Crucial Role for Outdoor Chemistry in Ultrafine ... - ACS Publications

Aug 24, 2015 - Crucial Role for Outdoor Chemistry in Ultrafine Particle Formation in. Modern Office Buildings. Nicola Carslaw,*,†. Mike Ashmore,. â€...
6 downloads 12 Views 800KB Size
Article pubs.acs.org/est

Crucial Role for Outdoor Chemistry in Ultrafine Particle Formation in Modern Office Buildings Nicola Carslaw,*,† Mike Ashmore,‡ Andrew C. Terry,† and David C. Carslaw§ †

Environment Department, University of York, York YO10 5DD, United Kingdom Stockholm Environment Institute, University of York, Heslington, York YO10 5DD, United Kingdom § Department of Chemistry, University of York, York YO10 5DD, United Kingdom ‡

ABSTRACT: In the developed world, we spend most of our time indoors, where we receive the majority of our exposure to air pollution. This paper reports model simulations of PM2.5 and ozone concentrations in identical landscape offices in three European cities: Athens, Helsinki, and Milan. We compare concentrations during an intense heatwave in August 2003 with a meteorologically more typical August in 2009. During the heatwave, average indoor ozone concentrations during office hours were 44, 19, and 41 ppb in Athens, Helsinki, and Milan respectively, enhanced by 7, 4, and 17 ppb respectively relative to 2009. Total predicted PM2.5 concentrations were 13.5, 3.6, and 17.2 μg m−3 in Athens, Helsinki, and Milan respectively, enhanced by 0.5, 0.4, and 6.7 μg m−3 respectively relative to 2009: the three cities were affected to differing extents by the heatwave. A significant portion of the indoor PM2.5 derived from gas-phase chemistry outdoors, producing 2.5, 0.8, and 4.8 μg m−3 of the total concentrations in Athens, Helsinki, and Milan, respectively. Despite filtering office inlet supplies to remove outdoor particles, gas-phase precursors for particles can still enter offices, where conditions are ripe for new particles to form, particularly where biogenic emissions are important outdoors. This result has important implications for indoor air quality, particularly given the current trend for green walls on buildings, which will provide a potential source of biogenic emissions near to air inlet systems.



INTRODUCTION There has been significant attention paid to the role of air pollution in causing adverse health effects in recent times. The International Agency for Research on Cancer (the IARC, part of the WHO) recently classified diesel exhaust as a carcinogen1 and Public Health England estimated that in the UK alone, 29 000 excess deaths arise each year from exposure to particulate air pollution.2 Further, the EU is taking the UK (and other member states) to court for failing to meet air quality guidelines.3 One issue that makes headlines less frequently is indoor air quality (IAQ), despite the fact that people in developed countries spend ∼90% of their time indoors, at home, at work, or commuting between the two. Many pollutants exist in the indoor environment through indoor emissions linked to human activities such as cleaning. In addition, outdoor air pollutants ingress through ventilation via open windows and doors, as well as through the building envelope. Consequently, concentrations of air pollutants indoors are often higher than outdoors, though legislation for exposure to air pollutants focuses on outdoor concentrations. For those who spend most of their time indoors, the current regulatory framework may provide inadequate protection from air pollutant exposure.4 Since the introduction of energy efficiency measures in the 1970s, particularly in office buildings, adverse health effects have been reported indoors frequently.5 These include a range © 2015 American Chemical Society

of building-related symptoms (BRS) such as eye, nose, throat and airway irritation, headaches, and fatigue, which have unknown etiology, but typically improve away from the workplace.6 Fiedler et al. (2005)7 reported that between 800 000 and 1.2 million buildings in the US may be associated with such symptoms, potentially affecting between 30 and 70 million workers and leading to significant economic loss. Although it is still unclear exactly what causes these symptoms, terpene oxidation products are one area of active research in this regard.8,9 It is likely, however, that many different factors contribute to BRS, with dampness/flooding, temperature, particles, and psychosocial factors potentially playing a role.5 Ozone is known to initiate chemical processing indoors (e.g., through reaction with terpenes indoors), to form a number of reaction products including oxygenated and nitrated carbon compounds and secondary organic aerosol (SOA).10−13 As SOA consists of ultrafine particles, UFPs, there are implications for health both in terms of mortality and morbidity.14 UFPs can act as a surface for potentially toxic species to sorb onto and their small size allows them (and the species sorbed upon them) to penetrate deep into the lungs.15 For a 10 μg m−3 Received: Revised: Accepted: Published: 11011

May 5, August August August

2015 20, 2015 24, 2015 24, 2015 DOI: 10.1021/acs.est.5b02241 Environ. Sci. Technol. 2015, 49, 11011−11018

Article

Environmental Science & Technology concentration, 2 μm particles have a surface area of 24 μm2 mL−1 compared to 3016 μm2 mL−1 for 20 nm diameter particles: smaller particles also have a longer lifetime so potentially lead to higher exposures.14 Although there is evidence emerging that the components of particles rather than their mass concentrations are more important for health effects (e.g., Rohr and Wyzga, 201216), a recent UK Government review concluded that the evidence was not yet strong enough to suggest a systematic reduction of one component of particulate matter over others.17 Terry et al. (2014)18 investigated the impact on the exposure of cleaners and office workers to such oxidation products indoors during a European heatwave event in August 2003. Many parts of Europe were affected by the 2003 heatwave, which resulted in record-breaking temperatures in France, Portugal, and Italy19 and 35 000 excess deaths across Europe.20,21 Temperatures reached their highest values for many years in Italy, with the northwest affected most strongly: in Milan, temperatures reached almost 40 °C and mortality increased by 21.1% for all ages (30.6% for 75 years and older) compared with the same period in 2002.19 The maximum apparent temperature (which is an index of human discomfort based on air temperature and dew point temperature) is estimated to have been 4.4 °C higher in the summer of 2003 when compared to a reference period from 1998 to 2002, the highest increase observed in Italy over this period.22 The measured ozone concentrations were also high over Europe during the heatwave period, with central Europe experiencing the worst of the pollution. In the UK, it has been estimated that between 21 and 38% of the 2045 excess deaths were caused by the high levels of air pollution rather than the heat.21 Such heatwaves are expected to become more frequent and/or intense with further climate change.23 Given that high outdoor ozone and PM concentrations lead to higher indoor concentrations, understanding the impacts of such events on office air quality is extremely important given the number of people who work in such environments around the world. For instance, Terry et al. (2014)18 found that in terms of exposure to secondary pollutants, cleaning mechanically ventilated offices first thing in the morning was better for cleaners (when outdoor and hence indoor ozone was lower), but afternoon cleaning was better for office workers, given that the secondary pollutants formed through the use of cleaning products had mostly dispersed by the following morning when they returned to the office. In the current study, we have used data from the 2003 European heatwave to initialize a detailed chemical model for indoor air chemistry to investigate the role of SOA indoors in detail. In particular, we investigate the importance of indoor and outdoor sources of PM2.5 in offices and the sensitivity of this relationship to outdoor biogenic volatile organic carbon (VOC) concentrations and particle filtration efficiency.

product data where available, or structure activity relationships in their absence.25 It considers chemistry initiated by key oxidants such as OH (hydroxyl) radicals, O3 (ozone), and NO3 (nitrate) radicals, and degradation of the subsequent reaction products. Unlike other models that describe indoor air chemistry,12,29,30 the detailed chemistry in the INDCM permits much greater insight into the chemical processing indoors. The model also contains 41 gas-to-particle conversions for limonene products, which was selected owing to its ubiquity indoors and also its propensity for forming secondary organic aerosol (SOA). Consequently, estimates of SOA formation through chemical reaction indoors will be lower limits if species other than limonene dominate particle formation. The absorptive partitioning theory of Pankow (1994)31 was used to represent the phase-partitioning of limonene oxidation products between gas and particles indoors. It was assumed that 30% of the outdoor particulate concentration was organic in nature,30 providing a surface for the indoor SOA to absorb onto once it ingresses indoors. The resulting box-model assumes a single well-mixed environment and the concentration of each species is calculated according to eq 1: Q dC i ⎛A⎞ = −νd⎜ ⎟C i + λrfCo − λrC i + i + ⎝ ⎠ dt V V

n

∑ R ij j=1

(1)

where Ci (Co) is the indoor (outdoor) concentration of a species; Vd is the deposition velocity of a species; A is the surface area of a room; V is the volume of a room; λr is the air exchange rate (AER) between indoors and outdoors (note that we assume the air moves once through the building and that there is no recirculation); f is the outdoor-to-indoor penetration factor (assumed to be equal to 1 for all gas-phase species and varied for PM2.5 as described later); Qi is the indoor emission rate for species i and Rij is the reaction rate between species i and j. The INDCM contains approximately 20 000 reactions that describe VOC emission rates, gas-phase chemical reactions, deposition loss to surfaces, exchange with outdoors and gas-toparticle partitioning for limonene oxidation products.10,24 Note that the deposition velocities used for O3 and PM2.5 are 0.0345 and 0.004 cm s−1 respectively10,30 and that the emission rate of limonene provides an indoor concentration of ∼2 ppb during the daytime based on measurements in the literature.32 Only one surface production rate is considered in the model for nitrous acid (HONO), where 2.9 × 10−3 m min−1 is used.24 Outdoor Concentration Data. IAQ was considered for identical offices in Athens, Helsinki, and Milan during the August 2003 heatwave period and for the same days in 2009, a more “typical” summer. Milan was in the center of the affected area, Athens was on the southern periphery, whereas Helsinki was not affected.18 Athens is a densely populated city with severe air pollution issues, exacerbated by poor city planning and the surrounding mountainous topography that prevents dispersion of air pollutants emitted within the city.33 Helsinki lies on a flat coastal plain adjacent to the Baltic. Traffic emissions dominate the air pollutant concentrations, with some contribution from long-range transport to local PM concentrations.33 Milan is the most densely populated and polluted city in Northern Italy: high levels of pollution are caused by a combination of emissions from traffic and domestic heating and poor dispersion.34 These three cities therefore represent very different locations in Europe.



MATERIALS AND METHODS Model. The model used in this work has been described in detail previously.24,10 Briefly, a detailed INdoor air Detailed Chemical box Model (INDCM) based on a comprehensive chemical mechanism (the Master Chemical Mechanism, MCM v3.2, http://mcm.leeds.ac.uk/MCM/) has been tailored for use in typical office environments. The MCM is a near explicit representation of the gas-phase degradation of 143 VOCs and has been described elsewhere.25−28 The MCM has been compiled using a strict protocol, which employs kinetic and 11012

DOI: 10.1021/acs.est.5b02241 Environ. Sci. Technol. 2015, 49, 11011−11018

Article

Environmental Science & Technology

midsummer were used as the basis for the model runs. Concentrations were either input directly into the model or smoothed using a polynomial fit as described in Carslaw (2007).24 Table 2 summarizes the mean outdoor volatile organic carbon (VOC) concentrations used in the model runs (for both

Outdoor concentration data for NOX (sum of NO (nitric oxide) and NO2 (nitrogen dioxide) concentrations) O 3 (ozone) and PM2.5 (particulate matter with an aerodynamic diameter of 2.5 μm or less) were extracted from the EU AirBase data set35 for the same 10 days in August of the two contrasting summers of 2003 and 2009. The 10 days were selected to represent the highest ambient concentrations of O3 in each location. PM2.5 data were unavailable for Athens, so were calculated using the relationship with PM10, based on data from five cities where both were available.18 The selected sites were classified as urban background, distant from busy roads and street canyons and typical of locations in these cities where most people work (Table 1).

Table 2. Mean Outdoor VOC Concentrations (μg m−3) for the Three Cities, Based on Passive Measurements Between 10:00 and 16:00 h over 5 days Outside 3−5 Office Buildings Per City during Summer 2012

Table 1. Summary Details of the Airbase Urban Background Monitoring Stations in Three European Cities from which the Mean Diurnal Profiles of Outdoor Pollutant Concentrations Were Generated for the Model Simulations station code

city

longitude

latitude

average August temperature (°C)a

FI00425 IT466A GR0039A

Helsinki Milan Athens

25.0°E 9.2°E 23.8°E

60.2°N 45.4°N 38.0°N

27.8 21.2 15.0

population (M)b 0.61 1.32 0.66

a

Data represent 43, 30, and 21 years of data for Athens, Milan, and Helsinki, respectively.30 bData represent population within city limits rather than wider metropolitan areas.31

Figure 1 illustrates the mean 24 h diurnal summer profiles of O3, NO, NO2, and PM2.5 concentrations outdoors at each of the three monitoring stations in 2003 and 2009, compiled by averaging the hourly concentrations over the same 10 days in August for each year. The highest daytime ozone concentrations in 2003 were recorded in Milan, although nighttime concentrations were higher in Athens. Helsinki did not experience the 2003 heatwave, and outdoor concentrations were much lower than those of the other two cities in both years (Figure 1). The mean outdoor concentration profiles for the two ten-day periods in

VOC

Athens

Milan

Helsinki

2-butoxyethanol acetaldehyde acrolein α-pinene benzaldehyde benzene ethylbenzene formaldehyde hexanal limonene n-hexane propionaldehyde styrene tetrachloroethylene toluene trichloroethylene xylenes

0.55 2.6 2.7 1.8 1.1 0.74 1.5 3.1 3.4 4.7 0.37 2.0 0.46 0.05 5.0 0.05 3.8

1.0 5.2 4.6 1.6 1.1 0.77 0.60 4.6 4.1 4.3 1.3 2.0 0.98 0.23 3.1 0.02 2.3

0.74 0.02 0.01 0.69 0.01 0.52 0.48 0.03 0.02 0.85 1.4 0.01 0.37 0.04 1.9 0.04 1.4

years), based on passive measurements made during an “OFFICAIR” measurement campaign of 3−5 buildings per city over 5 days in summer 2012 (Ioannis Sakellaris, University of Western Macedonia, Greece, personnel communication). The OFFICAIR Project was funded by the EU as part of the seventh framework program, and used an integrated approach (combining measurements, laboratory experiments and modeling) to evaluate the health risk to office workers from exposure to indoor air pollutants in modern office buildings across

Figure 1. Mean diurnal profiles of outdoor concentrations of NO, NO2, O3, and PM2.5 for 10 days in August for the two contrasting summers of 2003 (heatwave) and 2009 (average) in Athens, Helsinki, and Milan. Note that the concentrations of NO, NO2, and O3 are shown in units of ppb and the concentrations of PM2.5 in μg m−3. 11013

DOI: 10.1021/acs.est.5b02241 Environ. Sci. Technol. 2015, 49, 11011−11018

Article

Environmental Science & Technology Europe (http://www.officair-project.eu/). Other VOC concentrations used in the model were initialized as reported by Carslaw (2007).24 In reality, the concentrations of some of the VOCs reported in Table 2 are likely to have been much higher in August 2003 and we return to this issue in model sensitivity tests reported later on. Office Characteristics. We assumed the same characteristics for mechanically ventilated offices in each of the cities. Indoor temperature and relative humidity were assumed to be 300 K and 45% respectively, based on the mean values measured in an office in Athens during the OFFICAIR project (Ioannis Sakellaris, University of Western Macedonia, Greece, personnel communication). There were no measurements of indoor visible and UV radiation. Therefore, we used attenuated outdoor radiation values assuming a cloudless sky and transmission of 0.1 in the visible and 0.03 in the UV and assumed indoor lights were on from 09:00−17:00h.24 On the basis of modal values for landscape offices from the OFFICAIR project, we assumed a floor area of 169 m2, a total surface area of 645 m2, and volume of 467 m3. Such offices had a mean occupancy of 14 people (Corinne Mandin, Centre Scientifique et Technique du Bâtiment, Paris, France, personal communication). Mechanical ventilation with a mean value of 2.3 (range 0.4−5.0) ach−1 was used from 07:00−19:00 h, with a mean rate of 1.0 (range 0.1−1.8) ach−1 overnight when no mechanical ventilation operated.38 We assumed that outdoor particles were removed to some extent by filtration of the outdoor air supply. The efficiency of filters varies widely. There are a variety of MERV (minimum efficiency reporting value) ratings available, different models that reportedly have the same efficiency vary by manufacturer and loaded filters are more efficient than clean ones.39,40 There are also differences in removal efficiencies for different sized particles. Ultrafine particles (UFP) tend to be removed more efficiently than PM2.5: UFP are removed at efficiencies of 12.3% for a MERV 5 filter compared with 1.4% for PM2.5, though for higher efficiency filters, removal rates are similar.39 Riley et al. (2002)40 report that for an ASHRAE 85% filter, the removal efficiency for PM2.5 was 64% for typical urban air in office buildings, compared to 8% for an ASHRAE 40% filter. PM2.5 removal efficiency varied from ∼8−85% for a range of commonly used filters (MERV 8 to 13) and by ∼8−45% for a range of MERV 8 filters.40 Zhao et al. (2007)41 note that some filters can also remove ozone. The efficiency of this process varies, with clean filters removing very little ozone but loaded filters removing somewhat more.41 In this paper, we focus on particle removal only and assume for our baseline run that the removal efficiency of particles is 60% with no simultaneous loss of gaseous species (so f in eq 1 for PM2.5 is 0.4), but will test a range of potential values through sensitivity tests in the Discussion Section. The emission rates from office equipment (Table 3) were assumed to be continuous during office working hours (09:00− 17:00 h) and zero outside these times. The values for O3, PM2.5 and HCHO, key species for focus in OFFICAIR, are taken from Destaillats et al. (2008).43 Emissions from printers and photocopiers are dependent on usage (copy rate, occupancy). The average number of units (copies per working day) per landscape office derived from OFFICAIR data was assumed to be 1.5 (290) for laser printers, 0.1 (20) for inkjet printers and 0.65 (500) for photocopiers (so 1 inkjet printer was present in the office and working for 10% of the time). We assumed one PC for each occupant.

Table 3. Summary of Assumed Emission Rates Per Landscape Office (mg min−1) for O3, PM2.5, and HCHO, Per Photocopier, Printer, and Personal Computer (PC) emission rate (mg min−1) source

mode

photocopier laser inkjet

8.2 × 10−2 6.7 × 10−4 2.1 × 10−5

photocopier laser inkjet

8.9 × 10−4 1.1 × 10−4 5.2 × 10−6

photocopier PC

9.1 × 10−4 1.6 × 10−4

O3

PM2.5

minimum

maximum

3.3 × 10−3 2.3 × 10−6 0.0

0.8 8.7 × 10−3 4.0 × 10−4

3.6 × 10−5 3.8 × 10−7 0.0

8.9 × 10−3 1.5 × 10−3 9.8 × 10−5

3.6 × 10−5 1.14 × 10−5

9.1 × 10−3 3.6 × 10−4

HCHO

Running the Model. The model is run for 3 days to make sure that steady-state is achieved and that the dynamic profile remains consistent: results for the third day are then used for analysis. For each city, the model is first run in “outdoor mode” to assess the ambient concentrations of the gas-phase precursors (GPPs) for limonene-oxidation SOA species, given the reported ambient limonene concentration outside each of the offices.12 These concentrations then inform the indoor air model.



RESULTS Figure 2 shows the resulting modeled indoor concentrations of NO2, NO, and O3 for each of the three cities for the 2 years. Not surprisingly, the indoor concentrations of all species are higher in the Milan and Athens offices, reflecting the much higher concentrations outdoors. In 2003 (2009), indoor O3 concentrations averaged 44, 19, and 41 ppb (37, 15, and 24 ppb) from 09:00−17:00 h in Athens, Helsinki, and Milan, respectively. Clearly, the location of an office is a key determinant of the IAQ experienced by its occupants as the offices are identical except for the outdoor concentrations. In addition, the ambient concentrations have a large impact on indoor concentrations in all cities, but Milan in particular, experiencing enhanced concentrations indoors in 2003 compared to 2009. Note that there is a more pronounced increase in indoor O3 in the Athens office with the morning change in ventilation rate (Figure 2): there is less diurnal variation in ambient O3 compared to the other cities (Figure 1). Figure 3 shows the indoor diurnal PM2.5 concentrations for Milan for the August 2003 heatwave period. The predicted indoor PM2.5 concentration is in good agreement with a mean indoor value of PM2.5 of 18 μg m−3 from measurements made in the warm season in 2007 in offices in Milan.34 As well as the total PM2.5 concentration, Figure 3 shows the contribution of PM2.5 that has ingressed directly from outdoors through air exchange and that formed through indoor sources. The latter sources include direct emissions from office equipment (e.g., printers, photocopiers and computers) as well as a small amount of SOA production from the background concentrations of limonene (∼2 ppb) that exist in the offices (e.g., from fragrance used by office workers and cleaning products). Around 10−14 μg m−3 of the PM2.5 comes directly from outdoors during office hours, though the contribution from indoor sources over this time is very small (