Air Quality, Atmosphere & Health

, Volume 6, Issue 4, pp 747–757 | Cite as

Indoor–outdoor concentrations of particulate matter in nine microenvironments of a mix-use commercial building in megacity Delhi

Article

Abstract

Three naturally and six mechanically ventilated microenvironments (MEs) of a mix-use commercial building in Delhi are used to study indoor–outdoor (I/O) relationships of particulate matter ≤10 μm (PM10), ≤2.5 μm (PM2.5), and ≤1 μm (PM1). Effect of environmental and occupancy parameters on the concentrations of PM during working and non-working hours (i.e., activity and non-activity periods, respectively) are also investigated. Average outdoor concentration of PM10 and PM2.5 were found to exceed the 24-h averaged national standard values, showing a polluted environment surrounding the studied building. During working hours, indoor PM10 concentration was found 6–10 times, both PM2.5 and PM1 were 1.5–2 times, higher than the non-working hours in the selected MEs. The variations of indoor concentrations were highest (17.1–601.2 μg/m3) for PM10 compared with PM2.5 (16.9–102.6 μg/m3) and PM1.0 (10.6–63.6 μg/m3). The I/O for PM10, PM2.5, and PM1.0 varied from 0.37–3.1, 0.2–3.2, and 0.17–2.9, respectively. The results suggest highest I/O for PM10, PM2.5, and PM1 as 3.1, 2.15, and 1.76, respectively, in all the three natural-ventilated MEs (canteen, kitchen, reception). Irrespective of PM types, the average I/O was <1 for mechanically ventilated MEs compared with >1 for naturally ventilated MEs. As opposed to PM1, better correlation (r > 0.6) was noted between indoor PM10, PM2.5, and CO2 concentrations in most of the airtight MEs.

Keywords

Particulate matter Building microenvironment Environmental comfort parameters Occupancy I/O relationship Megacity Delhi 

Introduction

Indoor particulate matter (PM) is often linked with detrimental health impacts. Fractions of the PM (i.e., PM10, PM2.5, and PM1; the subscript indicating the upper cut-off particle diameter in micrometer) contains a complex mixture of solid and liquid particles that are made of both organic and inorganic substances (Heal et al. 2012). The size of the particles is vital for determining the duration for which the particles remain suspended in the air for human exposure. PM1 has not been studied enough yet to accumulate sufficient knowledge to regulate this size fraction, but there are evidences that these small particles cause adverse health effects (Polichetti et al. 2009; Heal et al. 2012). On the other hand, PM10 and PM2.5 have already been established as a cause for premature mortality and morbidity. For instance, Pope III and Dockery (2006) estimated a mortality increase in the order of 4–6 % with the increase in PM10 and PM2.5 concentrations by 20 and 10 μg/m3, respectively, in the ambient air of US cities.

Indoor activities such as walking, sweeping, and floor cleaning cause the generation of particles over 1-μm size, due to their resuspension from the dust deposited on floors and other interior surfaces (Thatcher and Layton 1995; Luoma and Batterman 2001). In particular, PM2.5 and PM1 are generated in substantial amounts during activities such as cooking (Morawska et al. 2003; Kumar et al. 2013a), heating and wood burning in fire places (Kleeman et al. 1999; Hussein et al. 2005, 2006), and tobacco smoking (Kleeman et al. 1999; Ott and Siegmann 2006; Miller and Nazaroff 2001).

Indoor air is also affected by outdoor air through infiltration (Colbeck et al. 2010; Massey et al. 2009; Chen et al. 2012). As a result, the geographical heterogeneity in indoor PM exposure can be expected due to inter-city differences in PM concentrations (Zhou et al. 2013). In the absence of indoor sources, the indoor PM10, PM2.5, and PM1.0 concentrations may show similar trends to those found in outdoor environments in naturally ventilated buildings, and can be estimated from the outdoor concentrations (Jones et al. 2000; Kumar and Morawska 2013). Therefore, indoor–outdoor (I/O) relationships are important to understand the real status of indoor air quality (IAQ). However, the case becomes complex in the presence of indoor sources (e.g., cooking, cigarette smoking, and sweeping) that can raise the indoor concentrations to notable levels (Morawska et al. 2001). The I/O relationship in residential buildings with the indoor sources have been found to be up to 2 or even higher in certain situations (Baek et al. 1997; Wallace 1996). One of the key factors that derives the indoor concentration levels is the atmospheric dispersion of pollutants around buildings (Santos et al. 2011), which, in turn, is affected by the land-use pattern of the area where a building in question is located (Kumar et al. 2013b).

Ventilation in naturally and mechanically ventilated buildings is another important parameter (Yamamoto et al. 2010). Air flow rate and its patterns are the two key indicator parameters for ventilation effectiveness in buildings. Higher ventilation flow rates generally result in lower average pollutant concentration. In a well-mixed condition, the average pollutant concentration will reduce linearly with the increase in ventilation flow rate (Memarzadeh 2009). Similarly, the air flow pattern inside the building is the driving force for building design and orientation to have effective ventilation. Several studies have reported the links between the IAQ and air flow rates in indoor environments by analyzing naturally and mechanically ventilated systems as well as the relation between indoor and outdoor air quality. For instance, Kukadia and Palmer (1998) studied the influence of atmospheric pollution on indoor pollution levels in two (naturally ventilated and air conditioned) buildings in the UK. They found that infiltration of outdoor pollution is higher in naturally ventilated buildings compared with air-conditioned buildings. Chaloulakou et al. (2003) studied the influence of outdoor CO concentration on the indoor CO concentration in two different naturally ventilated indoor environments (i.e., an office and a public school) in Athens, Greece. They concluded that outdoor concentrations can be used as a good estimator for indoor concentrations in naturally ventilated buildings. The I/O for an office building was higher (0.74 to 1.0) than those for a school building (0.53 to 0.89). The difference was attributed to factors such as variation in meteorological conditions, different dimensions, layout, and orientation of the buildings. Goyal and Khare (2009) studied the influence of outdoor PM10, PM2.5, and PM1.0 concentrations on the indoor classrooms concentrations of a naturally ventilated school building in Delhi, India. They concluded that the environmental parameters (temperature, relative humidity, wind speed, and direction) and ventilation rate in the building significantly influence the I/O of PM2.5 and PM1.0. They also noted a strong influence of the occupant’s activities on the I/O of PM10 in the classrooms. More recently, Habil et al. (2013) studied the I/O ratios of various PM fractions for the roadside and residentially located schools of Agra City in India. They found the highest I/O ratios during the summer season up to 1.31 (PM10), 1.20 (PM2.5), and 1.25 (PM1) for the residential schools, and up to 1.22 (PM10), 1.19 (PM2.5), and 1.24 (PM1) for the roadside schools. They attributed these highest ratios in summers to a much higher ventilation rate, which ranged between ∼74 and 100 m3 h−1, brought the outside polluted air indoors, and led to the accumulation of PM in classrooms.

A number of previous studies have also reported the influence of various indoor sources, occupant’s activities, outdoor infiltration, and ventilation rate on the PM concentrations in indoor environments, as highlighted by our recent review article (Goyal et al. 2012). However, the effect of building locations on the indoor PM concentrations, especially in a mix-use commercial building environment in polluted megacities like Delhi, are not yet fully understood. This study reports the results of PM10, PM2.5, and PM1.0 monitoring that was carried out in nine different microenvironments (MEs)—three naturally ventilated and six mechanically ventilated—of a mix-use commercial building (see “Site description” for details). The aim of this study is to assess the PM levels during working (activity period) and non-working (non-activity period) hours in selected MEs during the varying level of ventilation types (natural or mechanical) and conditions (frequency of openings of doors/windows), occupants’ activities, and thermal comfort parameters (temperature, pressure, and relative humidity). The study also aims to assess the I/O relationship of PM under these varying conditions during the working hours.

Methodology

Site description

The study building is located at Phase I of a mix-use (i.e., industrial and residential) Naraina Industrial Area in New Delhi (see Fig. 1). The commercial activities around it involve business centers like PVR cinema, hotels, restaurants, and office buildings. The residential area surrounding the studied building is a home for ∼75,000 inhabitants. A railway line carrying the diesel trains passes very close to the studied site. A slum area has also developed along the railway line and this area is a rich source of PM produced by wood burning. The studied building is a double story with a parking zone in its basement. The ground floor is occupied by a genomics lab. The first floor is used by the environmental lab and the different indoor MEs at this floor are selected for the IAQ study. The MEs include a chemical lab (referred hereafter as M1), instrumentation lab (M2), microbiology lab (M3), computational lab (M4), two office areas including administrative office (M5), and scientist’s working room (M6), reception area (M7), kitchen (M8), and canteen area (M9). Details of the volume of each ME, ventilation rates, and the occupancy levels are summarized in Table 1. The floor plan of the building is shown in Fig. 1b. In the chemical lab (M1), windows are kept open and the ceiling fans were continuously running to maintain the comfortable temperature during working hours. M1 is also equipped with split-type air conditioning (AC) system that was in use, but infrequently, when windows were closed and ceiling fans were switched off. Due to the prevalence of the mixed and complex nature of ventilation types, this can be considered as both natural/mechanically ventilated ME. M2 to M4 are also laboratories, but are categorized as mechanically ventilated MEs, because these MEs have split-type AC systems allowing the exchange of outside air into these rooms. The offices (M5 and M6) are also mechanically ventilated with both split and window type ACs working in them. M7, M8, and M9 are termed as naturally ventilated MEs, where both the ceiling and exhaust fans were running for maintaining the temperature and ventilation, respectively. The outdoor air quality monitoring site was located at the terrace of the building, as shown in Fig. 1b. The aerial distance of the ring road, which carries heavy traffic during the day time, from our ambient air monitoring locations is about 800 m (see Fig. 1a).
Fig. 1

Aerial view of the building showing a its location in mix-use area and b indoor–outdoor air monitoring locations. IAQ monitoring was carried out in all these MEs, as detailed in “IAQm monitoring” Section

Table 1

Details of various parameters of selected MEs for the IAQ monitoring in this study

Name of microenvironment (abbreviation)

Volume (m3)

Ventilation type

Monitoring duration (hours)

Occupancy levelb

Ventilation flow rate (cfm/person)c minimum–maximum (average)

Chemical lab (M1)

226.6

Mechanical/naturala

24

3–4

184–245 (214)

Instrumentation Lab (M2)

226.6

Mechanical

24

2–3

180–271 (225)

Microbiology lab (M3)

161.8

Mechanical

24

2–3

179–268 (224)

Computational lab (M4)

189.0

Mechanical

24

5–10

22–44 (33)

Admin office (M5)

291.37

Mechanical

24

8–15

14–26 (20)

Scientist room (M6)

189.0

Mechanical

24

3–8

48–127 (87)

Reception (M7)

269.5

Natural

8

2–3

301–452 (377)

Kitchen (M8)

105.0

Natural

24

2–4

135–270 (203)

Canteen (M9)

220.5

Natural

8

20–25

21–27 (24)

aAC is switched off for most of the time and windows remained open during experiments

bRange of the maximum number of people present at one time during working hours

cRepresents the fresh outdoor air available to each person inside the respective microenvironments—the details of cfm estimation method are available in SI Section S1. ASHRAE standards recommend the minimum required ventilation rate at breathing zone for office building and reception area as 5 cfm/person, computational and science lab as 10 cfm/person, and kitchen area as 75 cfm/person (ASHRAE 2003)

IAQ monitoring

IAQ monitoring was carried out between the month of July and August 2012. PM10, PM2.5, and PM1.0 measurements were made in each of the selected ME at a sampling rate of 1 min using environmental dust monitors (GRIMM make, Model 1.107). These PM monitors work on the principle of light scattering by drawing air with multiple particle sizes at a sample flow rate of 1.2 L/min through a flat-laser beam produced by a laser diode. This is capable of measuring particle mass concentrations in the 1–6,500 μg/m3 range. A 15-channel pulse height analyzer for size classification detects the scattering signals in the 0.3–25-μm size range. These counts from each precisely sized pulse channel are then converted to mass using well-established conversion equations (http://www.dustmonitor.com/Occupational/1107.htm). Two dust monitors were used for the measurements, and the sampling duration for each of the selected ME is mentioned in Table 1. One of the PM monitor was without the weather casing which was used for sequential measurements of indoor PM. The other was equipped with the weather housing, which was used for outdoor monitoring, for disregarding any possible effect of varying outdoor temperature and RH on particles. Concurrent sampling of both the indoor and outdoor PM could not be performed continuously for 24 h, because of the safety constraints raised by the adjacent slum area. Outdoor air quality monitoring was therefore carried out at the start and end of monitoring periods during the 0930–1730 hours (local time) and their average has been used to develop the I/O relationship during the period of occupancy (Fig. 2). Temperature, relative humidity (RH), pressure, and CO2 were also measured simultaneously with the indoor PM monitoring at a sampling rate of 1 min using an IAQ monitor (Testo make, Model X35). Occupancy levels were manually recorded at the time of monitoring in each of the MEs which are noted in Table 1.
Fig. 2

Outdoor PM concentration profile showing the number of exceedances over the NAAQS permissible limits

Data analysis

Exploratory analysis of PM10, PM2.5, and PM1.0 on an hourly, 8-hourly, and 24-hourly average basis was carried out together with the meteorological parameters to understand the influence of temporally varying environmental conditions on the PM concentrations. Indoor CO2 concentrations are monitored in all the selected MEs as a surrogate indicator of occupancy and ventilation conditions. Higher occupancy in an indoor space increases the CO2 concentration, resulting in reduced fresh air intake of occupants (measured as cubic feet per minute (cfm) per person) depending on the volume of the indoor space available. Detailed description of the methodology used to estimate cfm is presented in Supplementary Material (SI) Section S1 and the summary of results is provided in Table 1. I/O relationships of PM10, PM2.5, and PM1.0 have been calculated for all the selected MEs to understand the contribution from various indoor activities and the infiltration of particles from the outdoor environment. Fractional analysis of various PM sizes has also been carried out to understand the contribution of different PMs from various indoor and outdoor sources.

The SPSS package-16 has been used for performing the statistical analysis. This involved correlation and regression analysis between the pollutants and environmental parameters to understand their relationships. Pearson correlations are computed for understanding the significance of relationship between the hourly average PM concentrations and values of environmental and occupancy parameters. Difference in the mean concentrations of PM10, PM2.5, and PM1.0 has been computed using the t test.

Results and discussion

Exploratory data analysis

Figure 3 shows mean hourly concentrations of PM10, PM2.5, and PM1.0 as well as the environmental (temperature and RH) and occupancy (indoor CO2 concentrations) parameters for all the nine MEs. Irrespective of any ME, concentration levels of all the PM types are much higher during the working hours (0930–1730 hours; local time) compared with non-working hours. By looking at the different PM types separately, results show that the PM10 and PM2.5 concentrations were about 6–10 and 1.5–2 times higher during working hours than those during the non-working hours in all the MEs. Exception to the above observations were the cases of M6 and M8, where PM10 was only 2–3 times higher during working hours compared with non-working hours; no significant variation in PM2.5 and PM1.0 concentrations between the two periods were observed. Large variation in PM10 concentrations can be attributed to their aerodynamic properties (e.g., higher deposition and resuspension rates) compared with PM2.5 and PM1.0 during the presence of occupants in working hours and absence in non-working hours (Thatcher and Layton 1995; Raunemaa et al. 1989). Our results are in line with the findings of a study by Blondeau et al. (2005) for eight French schools. They concluded that occupancy is the dominant source of PM10 and their activities may lead to large variations in indoor PM10 concentrations.
Fig. 3

Diurnal variation of PM, indoor CO2 concentrations, and environmental parameters for the studied MEs. Note that the x-axis of the figures in rows 1 and 2 shows hourly PM concentrations over the period of 24 h, starting from 0900–1000 hours (previous day) to 0800–0900 hours (next day)

Inter-comparison of all the MEs indicates the maximum PM10, PM2.5, and PM1.0 concentrations in M9, M8, and M7, respectively. The highest PM10 concentrations in M9 can be attributed to larger number of occupants and their activities compared with any other ME. Further, the highest PM2.5 concentration in M8 can be due to cooking and frying activities in the kitchen (M8). In the case of PM1.0, highest concentrations were in the reception area (M7) which can be attributed to the outdoor infiltration as the large windows were mostly open during working hours. Cigarette smoking also takes place outdoors near the windows of M7 by the office staff, which may also contribute to the penetration of outdoor PM1.0 into reception area. Among the different laboratories (M1 to M4), the highest concentrations of all the PM10, PM2.5, and PM1.0 were observed in M1 during both working as well as non-working hours; this was despite the fact that there was not a great deal of difference in occupancy levels in these MEs (see Table 1). This day time increase appears to be due to the opening of windows that face towards the railway track and slum area, allowing outdoor particulates to penetrate in M1 to increase the concentration levels. During non-working hours, the windows of M1 were closed. This restricts both the entry and exit of PM in and out of the M1 to maintain the relatively high PM concentrations during these hours.

The CO2 concentration analysis for all the MEs shows that the concentration in laboratory MEs (M1 to M3), reception (M7) and kitchen (M8) varies from 350 to 450 ppm during day time and 250 to 300 ppm during night time. However, in the case of M4, which is a computational lab, the CO2 concentration during the day increases up to 760 ppm at the end of working hours (1700–1800 hours) and remains nearly unchanged for the next 4–5 h, even after the end of laboratory use (Fig. 3). These levels then started going down to the ambient levels to attain ∼300 ppm at about 0400 hours, indicating the effect of air tightness in the laboratory (Table 1).

Indoor PM10 and CO2 concentration profiles follow a similar shape during the day and night times in the labs (M1 to M3) and office (M5 and M6). This suggests a clear relationship between indoor PM10 and CO2 concentrations, because of their release from the common sources (i.e., occupants’ activities and exhalation). However, such a trend of PM10 and CO2 concentrations were not seen in the cases of M4, M8, and M9. For instance, the factor contributing to the higher CO2 concentration in the case of M4, compared with the other MEs, was the air tightness and poor ventilation conditions. Further, more intense, though discontinuous, activities of occupants in M8 and M9 led to large but short-term variations in PM10 concentration, which are not followed by increases in indoor CO2 concentrations (Goyal and Khare 2009).

SI Table S1 shows the 8-hourly average values of different PM types. A comparison of PM10, PM2.5, and PM1.0 concentrations in different lab and office MEs (M1 to M6) indicates the highest concentrations in M1 (159 ± 32.7, 41.8 ± 4.7, 26.8 ± 2.7 μg/m3, respectively) and the lowest in M4 (43.7 ± 12.9, 27.0 ± 4.0, 19.1 ± 1.9 μg/m3, respectively). All the MEs from M1 to M6 are air conditioned except M1 (see “Site description” Section). The 8-h average outdoor PM10, PM2.5, and PM1.0 concentrations were 116 ± 32.07, 60.7 ± 12.2, and 50.06 ± 11.6 μg/m3, respectively. The I/O ratio of PM10, PM2.5, and PM1.0 in M1 as 1.93, 1.16, and 0.94, respectively, suggests the outdoor infiltration contributing to all the PM types. The results are in line with the study conducted by Hopke and Martinac (1998). They concluded that indoor PM concentration in naturally ventilated buildings will be higher compared with conditioned buildings, if located in a high outdoor air pollution area. Our results support these findings since concentration of all PM types in naturally ventilated environments (M7, M8, and M9) were higher than those in air-conditioned MEs, due to the higher penetration of outdoor pollution and the presence of indoor sources. Overall, the highest 8-hourly average concentrations of PM10, PM2.5, and PM1.0 were measured in M9 (256.9 ± 194.7, 71.6 ± 17.6, 43.6 ± 6.4 μg/m3, respectively) and M8 (249.3 ± 65.8, 77.6 ± 27.4, 50.2 ± 22.1 μg/m3, respectively). The following two factors can explain these differences: outdoor infiltration which was common to all these three naturally ventilated MEs, and the indoor sources. PM10 was highest in M9 due to maximum number of occupants and their activities (Table 1). On the other hand, cooking and frying activities in M8 contributed to indoor PM2.5 and PM1.0 concentrations. The highest standard deviation (SD; ±194.7) from PM10 mean was observed in M9, indicating large variations in activities of occupants. The M9 was fully occupied (20–25 persons) between 0900 and 1000 hours (breakfast time), 1300 and 1600 hours (lunch time), and 1600 and 1700 hours (tea time)—this resulted in the generation of the highest PM10 concentration and consequently, the largest SD. In case of PM2.5 and PM1.0, the highest SD (±27.4 and ±22.1 μg/m3, respectively) was found in M8, indicating the variation in the intensity of emissions from sources (e.g., cooking and frying) at different hours (breakfast, lunch, and tea) that have led to variations in their concentrations. Similarly, smoking activity takes place in M7 and M9 that have caused the higher SD in PM1.0 values.

SI Table S.2 shows the 24-h average values of PM10, PM2.5, and PM1.0. There are no IAQ standards or guidelines yet available for indoor PM concentrations in India. Comparison with other standards shows that the 24-h average PM10 concentrations in M8 (198.3 ± 68.87 μg/m3) and M9 (204.8 ± 155.5 μg/m3) were found to be exceeding the US Environmental Protection Agency (USEPA) (150 μg/m3) and WHO (50 μg/m3) guidance values. The mean PM10 concentrations in M1 (83.99 ± 62.83 μg/m3), M5 (55.5 ± 28.9 μg/m3) and M7 (91.2 ± 22.8 μg/m3) were below the USEPA standards, but exceeded the WHO guideline values by up to a factor of two. The case was identical for 24-h average PM2.5 concentrations in M8 (60.95 ± 13.1 μg/m3) and M9 (59.2 ± 19.1 μg/m3), where these exceeded the WHO guideline value of 25 μg/m3. However, these are well within the USEPA standards (65 μg/m3) for 24-h average PM2.5 exposure. In the cases of M1, M5, M6, and M7, 24-h average PM2.5 concentrations also exceeded the WHO guidelines. No such comparisons can be made for PM1.0 due to the unavailability of standards and guidelines.

The 24-h average ambient outdoor PM10 and PM2.5 concentrations were found to be 115.76 ± 32.80 and 60.54 ± 11.6 μg/m3, respectively; these exceed the permissible limits of 100 and 60 μg/m3 of the National Ambient Air Quality Standards (NAAQS), India. These observations indicate that the studied building is located in an area having reasonably high outdoor particulate pollution. The PM emissions are generally caused by multiple sources such as biomass burning in nearby slum area (NSR 2010; Kulshreshtha et al. 2008), resuspension of dust due to large commercial vehicles on adjacent road to the building (NSR 2010), and exhausts from diesel engine trains (USEPA 2002) passing through the railway line located at the backside of studied building.

I/O relationship

Figure 2 shows the outdoor concentrations of various PM fractions, along with the permissible limits of NAAQS for PM10 and PM2.5. The hourly average concentrations have been used to calculate the I/O for different PM types in the studied MEs during working hours (Fig. 4). In case of the laboratories (M2–M4; except M1) and offices (M5–M6), the I/O for various PM types was below 1.0. Conversely, this was greater than 1 in the cases of M7, M8, and M9 for all PM types, clearly indicating the influence of higher outdoor air flow rate (cfm/person; Table 1) on the I/O since this allowed more particles to enter the naturally ventilated MEs. However, in case of canteen (M9), where occupancy is much higher than other naturally ventilated MEs, influence of outdoor airflow rate on I/O of PM was found to be overshadowed by the presence of indoor cooking sources and occupants activities.
Fig. 4

I/O ratio of PM10, PM2.5, and PM1.0 in different MEs

Detailed inspection of the individual PM types suggests that the I/O for PM10 was found to be varying from 0.37 to 3.1 in different MEs. The minimum (0.37) and maximum (3.1) I/O were for M3 and M9, respectively (see Table 2). These are presumably due to the combined influence of ventilation types and indoor sources. Further, I/O for M7 and M8 is 1.42 and 2.30, respectively. In M3, occupant’s entry is restricted and it is an air-conditioned ME. Therefore, contribution from outdoor infiltration as well as from indoor activities is the lowest. In case of M7, M8, and M9, highest I/O (1.6, 3.02, and 3.1) were due to the presence of intense emissions from indoor sources as well as higher infiltration from outdoors through open windows, doors, and ventilators (Goyal and Khare 2009). The results of numerous other studies on I/O relationship also showed them varying from 0.5 to 2.0 in different indoor environments in the absence and presence of indoor sources, respectively (Morawska et al. 2001; Hussein et al. 2005). Of course occupants activities is another important factor that results in the resuspension of coarser particles, as is also reported by previous studies (Gomes et al. 2007; Hu et al. 2007).
Table 2

I/O relationships of different PM types in studied MEs

MEs

I/O PM10

I/O PM2.5

I/O PM1.0

Mean

SD

Range

Mean

SD

Range

Mean

SD

Range

M1

1.93

1.14

1.5–2.4

1.16

0.63

1.05–1.44

0.94

0.37

0.76–1.23

M2

1.1

0.79

0.9–1.08

0.74

1.05

0.68–0.86

0.63

1.05

0.56–0.78

M3

0.62

0.71

0.61–0.71

0.75

1.27

0.8–1.1

0.64

0.98

0.75–0.92

M4

0.53

0.45

0.5–0.62

0.75

0.54

0.73–0.94

0.67

0.26

0.56–0.90

M5

0.99

0.73

0.83–1.07

0.81

0.61

0.78–0.92

0.66

0.43

0.61–0.78

M6

0.71

0.31

0.49–0.85

0.81

0.52

0.76–0.99

0.7

0.55

0.67–0.83

M7

1.6

1.2

1.3–1.44

1.65

2.73

0.13–2.78

1.51

1.86

1.4–1.57

M8

3.02

2.3

2.5–3.4

2.15

3.7

1.92–2.5

1.33

3.07

1.36–2.07

M9

3.1

6.8

2.89–5.65

1.98

2.3

2.25–2.5

1.47

0.89

1.44–1.87

The I/O for PM2.5 and PM1.0 varied from 0.2 to 3.2 and 0.17 to 2.9 in different MEs, respectively. Similar to the I/O for the PM10, the 24-h average minimum I/O (0.49 and 0.57) is found in M2 and M3 for both the PM2.5 and PM1.0. The maximum I/O for PM2.5 (1.9) and PM1.0 (1.51) is found in M9 and M7, respectively. These observations suggest that the outdoor infiltration and the indoor sources are responsible for the higher I/O in all the naturally ventilated MEs (M1, M7, M8 and M9).

The overall assessment of the I/O ratio of different PM types indicates that the variations in I/O are highest in case of PM10 compared with PM2.5 and PM1.0. Such a variation is expected given the more prominent settling and resuspension of PM10 compared with PM2.5 and PM1 (Thatcher and Layton 1995). In case of PM2.5 and PM1.0, the outdoor infiltration and building penetration factor may play a more significant role in their I/O ratios, depending on the ventilation type. A study by Dockery and Spengler (1981) on the I/O relationship of PM2.5 indicated that the mean infiltration rate of outdoor fine particulates is ∼70 % in case of naturally ventilated buildings and only ∼35–40 % in the case of fully air-conditioned buildings. In line with the previous findings (Kulmala et al. 1999), the key parameters, which are believed to control the I/O ratio of PM in our case, are the air exchange rate between the indoor and the outdoor air and the particle resuspension and settling.

Proportion of PM fractions in various MEs

Figure 5 shows the proportion of different PM fractions in all the selected indoor MEs and during outdoor measurements. Together, the PM2.5 (47 %) and PM1.0 (37 %) contributes ∼84 % of the total PM10 concentration in outdoor ambient air, leaving only 16 % for PM2.5–10. The higher fraction of smaller particles in the outdoor environment also indicates the dominance of contributions from biomass burning and fossil fuel combustion in road vehicles (Kumar et al. 2013b; NSR 2010). By looking at the indoor concentrations separately in different MEs, M7 shows nearly identical proportions of different PM fractions as were outdoors. The sum of PM2.5 and PM1.0 contributed ∼82 % of total PM as opposed to ∼84 % in outdoors—this can be expected given the frequent opening of doors/windows, allowing free movement of outdoor air into M7. In the cases of M1, M2, M8, and M9, PM2.5–10 contributes up to 56 % of the total PM10. This indicates that the main source of PM2.5–10 indoors are human activities such as walking and cleaning that lead to resuspension of previously deposited larger-sized particles. For instance, Almeida et al. (2011) and Majumdar et al. (2011) have carried out studies on school classrooms and found that concentration of PM2.5–10 increased by 50–100 % in the classrooms due to physical activities of students, resulting in resuspension of particles deposited on classroom floors. Likewise, Gomes et al. (2007) have carried out an experimental chamber study to simulate the influence of occupant’s walking on particle resuspension at various floor types. They found that aerodynamic disturbances dominate the particle resuspension behavior over the dust type, dust load, and floor types; i.e., the forces working on different size of the particles is most important over the other factors, which may lead to their resuspension. Furthermore, a review by Hu et al. (2007) on particle resuspension concluded that mechanical, aerodynamic, and electrostatic forces from human activity can lead to 100 % resuspension of particles in indoor environments.
Fig. 5

Proportion of PM concentrations in various size ranges in all the MEs and outdoor environment

PM1.0 and PM2.5 contribute more than or equal to 90 % of the total PM10 concentration in air-conditioned MEs (M3, M4, and M6). These three MEs have comparatively restricted physical activities of occupants, except the computational and printing activities that may contribute to fine particles (Horemans and Van Grieken 2010). The doors remain closed most of the time in these MEs due to the operation of ACs fitted with filters, suggesting that the particles, which enter from outdoor air through ACs, will keep on accumulating in these MEs. A recent review by Chen and Zhao (2011) reported that the room occupancy level influences the concentrations of different sizes of particle in indoor environments. The occupants also have an effect on transport of particulates by controlling the ventilation system and/or opening the windows/doors, and their activities may result in particle generation, or resuspension of previously deposit particles (Chen and Zhao 2011). In case of M3, M4, and M6, such activities are restricted that led to high concentration of PM2.5 and PM1.0 as oppose to PM10.

Relationships between PM types and environmental/occupancy parameters

Correlation analysis has been performed to understand the significance of relationship of different PM types with the environmental (temperature and RH) and occupancy parameters (CO2 concentration). The two–tailed Pearson’s correlation matrix has been drawn and significance of correlations coefficients is tested for two significant levels, i.e., 99 and 95 % confidence intervals (see SI Table S.3).

A positive correlation exists between CO2 and indoor PM concentrations; this correlation is more systematic and somewhat clearer for PM10, as shown in SI Table S.3. This indicates that sources of PM10 may be mostly the occupants and their activities (i.e., walking, cleaning, particle resuspension). These observations are in line with the findings of previous studies concluding that the more intense occupant’s activities result in, the higher the concentration of both CO2 and PM10 in indoor settings (Goyal and Khare 2009; Blondeau et al. 2005).

Correlations were established between the indoor temperatures and the PM types (SI Table S.3). Most of the mechanical-ventilated MEs showed negative correlations for temperature as opposed to positive correlations seen in the case of natural-ventilated MEs. Generally, when indoor temperature is high, particles tend to remain dry and hence contribute less towards increasing their sizes and mass concentrations (Massey et al. 2012). However, the variation in temperature during the experiments was modest (see Fig. 3) so the effect of temperature on PM type is hard to distinguish. Similar remains the case for relative humidity due to its small variations during the experiments (SI Table S.3).

Paired sample t test has been performed for comparing the means of different sizes of PM concentration in different MEs. Unlike other MEs, the hourly concentration of PM10 in M1, M7, and M9 does not show the significant difference in their means at 95 % confidence interval (p > 0.05 and t < 2.0; SI Table S.4). Likewise, in case of PM2.5, besides M1 and M7, no significant difference in the means of hourly concentrations are found (p > 0.05 and t < 2.0) in the rest of the MEs, presumably due to varying usage and occupants activities (SI Table S.5). There is, however, a significant difference (p < 0.05) between the means of hourly PM1.0 concentrations for all the MEs, except M7 and M9 (both affected by smoking), suggesting a common source for all of them such as the infiltrating from outdoors (see SI Table S.6).

Summary and conclusions

The IAQ monitoring was carried out inside a mix-use commercial building environment. As expected, the indoor concentration of PM10 during working hours (0930–1730 hours) were found to be 6–10 times higher than those during non-working hours (remaining hours) in all the MEs, except M6 and M8, where the differences were only 2–3 times. Indoor PM2.5 and PM1.0 concentrations during working hours were found to be ∼1.5–2 times higher than those during the non-working hours in most of the MEs. Indoor PM10 and CO2 concentration changed their values in tandem in the laboratories (M1 to M3) and offices (M5 and M6). However, this trend was not evident in MEs (M4, M8, and M9) influenced by other factors. For instance, poor ventilation conditions in M4 caused the higher CO2 concentrations. Intense but discontinuous activities of occupants in M8 and M9 were found to be responsible to generate variations in PM10 concentration which were not followed by the indoor CO2 concentrations.

The 24-h average data analysis of both outdoor and naturally ventilated indoor PM10 and PM2.5 concentrations indicated the violation of permissible limits for respective environments. However, their indoor concentrations in air-conditioned MEs were within the prescribed limits of USEPA. No standards values are available for PM1.0 concentration for comparison purposes. The results clearly suggests that if a building is located in a mix-use polluted area (commercial, industrial, slum, and transportation activities), the natural-ventilated MEs are likely to experience higher infiltration of PM pollution compared with air-conditioned, mechanical-ventilated, MEs. This is also reflected by the I/O relationships for all PM types. This was consistently less than 1 for mechanically ventilated MEs but much higher than 1 for naturally ventilated MEs (Fig. 4). The highest variations in the I/O relationship was found for the PM10 (0.37 to 3.1) and PM2.5 (0.2 to 3.2)—this was mainly due to the higher occupant activities leading to resuspension of coarse particles and the presence of indoor sources such as combustion and printing activities producing fine particles.

The proportion of different size fractions shows that PM2.5 and PM1.0 concentrations in outdoor air contribute to ∼84 % of the total PM10 concentrations—similar proportions were found in the reception (M7) area. However, these fractions change dramatically in some of the mechanically ventilated MEs (i.e., M3, M4, and M6) where the sum of both PM1.0 and PM2.5 contribute over 90 % of the total PM10 concentration—these were the MEs with ACs on and doors closed most of the time that did not allow the fine particles to escape once entered. Moreover, indoor sources such as computational and printing activities in these MEs exacerbated the levels of fine particles. Conversely, the MEs (M1, M2, M8, and M9) with greater physical activity of occupants resulted in the resuspension of previously deposited dust and showed a larger fraction (≥56 %) of coarse particles (PM2.5–10).

The statistical analysis of the data indicated a good correlation (r > 0.6) between indoor PM10, PM2.5, and CO2 concentrations and that the occupants and their activities are common indoor sources. The PM1.0 and CO2 were found to be poorly correlated, suggesting the dominance of outdoor infiltration on this relationship. Generally, indoor temperature and RH showed negative and positive correlations, respectively, with all the PM types. This relationship could not be verified due to small variations in the values of temperature and the RH. Comparison of the results of the paired sample t test shows that the means of indoor PM10 concentrations in different MEs are significantly different to those in outdoor environment. This strengthens our conclusion that occupants’ movement is important for determining the PM10 concentrations. The means of indoor and outdoor PM2.5 and PM1.0 are comparable, especially in all mechanically ventilated MEs, substantiating our observations that ventilation type is important to determine the indoor concentrations of finer-sized particles.

The study also has its explainable limitations. For instance, monitoring period was limited to one particular season due to practical constraints related to site access. Further data would have helped us to evaluate the influence of seasonal changes on the I/O concentrations of PM. However, given the uniqueness of the studied building in terms of its mixed use and location, the present study make a useful addition to the existing literature, in particular for a megacity like Delhi, where such measurements are yet under-represented.

The results of our study also have two important implications: one for the exposure assessment, and the other for future building design in megacities. Firstly, the derived I/O relationship provides important information for making exposure estimates and developing efficient control strategies to reduce health risks in mix-use, complex, building MEs such as those often forming part of non-domestic buildings in the polluted megacities. Secondly, it would be more appropriate to avoid natural ventilation, and use filter-fitted ACs in buildings that are situated in locations with significant outdoor PM pollution. If the latter option is not practically feasible, the use of recirculating air cleaners could be considered for decreasing PM levels and hence the associated exposure.

Notes

Acknowledgments

The authors would like to thank the Director of CSIR- National Environmental Engineering Research Institute (NEERI) for providing support for present research. They would also like to thank the Environmental Engineering Laboratory of Indian Institute of Technology, Delhi, for providing the instrumental support and Ms. Papiya Mandal, Scientist, NEERI, Delhi Zonal Laboratory for providing the help during the experimental campaigns to carry out the monitoring work.

Supplementary material

11869_2013_212_MOESM1_ESM.pdf (88 kb)
ESM 1(PDF 87 kb)

References

  1. Almeida SM, Canha N, Silva A, Freitas MC, Pegas P, Alves C, Evtyugina M, Adrião C (2011) Children exposure to atmospheric particles in indoor of Lisbon primary schools. Atmos Environ 45:7594–7599CrossRefGoogle Scholar
  2. ASHRAE (2003). ASHRAE standards - Ventilation for acceptable indoor air quality. American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc. ISSN 1041–2336Google Scholar
  3. Baek S, Kim YS, Perry R (1997) Indoor air quality in homes, offices, and restaurants in Korean urban areas indoor/outdoor relationships. Atmos Environ 31:529–544CrossRefGoogle Scholar
  4. Blondeau P, Lordache V, Popuard O, Genin D, Allard F (2005) Relationship between outdoor and indoor air quality in eight French schools. Indoor Air 15:2–12CrossRefGoogle Scholar
  5. Chaloulakou A, Mavroidis I, Douci A (2003) Indoor and outdoor carbon monoxide measurements at different microenvironments in the Athens area. Chemosphere 52(6):1007–1019CrossRefGoogle Scholar
  6. Chen C, Zhao B (2011) Review of relationship between indoor and outdoor particles: I/O ratio, infiltration factor and penetration factor. Atmos Environ 45:275–288CrossRefGoogle Scholar
  7. Chen C, Zhao B, Weschler CJ (2012) Indoor exposure to “outdoor PM10”: assessing its influence on the relationship between PM10 and short-term mortality in US cities". Epidemiology 23:870–878CrossRefGoogle Scholar
  8. Colbeck I, Nasir ZA, Ali Z (2010) Characteristics of indoor/outdoor particulate pollution in urban and rural residential environment of Pakistan. Indoor Air 20:40–51CrossRefGoogle Scholar
  9. Dockery DW, Spengler JD (1981) Indoor–outdoor relationships of respirable sulfates and particles. Atmos Environ 15:335–343CrossRefGoogle Scholar
  10. Gomes C, Freihaut J, Bahnfleth W (2007) Resuspension of allergen-containing particles under mechanical and aerodynamic disturbances from human walking. Atmos Environ 41(25):5257–5270CrossRefGoogle Scholar
  11. Goyal R, Khare M (2009) Indoor–outdoor concentrations of RSPM in classroom of a naturally ventilated school building near an urban traffic roadway. Atmos Environ 43:6026–6038CrossRefGoogle Scholar
  12. Goyal R, Khare M, Kumar P (2012) Indoor air quality: current status, missing links and future road map for India. J Civil Environ Eng 2:118. doi:10.4172/2165-784X.1000118
  13. Habil M, Massey D, Taneja A (2013) Exposure of children studying in schools of India to PM levels and metal contamination: sources and their identification. Air Qual Atmos Health 6:575–587CrossRefGoogle Scholar
  14. Heal MR, Kumar P, Harrison RM (2012) Particles, air quality, policy and health. Chem Soc Rev 41:6606–6630. doi:10.1039/C2CS35076A CrossRefGoogle Scholar
  15. Hopke P, Martinac I (1998) Indoor climate and air quality. Int J Biometeorol 42:1–7CrossRefGoogle Scholar
  16. Horemans B, Van Grieken R (2010) Speciation and diurnal variation of thoracic, fine thoracic and sub-micrometer airborne particulate matter at naturally ventilated office environments. Atmos Environ 44:1497–1505CrossRefGoogle Scholar
  17. Hu B, Freihaut J, Bahnfleth W, Aumpansub P, Brandolyn T (2007) Modeling the influence of human activity on particle resuspension and dispersion in a multi-zone indoor environment with HVAC system. ASCE J Archit Eng 13:4–10Google Scholar
  18. Hussein T, Hameri K, Heikkinen MSA, Kulmala M (2005) Indoor and outdoor particle size characterization at a family house in Espoo-Finland. Atmos Environ 39:3697–3709CrossRefGoogle Scholar
  19. Hussein T, Glytsos T, Ondráček J, Dohányosová P, Ždímal V, Hämeri K, Lazaridis M, Smolík J, Kulmala M (2006) Particle size characterization and emission rates during indoor activities in a house. Atmos Environ 40:4285–4307Google Scholar
  20. Jones NC, Thornton CA, Mark D, Harrison RM (2000) Indoor/outdoor relationships of particulate matter in domestic homes with roadside, urban and rural locations. Atmos Environ 34:2603–2612CrossRefGoogle Scholar
  21. Kleeman MJ, Schauer JJ, Cass GR (1999) Size and composition distribution of fine particulate matter emitted from wood burning, meat charbroiling, and cigarettes. Environ Sci Technol 33:3516–3523CrossRefGoogle Scholar
  22. Kukadia V, Palmer J (1998) The effect of external atmospheric pollution on indoor air quality: a pilot study. Energy Build 27(3):223–230CrossRefGoogle Scholar
  23. Kulmala M, Asmi A, Pirjola L (1999) Indoor air aerosol model: the effect of outdoor air, filtration and ventilation on indoor concentrations. Atmos Environ 33:2133–2144CrossRefGoogle Scholar
  24. Kulshreshtha P, Khare M, Seetharaman P (2008) Indoor air quality assessment in and around urban slums of Delhi city, India. Indoor Air 18:488–498CrossRefGoogle Scholar
  25. Kumar P, Morawska L (2013) Energy-pollution nexus for urban buildings. Environ Sci Technol 47:7591–7592CrossRefGoogle Scholar
  26. Kumar P, Pirjola L, Ketzel M, Harrison RM (2013a) Nanoparticle emissions from 11 non-vehicle exhaust sources—a review. Atmos Environ 67:252–277CrossRefGoogle Scholar
  27. Kumar P, Jain S, Gurjar BR, Sharma P, Khare M, Morawska L, Britter R (2013b) New directions: can a “blue sky” return to Indian megacities? Atmos Environ 71:198–201CrossRefGoogle Scholar
  28. Luoma M, Batterman SA (2001) Characterization of particulate emissions from occupant activities in offices. Indoor Air 11:35–48CrossRefGoogle Scholar
  29. Majumdar D, Gajghate DG, Pipalatkar P, Rao CVC (2011) Assessment of airborne fine particulate matter and particle size distribution in settled chalk dust during writing and dusting exercises in a classroom. Indoor Built Environ 000:1–11. doi:10.1177/1420326X11419691 Google Scholar
  30. Massey D, Masih J, Kulshrestha A, Habil M, Taneja A (2009) Indoor/outdoor relationship of fine particles less than 2.5 mm (PM2.5) in residential homes locations in central Indian region. Build Environ 44:2037–2045CrossRefGoogle Scholar
  31. Massey D, Kulshrestha A, Masih J, Taneja A (2012) Seasonal trends of PM10, PM5.0, PM2.5 & PM1.0 in indoor and outdoor environments of residential homes located in North-Central India. Build Environ 47:223–231Google Scholar
  32. Memarzadeh F (2009) Effect of reducing ventilation rate on indoor air quality and energy cost in laboratories. J Chem Health Saf 16:20–26CrossRefGoogle Scholar
  33. Miller SL, Nazaroff WW (2001) Environmental tobacco smoke particles in multizone indoor environments. Atmos Environ 35:2053–2067CrossRefGoogle Scholar
  34. Morawska L, He C, Hitchins J, Gilbert D, Parappukkaran S (2001) The relationship between indoor and outdoor airborne particles in the residential environment. Atmos Environ 35:3463–3473CrossRefGoogle Scholar
  35. Morawska L, He C, Hitchins J, Mengersen K, Gilbert D (2003) Characteristic of particle number and mass concentrations in residential houses in Brisbane, Australia. Atmos Environ 37:4195–4203CrossRefGoogle Scholar
  36. NSR (2010) Air quality monitoring, emission inventory and source apportionment study for Indian cities. National Science Report of Central Pollution Control Board, DelhiGoogle Scholar
  37. Ott WR, Siegmann HC (2006) Using multiple continuous fine particle monitors to characterize tobacco, incense, candle, cooking, wood burning, and vehicular sources in indoor, outdoor, and in-transit settings. Atmos Environ 40:821–843CrossRefGoogle Scholar
  38. Polichetti G, Cocco S, Spinali A, Trimarco V, Nunziata A (2009) Effects of particulate matter (PM10, PM2.5 and PM1) on the cardiovascular system. Toxicology 261:1–8CrossRefGoogle Scholar
  39. Pope CA III, Dockery DW (2006) Health effects of fine particulate air pollution: Lines that connect. J Air Waste Manag Assoc 56:707–742CrossRefGoogle Scholar
  40. Raunemaa T, Kulmala M, Saari H, Olin M, Kulmala MH (1989) Indoor air aerosol model: transport indoors and deposition of fine and coarse particles. Aerosol Sci Technol 11:11–25CrossRefGoogle Scholar
  41. Santos JM, Mavroidis I, Reis NC, Pagel EC (2011) Experimental investigation of outdoor and indoor mean concentrations and concentration fluctuations of pollutants. Atmos Environ 45:6534–6545CrossRefGoogle Scholar
  42. Thatcher TL, Layton DW (1995) Deposition, resuspension, and penetration of particles within a residence. Atmos Environ 29:1487–1497CrossRefGoogle Scholar
  43. USEPA (2002) Health assessment document for diesel engine exhaust. National Center for Environmental Assessment - Office of Research and Development Washington, DC EPA/600/8-90/057F Available from http://wwwepagov/iris. Accessed 22 March 2013
  44. Wallace L (1996) Indoor particles: a review. J Air Waste Manag Assoc 46:98–126CrossRefGoogle Scholar
  45. Yamamoto N, Shendell DG, Winer AM, Zhang J (2010) Residential air exchange rates in three major US metropolitan areas: results from the relationship among indoor, outdoor, and personal air study 1999–2001. Indoor Air 20:85–90CrossRefGoogle Scholar
  46. Zhou B, Zhao B, Guo X, Chen R, Kan H (2013) Investigating the geographical heterogeneity in PM10-mortality associations in the China Air Pollution and Health Effects Study (CAPES): a potential role of indoor exposure to PM10 of outdoor origin. Atmos Environ 75:217–223CrossRefGoogle Scholar

Copyright information

© Springer Science+Business Media Dordrecht 2013

Authors and Affiliations

  1. 1.CSIR-NEERI, Delhi Zonal LaboratoryNaraina Industrial AreaDelhiIndia
  2. 2.Department of Civil and Environmental Engineering, Faculty of Engineering and Physical Sciences (FEPS)University of SurreyGuildfordUnited Kingdom
  3. 3.Environmental Flow (EnFlo) Research Centre, FEPSUniversity of SurreyGuildfordUnited Kingdom

Personalised recommendations