Ragini Kumari, Ph.D.
Senior Programme Officer, Toxics Link, New Delhi Mobile: +91-9990746864
Deputy Director & Head
RDPU, NEERI, Nagpur-440020
Subject: Submission of Undertaking for the JESE paper no: 98/2010
We (Ragini Kumari, Arun K. Attri, Luc Int Panis, B. R. Gurjar) are happy to know that JESE has accepted our
manuscript titled- “Emission estimates of Particulate Matter and Heavy Metals from Mobile Sources in Delhi”. I am
sending the three copies of duly signed undertaking along with the original manuscript along with Tables and
As a corresponding author of this manuscript I agree with the terms and conditions as stated in the undertaking and
assure you that the same work has not been published elsewhere.
We thank you very much for considering our paper for the Journal of Environmental Science &
With warm regards,
Emission estimates of Particulate Matter and Heavy Metals from Mobile Sources in
Ragini Kumari1,2§, Arun K. Attri1, Luc Int Panis2, B. R. Gurjar3
1School of Environmental Sciences
Jawaharlal Nehru University (JNU)
New Delhi-110067, India
2Flemish Institute for Technological Research (VITO)
200- Boeretang, 2400 Mol
3Indian Institute of Technology Roorkee
Department of Civil Engineering
Roorkee- 247 667, India
§ Corresponding author: Dr. Ragini Kumari, Senior Programme Officer, Toxics Link, H2,
Jungpura Ext., New Delhi-14, Phone no: 011-24328006, Fax: 011-24321747
-An attempt has been made to make a comprehensive emission inventory of particulate matter
(PM) of various size fractions and also of heavy metals (HMs) emitted from mobile sources
(both exhaust and non-exhaust) from the road transport of Delhi (1991-2006). COPERT-III and 4
models were mainly used to estimate these emissions. Results show that the annual exhaust
emission of PM of size upto 2.5 micrometer (PM2.5) has increased from 3Gg to 4.5Gg during
1991-2006 irrespective of improvement in vehicle-technology and fuel use. PM emission from
exhaust and non-exhaust sources in general has increased. Heavy commercial vehicles need
attention to control particulate emission as it emerged as a predominant source of PM emissions.
Among non-exhaust emissions of total suspended particulate matter (TSP), road-surface wear
(~49%) has the prime contribution. As a result of introduction of unleaded gasoline Pb has
significantly reduced (~8 fold) whereas share of Cu and Zn are still considerable.
Among non-exhaust sources, Pb release was the most significant one from tyre-wear whereas
from break-wear, Cu release was found to be the most significant followed by Pb and Cr + Zn.
Because of public health concerns further policies need to be developed to reduce emissions of
PM and HMs from the road transport of megacity Delhi.
Key words: heavy metals; non-exhaust; particulate matter, road transport emissions; megacity,
Road transport is one of the major sources of particulate matter (PM) in urban areas (Gertler et
al., 2000)1. Transport generated PM may vary in size, shape and chemical composition (. For
example, it can be constituted of different size fractions [e.g., PM2.5, PM10 and total suspended
particulate matter (TSP)] and chemical signatures (like heavy metals) depending upon the source
(exhaust or non-exhaust), vehicle category, activity and/or fuel used (Furusjo et al., 2006)2.
Heavy metals (HMs), too, are emitted not only from exhaust but also from non-exhaust sources
like tyre-wear, road surface wear, brake wear, etc.
Importance of emission estimation of PM and HMs in an urban area is many fold, e.g., health
risk, pollution to ecosystem, soil and change in local to global climate. Personal exposure from
transport related air pollution is more because of level of inhalation matches with tailpipe
emissions (Kunzli et al., 2000)3. Inhalation of urban respirable PM is detrimental to human
health not only because of its size but also due to chemical composition (Dahl et al., 2006)4.
HMs emitted from transport, e.g., lead (Pb), nickel (Ni), chromium (Cr), and cadmium (Cd)
cause serious health problems (WHO, 2000)5. That is the reason India’s Central Pollution
Control Board (CPCB) has recently introduced ambient air quality standards (AAQS) of Nickel
(Ni), Vanadium (V), Mercury (Hg), and Arsenic (As) in addition to the existing AAQS of Pb.
Transport sector is one of the important sources of PM in urban areas such as megacity- Delhi
(Srivastava et al., 2009)6. Gurjar et al. (2008)7 have found that Delhi is 4th most polluted
megacity in the world in terms of levels of total suspended particles (TSP) in ambient air.
Various policy measures have been taken to reduce PM exhaust-emissions from the transport
sector, for Delhi. For example, Euro-I and Euro-II norms were implemented during 1999-2000
for cars and taxies followed by next phase for other vehicle-types during 2000-2001. Fuel related
policies were also implemented in Delhi to reduce exhaust emissions, e.g., lead (Pb)-reduction
programme for gasoline was started in 1994. Prior to June, 1994 the Pb content in gasoline was
0.56 gl-1. In June, 1994 low leaded (0.05 gl-1) gasoline was introduced and further became
unleaded (~ 0.013gl-1)] in Sep. 1998 (CPCB, 2001)8. Also, CNG was introduced in year 2001 for
the public transport system.
To see the impact of various policy measures such as listed above, we have carried out
inventorization for PM and heavy metals (HMs) in the present study. Current study has novelty
in many terms: (i) inclusion of PM of various size fractions (PM2.5, PM10, and TSP) from both
exhaust as well as non-exhaust sources (tyre-wear, brake-wear, and road surface-wear), and (ii)
estimation of range of heavy metals (HMs) [from exhaust (Pb, Cu, Cr, Se, Zn, Ni and Cd), and
non-exhaust sources (Pb, Cu, Cr, Zn, Ni, Cd, and As)] for the above mentioned period.
Section 2 briefly describes the methodologies for emission estimation and various policy
analyses. In Section 3, we describe and show results of emission estimation of various size
fractions of particulate matter and HMs from exhaust and non-exhaust sources. In Section 4
impact assessment of various policy measures on the exhaust emission of PM2.5 is quantified.
Section 5 discusses on the relative carcinogenic potency of HMs. In Section 6 we point out few
important limitations, which need to be taken care in future studies. In Section 7, we conclude
the present study, which has importance from the health perspective of inhabitants of megacity
We present a bottom-up emission inventory for exhaust emissions. Calculations are based on
vehicle usage i.e. vehicle kilometer traveled (VKT) for species of PM and emission estimates for
lead (Pb) and other HMs are fuel consumption dependent. We used COPERT-III9 and emission
factors from other sources for the exhaust emission estimation, whereas COPERT-410was used
for non-exhaust sources (tyre-wear, brake wear and surface wear). Emissions have been
estimated for following vehicular modes: motorized two-wheelers (TW’s) and three-wheeled
vehicles (THW’s), cars and taxis, buses, heavy-duty vehicles (HCV’s) and light commercial
vehicles (LCV’s). Fuels considered were gasoline (petrol), diesel and compressed natural gas
(CNG), depending on mode.
Vehicle number and annual average vehicle kilometer traveled
Emissions (exhaust and non-exhaust) from the road transport of Delhi are calculated by taking
into account vehicle number (ESD, 2005-2006)11 under each compliance class (uncontrolled,
Euro-I and Euro-II), fuel use (gasoline, diesel and CNG) and annual average vehicle kilometer
traveled (Mashelkar et al., 200212; CRRI, 200613-personal communication). Data used for vehicle
number and annual average vehicle kilometer traveled (AVKT) are given in Fig. 1. Vehicle sub
grouping was done on the basis of technology and fuel use. Euro-I emission norm for cars and
taxies were introduced in year 1999 and Euro-II in the next year. Same norms for other vehicle
categories were introduced in year 2000 and 2001 respectively. CNG was introduced in the
public transport system from 2001 on the order of the Supreme Court of India. Further details
about composition (%) per fuel use are given in supporting Table S 1, and proportion of 2-
stroke and 4-stroke TW’s and THW’s for the period 1991 to 2006 in supporting Table S 2.
Emission estimation of PM
Exhaust emission of PM2.5 and PM10
Detailed bottom-up methodology has been adopted from the COPERT-III9 model
(http://lat.eng.auth.gr/copert/files/tech49.pdf) for the estimation of PM2.5 from the road transport
sector of Delhi for the period from 1991 to 2006. The annual emission of fine particulate matter
(PM2.5) in a given year ‘i’ was calculated by using the following general formula
Where, subscripts ‘i’ represent the year; ‘j’ is the respective vehicle category, ‘k’ is the
compliance class (uncontrolled, Euro-I, Euro-II) for the vehicle categories ‘j’.
Eijk= annual emission estimates of PM2.5 ‘p’ (in kg).
Nijk = number of vehicles in year ‘i’, of category ‘j’ and falling in compliance class ‘k’.
AVKTijk= average annual vehicle kilometer traveled (in km).
EFjkp= emission factor of vehicle category ‘j’ falling in compliance class ‘k’ (g km-1).
Emission factors (g km-1) used for the estimation of PM2.5 from vehicular exhausts is given in
Table 1. Further, Eijk (PM2.5) estimated in kg is finally converted into Mg.
COPERT-III9 or 410 model does not include PM10 emission factor from vehicle exhaust. But,
ARAI (2007)16 has computed PM18 emission factors for Indian vehicles (Table 1). Since, PM18
includes PM10 fractions and share of PM10 is much larger in PM18, so, we have used these
emission factors for the emission estimation of PM10 from exhaust sources using methodology
similar to that of PM2.5 (Eqn. 1).
Non-exhaust emission of PM10 and total suspended particulate matter (TSP)
Non-exhaust emission sources considered for estimation of PM10 and TSP emissions include tyre
wear, brake wear and road surface wear. Though the climatic as well as road conditions in India
are different than most of European cities, we have used corresponding emission factors given in
COPERT 410 to estimate PM10 (Table 2) and TSP (Table 3) as India specific emission factors
are not available for non-exhaust sources. This could introduce uncertainties in our emission
estimates that should take into account while interpreting our results.
Emission estimation of heavy-metals (HMs)
Exhaust emission- lead (Pb) and other HMs
Bottom-up approach (using COPERT-III)9 was applied to estimate HM’s from vehicular
exhaust. Since fuel consumption (FC) statistics available for megacity Delhi (Delhi Statistical
Handbook, 2000)20 does not differentiate between fuel used in the industry, transport or domestic
sector (Gurjar et al., 2004)21, FC in the road transport was estimated using COPERT-III9 model.
Eqn. 2 was used to calculate the annual FC for year ‘i’, fuel-type ‘f’ (gasoline, diesel and CNG).
Further, estimated FC was taken as base to estimate HMs.
FCF AVKTN FC
Where subscripts i= year; j= vehicle category (I-VII), k= compliance class (uncontrolled, Euro-I,
Euro-II) for the vehicle categories ‘j.
FCi= annual fuel consumption estimated in year ‘i’ (in kg).
Njk = number of vehicles of category ‘j’ and falling in compliance class ‘k’.
AVKTjk= average annual vehicle kilometer traveled (in km) for vehicle category ‘j’ and falling
in compliance class ‘k’.
FCFjkf= Fuel consumption factor of vehicle category ‘j’ under compliance class ‘k’ for fuel type
‘f’ (g km-1).
Details about vehicle specifications-fuel use, compliance class (uncontrolled, Euro-I and Euro-
II), weight of the respective vehicle category (tonnes), annual average vehicle speed (km h-1) and
values used for FCFjkf is given in Table 1.
Pb content (w/w) in gasoline over the period of study is given in Table 4. The assumption behind
calculating the emission of Pb was that, only 75% (w/w) of Pb present in the gasoline was
emitted into the air (Hassel et al., 1987)22. It is speculated that 25% of it either end up as deposits
of Pb or Pb-oxides along the exhaust pipe surfaces. The mathematical formulation used for the
calculation of Pb is given in the Eqn. 3.
Where, subscripts i,j,k stands for the same as discussed before,
EPb= Emission of Pb (kg) to air,
WPb = amount of Pb (g kg-1) present in gasoline in the year ‘i'.
FC= annual fuel consumption (Mg).
Emissions of other HMs, i.e. Cd, Se, Cu, Cr, Ni and Zn, were calculated by top down method.
Their respective EFs in mg kg-1 of fuel consumed are as follows: Cd and Se (0.01); Cu (1.70); Cr
(0.05); Ni (0.07) and Zn (1.00), which are taken from COPERT 4. Their emissions were
calculated by using Eqn. 4.
f HMf ijk HMs
Where, EHMs= Emission of heavy metals except Pb (g year-1)
EFHMs = Emission factor (mg kg-1).
Non-exhaust emission (Tyre- wear and Brake-wear)
Heavy metals like As, Cd, Cr, Ni, and Pb were estimated from non-exhaust sources (tyre-wear
and brake-wear) by using EFs (µg veh. km-1) from Kummer et al. (2009)23, whereas Cu and Zn
were estimated merely from brake wear (Westerlund and Johansson, 200224 – also cross referred
by Johansson et al., 200825) under non-exhaust category (Table 5).
Impact analysis of various policy measures (CNG, Euro-I and II, and 2-stroke to 4-stroke) on
PM2.5 exhaust emission
To quantify the impact of the various policies we had run the model with and without the effect
of policies. For example, to quantify the impact of CNG fuel alone, we first estimated the PM2.5
exhaust emission assuming that the CNG-fuel was not introduced. From this result we then
subtracted the actual emissions (taking into account the CNG fuel) to obtain the total amount of
pollutants that were avoided due to CNG. This methodology was also extended to assess the
effectiveness of Euro I and II norms in the road transport sector and shift of 2-stroke TW’s and
THW’s to 4-strokes in megacity Delhi.
Results and discussion
Estimation of annual emissions of PM
PM2.5 exhaust emissions
Total annual PM2.5 exhaust emission has increased from 3Gg to 4.5Gg during1991-2006.
Contribution from diesel fuelled HCV’s and gasoline driven TW’s increased from 41% to 63%
and 17% to 26% respectively between year 1991 to 2006; whereas significant decline was
observed as a result of CNG introduction in buses (15% to 4%), LCV’s (21% to 4%), and
THW’s (4% to 0.12%) (Fig. 2).
PM10 exhaust and non-exhaust emissions
Total annual PM10 (equivalent to PM18) exhaust emission has increased ~ 1.5 times (i.e. 4 Gg to
6 Gg) from 1991 to 2006. The contribution from diesel fuelled HCV’s has increased between
year 1991 to 2006 (from 48% to 77%) whereas emission from bus, LCV and THW’s has
declined, as a result of CNG switchover since year 2001 (Fig. 3, I). The total annual contribution
of non-exhaust PM10 emission increased ~ 4-times from road surface wear, brake wear, and tyre
wear increased from 110 Mg to 424 Mg, 102 Mg to 414 Mg, and 93 Mg to 365 Mg, respectively
during 1991-2006 (Fig. 3, II, III, and IV). Emission contribution from TW’s and cars has also
increased through road surface wear as (25% to 38%) and (18% to 22%) respectively, whereas it
is about 12% throughout the study period from buses, and declined from other vehicle categories
(Fig. 3, II). Share of PM10 from brake wear also increased from TW’s (34% to 48%), cars (16%
to 19%), but decreased from HCV’s (26% to 16%) and LCV’s (10% to 6%) (Fig. 3, III).
Likewise contribution of TW’s and cars in tyre wear has increased from 28% to 41% and 18% to
22%, respectively. Whereas, emission from buses remained about 10% and declined in case of
HCV’s (28% to 18%) between 1991 to 2006 (Fig. 3, IV).
TSP non-exhaust emissions
In non-exhaust TSP category, road surface wear contributes the most (~50%) followed by tyre
wear (29%) and brake wear (22%). Its contribution increased from 221 Mg to 853 Mg from road
surface wear followed by tyre wear (125 Mg to 525 Mg), and brake wear (102 Mg to 422 Mg),
as shown in Fig. (4. I, II, and III). TSP emission from tyre wear increased from 34% to 47%
and 22% to 25% from TW’s and cars, respectively. Non-exhaust emissions of TSP from all
sources followed similar trend.
Emission estimation of HMs
Pb emission has considerably decreased from 246 Mg to 29 Mg between 1991 and 2006 (Fig. 5,
II). This decrease underlines the effectiveness of the fuel policy to introduce unleaded gasoline
fuel in Delhi. Other HMs estimated from exhaust were Cu, Cr, Se, Zn, Ni and Cd. Emission load
of these HMs (including Pb) increased ~3-times (from 3164 Mg to 8411 Mg) between 1991 to
2006. Cu and Zn contribute significant (60%) and (35%) respectively (Fig. 5, I), whereas rest
(Ni, Cr, Cd, Pb and Se) contributes only 5% (Fig. 5, II).
Range of HMs emerge from brake wear and tyre wear. Annual HM emission from tyre-wear has
increased by >4 times (from 40 Kg to 180 Kg) over the period of study. Pb emission from tyre-
wear was largest (70%) followed by Ni (13%), Cr (10%) and Cd (2%). HM contribution from
brake wear increased about five-times (8.5 Mg to 41 Mg), Cu (86%) and Pb (9%) were mainly
emitted from brake-wear to the ambient air 1991 to 2006 (Fig. 5, IV).
Impact of various policy measures on PM2.5 emission
We compared the results after running the model with and without the effect of pollution
abatement policies. The mandatory implementation of CNG-fuel in public transport resulted in
the reduction (1.3 times), Euro-I and II (1.2 times) and economy driven shift in technology from
2-stroke towards 4-stroke also contributed in reducing PM2.5 emission from exhaust by 1.5 times
(Fig. 6). So, the decline observed in the particulate ambient air quality (CPCB, 2007)26 is not
solely CNG policy driven rather it’s a cumulative impact of these transport related changes in
Relative carcinogenic potency of HMs
Quantification of emissions with respect to health risk is important as illustrated in Table 6 that
the unit risk of Cr (in terms of risk/ng/m3) is more than As and Ni, whereas risk from Pb is less
than Ni and As. So, just by introducing unleaded-gasoline will not make inhabitants of urban
areas risk-free from all types of HMs (Fig. 5, I, II). Our results show that exhaust emission load
of Cr (51 Mg to 147 Mg) and Ni (from 72 Mg to 207 Mg), keep increasing (~3-times) from 1991
to 2006 (Fig. 5, II). Non-exhaust emissions also contribute further in it. Emission of Cr from
(brake wear and tyre wear) has substantially increased (273 kg to 1335 kg) followed by Pb (761
kg to 3715 Kg), Ni (44 kg to 213 kg), and As from 12 kg to 51 kg from 1991 to 2006. The most
cancer potent HM, Cr substantially emitted by brake wear (1317 kg) followed by that was Pb
(3583 kg), in 2006 (Fig. 7). So, not only exhaust but also non-exhaust emissions affect human-
health in Delhi. With rising congestion it will further aggravate HM emissions aggravating
human health risk.
Limitations and further scope of the study
Emission factor used for the quantification of PM2.5 fraction was from COPERT-III (Europe)
whereas for PM10 from ARAI (2007)17 (India), so we expect some degree of uncertainty because
of differences in road conditions, driving pattern, and climate conditions. Emission factors used
for the estimation of non-exhaust emissions (PM and HM’s) were from European model
(COPERT 4)10 due to unavailability of indigenous emission factors, hence its application for
Delhi need to be revised in future when emission factors under Indian situation would be
First attempt has been made to make a comprehensive emission inventory of particulate
matter (PM) of various size fractions and heavy metals (HMs) from both exhaust and non-
exhaust sources from the road transport of Delhi (1991-2006). Total annual exhaust emission of
PM2.5 has increased from 3Gg to 4.5Gg during 1991-2006 irrespective of improvement in
technology and fuel use. Prime contributor of PM2.5 in year 2006 was diesel fuelled HCV’s
(63%) followed by gasoline driven TW’s (26%). We estimated the reduction in emission as a
result of implementation of CNG-fuel in public transport by 1.3 times followed by Euro-I and II
(1.2 times). Also the economy driven shift in technology of TW’s and THW’s (from 2-stroke
towards 4-stroke) contributed about 1.5 times reduction in PM2.5 emissions from exhaust.