and highest values of the I/O ratio was 0.17 and 1.44 observed at 15-20 and 480 nm, respectively. There is a sharp drop in the I/O ratio from 0.59 to 0.17 in the particle size range of 10-17 nm. The average size distributions of both indoor and outdoor UFP and size-resolved I/O ratios during 12:00-17: 00, 07:00-09:00, 11:00-13:00, and 16:00-19:00 are provided in Figure S3. It can be also seen that there were differences between indoor and outdoor size distributions, especially during the 11:00-13:00 period. More study is needed to identify the possible indoor UFP sources. Monthly Variations of Indoor and Outdoor Particle Number Concentrations. The monthly average number concentrations and standard variations of particles both indoor and outdoor in three size bins are presented in Figure S4. The mean outdoor number concentrations of 10-50 nm particles in winter months tended to be higher than the values in summer months, as seen in prior work (18, 19). Colder temperature may result in increased nucleation of combustion exhaust such as from vehicles and lower average mixing heights or inversions that occur more often in winter. There were lower temporal variations for larger particles. For indoor particles, there was no evidently significant difference among different seasons. Monthly Variations of Indoor/Outdoor Concentration Ratios and Indoor-Outdoor Correlation Coefficients. Results of the calculations of the monthly average I/O ratios in the three size bins are shown in Figure 2. The average I/O ratio distributions in three particle size bins in four different seasons are shown in Figure S5. In Figure 2, the ensemble average I/O ratios were 0.34, 0.41, and 0.46 for the particles in the size ranges of 10-50, 50-100, and 100-500 nm, respectively. The I/O ratio of UFP number concentrations varied from month to month ranging from a minimum of 0.15 to a maximum of 0.61 in 10-50 nm size range observed in January and July, respectively. The I/O ratios in summer months (Jun-Aug) were greater than those in the winter months (Dec-Feb) for all particles sizes. This finding indicates that the UFP number concentrations indoors generally followed the concentrations outdoors more closely during summer months probably because of the side door being open more frequently (AER changing). To see the seasonal dependence of number concentrations of indoor particles in three size bins on the outdoor concentrations, Pearson correlation coefficients between
FIGURE 2. Average monthly variations of I/O ratios in three size bins. R1, R2, and R3 represent I/O ratios of particles in three size bins.
FIGURE 3. Monthly variations of correlation coefficients between indoor and outdoor particle number concentrations in three particle size ranges. Pearson correlation coefficients of particles in three size bins are represented by r1, r2, and r3. indoor and outdoor concentrations were calculated for each month (Figure 3). The consistent positive correlations reflect an increase in indoor particle concentrations as the outdoor values increase. The average correlation coefficients for the three size ranges are 0.57, 0.66, and 0.79, respectively. It can be seen that the lower correlation coefficients were generally found in warmer weather periods for all particle sizes. Previous work found that rapid changes in the AER may lead to lower correlations between indoor and outdoor particle number concentrations (20). In general, these results indicated that changes of the AER may affect the correlation between indoor and outdoor UFP number concentrations. Poorer correlation between indoor and outdoor particles in the smaller size range suggested the potential presence of indoor sources and higher losses in transit into the building. Diurnal Variations of Indoor and Outdoor Particle Number Concentrations and I/O Ratios. The diurnal behavior of the number concentrations and the I/O ratios are presented separately for three different size ranges in Figure 4. The outdoor UFP number concentration at 07:00 was observed to be elevated corresponding to the morning traffic. Two hours later, a spike showed up in the I/O ratio. The early afternoon concentration peak of particles in the 10-50 nm bin was likely due to secondary aerosol formation. Again two hours later, spikes can be seen in both the I/O ratio and indoor particle profile. The observation of the same lag time between outdoor and indoor spikes at multiple times during the day shows consistent particle transit times. The I/O ratios at noon are clearly higher, especially for particles in the 10-100 nm size range. Also a spike of the I/O ratios during midnight-early morning was observed for all particle sizes.
FIGURE 4. Diurnal variations of indoor and outdoor particle number concentrations and I/O ratios in three size bins (a) 10-50 nm, (b) 50-100 nm, and (c) 100-500 nm during all days. A time series analyses of the indoor particle number concentrations and indoor activities shows that food preparation during the lunch hour (11:00-13:00) in the break room may generate a large number of particles, mostly in the ultrafine size mode (
