Applied Energy xxx (2016) xxxxxx

Contents lists available at ScienceDirect

Applied Energy
journal homepage: www.elsevier.com/locate/apenergy

Passenger vehicles that minimize the costs of ownership and

environmental damages in the Indian market
Elisabeth A. Gilmore , Anand Patwardhan
School of Public Policy, University of Maryland, College Park, MD 20742, USA

h i g h l i g h t s

Full costs (private and social) are evaluated for Indian passenger cars.

Diesel has low ownership costs, but higher climate and health damages.

Compressed natural gas cars have lower costs and damages than petrol cars.

Electric cars have higher damages due to electricity generation emissions.

CNG and less carbon intensive electricity minimizes Indian cars full cost.

a r t i c l e i n f o a b s t r a c t

Article history: Rapid expansion of population and income growth in developing countries, such as India, is increasing
Received 10 February 2016 the demand for many goods and services, including four-wheeled passenger cars. Passenger cars provide
Received in revised form 12 August 2016 personal mobility; however, they also have negative implications for human wellbeing from increased air
Accepted 26 September 2016
pollutants and greenhouse gases (GHG). Here, we evaluate the range of passenger vehicles available in
Available online xxxx
the Indian market to identify options that minimize costs, human health effects and climate damages.
Our approach is to compare alternative fuel/powertrain vehicles with similar conventional gasoline
Keywords:
fueled vehicles and assess the differences in full (private and societal) costs for each pair. Private costs
Full cost
Greenhouse gases
are the combination of capital costs and the discounted expected future fuel costs over the vehicle life-
Air quality time. The costs to human health from air quality are calculated using intake fractions to estimate expo-
Human health sure and literature values for the damage costs adjusted by benefits transfer methods. We use the Social
Passenger vehicles Cost of Carbon to estimate climate damages. We find that, on average, the net present value (NPV) of the
India full costs of compressed natural gas (CNG) vehicles are lower than comparable gasoline vehicles, while,
diesel vehicles have higher costs. Presently, electric vehicles have higher private costs (due to high capital
costs) and societal costs (due to electricity generation emissions). Either a less carbon intensive electricity
grid or an increase in the CNG fleet would minimize total costs, human health effects and GHG emissions
from the passenger vehicle fleet. Policy makers should consider promoting the use of CNG vehicles,
although this would require supporting infrastructure improvements.
2016 Elsevier Ltd. All rights reserved.

1. Introduction
The ubiquity of four-door passenger vehicles in developed
countries speaks to the substantial benefits provided to their own-
ers and their passengers, such as comfort and convenience. These
vehicles have also introduced major societal-level risks, namely
premature mortality from exposure to adverse air quality and the
widespread impacts of climate change from greenhouse gas (GHG)
emissions. Higher fuel economy, more stringent tailpipe emission
Corresponding author.
E-mail address: gilmore@umd.edu (E.A. Gilmore).
standards and improved fuel quality have reduced these impacts
on a per-vehicle basis, but overall emissions continue to rise due
to an increase in demand [1]. Currently, much of the growth in
ownership of passenger vehicles is found in developing countries
[2]. While these vehicles substantially improve individual mobility,
the continued addition of new vehicles will further exacerbate the
poor air quality that is found in many urban centers in developing
countries [3] and present a challenge for climate mitigation efforts
[4].
One way to reduce societal impacts from passenger vehicles is
to promote alternative fuel/powertrain vehicles, such as hybrid
and electric vehicles. However, these vehicles generally have a

http://dx.doi.org/10.1016/j.apenergy.2016.09.096
0306-2619/ 2016 Elsevier Ltd. All rights reserved.

Please cite this article in press as: Gilmore EA, Patwardhan A. Passenger vehicles that minimize the costs of ownership and environmental damages in the
Indian market. Appl Energy (2016), http://dx.doi.org/10.1016/j.apenergy.2016.09.096
2 E.A. Gilmore, A. Patwardhan / Applied Energy xxx (2016) xxxxxx
higher cost of ownership than the conventional gasoline or diesel
options. Systematic benefit-cost analysis (BCA) of passenger vehi-
cle options can provide guidance to policymakers who aim to bal-
ance these private benefits and societal costs [5]. Here, we employ
a benefit-cost analysis (BCA) framework to evaluate the costs of
ownership and the environmental and human health effects of
alternative fuel/powertrain compared to gasoline (petrol) four-
door passenger vehicles for developing countries, using India as a
case study. Specifically, we identify vehicles that minimize both
the private costs and environmental burden by comparing the total
costs of each vehicle, comprised of the private costs of vehicle own-
ership and the social costs of the environmental attributes. Existing
benefit-cost and other analyses of the societal costs of vehicle
emissions have been conducted primarily for fuel/powertrain
options in the United States (US) and Europe [6] which may not
translate to developing countries. First, the vehicle markets in
developing countries often have a wider range of vehicles targeting
a more diverse consumer market. Second, some methods that are
employed for analysis in developed countries may be inappropri-
ate or less amenable to developing countries due to different con-
ditions or data availability. Additionally, there is a pressing need
for policy guidance for developing countries. Since vehicles and
infrastructure are durable goods, decisions made during this period
of growth will define the environmental attributes of the vehicle
fleet for decades.
In a recent review, Roosen et al. [6] identified 12 studies that
evaluated the private and societal costs of vehicles, monetizing
GHG and/or air pollutants from tailpipe emissions or electricity
production for electric vehicles. Overall, these studies show that
while conventional gasoline fueled vehicles have the lowest pri-
vate costs of ownership, alternative fuel/powertrain vehicles can
dominate on their full costs due to the reductions in the societal
costs from the GHG and other air pollutants. The studies, however,
draw a wide range of conclusions depending on assumptions about
capital costs, fuel prices and the effectiveness of emission controls.
For example, Keefe et al. [7] conclude that diesel passenger vehi-
cles are preferred on their full costs when compared to gasoline
and gasoline hybrid electric vehicles (HEV), assuming that the die-
sel vehicles attained the stated US Environmental Protection
Agency (USEPA) tailpipe emission standards [7]. By contrast, Lip-
man and Delucchi [8] find that hybrid vehicles outperform the die-
sel and gasoline fueled vehicles by making more favorable
assumptions about hybrids capital costs and fuel economy [8].
Expanding the alternatives to include the wider range of vehicles
that are available in the European market, Torchio and Santarelli
[9] found that the natural gas fueled vehicles had the lowest full
costs [9]. There has also been a growing interest in the potential
benefits of plug-in electric (PHEV) and all electric vehicles (EV).
Previous analysis has suggested that the monetized damages of
the health impacts and the GHG emissions are smaller than the dif-
ference in capital costs between PHEVs and conventional gasoline
vehicles [10]. These benefits, however, are also highly dependent
on the electricity generation source for both air quality and GHG
emissions. For example, Tessum et al. [11] found that charging
EVs with coal-based or the US grid average would increase mone-
tized health impacts from air pollution by 80% or more relative to
using conventional gasoline, although natural gas and other
renewables would reduce health impacts by 50% or more [11]. Fur-
ther, Tamayao et al. [12] concluded that the GHG emissions from
PHEVs could exceed those of HEVs depending on the location in
the US [12].
Given the range of factors that may influence vehicle selection,
these analyses may not provide sufficient guidance for policymak-
ers in developing countries. We select India as an important case
study of a rapidly developing country. Sales of four-wheel passen-
ger vehicles more than doubled from approximately 1.14 million in
Please cite this article in press as: Gilmore EA, Patwardhan A. Passenger vehicles that minimize the costs of ownership and environmental damages in the
Indian market. Appl Energy (2016), http://dx.doi.org/10.1016/j.apenergy.2016.09.096
20052006 to 2.6 million in 201415 [13]. Thus, an evaluation of
the benefits and costs of these vehicles is timely for policymakers
seeking to achieve air quality and GHG goals. The Indian case study
also provides both opportunities and challenges in terms of
approaches for this type of analysis. Unlike the more mature US
market that is dominated by a few vehicle types, the Indian car
market supports a wide range of vehicles: gasoline spark injection,
liquefied petroleum gas (LPG, also known as AutoGas) spark injec-
tion, compressed natural gas (CNG) spark injection, compression
injection diesel, and electric vehicles (EV). As a result, we can use
actual data on the retail prices observed by Indian consumers.
Comparing vehicles based on their actual prices for different size
classifications, rather than more generic techno-economic esti-
mates, is important since individuals make their purchase deci-
sions based on the vehicle prices that they observe in the market
as well as observed fuel prices. Due to the limited availability of
data, many assessments of alternative vehicles are based on
techno-economic estimates (e.g. [7]). While these modeling
approaches are often employed, especially for less mature tech-
nologies like PHEVs (e.g. [14]), it is generally preferable to use
actual retail prices as they reflect the prices faced by the con-
sumers (e.g. [15]).
We also investigate the influence of current policies promul-
gated by the national government on the present and future costs
and environmental attributes of the vehicle fleet, using benefit val-
uations that are appropriate for an Indian context. Indian tailpipe
emission standards have generally followed the policies developed
by the European Union [16]. Currently, new vehicles sold in urban
areas (13 major cities) must achieve Bharat IV standards (EURO 4)
with the rest of the country meeting Bharat III (EURO 3) standards.
It is expected that the Bharat IV will apply nationwide by 2017
[17]. Similar to other large urban centers in developing countries,
there are high levels of fine particulate matter (PM2.5) and other
air pollutants in India [18]. To address persistent air quality prob-
lems, Bharat VI (EURO 6) standards are to come into effect in 2020.
Further, the government has outlined a number of policies in its
Intended Nationally Determined Contribution (INDC) as part of
lead up to the Paris Climate Agreement. Notably, the Government
of India included its National Electric Mobility Mission Plan
(NEMMP) [19]. This plan envisions 57 million hybrids, PHEVs
and all electric vehicles in the Indian market by 2020.
2. Data and methods
We compare the private and societal costs of owning and oper-
ating alternative fuel/powertrain vehicles to conventional gasoline
vehicles. The vehicle with the lowest full cost is the preferred
option. In this analysis, we focus on the usage phase of the vehicle,
treating consumer behavior related to driving patterns and vehicle
lifetime as well as fuel prices, infrastructure and the electricity
generation sector parametrically through a sensitivity analysis.
Our approach is to pair vehicles of the same manufacturer, model
and variant that differ only in their fuel/powertrain. This allows us
to account for as many features (e.g. trim, air conditioning, etc. . .)
as possible that could potentially explain the difference in the
observed market prices for the vehicles.
In this section, we first describe the method for private and
social cost calculations. Second, we describe the collection of a
comprehensive dataset of four door passenger vehicles available
in India, and our approach to the pairing of vehicles. Third, we pre-
sent our estimates for the private costs of purchasing and operat-
ing the vehicles. Finally, we describe the social cost calculations,
including tailpipe emissions and our approach for estimating the
value of the emissions. The costs and emissions for other sectors
are beyond the scope of this analysis, such as expanding the fueling
 E.A. Gilmore, A. Patwardhan / Applied Energy xxx (2016) xxxxxx
infrastructure associated with the alternative fuel/powertrains or
the electricity generation needs for a full electric fleet. However,
we will qualitatively address these issues in the discussion.
2.1. Calculating the full cost of passenger vehicles
The full cost is comprised of the private costs that accrue to the
consumer as well as the societal costs that arise from the human
health and climate impacts discounted over the lifetime of the
vehicles. In Fig. 1, we show how we calculate the net present value
(NPV) of the full cost for each vehicle. We follow the conventions of
an influence diagram for our notation. The objective is to select the
vehicle that for a given set of conditions (or input parameters) has
the lowest NPV full cost which in this model is the private costs
of vehicle ownership and operation and the societal costs from cli-
mate emissions and adverse health impacts from exposure to
ambient concentrations of fine particulate matter (PM2.5). In our
analysis, we set the gasoline vehicle as the baseline and show
the incremental NPV for the alternative vehicles, where an incre-
mental NPV of less than zero means that the vehicle is preferred
to the gasoline vehicle. The lowest incremental NPV is the pre-
ferred vehicle. For the NPV calculations, we employ a private dis-
count rate of 10% and a social discount rate of 4%.
For electric vehicles, we highlight that the assumptions about
the electricity mix will influence the air emissions and, for conven-
tional air pollutants, it will also affect the exposed population. We
do not explicitly show the costs associated with the infrastructure
investment. However, we conduct a screening level analysis later
in Supplemental Information (Section S4).
The private costs (PC) are calculated as the sum of the initial
capital cost of the vehicle and the discounted sum of the cumula-
tive fuel costs, operating and maintenance costs (O&M), and resale
value, according to Eq. (1).
Xn
FC t OMt RS
PC CC o t 
t1 1 r 1 r n
where CCo is the capital cost in Indian Rupees (INR);
Fig. 1. Graphical representation of the full cost calculations.
Please cite this article in press as: Gilmore EA, Patwardhan A. Passenger vehicles that minimize the costs of ownership and environmental damages in the
Indian market. Appl Energy (2016), http://dx.doi.org/10.1016/j.apenergy.2016.09.096
3
FCt is the cost of fuel in year t in INR;
OMt is the cost of maintenance and repairs in year t in INR;
RS is the resale value of the vehicle at year n in INR;
r is the real private discount rate; and,
n is the number of years of ownership.
The fuel costs in year t (FCt) are given by Eq. (2).
FPt
FC t  VKT t 2
FE
where FE is the fuel economy of the vehicle in km per L (or km per
kg for CNG);
VKTt is the vehicle kilometers travelled in year t in km; and,
FPt is the fuel price in year t in INR per L (or kg for CNG).
For the societal costs, we focus on premature mortality from
exposure to fine particulate matter (PM2.5), where PM2.5 is com-
prised of directly emitted primary PM2.5 as well as secondary
PM2.5 which is formed from nitrogen dioxide (NOx) and sulfur
dioxide (SO2) that react chemically after being emitted from the
tailpipe. We also include the climate impacts from carbon dioxide
(CO2). The emissions in year t are then estimated using Eq. (3). For
electric vehicles, we multiply the emission factors from the charg-
ing plant by the required charging.
ET it EF i  VKT t 3
where ETit is the total emissions of pollutant i in year, t in grams;
and,
EFi is the emission factor of pollutant i in grams per km.
The societal cost (SC) is then calculated for summing the
discounted value of the damages caused by these emissions. We
discuss the methods in more detail in Section 2.4.
1 2.2. Vehicle dataset and pairing approach
One of the strengths of this analysis is that we use the prices
that are available in the market. We start by compiling a compre-
4 E.A. Gilmore, A. Patwardhan / Applied Energy xxx (2016) xxxxxx

hensive dataset of all passenger vehicles presently available in the
Indian market. This list is based on information from AutoCar India
and verified using the list of vehicles that have reported fuel econ-
omy data to the Society for Indian Automobile Manufacturers
(SIAM) for 2014 [20]. For each vehicle, we also compile data on
the vehicle style (e.g. sedan, hatchback or SUV), engine displace-
ment, torque, transmission (e.g. automatic or manual) and fuel
type. For these vehicles, we pair fuel/powertrain alternatives and
the conventional gasoline option, matching on manufacturer,
model and variant. This yields 84 pairs that are summarized in
Table 1 by fuel pair type and size category. The vehicles are classi-
fied according to Indian size categories of mini-car, compact, mid-
sized, executive, premium, and luxury. We group executive, pre-
mium and luxury into to same category Premium. We also include
a separate category for utility vehicles, which includes sport utility
vehicles (SUV) and multi-utility vehicles (MUV).
Presently, in the domestic Indian market, there is one electric
vehicle option and one hybrid-electric (HEV) option. The Mahindra
e20 is the first all-electric vehicle. Originally, it was sold at an aver-
age price of 750,000 INR. The Government of India has subse-
quently introduced a subsidy of 100,000 INR through its Faster
Adoption and Manufacturing of Electric Vehicles (FAME) policy.
We evaluate the costs of the vehicle on its unsubsidized cost and
then consider whether the Government subsidy is appropriate.
Further, Mahindra introduced a program named e20 care protec-
tion plan that includes an initial down-payment of approximately
479,000592,000 INR depending on the city followed by monthly
payments of 2999 INR per month that covers up to 50,000 km or
5 years with an overage charge of 2.5 INR per km [21]. For our
model, we assume that at the end of the 5 year period, there is
an option to purchase another package for 170,000 INR, which is
the difference in the capital cost. We estimate that the electric
vehicle will consume 84 Wh/km for city driving and 133 Wh/km
on the highway [22]. Since there is no direct match for the e20,
we pair it with the Tata Nano as well as the Maruti Suzuki Wagon
R as a sensitivity analysis.
For hybridized drive trains, there is a diesel hybrid vehicle, the
Maruti Suzuki Ciaz Hybrid, which is produced domestically and
priced at 823,0001,017,000 INR with a fuel economy of 28 km/L.
While the Toyota sells its Prius and the Camry hybrid in the Indian
market, they are taxed as foreign imports and are priced
significantly above domestic hybrids and we do not consider them
in this analysis. For the diesel hybrid, we pair it with a compact
diesel/gasoline combination from Maruti Suzuki.
2.3. Private costs
2.3.1. Capital costs
For our analysis, we employ the manufacturers suggested retail
price (MSRP) as our estimate of the capital costs for the vehicle.
The MSRP may not always be an accurate reflection of the costs
of manufacturing and selling the vehicle because manufacturers
also set vehicle prices based upon their expectations of consumers
willingness to pay. Assuming that the manufacturers estimates of
consumers willingness to pay may vary systematically by model
and make, and not by fuel type, our pairing approach accounts
for differences between MSRP and actual market price. This pairing
approach also allows us to control for a number of features that are
not directly captured in our dataset, such as the interior trimming,
air-conditioning and safety features. This approach, however, can-
not account for all features. For example, uncontrolled differences
may remain due to engine displacement and torque variances that
stem from differences in fuel/powertrain, especially for diesel vehi-
cles. We also considered developing hedonic equations to control
for differences in vehicles. For example, Chugh et al. [23] applied
this approach to generate estimates of capital costs that isolated
the value of fuel economy, controlling for differences in torque
[23]. This approach, however, requires a large set of vehicles. Since
the options for electric, LPG and CNG vehicles are more limited in
India, a statistical approach cannot be employed for these vehicles.
Thus, in this case, the benefits of using the full dataset of actual
vehicles outweighs the potential concern of not completely con-
trolling for the value that consumers may place on the different
uncaptured attributes of these vehicles. We will examine the
implications of these pairings further in the discussion section.
Initial capital costs can also be offset by the potential resale
value of the vehicle. Previous analysis conducted in the US market
suggests that diesel and hybrid vehicles retain a greater share of
their value at resale [15]. However, the resale market in India is
still relatively thin. Thus, we assume that there is no difference

Table 1
Summary statistics for the vehicle pairs by size classification. We show the number of pairs, the median, minimum and maximum difference in the capital costs in Indian Rupees
(INR) (2014) for the paired alternative fuel/powertrain (e.g. Diesel, LPG and CNG) and conventional gasoline vehicles, and the median, minimum and maximum fuel economy for
the alternative vehicles and the conventional vehicles. Fuel economy is in km/L for gasoline, diesel and LPG and in km/kg for CNG.

Size Pairing N D Capital cost for alternative to Fuel economy alternative vehicle Fuel economy gasoline vehicle
gasoline vehicle (in INR) (in km/L or km/kg for CNG) (in km/L)
Mini CNG 2 35,000 36.22 25.2
(31,00040,000)
Compact CNG 4 55,000 25.0 19.4
(50,00062,000) (23.826.3) (18.919.7)
Diesel 33 110,000 21.1 16.8
(52,000154,000) (17.723.1) (14.018.6)
LPG 7 31,000 13.2 18.3
(23,00037,000) (11.913.3) (16.818.9)
Mid-sized CNG 2 64,000 21.4 15.5
Diesel 20 117,000 20.0 14.1
(87,000152,000) (19.921.7) (13.717.9)
LPG 1 40,000 11.3 16.4
Premium Diesel 10 107,000 19.9 14.2
(69,000193,000) (16.521.4) (12.914.5)
SUV CNG 1 59,000 20.0 15.1
Diesel 3 73,000 13.0 10.4
(68,000106,000)
LPG 1 22,000 10.9 14.7

Please cite this article in press as: Gilmore EA, Patwardhan A. Passenger vehicles that minimize the costs of ownership and environmental damages in the
Indian market. Appl Energy (2016), http://dx.doi.org/10.1016/j.apenergy.2016.09.096
 E.A. Gilmore, A. Patwardhan / Applied Energy xxx (2016) xxxxxx 5

in the resale value of the vehicle pairs. We conduct a sensitivity Table 2

analysis to test the implications on this assumption in Supplemen- Base case, low and high fuel prices. In brackets, we also show the equivalent figures in
USD/gallon for petrol, diesel and LPG, in USD/kg for CNG and USD/kWh for electricity
tary Information (Section S4). at an exchange rate of 1 USD to 65 INR.

2.3.2. Operating costs and fuel prices Variable Base case Low High

Fuel and operating and maintenance (O&M) are the main vari-
able costs associated with a vehicle. Due to a lack of publicly avail-
able data, we assume that the O&M costs are the same for all of the
paired vehicles. We focus on fuel costs, which depend on driving
patterns, the fuel economy of the vehicle, and the fuel prices. For
this analysis, we assume a base case with a vehicle lifetime of
10 years and 10,750 km travelled per year. In their comprehensive
literature review, Baidya and Borken-Kleefeld [24] found estimates
for Indian drivers ranging from 7500 to 15,000 vehicle kilometers
travel per year (VKT), with final estimates of 930012,200 VKT
for large urban centers and 920010,400 VKT for the rest of India
[24]. We employ the average of the urban center VKT estimates
of 10,750 km per year as our central estimate for urban driving.
We also conduct a sensitivity analysis with a driving profile that
reflects the trend that the number of kilometers driven decrease
with vehicle age. An analysis of vehicles in a mid-sized city (Pune)
in India found that passenger cars start at approximately
14,000 km per year and decline by approximately 790 km per year
[25]. Over a 10 year ownership period, this leads to approximately
100,000 km with an average of approximately 10,000 km per year.
Some vehicles, specifically electric vehicles, may be favored by
consumers who drive short trips in urban centers and thus, the
annual VKT could be closer to the lower end of the range. By con-
trast, diesel vehicles may be preferred by consumers who drive
longer distances. We account for these potential differences in
driving behaviors by conducting a sensitivity analysis in annual
VKT. For our base case analysis, we assume that the consumers
are insensitive to trip length and have equal access to both conven-
tional and alternative refueling facilities.
Fuel economy in India is established using the modified Indian
Driving Cycle (IDC). However, anecdotal evidence and peer-review
literature suggest that the IDC does a poor job of capturing the
actual fuel economy observed by drivers in India [26]. This occurs
because the actual driving patterns (e.g. acceleration and deceler-
ation) and conditions (e.g. % highway compared to % urban driving,
amount of congestion, etc.) are not well represented in the IDC. In
the absence of specific information about driving patterns, we can-
not directly correct for these conditions. If we assume that each
vehicle type has the same error between the IDC and the real-
world performance, then our pairing strategy helps to correct for
this error. The validity of this assumption depends on the perfor-
mance of the different vehicle types on the actual driving cycle.
Thus, we examine the effect of the fuel economy values on the final
calculations through a sensitivity analysis. To calculate the operat-
ing costs for electric vehicles, we estimate the amount of charging
that is required for the same annual driving patterns.
In Table 2, we show the fuel prices used in this work. We
employ the average of the median fuel prices observed in 2015
in major Indian cities as our base case and conduct extensive sen-
sitivity analysis on these values to account for uncertainty about
future fuel prices over the course of the vehicle lifetime. The
state-owned fuel retailers revise the prices two times per month
based on the costs of imports and the INR to US dollar exchange
rate [27]. To develop a time series for the fuel prices over the vehi-
cle lifetime, we start with the forecasts for crude oil prices from the
Energy Information Agency for 20152030 [28]. We then trans-
form the crude oil prices into the price of gasoline in Delhi, using
the relationship from the 2010 Expert Report which found that
the price of gasoline in Delhi should increase by 20 INR/L for a rise
in crude oil prices from 70 USD/barrel to 120 USD/barrel [29]. We
assume that this relationship is linear over the forecasted oil
Petrol prices in INR/La 64.0 48.0 82.0
(3.72) (2.80) (4.78)
b
Diesel prices in INR/L 46.00 28.0 62.0
(2.74) (1.63) (3.61)
Liquefied propane gas prices in INR/La 40.7 24.6 54.6
(2.37) (1.43) (3.18)
Compressed natural gas prices in INR/kga 38.0 28.0 47.5
(1.52) (1.12) (1.90)
Electricity prices in INR/kWhc 5.65 2.55 13.8
(0.09) (0.04) (0.21)
a
Government of India [27].
b
Derived from relationship in Anand et al. [30].
c
Bijli Bachao [32].
prices. In 2014, the Government of India ended subsidies on diesel
fuel. We employ an estimate from Anand et al. [30] that, in the
absence of the subsidies, there should be an approximate 20 INR/
L difference between diesel and gasoline [30]. In our base case,
we assume that this price difference will hold for the entire time
period of our analysis (e.g. 20152025).
The price of CNG and LPG for automobiles is also market based.
We use the 2015 average city price of 40.7 INR/L and 38 INR/kg for
CNG and LPG, respectively [27]. Projecting future natural gas prices
is complicated. Presently, natural gas is subsidized as an input to
fertilizers in the agricultural sector [31], although there is the
potential for a more integrated natural gas market. There are also
shale reserves that may alter the supply of natural gas. Thus, we
explore natural gas prices through a sensitivity analysis for low
and high values at 75% and 125% of the base case. We assume that
LPG prices will trend with petrol and diesel prices. Electricity
prices are regulated by the State Electricity Regulatory Commis-
sions, and there are large differences in prices throughout the
country. Electricity prices and reliability can vary significantly even
within an urban region. In 20142015, the median electricity price
in India was approximately 5.65 INR/kWh for higher electricity
consumption with a range of 3.9 INR/kWh to 13.85 kWh/INR [32].
2.4. Costs of damages from air quality and climate change
Vehicles impose damages at the societal level that are not inter-
nalized into the private cost calculations, namely premature mor-
tality from air emissions and climate damages from GHG
emissions. To evaluate the costs of these damages, we compile
emission characteristic data for all new four-wheel passenger cars
available in the Indian market place. We then combine these emis-
sion factors with driving patterns and damage valuation caused by
the air pollutants and GHG emissions to calculate the difference in
societal cost of each vehicle pair. To transform the emissions into
human health endpoints, we follow the approach described in a
working paper by Cropper et al. [33] and further elaborated in
Greco et al. [34]. This approach has been developed as a way to cir-
cumvent the need for comprehensive air quality modeling to con-
vert the emissions into ambient concentrations when there are
insufficient data or for a screening level [33,34].
2.4.1. Greenhouse gas and conventional air pollutant emissions
We consider the emissions of primary fine particulate matter
(PM2.5), nitrogen oxides (NOx), sulfur dioxide (SO2) and carbon
dioxide (CO2). We show the emissions factors related to the operat-
ing phase either tailpipe or from vehicle charging in g/km for our

Please cite this article in press as: Gilmore EA, Patwardhan A. Passenger vehicles that minimize the costs of ownership and environmental damages in the
Indian market. Appl Energy (2016), http://dx.doi.org/10.1016/j.apenergy.2016.09.096
6 E.A. Gilmore, A. Patwardhan / Applied Energy xxx (2016) xxxxxx
Table 3
Emission factors for NOx and PM2.5 for vehicles in g/km and coal fired power plants in
g/kWh.
Source NOx
Gasoline vehiclea 0.08
Diesel vehiclea 0.25
LPG vehicleb 0.076
CNG vehicleb 0.064
Coal power plantc 2.0
(1.82.3)
a
b
DieselNet [17], Bharat IV standards.
International Vehicle Emissions Model V. 2.0, 2008.
c
Cropper et al. [33], Mittal et al. [37].
vehicle types in Table 3. For conventional air pollutants, new
vehicles in major Indian cities are required to meet Bharat Stage
IV equivalent to EURO IV emission standards with 50 ppm sulfur
diesel fuel [17]. There are separate standards for gasoline and diesel
fueled vehicles. For CNG and LPG vehicles, we adopt the tailpipe
emission factors from the International Vehicle Model V.2 [35].
For electric vehicles, the net emissions depend on the type of
generation used to charge the vehicle [36]. A detailed analysis of
the generation mix at the location of charging and its dispatch at
the time of vehicle charging is beyond the scope of this analysis,
as we aim to provide broad guidance on vehicle choices for Indian
cities. Rather, we address this issue by using a bounding analysis of
a renewable energy source (e.g. zero emissions) and a coal power
plant, which represents the main source of the electricity in India.
We use emission factors for coal plants from [33,37]. The coal gen-
eration plants have a median SO2 emission factor of 6.9 g/kWh, and
we evaluate a range of 6.37.9 g/kWh. These plants have CO2 emis-
sion factors of 940 g/kWh. We evaluate a range of 870980 g/kWh.
2.4.2. Valuation of emissions
To complete the societal cost calculations, we translate the
emissions to damages and then apply estimates of monetary val-
ues of damages caused by the pollutants. To estimate the valuation
of conventional air pollutants, the emissions are first transformed
into their equivalent ambient concentrations. Second, the ambient
concentrations are translated into health effects by using epidemi-
ological relationships of exposure to health endpoints combined
with the exposed population and population specific baseline
health rates. Third, the human health endpoints are monetized
using population appropriate estimates of willingness-to-pay
(WTP) to avoid these damages. Here, we employ a screening
approach to estimate societal costs that we first encountered in
Cropper et al. [33] and was subsequently examined in greater
detail by Greco et al. [34].
To evaluate the societal costs from air emissions that affect
human health, emissions need to be transformed into ambient con-
centrations and exposure. This is generally achieved with a chem-
ical transport model (CTM). These models, however, require
specialized expertise and are computationally intensive; thus,
there is often a desire for screening level tools. For the United
States, there are now several reduced form societal cost values
(expressed in $ per ton of emission) that allow analysts to bypass
the CTM (e.g. [3840]). For developing countries, there are no
available models; thus, Cropper et al. [33] and Greco et al. [34]
developed an approach using previously modeled intake fractions.
Intake fractions (iF) is the ratio of the mass of pollutant inhaled to
the mass of pollutant emitted integrated over a population and a
period of time [41]. We provide additional details regarding intake
fractions and how they are modeled in Supplemental Information
(Section S1). Air quality models, including CTMs, are used to
determine the change in concentration attributable to a change
Please cite this article in press as: Gilmore EA, Patwardhan A. Passenger vehicles that minimize the costs of ownership and environmental damages in the
Indian market. Appl Energy (2016), http://dx.doi.org/10.1016/j.apenergy.2016.09.096
in emissions. These concentrations are mapped to the relevant
population at the spatial resolution of the model. The intake frac-
tions (iF) can be transformed into $ per ton estimates with popula-
PM2.5 tion data and valuations that are appropriate for an Indian context.
0.001 We convert the intake fractions in ppm (or lg of pollutant inhaled
0.025 per g of pollutant emitted) into their equivalent ambient concen-
0.001 trations using Eq. (3):
5  105
0.14 IF  Q
(0.100.21) Dc 3
BR  P
where Dc is the change in ambient concentration in lg/m3;
iF is the intake fraction in ppm;
Q is the emission rate in g/day;
P is the population exposed to the change in ambient
concentration;
BR is the breathing rate (estimated at 14.5 m3 per day).
For primary PM2.5, we follow the approach outlined in Greco
et al. [34] and employ the intake fractions calculated in Apte
et al. [42]. Over 340 urban centers in India, there is a population
weighted intake fraction of 51 ppm; however, three major Indian
cities experience much higher intake fractions for primary PM2.5
of 79, 105, and 151 ppm for Mumbai, Delhi and Kolkata, respec-
tively [42]. We show the societal costs for these cities in a sensitiv-
ity analysis in Supplemental Information (Section S2). For the
secondary PM2.5 that is formed chemically from NOx and SO2, we
employ the generic values of 0.20 ppm and 0.99 ppm recom-
mended by Humbert et al. [41].
We then convert PM2.5 concentrations to premature mortality
using the concentration-response (C-R) functions developed by a
range of sources [43]. We consider three C-R functions for our anal-
ysis. The first two are the all-cause mortality relationships for the
population over 30 years old that are employed by the U.S. Envi-
ronmental Protection Agency for its benefits analysis for air quality
regulations: Krewski et al. [44] as the primary estimate and
Lepeule et al. [45] as a sensitivity analysis. The change in popula-
tion level incidence is given by Eq. (4).
 
Dy y0  P  ebDc  1 4
where y0 is the baseline incidence for the population (e.g. pre-
mature mortality events per 100,000 people);
P is the population exposed to the change in ambient
concentration;
b is the effect estimate; and,
Dc is the change in ambient PM concentrations.
For the b, we employ 0.00583 as our primary estimate [44] and
0.0131 as a sensitivity analysis [45]. We use a country-level base-
line mortality risk for the population aged 30 years or older from
the World Health Organization life tables of 1481 per 100,000.
Additionally, we consider a logarithmic form of the C-R function
where the magnitude of the response is a function of the back-
ground ambient concentrations. This follows the guidance of Ostro
[46] based on the observation that the risk of premature mortality
from exposure to PM2.5 from the US studies may not be appropriate
for developing countries which experience much higher back-
ground ambient PM2.5 concentrations. More recently, a C-R func-
tion, known as the integrated exposure response (IER) function,
was developed to cover the risks of exposure over a wider range
of ambient concentrations through the Global Burden of Disease
(GBD) project [47]. The IER function, however, uses ischemic heart
disease as its health endpoint; thus, the results from the IER func-
tion are not directly comparable to the all-cause mortality
function.
 E.A. Gilmore, A. Patwardhan / Applied Energy xxx (2016) xxxxxx
Finally, we monetize this endpoint using the Value of a Statisti-
cal Life (VSL). We generate our core VSL estimate of 14.4 million
INR using the approach outlined in Hammitt and Robinson [48].
We start with the VSL from the US Environmental Protection
Agency (USEPA) of $7.4 million USD and adjust by income, using
GDP per capita as a proxy, with an income elasticity of 1.5. We
show this relationship in Eq. (4). We draw on a survey of the liter-
ature in Cropper et al. [33] to generate bounds of 2.5 million INR to
50 million INR. We present additional information in the Supple-
mental Information (Section S2).
 e
IncomeIndia
VSLIndia VSLUSA 
IncomeUSA
where
VSLIndia is the estimated value of a statistical life in India;
VSLUSA is the USEPA value of a statistical life;
IncomeIndia and IncomeUSA are the GDP/capita in India and USA,
respectively; and, e is the elasticity.
For charging the electric vehicle, we also have to account for the
emissions from the power plant. To estimate the damages that
would arise from using a coal plant to charge the electric vehicle,
we adopt the values from Cropper et al. [33] of 23 mortalities
per 1000 metric tons of PM2.5, 10 mortalities per 1000 metric tons
of SO2 and 9 mortalities per 1000 metric tons of NOx [33]. In this
analysis, this would represent a generic coal plant. We also show
a bounding calculation for the damages for the electric vehicle in
Supplemental Information (Section S3).
To estimate the societal costs due to CO2, we use the values of
the global damages developed for the US government, referred to
as the Social Cost of Carbon (SCC) [49]. The central SCC of $36/
metric ton (2007 USD) represents the global damages from one
metric ton of CO2 emitted in 2015. It was developed by averaging
of the estimates of damages from three models using a 3% discount
rate. We follow the guidance of the Interagency Working Group
report, conducting a sensitivity analysis with the lower value of
$11/metric ton (average and discount rate of 5%) and a higher value
of $56/metric ton (average and discount rate of 2.5%). We also
consider the 95th percentile SCC estimate across all three models
at a 3% discount rate. This value of $105/metric ton represents a
higher end value for the damages of climate impacts.
3. Results and discussion
3.1. Private and societal costs for diesel, CNG and LPG vehicles
In Fig. 2, we show the median difference in the costs of owning
and operating diesel, CNG and LPG passenger cars in India com-
pared to the paired petrol vehicle by size category. We also show
the minimum and maximum values over the capital costs, fuel
costs and driving patterns. The high fuel economy and low CNG
fuel costs make the CNG vehicles the least-cost options in terms
of private costs. The LPG version is more expensive, despite total
fuel costs that are lower than comparable gasoline vehicles. In
the premium vehicles category, gasoline fueled cars have the most
expensive private costs. This is because of higher capital costs,
higher fuel prices, and lower fuel economies of petrol vehicles
compared to the other options.
In Table 4, we show the private costs as well as the costs of the
damages from air quality and climate change for the average of all
cities. For the average Indian metropolitan area, we find that all
vehicles have a lower full cost than their gasoline equivalents.
We find that CNG has the largest benefit over a petrol vehicle in
all size categories in which the option exists. The social costs due
Please cite this article in press as: Gilmore EA, Patwardhan A. Passenger vehicles that minimize the costs of ownership and environmental damages in the
Indian market. Appl Energy (2016), http://dx.doi.org/10.1016/j.apenergy.2016.09.096
7
to air quality emissions for a CNG vehicle are 3754% lower due
to lower emission factors and the absence of sulfur in CNG. Like-
wise, the social costs due to CO2 for a CNG vehicle are 5682%
lower than a conventional petrol vehicle belonging to the same
size category because of the lower carbon content in CNG com-
pared to petrol. The social costs due to CO2 emissions for an LPG
vehicle are also lower because of the lower carbon content in
LPG. While the air quality costs for diesel vehicles are higher than
the equivalent gasoline vehicle, the overall full costs are lower due
to the fuel savings as well as the lower CO2 emissions. If diesel
prices were to be closer to parity with gasoline as is the case in
the United States, then the full costs for diesel could exceed those
5
of a gasoline fueled vehicle. Additionally, this analysis was con-
ducted assuming that the diesel fueled vehicles attained the Bharat
standards. Higher emissions would further increase the full cost of
the diesel compared to gasoline vehicles. Presently, there is only
one domestic hybrid powertrain vehicle in the Indian market.
The fuel savings over a petrol vehicle are substantial. A more
appropriate pairing, however, is a conventional diesel vehicle.
Here, the fuel savings are more modest and do not justify the price
premium.
3.2. Private and societal costs for electric vehicles
The final passenger vehicle option that we evaluate is the
potential benefits for the plug-in electric vehicle, the Mahindra
e20. At the unsubsidized vehicle price, we find that we find that
the fuel savings of shifting to electric vehicle are insufficient to
recover the initial price premium over the Tata Nano however,
the Maruti wagon may breakeven at higher gasoline prices and
lower electricity prices. With the subsidy, the fuel savings are suf-
ficient to cover the difference in the capital costs for both our more
expensive (Maruti Suzuki Wagon R) and our less expensive (Tato
Nano) pairings. Thus, we investigate whether the 100,000 INR
governmental subsidy is justified from the perspective of the soci-
etal costs. We bound the calculation with a zero carbon electricity
source and a coal power plant, as coal dominates the present
electricity generation fleet in India. We show the societal cost
calculations in Supplemental Information (Section S3). In the urban
center, there are benefits in the reduction of human health impacts
by displacing a gasoline vehicle. However, if charged with a coal
power source, the total air quality costs are higher despite the
urban benefit. From a perspective of climate damages, charging
from either a renewable or coal power plant produces net benefits.
Overall, we calculate a median societal benefit for charging with a
renewable power source of 42,000 INR with a range of 13,000
72,400 INR. At the high end, this begins to approximate the
100,000 INR subsidy. Charging with an indicative coal plant in
India, however, reduces the benefits of electric vehicles to almost
zero or even slightly positive at higher VSLs. This is consistent
with the findings of Michalek et al. [10] who concluded that
plug-in electric vehicles with large battery packs may increase
damages depending largely on GHG and SO2 emissions from
electricity [10].
An additional consideration for electric vehicles is how they will
retain their value over time compared to their gasoline equivalent.
We assumed that there is no difference in the resale prices
between the paired vehicles. We justified this assumption for our
core analysis as the resale market in India is thin, making it difficult
to develop stable estimates. We investigated the implications of
this assumption in Supplemental Information (Section S4). For
electric vehicles compared to gasoline, however, this assumption
may be weaker as newer vehicles tend to depreciate faster than
more conventional options [15]. This may further reduce the desir-
ability of the electric vehicle from a private perspective.
8 E.A. Gilmore, A. Patwardhan / Applied Energy xxx (2016) xxxxxx

Fig. 2. Median values for the difference in capital costs and fuel costs for the paired vehicles by vehicle size and fuel type at a 10% discount rate. We also show the minimum
and maximum values over the paired values for the capital costs, and the minimum and maximum values for the difference in fuel costs accounting for the range of fuel prices
and driving patterns.

Table 4
Median values with lowest and highest values in brackets for the difference in the private and societal costs for air quality and carbon dioxide the paired vehicles by vehicle size
and fuel type in 1000 s of INR. A VSL of 14.4 million INR is used for the median value with bounds of 650 million. The central estimate for climate damages is calculated at 2670
INR/metric ton of CO2 with bounds of 8204160 INR/metric ton.

Vehicle size Fuel type D Private costs (in 1000 INR) D Air quality costs (in 1000 INR) D CO2 costs (in 1000 INR) D Full costs (in 1000 INR)
Mini-Car CNG 81 0.65 13.6 95.4
(31.7, 185) (0.27, 2.27) (4.16, 39.8) (36.1, 227)
Compact Diesel 12 10.2 7.1 9.22
(57.5, 184) (4.2435.3) (2.16, 20.7) (59.6, 169)
CNG 85 0.65 14.0 99.3
(24.2, 223) (0.27, 2.27) (4.29, 41.0) (28.8, 266)
LPG 0.6 0.01 10.1 10.7
(16.0, 71.9) (0, 0.04) (3.09, 29.5) (12.9, 101)
Mid-sized Diesel 55 10.2 15.6 60.4
(25.0, 259) (4.2435.3) (4.76, 45.4) (24.5, 269)
CNG 126 0.65 17.3 144
(53.4, 304) (0.27, 2.27) (5.28, 50.4) (58.9, 356)
LPG 37.5 0.01 18.5 56.0
(21.1, 110) (0, 0.04) (5.66, 54.0) (26.8, 164)
Luxury Diesel 62 10.2 14.9 66.5
(67.9, 254) (4.2435.3) (3.56, 43.5) (67.6, 262)
SUV Diesel 127 10.2 11.7 128
(51.2, 300) (4.2435.3) (3.56, 34.0) (50.4, 299)
CNG 123 0.65 19.2 143
(53.7, 291) (0.27, 2.27) (5.85, 55.9) (59.8, 349)
LPG 74.1 0.01 22.0 96.1
(52.4, 128) (0, 0.04) (6.73, 64.2) (59.1, 192)

3.3. Implications for passenger vehicle policy
These results have important implications in the context of
Indian transport policy. Historically, smaller cars have been popu-
lar in India, as the population shifts from two wheel motorcycles
and scooters to four wheel passenger cars. The Tata Nano (the
one lakh car) as well as other small vehicles suggests that there
may be sharp increases in the sales of passenger cars in the near
future. On the fleet level, small cars would present lower costs
not only for the private consumer but also for the society in terms

Please cite this article in press as: Gilmore EA, Patwardhan A. Passenger vehicles that minimize the costs of ownership and environmental damages in the
Indian market. Appl Energy (2016), http://dx.doi.org/10.1016/j.apenergy.2016.09.096
 E.A. Gilmore, A. Patwardhan / Applied Energy xxx (2016) xxxxxx
of avoided externality costs. Regardless of the vehicle size, CNG
vehicles would result in both lower ownership costs to consumers
as well as lower social costs compared to conventional fuel types
such as petrol or diesel. The benefits of CNG have already been
observed in Indian urban centers. Shifting public buses from diesel
to CNG has been credited with some improvements in urban air
quality in India [50]. However, presently CNG passenger vehicles
comprise less than 1% of the Indian vehicle fleet with approxi-
mately 900 filling stations primarily in urban cities where improv-
ing air quality is a priority [32]. The continued growth of natural
gas use in the transportation sector depends on the expansion of
the gas transportation network, the construction of the necessary
infrastructure within the country, and the availability of natural
gas for CNG fueled vehicles. It is beyond the scope of this effort
to fully model the potential trade-offs of using natural gas as a
vehicle fuel compared to other potential uses either for passenger
vehicles (e.g. electrification of the vehicle fleet with natural gas
generation) or other applications, specifically in the agricultural
sector.
With growing income levels, however, the average size of pas-
senger cars in the fleet is likely to increase. Based on the vehicles
that are currently available in the Indian market, a fleet dominated
by mid-sized and premium sized vehicles may mean diesel vehi-
cles comprise a larger proportion of the fleet. The social costs asso-
ciated with a diesel dominated fleet could be significantly higher
than CNG or petrol. Previous work has addressed the trade-off
between adverse human health effects and climate benefits for die-
sel vehicles and has noted that strong emission controls are needed
to manage the health risks [51]. Difficulties in attaining the more
stringent emission standards, including Bharat VI, however, may
further restrict the use of diesels. Presently, there are few hybrid
options that provide fuel economy and air quality improvements;
and thus, in absence of hybrid options, CNG or LPG are preferred
even at larger vehicle sizes over gasoline.
Finally, we find that electrification of the vehicle fleet could pro-
duce large benefits, especially in large urban centers. However, the
avoided damages to human health and climate impacts only just
approximate the subsidy that is required to make these vehicles
viable from a private perspective. Moreover, if the goal is to reduce
GHG emission globally, electrification of the vehicle fleet in India
only supports this goal if the carbon intensity of the power system
can be reduced substantially [52]. We investigated this possibility
with a bounding scenario of renewable energy, such as photo-
voltaic solar plant. We note, however, that at a fleet level, this
would require a substantial expansion of generation capacity.
While it is beyond the scope to conduct a detailed analysis of these
costs, we conduct a back of the envelope calculation which we
show in Supplemental Information (Section S5). For each electric
vehicle, we approximate 60,000 INR of new electricity generation
infrastructure. For our analysis, this expansion would be difficult
to justify solely for electric vehicles; however, there are additional
benefits to the electricity generation sector that may offset part of
the infrastructure costs.
Our findings are based on several assumptions. First, we have
used estimates of how air quality emissions are translated into
exposure and monetization that are adopted from a US context.
The actual social costs of the damages caused by the pollutants
in Indian cities may differ from our estimates. However, our rela-
tive rankings of the vehicles is robust to different valuations. Sec-
ond, we assume that consumers have the same driving patterns
regardless of the type of vehicle. This is less likely since consumers
tend to select vehicles that are consistent with their driving needs.
For example, electric vehicles may be more appealing to urban res-
idents while diesel vehicles may be more suited for highway driv-
ing. Such preferences also influence the benefits and distributions
of the social costs. Finally, we assume that the relative fuel prices
Please cite this article in press as: Gilmore EA, Patwardhan A. Passenger vehicles that minimize the costs of ownership and environmental damages in the
Indian market. Appl Energy (2016), http://dx.doi.org/10.1016/j.apenergy.2016.09.096
9
remain constant over the next decade, although climate change
policy as well as other political events may lead to changes in
the relative price of these fuels.
4. Conclusions
It is critical to identify passenger vehicles that reduce environ-
mental and human health burdens, especially in the context of
developing countries. We have presented a full cost analysis of
new passenger cars in India. We find that the full costs of owning
and operating small vehicles are lower than other options. While
both CNG and LPG are both attractive options in terms of societal
costs, the higher LPG private costs make CNG vehicles the most
attractive option in terms of both private and societal costs for
all size-categories. The social costs of operating diesel vehicles
are the highest. Consistent with findings in the US, we find that
the savings in fuel and benefits from lower social costs of electric
vehicles do not compensate for the high capital cost premium
[10,11]. However plug-in electric vehicles have the potential to
reduce externality costs if emissions from electricity generation
are reduced substantially. We conclude that the promotion of the
use of CNG vehicles in cities is the best option for Indian policy
makers to reduce GHG and air pollution with sustained and eco-
nomically feasible supporting infrastructure improvements in
cities and along major highways and less urban areas. As the vehi-
cle fleet in India is expected to rapidly increase, this work makes a
timely contribution to the literature by producing relevant esti-
mates that can be used to inform India policy decisions.
Acknowledgements
We acknowledge the assistance of Gokul Iyer and Eric Lui for
compiling the car price dataset, and Delia Belausteguigoitia and
Elizabeth Tennant for editing and preparing the text for
publication.
Appendix A. Supplementary material
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.apenergy.2016.
09.096.
References
[1] National Research Council. Implementing emission controls on mobile sources,
Chapter 4. In: Air quality management in the United States. Washington,
DC: The National Academies Press; 2004. p. 13373. doi: http://dx.doi.org/
10.17226/10728.
[2] Sims R, Schaeffer R, Creutzig R, Cruz-Nez X, DAgosto M, Dimitriu D, et al.
Transport. In: Edenhofer O, Pichs-Madruga R, Sokona Y, Farahani E, Kadner S,
Seyboth K, Adler A, Baum I, Brunner O, Eickemeier P, Kriemann B, Savolainen J,
Schlmer S, von Stechow C, Zwickel T, Minx JC, editors. Climate change 2014:
mitigation of climate change. Contribution of working group III to the fifth
assessment report of the intergovernmental panel on climate
change. Cambridge, United Kingdom and New York, NY, USA: Cambridge
University Press; 2014.
[3] Bhalla K, Shotten M, Cohen A, Brauer M, Shahraz S, Burnett R, et al. Transport
for health: the global burden of disease from motorized road
transport. Washington, DC: World Bank Group; 2014.
[4] International Energy Agency (IEA). Energy technology perspectives 2012. IEA:
Paris, France; 2012. <http://www.iea.org/Textbase/npsum/ETP2012SUM.pdf>.
[5] Harrington W, McConnell V. Motor vehicles and the environment. Washington
DC: Resources for the Future; 2003. <http://www.rff.org/files/sharepoint/
WorkImages/Download/RFF-RPT-carsenviron.pdf>.
[6] Roosen J, Marneffe W, Vereeck L. A review of comparative vehicle cost analysis.
Transport Rev 2015;1647(September):129. doi: http://dx.doi.org/10.1080/
01441647.2015.1052113.
[7] Keefe R, Griffin JP, Graham JD. The benefits and costs of new fuels and engines
for light-duty vehicles in the United States. Risk Anal: Off Publ Soc Risk Anal
2008;28(5):114154. doi: http://dx.doi.org/10.1111/j.1539-6924.2008.01099.
x.
10 E.A. Gilmore, A. Patwardhan / Applied Energy xxx (2016) xxxxxx
[8] Lipman TE, Delucchi MA. A retail and lifecycle cost analysis of hybrid electric
vehicles. Transport Res Part D: Transport Environ 2006;11(2):11532. doi:
http://dx.doi.org/10.1016/j.trd.2005.10.002.
[9] Torchio MF, Santarelli MG. Energy, environmental and economic comparison
of different powertrain/fuel options using well-to-wheels assessment, energy,
and external costs European market analysis. Energy 2010;35:415671. doi:
http://dx.doi.org/10.1016/j.energy.2010.06.037.
[10] Michalek JJ, Chester M, Jaramillo P, Samaras C. Valuation of plug-in vehicle life-
cycle air emissions and oil displacement benefits 2011; 108(40). http://dx.doi.
org/10.1073/pnas.1104473108.
[11] Tessum CW, Hill JD, Marshall JD. Life cycle air quality impacts of conventional
and alternative light-duty transportation in the United States. Proc Natl Acad
Sci 2014;111(52):184905. doi: http://dx.doi.org/10.1073/pnas.1406853111.
[12] Tamayao MA, Michalek JJ, Hendrickson C, Azevedo IM. Regional variability and
uncertainty of electric vehicle life cycle CO emissions across the United States.
Environ Sci Technol 2015;49(14):884455. doi: http://dx.doi.org/10.1021/acs.
est.5b00815.
[13] Society of Indian Automobile Manufacturers (SIAM). Automobile domestic
sales trends; 2015. <http://www.siamindia.com/statistics.aspx?mpgid=8&
pgidtrail=14>.
[14] Al-Alawi BM, Bradley TH. Total cost of ownership, payback, and consumer
preference modeling of plug-in hybrid electric vehicles. Appl Energy
2013;103:488506. doi: http://dx.doi.org/10.106/j.apenergy.2012.10.009.
[15] Gilmore EA, Lave LB. Comparing resale prices and total cost of ownership for
gasoline, hybrid and diesel passenger cars and trucks. Transp Pol 2013;27
(May):2008.
[16] Bansal G, Bandivadekar A. Overview of Indias Vehicle Emission Control
Program; 2013. <http://www.theicct.org/sites/default/files/publications/ICCT_
IndiaRetrospective_2013.pdf>.
[17] DieselNet. Emission Standards India; 2015. <https://www.
dieselnet.com/standards/in/>.
[18] Brauer M, Freedman G, Frostad J, et al. Ambient air pollution exposure
estimation for the Global Burden of Disease 2013. Environ Sci Technol
2016;50:7988. doi: http://dx.doi.org/10.1021/acs.est.5b03709.
[19] Department of Heavy Industry (Government of India). National electric
mobility mission plan 2020; 2012. <http://dhi.nic.in/writereaddata/Content/
NEMMP2020.pdf>.
[20] Society of Indian Automobile Manufacturers (SIAM). SIAM 6th fuel economy
declaration (2014) updated as on April 2014; 2014. <www.siam.in/
uploads/filemanager/246th-2W-FE-Data-Declaration.pdf>.
[21] Mahindra. Government FAME Demand Incentive; 2015. <http://
mahindrareva.com/buy/Govt-FAME-Demand-Incentive>.
[22] Saxena S, Gopal A, Phadke A. Electrical consumption of two-, three- and four-
wheel light-duty electric vehicles in India. Appl Energy 2014;115:58290. doi:
http://dx.doi.org/10.1016/j.apenergy.2013.10.043.
[23] Chugh R, Cropper M, Narain U. The cost of fuel economy in the Indian
passenger vehicle market. Energy Pol 2011;39(11):717483. doi: http://dx.doi.
org/10.1016/j.enpol.2011.08.037.
[24] Baidya S, Borken-Kleefeld J. Atmospheric emissions from road transportation
in India. Energy Pol 2009;37(10):381222. doi: http://dx.doi.org/10.1016/j.
enpol.2009.07.010.
[25] International Vehicle Emission Model. Pune vehicle activity study; 2004.
[26] Nesamani KS, Subramanian KP. Impact of real-world driving characteristics on
vehicular emissions. JSME Int J Ser B 2006;49(1):1922.
[27] Government of India, Ministry of Petroleum & Natural Gas, Petroleum
Planning & Analysis Cell; 2015. <http://ppac.org.in/content/149_1_
pricespetroleum.aspx>.
[28] Environmental Information Agency. Annual energy outlook; 2015. <http://
www.eia.gov/forecasts/aeo/>.
[29] Government of India. Report of the expert group on a viable and sustainable
system of pricing petroleum products; 2010. <http://petroleum.nic.in/docs/
reports/reportprice.pdf>.
[30] Anand R, Coady D, Mohommad A, Thakoor VV, Walsh JP. The fiscal and welfare
impacts of reforming fuel subsidies in India. IMF working papers 2013;13
(128):1. http://dx.doi.org/10.5089/9781484305164.001.
[31] Sen A. Gas Pricing Reform in India: implications for the Indian gas landscape;
2015. <http://www.oxfordenergy.org/wpcms/wp-content/uploads/2015/04/
NG-96.pdf>.
[32] Bijli Bachao. Domestic electricity tarrifs; 2015. <https://www.bijlibachao.com/
news/domestic-Electricity-Lt-Tariff-Slabs-and-Rates-for-All-States-in-India-
in-2015.html>.
Please cite this article in press as: Gilmore EA, Patwardhan A. Passenger vehicles that minimize the costs of ownership and environmental damages in the
Indian market. Appl Energy (2016), http://dx.doi.org/10.1016/j.apenergy.2016.09.096
[33] Cropper M, Gamkhar S, Malik K, Limonov A, Partridge I. The health effects of
coal electricity generation in India. RFF discussion paper 2012;1225. doi:
http://dx.doi.org/10.2139/ssrn.2093610.
[34] Greco SL, Belova A, Huang J. Benefits of decreased mortality risk from
reductions in primary mobile source fine particulate matter: a limited data
approach for urban areas worldwide. Risk Anal 2016. doi: http://dx.doi.org/
10.1111/risa.12612.
[35] IVE Model V. 2.0. International vehicle emissions model attachment A.
Development of the base emission rates for use in the IVE model (April); 2008.
[36] Sioshansi R, Denholm P. Emissions impacts and benefits of plug-in hybrid
electric vehicles and vehicle-to-grid services. Environ Sci Technol 2009;43
(4):1199204. doi: http://dx.doi.org/10.1021/es802324j.
[37] Mittal ML, Sharma C, Singh R. Decadal emission estimates of carbon dioxide,
sulfur dioxide, and nitric oxide emissions from coal burning in electric power
generation plants in India. Environ Monit Assess 2014;68576866. doi: http://
dx.doi.org/10.1007/s10661-014-3894-3.
[38] Fann N, Fulcher CM, Hubbell BJ. The influence of location, source, and emission
type in estimates of the human health benefits of reducing a ton of air
pollution. Air Qual Atmos Health 2009;2(3):16976. doi: http://dx.doi.org/
10.1007/s11869-009-0044-.
[39] Heo J, Adams PJ, Gao HO. Reduced-form modeling of public health impacts of
inorganic PM2.5 and precursor emissions. Atmos Environ 2016;137:809. doi:
http://dx.doi.org/10.1016/j.atmosenv.2016.04.026.
[40] Muller NZ, Mendelsohn R. Measuring the damages of air pollution in the
United States. J Environ Econ Manage 2007;54(1):114. doi: http://dx.doi.org/
10.1016/j.jeem.2006.12.002.
[41] Humbert S, Marshall JD, Shaked S, Spadaro JV, Nishioka Y, Preiss P, et al. Intake
fractions for particulate matter: recommendations for life cycle assessment.
Environ Sci Technol 2011;45(11):480816.
[42] Apte JS, Bombrun E, Marshall JD, Nazaroff WW. Global intraurban intake
fractions for primary air pollutants from vehicles and other distributed
sources. Environ Sci Technol 2012;46(6):341523. doi: http://dx.doi.org/
10.1021/es204021h.
[43] Fann N, Gilmore EA, Walker KD. Characterizing the long-term PM2.5
concentration response function: comparing the strengths and weaknesses
of research synthesis approaches. Risk Anal 2015. doi: http://dx.doi.org/
10.1111/risa.12435.
[44] Krewski D, Jerrett M, Burnett RT, Ma R, Hughes E, Shi Y, et al. Extended follow-
up and spatial analysis of the American cancer society study linking
particulate air pollution and mortality. Cambridge, MA: Health Effects
Institute; 2009.
[45] Lepeule J, Laden F, Dockery D, Schwartz J. Chronic exposure to fine particles
and mortality: an extended follow-up of the Harvard Six Cities Study from
1974 to 2009. Environ Health Perspect 2012;120:96570. doi: http://dx.doi.
org/10.1289/ehp.1104660.
[46] Ostro B. Outdoor air pollution: assessing the environmental burden of disease
at national and local levels. Geneva: World Health Organization; 2004.
[47] Burnett RT, Pope III CA, Ezzati M, Olives C, Lim SS, Mehta S, et al. An integrated
risk function for estimating the global burden of disease attributable to
ambient fine particulate matter exposure. Environ Health Perspect
2014;122:397403.
[48] Hammitt JK, Robinson LA. The income elasticity of the value per statistical life:
transferring estimates between high and low income populations. J Benefit-
Cost Anal 2011;2(1). doi: http://dx.doi.org/10.2202/2152-2812.1009.
[49] Interagency Working Group on Social Cost of Carbon (United States
Government). Technical support document - technical update of the social
cost of carbon for regulatory impact analysis - under executive order 12866;
2013. <https://www.whitehouse.gov/sites/default/files/omb/
inforeg/social_cost_of_carbon_for_ria_2013_update.pdf>.
[50] Takeuchi A, Cropper M, Bento A. The impact of policies to control motor
vehicle emission in Mumbai India. J Reg Sci 2007;47(1):2746. doi: http://dx.
doi.org/10.1111/j.1467-9787.2007.00498.x.
[51] Mazzi EA, Dowlatabadi HB. Air quality impacts of climate mitigation: UK
policy and passenger vehicle choice. Environ Sci Technol 2007;41(2):38792.
doi: http://dx.doi.org/10.1021/es060517w.
[52] Deetman S, Hof AF, Girod B, van Vuuren DP. Regional differences in mitigation
strategies: an example for passenger transport. Reg Environ Change 2015;15
(6):98795. doi: http://dx.doi.org/10.1007/s10113-014-0649-1.