Primary emissions of fine carbonaceous particles in Europe

ARTICLE IN PRESS

Atmospheric Environment 41 (2007) 2156–2170 www.elsevier.com/locate/atmosenv

Primary emissions of fine carbonaceous particles in Europe Kaarle Kupiainena,b,, Zbigniew Klimonta a

IIASA—International Institute for Applied Systems Analysis, Schlossplatz 1, A-2361 Laxenburg, Austria b Finnish Environment Institute (SYKE), P.O. Box 140 FI-00251 Helsinki, Finland Received 24 May 2006; received in revised form 27 October 2006; accepted 31 October 2006

Abstract The European emissions of BC and OC in fine particles are calculated for the years 1990, 1995 and 2000 applying the RAINS model that, beyond fuel-sector distinction, explicitly includes various combustion technologies and the penetration of abatement options. The emission factors used are developed considering specific European conditions. The main sources of carbonaceous aerosols in Europe are emissions from traffic and residential combustion of solid fuels. Between 1990 and 2000, the BC and OC emissions are estimated to decline from 0.89 to 0.68 Tg and from 1.4 to 1.0 Tg, respectively. Most of the reduction occurred in the early 1990s in Eastern Europe owing to structural changes that resulted in energy efficiency improvements in industry and lower consumption of solid fuels in residential–commercial sector; the latter having strong impact on BC and OC emissions. Furthermore, the growth in transport volumes, and expected increase in emissions, was offset by introduction of stricter legislation for road transport from 1995. Focusing on the most important sectors, transport and residential combustion, the variation in measured carbonaceous emission shares and its impact on total emissions was evaluated. This analysis indicates a range of about 25% to +20% for BC and 7% and +15% for OC, compared to the central case. r 2006 Elsevier Ltd. All rights reserved. Keywords: Black and organic carbon; Emission modelling; Combustion; Anthropogenic sources

1. Introduction Primary carbonaceous particles are emitted mainly from combustion processes. They are classified into organic and black carbon. Organic carbon (OC) comprises the numerous organic compounds and black carbon (BC) refers to the ‘black’, light-absorbing aerosol that composes Corresponding author. Finnish Environment (SYKE), P.O. Box 140 FI-00251 Helsinki, Finland. Tel.: +358 9 40300384; fax: +358 9 40300390. E-mail address: kaarle.kupiainen@ymparisto.fi (K. Kupiainen).

Institute

1352-2310/$ - see front matter r 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.atmosenv.2006.10.066

mainly from elemental carbon and is commonly known as soot. Both BC and OC are most abundant in the accumulation mode (Gray and Cass, 1998), which also implies that they have rather long atmospheric residence times. Carbonaceous particles have been linked with adverse health effects as well as climatic impacts. According to Jacobson (2002), BC could be the second most important direct-forcing component after CO2. Considering that it also has a relatively shorter lifetime in the atmosphere than the greenhouse gases controls of BC emissions could provide a faster way to combat global warming in the short-term (Hansen et al., 2000; Jacobson, 2002;

ARTICLE IN PRESS K. Kupiainen, Z. Klimont / Atmospheric Environment 41 (2007) 2156–2170

Bond and Sun, 2005). However, a more profound quantitative understanding on the role of absorbing aerosols in the global climatic system and climate change is needed to formulate reliable policy recommendations. This applies also to emission estimates of carbonaceous particles where recent studies indicate much lower global emission strengths compared to the earlier assessments (Bond et al., 2004) while underlining the importance of their uncertainties. Although European emissions have been dealt with in the global emission inventories of carbonaceous particles (Penner et al., 1993; Cooke and Wilson, 1996; Cooke et al., 1999; Bond et al., 2004) and Schaap et al. (2004) estimated European BC emissions for 1995 applying BC fractions in PM1 developed within a study for China (Streets et al., 2001), a detailed approach considering countryspecific characteristics within the European domain is missing. This paper presents estimates of the primary fine BC and OC emissions in Europe for 1990, 1995 and 2000. For that purpose, the RAINS model for particulate matter (Klimont et al., 2002) was extended to calculate carbonaceous particles (Kupiainen and Klimont, 2004). 2. Method The sectoral structure for calculation of emissions remains the same as in the current implementation of the RAINS model (Klimont et al., 2002); also available from the Internet (www.iiasa.ac.at/rains). The RAINS methodology includes three steps: (i) region (i), sector (j) and fuel (k) specific ‘‘raw gas’’ emission factors (ef) for total suspended particles (TSP) are derived, (ii) ‘‘raw gas’’ emission factors for each of the size fractions and chemical species (y) are estimated considering source specific profiles, and (iii) PM emissions (E) are calculated for the distinguished size fractions and chemical species taking into account size, chemical species’ and abatement (m) specific removal efficiencies (eff) as well as penetration of control technologies (X)1 (Eq. (1)). If no emission controls are applied, the abatement efficiency equals zero (eff ¼ 0) and the application rate is one (X ¼ 1). In that case the emission calculation is reduced to a simple multiplication of the activity rate (A) by the ‘‘raw 1 Actual implementation rate of the considered abatement, e.g., percent of total coal used in power plants that are equipped with electrostatic precipitators.

2157

gas’’ emission factor (ef). X E i;j;k;m;y E i;y ¼ j;k;m

¼

X

Ai;j;k ef i;j;k;y ð1  eff m;y ÞX i;j;k;m .

ð1Þ

j;k;m

For mobile sources, shares of BC and OC in measured PM (various size fractions or total carbon) were reviewed and further recalculated into shares of TSP. Although, BC and OC shares are related to TSP, only BC and OC in fine particles are considered. A mass balance check was performed that the sum of BC and total organic matter (OM) is not greater than RAINS PM2.5. OM was estimated as OC times a factor of 1.3–1.7 based on the discussion in Turpin et al. (2000). The resulting absolute emission factors (EFs) for BC and OC were compared with the EFs available in the literature. Similarly, for stationary sources burning biomass, liquid, gaseous, and other solid fuels, shares of carbonaceous particles were used to derive EFs. Since BC and OC EFs are largely independent from ash content of fuel, the EFs for coal combustion were derived directly from measurement data. Developing EFs we have reviewed results of measurements performed with all types of devices, however, we attempted to rely on results of thermal–optical methods, when available. The variation due to differences in thermal protocols (Turpin et al., 2000; Chow et al., 2001; Watson and Chow, 2002) used is included in the analysis in Section 4.1. A major source of variation in emission characteristics is the variation in combustion conditions due to differences in fuel characteristics, designs of the combustion devices, and operation practice. Although, it is difficult to take all of these parameters into account in a large-scale model like RAINS, we have made use of knowledge on design and typical operation practices of particular devices (e.g. stoves) in various countries in Europe when developing EFs for this work. The penetration of control technology is directly included in the calculation and therefore specific removal efficiencies for carbonaceous species needed to be developed considering changing size and chemical speciation of controlled mobile and stationary sources. Very few studies reported removal efficiencies of carbonaceous particles for individual control technologies. Therefore, they had to be derived based on the analysis of emission characteristics for varying level of abatement and prescribed standards.

ARTICLE IN PRESS 2158

K. Kupiainen, Z. Klimont / Atmospheric Environment 41 (2007) 2156–2170

The activity data are derived from international energy and production statistics (International Energy Agency (IEA), 2005a, b; Eurostat, 2003; UN, 2003), national sources and consultations carried out within the clean air for Europe programme (CAFE) (EC, 2001) where national experts reviewed RAINS activity and technological databases. The latter led to a number of very important improvements, e.g. filling in gaps with respect to non-commercial fuel consumption (biomass) in residential sector, updating information on the technological splits and penetration of abatement measures. 3. Emission factors This section discusses development of BC and OC emission factors for the RAINS model. Significantly more attention is paid to key sources of BC and OC, i.e. residential combustion of solid fuels and mobile sources. For other sources that contribute small amounts of BC and OC, details of EFs used are not shown here but can be downloaded with the background documentation where detailed discussion and original sets of EFs are available (Kupiainen and Klimont, 2004 available from http://www.iiasa.ac.at/rains/reports/ir-04-079.pdf). 3.1. Large-scale stationary combustion of solid fuels Advanced technology applications like pulverized coal (PC) combustion and fluidized bed combustion (FBC), have low emissions of elemental and OC because of the high combustion temperatures, oxidizing conditions and long residence times (Ohlstro¨m et al., 2000). Furthermore, vast majority of plants are equipped with high-efficiency PM abatement devices. Typically reported carbonaceous shares are below 1% in fine PM resulting in very small EFs for BC and OC. In case of smaller industrial installations, e.g. grate boilers, the emissions may include more carbonaceous particles due to less efficient combustion. The EFs used in this study for the latter category and review of underlying measurements is shown in Tables 1 and 2, accordingly. 3.2. Stationary combustion of liquid and gaseous fuels Particulate emissions from burning of liquid fuels depend, similarly to solid fuels, on combustion

conditions and fuel type. Generally, emissions of PM from these sources are small. For heavy fuel oil, particulate emissions are influenced by the ash and sulfur content of the oil. Ash and carbon-rich cenospheric particles are coarse and may account for a major part of the mass emitted, specifically in small boilers. However, owing to their relatively short atmospheric lifetime, coarse particles are not of primary interest for this analysis. In larger boilers combustion is more efficient resulting in smaller carbon shares although almost all emitted particles are in fine or submicron mode. Particles that are formed during combustion of light fuel oil are mainly carbonaceous, but may contain inorganic components, e.g. sulfates. The ash content is, however, much smaller than that of heavy fuel oil. PM emissions from natural gas combustion are very low, primarily influenced by the quality of combustion and consist mainly of unburned OC found in submicron mode. The estimated BC and OC EFs from fuel oil and natural gas combustion are shown in Table 3. 3.3. Residential combustion Small-scale combustion of solid fuels takes place in relatively cool, poorly mixed conditions and carbonaceous species, especially OC, dominate PM emissions. In higher temperatures and more efficient combustion, the share of BC rises and becomes dominant (Rau, 1989). In RAINS, this sector is represented by a number of source categories, i.e., fireplaces, stoves, single house boilers and medium sized boilers. Table 1 provides background on measurement data and Table 2 shows EFs used in the model. 3.3.1. Coal The primary factors influencing PM emissions from residential coal combustion are coal’s volatile matter content and composition as well as design of the combustion equipment. The volatile matter of brown coal (lignite) includes lighter hydrocarbons than hard coal (Bond et al., 2004) and therefore less BC is formed (Table 2). In larger installations, combustion conditions are expected to be more stable and efficient than in stoves resulting in much lower emissions (Table 2). Since there were no direct measurements available for combustion of coal in single house boilers, the high temperature phase measurements from stoves reported by Pinto et al. (1998) and Watson et al. (2001) were used to derive

ARTICLE IN PRESS K. Kupiainen, Z. Klimont / Atmospheric Environment 41 (2007) 2156–2170

2159

Table 1 BC and OC shares (% in PM2.5) and emission factors (mg MJ1) from literature for stationary combustion sources Shares

Emission factors

Literature

BC

OC

BC

OC

Industrial combustion Biomass Grate firing

0–8

5–19

—

—

Brown coal Hard coal

o1a o1a

o3–7a 0–3a

— —

— —

Heavy fuel oil Light fuel oil

0.2–13 29

0.8–9 5

— —

— —

Residential combustion Biomass Fireplaces

1–8b; 5–36c

38–52

5–50b; 15–186c

160–390

5–14; 16–38d

47–67; 14–20d

39–43

70–390

Hard coal Stoves

2–26

70

—

—

Watson et al. (2001), Engelbrecht et al. (2002)

Brown coal Stoves Heavy fuel oil Gas

2–6 — 7

62–68 — 75e

— 1 0.007–0.02

— 0.4 0.03-0.1

Pinto et al. (1998) Bond et al. (2004) Hildemann et al. (1991)

Stoves

US EPA (2002), Wierzbicka et al. (2005) Pinto et al. (1998), Bond et al. (1999) Fisher et al. (1978, 1979), Henry and Knapp (1980), Griest and Tomkins (1984), Olmez et al. (1988), Ge et al. (2001), Watson et al. (2001) Olmez et al. (1988), US EPA (2002) Hildemann et al. (1991)

Butcher and Sorenson (1979), Muhlbaier Dasch (1982), Muhlbaier and Williams (1982), Rau (1989), Hildemann et al. (1991), Watson et al. (2001), Fine et al. (2002) Butcher and Sorenson (1979), Cooper (1980), Rau (1989), Watson et al. (2001), Environment Australia (2002), US EPA (2002)

a

Shares in abated emissions of mostly pulverized combustion of coal. Hardwoods. c Softwoods. d Hot burning phase results measured by Rau (1989) in parenthesis. e Scaled down so that BC+OM comprises 97% of the total mass. b

Table 2 Emission factors for combustion of coal and biomass (mg MJ1) Combustion technique

Fireplace Stove Boiler o50 kWth (manual feed) Boiler o1 MWth (manual feed) Boiler o50 MWth (automatic feed) Pellet boiler o50 MWth (automatic feed) Industrial grate firing a

Control technology

Uncontrolled Old New Old New Uncontrolleda Uncontrolleda Uncontrolleda Uncontrolleda

Biomass

Brown coal

Hard coal

BC

OC

BC

OC

BC

OC

75–100 75–105 56–79 75 75 35 7.5 0.83 9.6

375–500 225–315 11–16 113 113 25 8 0.83 14

— 24 24 29 26 2.3 2.3 — 4.5

— 312 312 300 120 5.0 5.0 — 18

— 130 130 215 194 4.0 4.0 — 6.0

— 200 200 175 70 2.0 2.0 — 3.0

For installations equipped with low efficiency PM abatement, e.g., cyclones, EFs are reduced by 10%; in case high efficiency abatement is present EFs will be well below 1 mg MJ1.

ARTICLE IN PRESS 2160

K. Kupiainen, Z. Klimont / Atmospheric Environment 41 (2007) 2156–2170

Table 3 Emission factors for combustion of liquid fuels and gas (mg MJ1) Source sector

Industry and power plants Small residential boilers

Control technology

Uncontrolled Regular maintenance Uncontrolled Regular maintenance

EFs. Results from Bond et al. (1999) and Ge et al. (2001) were used to estimate emission characteristics of medium size boilers. 3.3.2. Wood and biomass Major factor affecting the carbonaceous emission profile from biomass burning is combustion temperature that in turn depends on air supply, burn cycle, stove design, burn rate, as well as species and moisture content of the fuel. Several studies reported lower shares of BC and higher shares of OC in emissions from hardwoods (e.g. poplar, white ash) than softwoods (e.g. pine) when burned in similar appliances. Emission factors for BC and OC from wood burning in fireplaces (Table 2) were derived based on studies referred to in Table 1, assuming that the shares of hardwood and softwood used are 70/30 (not verified with data). Comparison of RAINS EFs with the EFs reported in reviewed studies shows a good agreement for BC while for OC the derived values are higher but within the reported range in e.g. Hildemann et al. (1991) and Fine et al. (2002). For stoves, a similar procedure as for fireplaces was followed to develop EFs for this work, assuming equal use of hardwood and softwood. Rau (1989) has shown for wood stoves that hot burning vs. cool burning results in a smaller share of total carbon in emitted PM as well as in a higher share of BC in total carbon. Since there were no direct measurements available for single house boilers, the high temperature phase measurements from stoves reported by Rau (1989) for wood and biomass were used to derive EFs for single house boilers (Table 2). For medium size boilers a distinction is made between manually and automatically operated installations. Since no measurements of carbonaceous species were found, for manually operated boilers a profile similar to the single house boilers was used assuming that due to higher combustion

Light fuel oil

Heavy fuel oil

BC

OC

BC

OC

110–180 100–170 530 500

20–30 18–20 130 120

670–840 640–800 6800 6500

290–370 260–330 380 340

efficiency the BC share will be higher and OC lower than for stoves. The automatically operated boilers are assumed to be similar to grate boilers used in industrial applications for which the SPECIATE 3.2 (US EPA, 2002) emission profiles were used. The derived emission factors are presented in Table 2. 3.4. Other sources Measurements of carbonaceous emissions from industrial processes are scarce and were found for few sources. However, based on the knowledge about different stages of the process and the raw materials used, an attempt was made to estimate their share in emitted particles. For several categories in iron and steel industry, non-ferrous metal industry, carbon black, and cement and lime production EFs were derived from the US EPA SPECIATE 3.2 database (US EPA, 2002), Engelbrecht et al. (2002) and Olmez et al. (1988). In general industrial processes are not expected to be a major source of BC and OC. There are several other miscellaneous sources that emit carbonaceous particles, such as burning of agricultural residues, cigarette smoking, and barbeques, which are included in the model. The calculation details can be found from Kupiainen and Klimont (2004). Although they might have a significant contribution locally, their contribution to overall European emissions of BC and OC is small. 3.5. Mobile sources Exhausts are the main contributor of carbonaceous particles from mobile sources. Non-exhaust sources (tire wear, brake lining and road wear) emit some carbonaceous PM, but their contribution to total is small. Exhaust particles are mostly submicron agglomerates of small (10–80 nm in diameter) carbonaceous particles. The exhaust emission sources have been further divided according to the

ARTICLE IN PRESS K. Kupiainen, Z. Klimont / Atmospheric Environment 41 (2007) 2156–2170

fuel and engine type, i.e. diesel and gasoline vehicles. Particulate emissions from diesel engines are in general higher than the emissions from spark ignition engines. Diesel engines have been found emitting usually a larger BC fraction (BC/OC41) than gasoline engines (BC/OCo1), although there is considerable variation between individual vehicles (Gillies and Gertler, 2000). Due to introduction of after-treatment of the exhaust gases the emission rates are lower for the newer diesel vehicles (Table 5). Table 5 shows ranges of EFs that represent variability in fleet composition (age, size) and driving patterns across Europe. 3.5.1. On-road diesel vehicles For vehicles without emission controls (per-1990 engines without oxidation catalyst) the reviewed studies show large variability in carbonaceous emission factors (Table 4). This large spread can be explained by the variety of vehicle types tested, driving conditions, engine maintenance level, and load observed in the studies. A commonly observed feature is that the BC/OC ratios were typically larger than one, which is also reflected in the RAINS EFs (Table 5).

2161

Only limited data was available on emissions of carbonaceous particles from controlled vehicles. The reported values for heavy duty vehicles (Table 4) were assumed to represent best EURO II vehicles. Emission factors for EURO I heavy duty vehicles were assumed to be approximately twice for both BC and OC (Table 5). Lambrecht and Ho¨pfner (2000) indicated that diesel light duty vehicles equipped with advanced emission control technology have overall lower emissions of PM but a higher share of BC. The same study showed that emissions of organics decline with the introduction of oxidation catalysts, which is confirmed by Kerminen et al. (1997). 3.5.2. Spark ignition engines Although diesel engines are characterized by significantly higher unit emissions of PM than spark-ignition engines, the vehicle kilometers traveled by vehicles equipped with the latter can exceed those of diesel vehicles. The situation varies between countries but the contribution of gasoline-powered vehicles to carbonaceous PM must not be forgotten. Emissions from unleaded-non-catalyst fourstroke gasoline vehicles were assessed based on

Table 4 BC and OC shares (% in PM2.5) and emission factors (mg MJ1) from literature for on road mobile sources Shares

Emission factors

Literature

BC

OC

BC

OC

22–54

16–45

15–49

10–69

40–60

14–35

3–6

1–4

50–70

20–30

20–100

21–49

75

15

—

—

Gasoline vehicles Light duty Leaded

5–10

27–42

0.5–1

2–10

No control

10–24

21–70

0.5–2

2–10

15–30

22–45

0.6–1.5

Diesel vehicles Heavy duty No control

Controlled (~EURO II) Light duty No control

Controlled (~EURO I/II)

Controlled (three way catalyst)

0.3–1.6

Williams et al. (1989), Hildemann et al. (1991), Lowenthal et al. (1994), Israe¨l et al. (1996), Norbeck et al. (1998b), Kirchstetter et al. (1999), Gillies and Gertler (2000) Norbeck et al. (1998b), Schauer et al. (1999), Shi et al. (2000) Williams et al. (1989), Watson et al. (1990), Norbeck et al. (1998a), Cadle et al. (1999), Lambrecht and Ho¨pfner (2000), US EPA (2002) Lambrecht and Ho¨pfner (2000)

Williams et al. (1989), Hildemann et al. (1991), Gillies and Gertler (2000) Sagebiel et al. (1997), Norbeck et al. (1998a, c), Cadle et al. (1999) Sagebiel et al. (1997), Norbeck et al. (1998a), Kirchstetter et al. (1999), Cadle et al. (1999)

ARTICLE IN PRESS 2162

K. Kupiainen, Z. Klimont / Atmospheric Environment 41 (2007) 2156–2170

Table 5 Emission factorsa for on-road mobile sources (mg MJ1) Source

Control technology

Diesel

Gasoline

BC

OC

BC

OC

Heavy duty vehicles

Uncontrolled EURO I EURO II

24–34 16–22 6–9

15–21 8–11 3–4

— — —

— — —

Light duty vehicles (GVWo3500 kg)

Uncontrolled EURO I EURO II Uncontrolled, 2-stroke engine

51–77 41–62 17–25 —

20–29 8–12 3–5 —

0.99 0.91 0.91 10–20

3.3 1.62 1.62 135–270

a

Exhaust emissions only.

measurements of pre- and early 1980s vehicles (Table 4). As for diesel vehicles there was a large variation between the reported values owing to variation in vehicle types, driving conditions, and engine maintenance level. The reviewed studies showed consistently BC/OC ratios lower than one. The carbonaceous emission factors for unleaded gasoline vehicles are assessed at almost ten times lower than for diesel vehicles (Table 5). Based on studies by Norbeck et al. (1998a, c) and Cadle et al. (1999) the BC/OC ratio of the modern (post 1991) gasoline vehicles can be closer to or even larger than one and PM emissions are generally lower. Leaded gasoline increases significantly the overall PM emission factors but based on the reviewed studies it seems not to have a significant effect on carbonaceous emission factors when compared with unleaded fuel (Table 4). Durbin et al. (1998) reported that emissions from engines fueled with CNG and methanol were comparable to those of their gasoline counterparts. PM emissions from two-stroke gasoline engines are higher than from four-stroke engines due to a large proportion of unburned oil droplets (Kojima et al., 2000). A large share of the emissions can be expected to be carbonaceous. For example Smith (2000) observed a decrease of 98–99% in mass after removal of organics from PM10 samples. No studies about BC content were found, but drawing on the results of Smith (2000) the BC emissions are estimated as small. 3.5.3. Off-road transport No specific data on BC and OC emission factors or shares for off-road diesel engines was found. There are several similarities between heavy duty and off-road engines, but the typical operating

Table 6 Emission factors for off road transport (mg MJ1) Source

Heavy fuel oil

Diesel

BC

BC OC BC

OC

Off-road; land based Agriculture Construction Rail Other: four-stroke Other: two-stroke Shipping Inland Marine: medium vessels 49 Marine: large vessels 49

33 33

58 65 44 49

41 30 25 35

43 11

30 7

Gasoline OC

5 5

17 17

5 21

17 286

conditions usually vary, which affects the BC/OC ratio (Shi et al., 2000; Lambrecht and Ho¨pfner, 2000). Shi et al. (2000) measured BC and OC emitted from a diesel test engine under changing operating conditions, specifically load and rpm. The conditions that represent emissions from an off-road engine were assumed to be similar to the average of the tests with 50% load reported in that study. For rail, it was assumed that typical operating conditions would be characterized by somewhat higher rpm, leading to a different BC/OC ratio than for other off-road engines. For construction machinery, a higher average load as well as slightly higher rpm than other off-road machinery, e.g. tractors, was assumed. For marine vessels running on diesel, the same BC and OC shares were assumed as for off road machinery. For medium and large ships using fuel oil the BC/OC ratio takes into account char formation. Emission factors are summarized in Table 6.

ARTICLE IN PRESS K. Kupiainen, Z. Klimont / Atmospheric Environment 41 (2007) 2156–2170

4. Results and discussion Table 7 shows total European emissions of BC and OC for the years 1990, 1995 and 2000. Western Europe refers to the EU-15, Switzerland and Norway while Eastern Europe to other countries, including Turkey and the European part of Russia. The overall European emissions of BC and OC decline between 1990 and 2000 from 0.89 to 0.68 Tg year1 for BC and from 1.4 to 1 Tg year1 for OC. This is primarily due to developments in Eastern Europe where structural changes in the early 1990s led to significant emission reductions between 1990 and 1995. Much smaller variation in emissions was calculated for Western Europe; while BC emissions increased slightly between 1990 and 1995 they declined by 2000 following introduction of stricter legislation for road transport. The latter is well illustrated in Fig. 1 where sectoral distribution of emissions and the amount removed through introduction of abatement technologies are shown. We estimate that about 60–67% of BC emissions in Western Europe originate from transport sector (Fig. 1) while another 25–30% come from fossil fuel combustion in residential sector. Road transport alone contributes 43% in 1990 and around 50% in 1995 and 2000. The slight increase in the beginning of the studied period is associated with growing numbers of diesel vehicles (compare Fig. 2). The picture is different in Eastern Europe where residential combustion accounts for over 50% of the BC emissions (Fig. 1); its share declining steadily due to lower use of solid fuels and growth of transport sector (Fig. 2). The switch from coal to gas in residential sector is the biggest factor contributing to the sharp decline in BC and OC emissions in this region in the beginning of 1990s (Fig. 2).

2163

The fuel structure of BC emissions in Western and Eastern Europe is significantly different (Fig. 2). Higher car ownership and larger penetration of diesel vehicles in Western Europe makes diesel use in transport the most important emission source. While emissions from coal use in residential sector are one of the largest BC sources in Eastern Europe in 1990 (nearly 40%), its share declines with time and it is biomass that becomes more important in relative terms; Western European emissions from this sector are already dominated by biomass combustion. Emissions of OC in Europe show similar features to BC, however, residential combustion is even more important while share of transport is smaller (Fig. 1). Solid fuel use, especially biomass, in domestic combustion is a source of about 50% and 65% of OC emitted in Western and Eastern Europe, respectively. Most of the remaining OC originates from transport with a significant share of off-road. However, in Eastern Europe open burning of agricultural residues and other waste is nearly as important (category ‘Other’ in Fig. 2). An important feature of European BC and OC emissions in this period is continuously growing contribution from biomass combustion (Fig. 2). By 2000, it was a source of about 25% of BC and 50% of OC aerosols emitted in Europe (excluding emissions from forest fires that are estimated in Europe at about 50–70 Gg year1 BC and 700–900 Gg year1 OC; derived from Bond et al., 2004). Furthermore, relative importance of emissions from gasoline is higher for OC than for BC, which is due to a lower BC/OC-ratio in emissions from gasoline vs. diesel engines. The estimates of total emissions of BC and OC in Europe vary largely (Table 8), for example for BC between 0.5 Tg year1 (Bond et al., 2004) up to over 2.5 Tg year1 (Cooke and Wilson, 1996).

Table 7 Europeana emissions of primary anthropogenic carbonaceous particles (low and high values given in brackets) (Gg year1) Region

Western Europe Eastern Europe Total Europe Sea regions a

Black carbon (BC)

Organic carbon (OC)

1990

1995

2000

1990

1995

2000

361 (281–424) 526 (365–643) 887 (646–1068) 47

365 (280–424) 352 (249–428) 717 (529–852) 47

337 (252–391) 343 (245–412) 680 (497–803) 47

515 (479–618) 865 (820–942) 1380 (1299–1560) 31

433 (398–538) 620 (581–687) 1053 (979–1225) 31

394 (357–495) 602 (557–675) 996 (914–1170) 31

Sea regions include Atlantic Ocean, Baltic Sea, Black Sea, Mediterranean Sea, and North Sea. Low and high cases were not estimated for shipping.

ARTICLE IN PRESS K. Kupiainen, Z. Klimont / Atmospheric Environment 41 (2007) 2156–2170

2164

Western Europe

600

Eastern Europe Black carbon [Gg C]

500 400 300 200 100 0 1200

Organic carbon [Gg C]

1000 800 600 400 200 0 1990

1995

2000

1990

1995

Power plants & industrial comb.

Domestic

Road transport

Off-road transport

Industrial processes

Other

2000

Removed through technology Fig. 1. Sectoral black carbon and organic carbon emissions in Europe.

The variation can be partly explained by the different base years and different geographical coverage but the main factor is that the earlier estimates used higher emission factors (Bond et al., 2004; Schaap et al., 2004). The difference is clearly seen in a comparison made by Bond et al. (2004), where they applied Cooke et al. (1999) emission factors to their activity data (1996 fuel use) and arrived at almost three times higher emissions of BC than in their own calculation. Bond et al. (2004) stated that they often found no evidence from

measurement data to support as high emission factors as used by Cooke et al. (1999). As striking the difference is, one has to note that there are large uncertainties linked to the BC emission estimates; Bond et al. (2004) suggests they are likely to be in the order of factor two, largely due to uncertainties in emission factors. Schaap et al. (2004) calculated the European (excluding Former Soviet Union) BC emissions to be 0.47 Tg BC in fine particles, which is very close to the estimate by Bond et al. (2004) (Table 8).

ARTICLE IN PRESS K. Kupiainen, Z. Klimont / Atmospheric Environment 41 (2007) 2156–2170

Western Europe

600

2165

Eastern Europe Black carbon [Gg C]

500

400 300

200

100 0 1000

Organic carbon [Gg C]

800

600

400

200

0 1990 Coal

1995

2000

Biomass

1990 Diesel

1995 Gasoline

2000 Other

Fig. 2. Fuel structure of black carbon and organic carbon emissions in Europe.

Derwent et al. (2001) used dispersion modelling and daily air mass sector allocation methods to estimate European BC emission strength and arrived at about 0.5 Tg a1. The European BC and OC emissions calculated in this study (0.52 and 0.71 Tg, for central case in 1995, respectively) are fairly close to the central estimates by Bond et al. (2004) and Schaap et al. (2004) as well as the estimate by Derwent et al. (2001) (Table 8). However, compared to Cooke et al. (1999), our value is almost 60% lower. The difference seems to be bigger for the non-EU-15 countries since for EU15 Cooke et al. (1999) and this study agree for BC and OC within about 20%. Obviously there is a

difference in energy use between 1984 and 1995 but as stated by Bond et al. (2004) the bulk of the difference could be attributed to different set of emission factors used. Although the recent studies arrive at lower European emissions of BC, the previous estimates seem to match better with observational data (Schaap et al., 2004). This trend is also observed on the global level (Bond et al., 2004). For example, Schaap et al. (2004) compared the calculated concentrations of BC in Europe with available observations, the model showed an underestimation by a factor of two. However, both Schaap et al. (2004) and Bond et al. (2004) note that recent

ARTICLE IN PRESS 2166

K. Kupiainen, Z. Klimont / Atmospheric Environment 41 (2007) 2156–2170

Table 8 Various estimates of European (excluding FSU) emissions of BC and OC (Tg C year1) Study

BC

Penner et al. (1993)

1.97 2.14 2.65 1.74 1.26 0.55 0.47 0.47 (0.34–0.84)

Cooke and Wilson (1996) Cooke et al. (1999)

Bond et al. (2004)

1.09

OC

1.64 1.36 0.68 0.51 0.57 (0.32–1.1) 2.03

Base year

Remarks

1980

Diesel fuel and coal Additionally wood and bagasse burning Fossil fuel and some biomass sources Fossil fuel emissions in total PM Fossil fuel emissions in submicron PM EU-15, in total PM EU-15, in submicron PM Excluding open burning sources

1984 1984

1996 1996

1995

Bond et al., using Cooke et al. (1999) emission factors, excluding open burning sources Europe, excluding Former Soviet Union (FSU), emissions in PM2.5 Europe, based on Lagrangian dispersion modelling Europe, based on daily air mass source sector allocation Europe, excluding FSU and wildfires

1995

EU-15, excluding wildfires

Schaap et al. (2004)

0.47 (0.23–0.70)

1995

Derwent et al. (2001)

0.48 (0.34–0.62)

1995–1998

0.51 (0.38–0.64)

1995–1998

This study

0.52 (0.39–0.61) 0.35 (0.27–0.40)

0.71 (0.65–0.85) 0.41 (0.38–0.51)

modelling efforts represent an improvement as they rely more on measurement data. The discussed gap calls for reducing the uncertainties in estimates of carbonaceous emissions by, i.a., improvement of activity statistics and more measurements of emission sources as well as ambient levels of carbonaceous aerosols. 4.1. Analysis of variation in BC and OC emission shares As was shown in previous sections, there is often a considerable variation in observed (measured) values of BC and OC shares. Since most of the source specific measurement data available reported rather BC and OC shares than EFs we analyzed the possible impact of the variation on resulting total emissions. For the most important source sectors,2 i.e. biomass burned in small domestic appliances and on- and off-road diesel vehicles, a new set of EFs (low and high) was derived. The available measurement data were grouped by sector and abatement level and the distribution was analyzed to derive high and low emission values. The data distributions were evaluated with 2

Selected sources account for approximately 80% of total emissions.

Wilk–Shapiro and Kolmogorov–Smirnov goodnessof-fit tests. Due to the low number of cases in each population, approximately ten to 35 cases, the results of the tests can be treated only as indicative. For diesel vehicles with emission controls the data followed the normal distribution reasonably well and the 95% confidence intervals were used to represent high and low EFs ranges (Fig. 3). For woodstoves and fireplaces several factors influence emission results (see Section 3.3.2) while the reviewed studies were not designed to compile precise datasets on how, in what kind of installations and with what fuel wood exactly the combustion is done in households (Fine et al., 2001). To derive the high and low ranges the data compiled from the measurements were used as a base, but the distribution parameters were not used (Fig. 4). The BC data ranges were estimated to be 5–20% for both stoves and fireplaces. The OC ranges were estimated at 37–52% for stoves and 45–75% for fireplaces. We have performed similar evaluation for residential coal combustion and gasoline vehicles that allowed estimating low and high shares. For other sources, such analysis was not possible due to small number of measurements; however, they contribute relatively small proportion of total emissions and therefore would not affect the general conclusions.

ARTICLE IN PRESS K. Kupiainen, Z. Klimont / Atmospheric Environment 41 (2007) 2156–2170

OC

BC 9 Frequency

8 Frequency

2167

6 4 2

6 3 0

0 5

15

25

35 45 Shares

55

65

75

5

15

25

35 45 Shares

55

65

75

Fig. 3. Frequency distributions of carbonaceous particle shares for heavy-duty diesel vehicles without emission abatement.

BC

OC 3 Frequency

Frequency

8 6 4 2

2 1 0

0 0 5 10 15 20 25 30 35 40 45 50 55 60 Shares

30 35 40 45 50 55 60 65 70 75 80 85 90 Shares

Fig. 4. Frequency distributions of carbonaceous particle shares for fireplaces burning wood.

Our analysis of the impact of the variation in BC and OC EFs can explain at most about 17% and 23% higher BC and OC emissions for Western Europe and 21% and 11% higher BC and OC emissions for Eastern Europe, compared to the central case. Using low EFs results in about 25–30% lower BC and about 10% lower OC. Schaap et al. (2004) specifically pointed out traffic sector and proposed that the underestimation would be caused by PM2.5 particle emissions and not their BC fraction. Our result show that there is some potential (max 20%) for higher BC emissions if the shares would be at the high end of variation, but such an increase would not be enough to explain as high difference to observations as indicated by Schaap et al. (2004). This implies, as suggested also by Bond et al. (2004), a need for careful evaluation of other large sources, especially EFs and activity levels (recognizing specific combustion techniques) in domestic sector and small industrial installations.

domestic appliances contribute to the emissions of BC and OC. Coal is important in 1990 in Eastern Europe but its share declines towards 2000. Although fuel use in traffic increased, the European emissions in 2000 are estimated to remain at about 1990 levels due to introduced abatement. After a sharp decline between 1990 and 1995, the emissions in Eastern Europe remain fairly constant. Our estimate of European BC and OC emissions are in line with the more recent inventories on carbonaceous particles indicating lower emissions than previous studies. Based on our analysis, the variation in BC and OC EFs shares (reported in reviewed measurement campaigns) could explain at most about 20% higher emissions of BC and OC. Considering the importance of carbonaceous aerosols further work is needed to understand better their sources and reasons for mismatch between emissions and observations as well as to assess the likely future evolution of emissions.

5. Conclusions

Acknowledgements

The results of this study show that the main sectors contributing to European emissions of BC and OC are transport and domestic combustion. Especially diesel vehicles and biomass burned in

The authors thank Tami Bond for free and fruitful discussions and Imrich Bertok for programming impossible requests. Kaarle Kupiainen thanks Niko Karvosenoja from SYKE and Matti

ARTICLE IN PRESS 2168

K. Kupiainen, Z. Klimont / Atmospheric Environment 41 (2007) 2156–2170

Johansson from UNECE for valuable discussions during the process, Finnish NMO supporting his stay at IIASA during YSSP 2001 programme, and Nordic Envicon Oy where he has done part of the work. Zbigniew Klimont was supported by the European Network of Excellence ACCENT (Atmospheric Composition Change) of the European Commission.

References Bond, T.C., Sun, H., 2005. Can reducing black carbon emissions counteract global Warming? Environmental Science and Technology 39, 5921–5926. Bond, T.C., Bussemer, M., Wehner, B., Keller, S., Charlson, R.J., Heintzenberg, J., 1999. Light absorption by primary particle emissions from a lignite burning plant. Environmental Science and Technology 33 (21), 3887–3891. Bond, T.C., Streets, D.G., Yarber, K.F., Nelson, S.M., Woo, J.-H., Klimont, Z., 2004. A technology-based global inventory of black and organic carbon emissions from combustion. Journal of Geophysical Research 109, D14203. Butcher, S.S., Sorenson, E.M., 1979. A study of wood stove particulate emissions. Journal of the Air Pollution Control Association 29 (7), 724–728. Cadle, S.H., Mulawa, P.A., Hunsanger, E.C., Nelson, K., Ragazzi, R.A., Barrett, R., Gallagher, G.L., Lawson, D.R., Knapp, K.T., Snow, R., 1999. Composition of light-duty motor vehicle exhaust particulate matter in the Denver, Colorado Area. Environmental Science and Technology 33 (14), 2328–2339. Chow, J.C., Watson, J.G., Crow, D., Lowenthal, D.H., Merrifield, T., 2001. Comparison of IMPROVE and NIOSH Carbon Measurements. Aerosol Science and Technology 34, 23–34. Cooke, W.F., Wilson, J.J.N., 1996. A global black carbon aerosol model. Journal of Geophysical Research 101 (D14), 19395–19409. Cooke, W.F., Liousse, C., Cachier, H., Feichter, J., 1999. Construction of a 1  1 fossil fuel emission data set for carbonaceous aerosol and implementation and radiative impact in the ECHAM4 model. Journal of Geophysical Research 104 (D18), 22137–22162. Cooper, J.A., 1980. Environmental impact of residential wood combustion emissions and its implications. Journal of the Air Pollution Control Association 30 (8), 855–861. Derwent, R.G., Ryall, D.B., Jennings, S.G., Spain, T.G., Simmonds, P.G., 2001. Black carbon aerosol and carbon monoxide in European regionally polluted air masses at Mace Head, Ireland during 1995–1998. Atmospheric Environment 35, 6371–6378. Durbin, T.D., Truex, T.J., Norbeck, J.M., 1998. Particulate measurements and emissions characterization of alternative fuel vehicle exhaust. Final Report for National Renewable Energy Laboratory. Center for Environmental Research and Technology, University of California, Riverside, CA, 1998. Engelbrecht, J.P., Swanepoel, L., Chow, J.C., Watson, J.G., Egami, R.T., 2002. The comparison of source contributions from residential coal and low-smoke fuels, using CMB

modeling, in South Africa. Environmental Science and Policy 5 (2), 157–167. Environment Australia, 2002. Technical Report No. 5: Emissions from Domestic Solid Fuel Burning Appliances. March 2002, ISBN:0642548676 /http://www.deh.gov.au/atmosphere/ airtoxics/publications/report5/index.htmlS. European Commission (EC), 2001. Communication from the Commission. The Clean Air for Europe (CAFE) Programme: Towards a Thematic Strategy for Air Quality. COM(2001) 245 final, Brussels. Eurostat, 2003. Energy balance sheets-data 2000–2001, Luxembourg. Fine, P.M., Cass, G.R., Simoneit, B.R.T., 2001. Chemical charactrization of fine particle emissions from the fireplace combustion of woods grown in the Northeastern United States. Environmental Science and Technology 35 (13), 2665–2675. Fine, P.M., Cass, G.R., Simoneit, B.R.T., 2002. Chemical characterization of fine particle emissions from the fireplace combustion of woods grown in the Southern United States. Environmental Science and Technology 36 (7), 1442–1451. Fisher, G.L., Prentice, B.A., Silberman, D., Ondov, J.M., Biermann, A.H., Ragaini, R.C., McFarland, A.R., 1978. Physical and morphological studies of size-classified coal fly ash. Environmental Science and Technology 12 (4), 447–451. Fisher, G.L., Chrisp, C.E., Hayes, T.L., 1979. Carbonaceous particles in coal fly ash. In: Proceedings, Carbonaceous Particles in the Atmosphere. March 20–22, 1978. Lawrence Berkeley Laboratory, University of California. Ge, S., Bai, Z., Liu, W., Zhu, T., Wang, T., Qing, S., Zhang, J., 2001. Boiler briquette coal versus raw coal: Part I—Stack gas emissions. Journal of the Air & Waste Management Association 51 (4), 524–534. Gillies, J.A., Gertler, A.W., 2000. Comparison and evaluation of chemically speciated mobile source PM2.5 particulate matter profiles. Journal of the Air & Waste Management Association 50, 1459–1480. Gray, H.A., Cass, G.R., 1998. Source contributions to atmospheric fine carbon particle concentrations. Atmospheric Environment 32 (22), 3805–3825. Griest, W.H., Tomkins, B.A., 1984. Carbonaceous particles in coal combustion stack ash and their interaction with polycyclic aromatic hydrocarbons. Science of the Total Environment 36, 209–214. Hansen, J., Sato, M., Ruedy, R., Lacis, A., Oinas, V., 2000. Global warming in the twenty-first century: an alternative scenario. Proceedings of the National Academy of Sciences USA 97 (18), 9875–9880. Henry, W.M., Knapp, K.T., 1980. Compound forms of fossil fuel fly ash emissions. Environmental Science and Technology 14 (4), 450–456. Hildemann, L.M., Markowski, G.R., Cass, G.R., 1991. Chemical composition of emissions from urban sources of fine organic aerosol. Environmental Science and Technology 25 (4), 744–759. International Energy Agency (IEA), 2005a. Energy statistics of OECD countries, Paris. International Energy Agency (IEA), 2005b. Energy statistics of non-OECD countries, Paris. Israe¨l, G.W., Schlums, C., Treffeisen, R., Pesch, M., 1996. Rubimmission in Berlin. Herkunftsbestimmung—KFZ-Flottenemissionsfaktoren—Vergleichbarkeit von Probenahme-

ARTICLE IN PRESS K. Kupiainen, Z. Klimont / Atmospheric Environment 41 (2007) 2156–2170 methoden. VDI Fortschrittberichte, Reihe 15 Nr. 152. VDI Verlag, Du¨sseldorf, Germany. Jacobson, M.Z., 2002. Control of fossil-fuel particulate carbon and organic matter, possibly the most effective method of slowing global warming. Journal of Geophysical Research 107 (D19), 4410. Kerminen, V-M., Ma¨kela¨, T.E., Ojanen, C.H., Hillamo, R.E., Vilhunen, J.K., Rantanen, L., Havers, N., von Bohlen, A., Klockow, D., 1997. Characterization of the particulate phase in the exhaust from a diesel car. Environmental Science and Technology 31 (7), 1883–1889. Kirchstetter, T.W., Harley, R.A., Kreisberg, N.M., Stolzenburg, M.R., Hering, S.V., 1999. On-road measurement of fine particle and nitrogen oxide emissions from light- and heavy-duty motor vehicles. Atmospheric Environment 33, 2955–2968. Klimont, Z., Cofala, J., Bertok, I., Amann, M., Heyes, C., Gyarfas, F., 2002. Modelling particulate emissions in Europe. A Framework to Estimate Reduction Potential and Control Costs. Interim Report IR-02-076. IIASA, Austria, 169pp. Kojima, M., Brandon, C., Shah, J., 2000. Improving Urban Air Quality in South Asia by Reducing Emissions from Two-Stroke Engine Vehicles. World Bank, Washington, DC, USA. Kupiainen, K., Klimont, Z., 2004. Primary emissions of submicron and carbonaceous particles in Europe and the potential for their control. Interim Report IR-04-079. IIASA, Austria, 115pp. Lambrecht, U., Ho¨pfner, U., 2000. Mo¨gliche Diskrepanzen bei der Bestimmung von Partikeln. Untersuchung zum Zusammenhang zwischen Emissionsberechnung und Immissionskonzentration. IFEU Heidelberg im Auftrag des Ministeriums fu¨r Umwelt und Verkehr des Landes BadenWu¨rttemberg. 45pp (In German). Lowenthal, D.H., Zielinska, B., Chow, J.C., Watson, J.G., Gautam, M., Ferguson, D.H., Neuroth, G.R., Stevens, K.D., 1994. Characterization of heavy-duty diesel vehicle emissions. Atmospheric Environment 28 (4), 731–743. Muhlbaier Dasch, J., 1982. Particulate and gaseous emissions from wood-burning fireplaces. Environmental Science and Technology 16 (10), 639–645. Muhlbaier, J.L., Williams, R.L., 1982. Fireplaces, furnaces and vehicles as emission sources of particulate carbon. In: Wolff, G.T., Klimisch, R.L. (Eds.), Particulate Carbon. Atmospheric Life Cycle. Plenum, , New York, USA, pp. 185–205. Norbeck, J.M., Durbin, T.D., Truex, T.J., 1998a. Measurement of primary particulate matter emissions from light-duty motor vehicles. Final Report CRC Project No. E-24-2. Center for Environmental Research and Technology, University of California, Riverside, CA, 1998. Norbeck, J.M., Durbin, T.D., Truex, T.J., Smith, M.R., 1998b. Characterizing particulate emissions from medium- and light heavy duty diesel-fueled vehicles. Final Report for South Coast Air Quality Management District, Contract no. 97031. Center for Environmental Research and Technology, University of California, Riverside, CA, 1998. Norbeck, J.M., Durbin, T.D., Truex, T.J., 1998c. Characterization of particulate emissions from gasoline-fueled vehicles. Final Report for California Air Resources Board, Contract 94-319. Center for Environmental Research and Technology, University of California, Riverside, CA, 1998.

2169

Ohlstro¨m, M.O., Lehtinen, K.E.J., Moisio, M., Jokiniemi, J.K., 2000. Fine-particle emissions of energy production in Finland. Atmospheric Environment 34, 3701–3711. Olmez, I., Sheffield, A.E., Gordon, G.E., Houck, J.E., Pritchett, L.C., Cooper, J.A., Dzubay, T.G., Bennett, R.L., 1988. Compositions of particles from selected sources in Philadelphia for receptor modeling applications. Journal of the Air Pollution Control Association (JAPCA) 38, 1392–1402. Penner, J.E., Eddleman, H., Novakov, T., 1993. Towards the development of a global inventory for black carbon Emissions. Atmospheric Environment 27A (8), 1277–1295. Pinto, J.P., Stevens, R.K., Willis, R.D., Kellogg, R., Mamane, Y., Novak, J., Sˇantroch, J., Benesˇ , I., Lenicˇek, J., Buresˇ , V., 1998. Czech air quality monitoring and receptor modeling study. Environmental Science and Technology 32 (7), 843–854. Rau, J.A., 1989. Composition and size distribution of residential wood smoke particles. Aerosol Science and Technology 10, 181–192. Sagebiel, J.C., Zielinska, B., Walsh, P.A., Chow, J.C., Cadle, S.H., Mulawa, P.A., Knapp, K.T., Zweidinger, R.B., Snow, R., 1997. PM-10 exhaust samples collected during IM-240 dynamometer tests of in-service vehicles in Nevada. Environmental Science and Technology 31 (1), 75–83. Schaap, M., Van Der Gon, H.A.C.D., Dentener, F.J., Visschedijk, A.J.H., Van Loon, M., ten Brink, H.M., Putaud, J.-P., Guillaume, B., Liousse, C., Builtjes, P.J.H., 2004. Anthropogenic black carbon and fine aerosol distribution over Europe. Journal of Geophysical Research 109, D18207. Schauer, J.J., Kleeman, M.J., Cass, G.R., Simoneit, B.R.T., 1999. Measurement of emissions from air pollution sources. 2. C1 through C30 organic compounds from medium duty diesel trucks. Environmental Science and Technology 33 (10), 1578–1587. Shi, J.P., Mark, D., Harrison, R.M., 2000. Characterization of particles from a current technology heavy-duty diesel engine. Environmental Science and Technology 34 (5), 748–755. Smith, A.P., 2000. Assessment of particulate sampling for twostroke vehicles. AEA Technology Report submitted to the World Bank, Washington, DC. Streets, D.G., Gupta, S., Waldhoff, S.T., Wang, M.Q., Bond, T.C., Yiyun, B., 2001. Black carbon emissions in China. Atmospheric Environment 35 (25), 4281–4296. Turpin, B.J., Saxena, P., Andrews, E., 2000. Measuring and simulating particulate organics in the atmosphere: problems and prospects. Atmospheric Environment 34 (18), 2983–3013. UN, 2003. 2003 Industrial Commodity Statistics Yearbook. United Nations Statistics Division, New York US EPA, 2002. SPECIATE3.2 US EPA’s repository of total organic compound (TOC) and particulate matter (PM) speciated profiles for a variety of sources for use in source apportionment studies, /http://www.epa.gov/ttn/ chief/software/speciate/S. Watson, J.G., Chow, J.C., 2002. Comparison and evaluation of in situ and filter carbon measurements at the Fresno supersite. Journal of Geophysical Research 107 (D21), 8341. Watson, J.G., Chow, J.C., Pritchett, L.C., Houck, J.A., Ragazzi, R.A., Burns, S., 1990. Chemical source profiles for particulate motor vehicle exhaust under cold and high altitude operating conditions. Science of the Total Environment 93, 183–190.

ARTICLE IN PRESS 2170

K. Kupiainen, Z. Klimont / Atmospheric Environment 41 (2007) 2156–2170

Watson, J.G., Chow, J.C., Houck, J.E., 2001. PM2.5 Chemical source profiles for vehicle exhaust, vegetative burning, geological material, and coal burning in Northwestern Colorado during 1995. Chemosphere 43, 1141–1151. Wierzbicka, A., Lillieblad, L., Pagels, J., Strand, M., Gudmundsson, A., Gharibi, A., Swietlicki, E., Sanati, M., Bohgard, M., 2005.

Particle emissions from district heating units operating on three commonly used biofuels. Atmospheric Environment 39, 139–150. Williams, D.J., Milne, J.W., Quigley, S.M., Roberts, D.B., Kimberlee, M.C., 1989. Particulate emissions from ‘in-use’ motor vehicles—II. diesel vehicles. Atmospheric Environment 23 (12), 2647–2661.