Vous êtes sur la page 1sur 13

International Journal of Green Energy

ISSN: 1543-5075 (Print) 1543-5083 (Online) Journal homepage: http://www.tandfonline.com/loi/ljge20

Evaluating the efficiency of green vehicles and


diesel vehicles

Fu-Kwun Wang & Masaya Saito

To cite this article: Fu-Kwun Wang & Masaya Saito (2016) Evaluating the efficiency of green
vehicles and diesel vehicles, International Journal of Green Energy, 13:11, 1163-1174, DOI:
10.1080/15435075.2016.1183206

To link to this article: http://dx.doi.org/10.1080/15435075.2016.1183206

Published online: 12 Sep 2016.

Submit your article to this journal

Article views: 49

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at


http://www.tandfonline.com/action/journalInformation?journalCode=ljge20

Download by: [Imperial College London Library] Date: 17 March 2017, At: 23:42
INTERNATIONAL JOURNAL OF GREEN ENERGY
2016, VOL. 13, NO. 11, 11631174
http://dx.doi.org/10.1080/15435075.2016.1183206

Evaluating the efficiency of green vehicles and diesel vehicles


Fu-Kwun Wang and Masaya Saito
Department of Industrial Management, National Taiwan University of Science and Technology, Taipei City, Taiwan

ABSTRACT KEYWORDS
One of the major interests of governments in the area of environmental protection is to reduce green- Data envelopment analysis;
house gas (GHG) emissions and air pollutants emissions (HC, CO, NOx, PM, and SOx) from vehicles. This efficiency; green vehicles;
truncated regression
paper aims at evaluating the efficiency of green and diesel vehicles. Two hundred fifty vehicles were
selected for comparison study. The results show that the efficiency of each type of vehicle hardly differed
in 2008 and 2009, but the gap between green and diesel vehicles became significant in 20102012. The
results also indicate that larger automotive companies and smaller green vehicles show higher efficiency in
recent years, whereas the type of green vehicles and the origin of automotive companies are not
significantly related to the efficiency score of green vehicles. These findings provide more information
for understanding differences in green vehicles.

Introduction improving environmental performance to efficient perfor-


mances, including both functional and environmental
In response to the elevation of environmental protection
attributes.
campaigns in public, governments have begun establishing
Customers are also concerned with environmental issues
laws and regulations against firms to limit their impact on
and the rapid rise in gas prices. They are more careful about
the environment. One of the major interests of governments
their decisions when purchasing a vehicle. For example,
in the area of environmental protection is to reduce green-
according to the results of the 2013 Car Brand Perception
house gas (GHG) emissions and air pollutants emissions
Survey (Consumer Report Org 2013), 62% of the customers
(HC, CO, NOx, PM, and SOx) from vehicles. Nakaya
recognized green and environment-friendly as important
(2003) reported that fuel cell vehicles that were introduced
factors in purchasing a new vehicle. According to a report by
into the passenger transportation sector in Japan signifi-
The Society of Motor Manufacturers and Traders (2013), the
cantly affect reduction in CO2 emission. In the United
number of alternative fuel vehicles registered in the United
States, the Environmental Protection Agency (EPA) and
Kingdom in 2012 and 2013 reached 25,000 and 32,500,
National Highway Traffic Safety Administration are plan-
respectively.
ning to set CO2 emission targets for vehicles depending on
In this study, the efficiency of green and diesel vehicles is
the vehicle size (Hwang et al. 2013). Furthermore, the
evaluated using a two-stage data envelopment analysis (DEA)
European Commission had proposed a law that set fuel
model. It is helpful for the automotive industry to know the
economy standards, depending on the size and type of vehi-
environmental impact of their vehicles in comparison to their
cles, which new cars in the market had to satisfy by the end
competitors, and what should be improved to increase the effi-
of 2007 (Chen et al. 2012).
cient performance of green vehicles. Moreover, the truncated
Meanwhile, many firms in the automotive industry have
regression model was used to analyze relationship between the
recognized the importance of environmental protection (i.e.
efficiency score and the variable of interest, such as vehicle type,
complying with environmental laws and regulations for vehi-
origin of maker, vehicle size, and type of company, which was
cles and fulfilling the social responsibility required by the
calculated using the two-stage DEA. The results would help
society) as a key for long-term success. To achieve these
consumers to make better decisions in selecting available green
goals, sustainable product design has gained much attention
vehicles in the market. This paper is organized as follows: Green
from automotive manufacturers. Fiksel (2009) reported that
vehicles and data envelopment analysis are outlined in Section 2.
sustainable product design, which is also known as design for
In Section 3, a two-stage DEA model is demonstrated. In Section
the environment, was the systematic consideration of design
4, the efficiency evaluation of green vehicles using a two-stage
performance with respect to the environment, health, safety,
DEA model is shown. Furthermore, the results are statistically
and sustainability of objectives over the full product and
compared and analyzed. Section 5 discusses the findings of this
process life cycle. Although there is a trend toward aggressive
study. Finally, we make conclusions and provide some directions
involvement with green vehicles, there is simultaneously a
for the future research.
shift in sustainable product design from focusing on

CONTACT Fu-Kwun Wang fukwun@mail.ntust.edu.tw 43 Keelung Road, Section 4, Taipei City 10607, Taiwan.
2016 Taylor & Francis Group, LLC
1164 F.-K. WANG AND M. SAITO

Green vehicles and data envelopment analysis environmental problems, from purchasing green vehicles.
Through a focus group discussion, they discovered some fac-
Research on green vehicles
tors as the constraints: features of green vehicles, uncertainty of
According to Beltramello (2012), green vehicles can be environmental impact, and work responsibility. Sanitthangkul
defined as vehicles that can run on alternative fuels other et al. (2012) found that because of the rising price of gas, better
than gasoline, which include bio-fuels, natural gas, hydrogen, performance of fuels would positively affect the attitudes
and electricity. Green vehicles are more efficient and less toward green vehicles and promote the use of green vehicles.
polluting. Green vehicles include flexible fuel vehicles Chen et al. (2013) compared the performance and emissions of
(FFVs), electric vehicles (EVs), plug-in hybrid electric vehicles a diesel engine and the blend of diesel and butanol, which is
(PHEVs) or hybrid vehicles (HEVs), compressed natural gas one of the potential biofuels. Sushandoyo and Magnusson
(CNG) vehicles, and fuel cell vehicles (FCVs) (US (2014) analyzed the experiences from extensive field tests in
Environmental Protection Agency (USEPA) 2013). FFVs can London, which involved different manufacturers, hybrid sys-
use gasoline or E85 (a mixture of 85% ethanol and 15% tems, and vehicles. Correia et al. (2014) provided a good review
gasoline). EVs contain a rechargeable electric motor. of four existing environmental rating methodologies covering
Compared to regular gasoline vehicles, electric vehicles run the United States, Europe (Belgium and Germany), and
without any noise, are independent of gasoline, and produce Australia geographical regions for green vehicles based on
no emissions during its use; thus, these are highly environ- emissions of greenhouse gases (CO2, N2O, and CH4) and
mental-friendly. PHEVs are powered with electricity and pollutants (NOx, VOC, CO, SOx, PM 10, and PM 2.5) using
gasoline and can help in saving approximately 32% of GHG life cycle assessment (LCA) principles.
emissions compared with other gasoline vehicles (Samaras
and Meisterling 2008). CNG vehicles have specialized fittings
for a leak-free connection to natural gas vehicle. FCVs use Data envelopment analysis
pressurized hydrogen, which one pumps into the car through Data envelopment analysis has been widely used to measure
a special leak-free connection. The hydrogen powers a fuel the efficiency score of each decision-making unit (DMU),
cell, which subsequently generates electricity to power the which is calculated for multiple inputs and outputs (Charnes,
vehicle. Table 1 provides the pros and cons of green vehicles. Cooper, and Rhodes 1978; Emrouznejad et al. 2008). DEA has
Lane and Potter (2007) reported the initial findings of two several advantages. First, the relative weights for inputs and
research projects that identified the attitudinal barriers that outputs need not to be provided. Second, since the inputs
inhibited the adoption of cleaner vehicles in the United and outputs in different units can be calculated independently,
Kingdom. Choi and Oh (2010) recommended that effective data not necessarily be transformed into one single unit
governmental supports, which include multi-incentives, should (Shimshak, Lenard, and Klimberg 2009). Recently, DEA has
waive difference in performance between traditional vehicles been widely used for performance evaluation in different
and hybrid vehicles. Kang and Park (2011) reported that gov- industries. The model developed by Banker, Charnes, and
ernments should provide some types of incentives that would Cooper (1984) was a single-stage model. However, the problem
convince and educate consumers to purchase green vehicles. with a single DEA model is that different types of inputs
Wang and Chen (2011) provided the development of Chinas and outputs are mingled together, and the factors that cause
remanufacturing policies based on extended producer respon- inefficiency cannot be identified (Chang, Chung, and Hsu
sibility, and discussed the recycling system in line with coun- 2012). To address this issue, Seiford and Zhu (1999) divided
trys actual conditions. Browne, OMahony, and Caulfield the single DEA model into a two-stage DEA model. In the
(2012) presented a framework which can be modified and model, the outputs of the first stage are used for the inputs of
used by policy makers to identify and qualitatively evaluate the second stage. Seiford and Zhu (1999) measured the
the barriers of promoting alternative fuels and vehicles efficiency of commercial banks in the United States using the
(AFVs) and potential policies that may be implemented to two-stage DEA model by separately evaluating profitability and
address these barriers. Flamm and Agrawal (2012) identified marketability. In this model, the efficiency scores in the
the constraints that prevent people, who have concerns about first and second stages were evaluated in independent DEA.

Table 1. The advantages and disadvantages of green and diesel vehicles.


Items Diesel vehicles EVs HEVs and PHEVs CNG vehicles FFVs FCVs
Fuel economy city/highway 30/45 103/95 98 (electricity) 25/35 21/29 53/61
unit = miles/gallon 38 (gasoline)
Greenhouse gas rating and smog ratinga 8 and 6 10 and 10 10 and 8 10 and 8 7 and 6 10 and 10
Maintenance costs Regular Low except battery Expensive Low Regular Low
replacement
Safe to drive Normal Good Normal Normal Normal Good
Stations or recharge points Normal Limited Normal Few Few Limited
Range and speed Regular Limited Regular Shorter and slower Regular Limited
Recharge time None 8h 4h Regular Regular Regular
Electricity cost None Yes Yes None None Yes
Purchase price Regular Expensive Higher Regular Regular Expensive
Note: aRating scale from 0 to 10, 10 = the best emits 0 grams per mile.
INTERNATIONAL JOURNAL OF GREEN ENERGY 1165

Chen et al. (2012) developed a two-stage DEA model to performances of each DMU. According to Kao and
evaluate sustainable product designs in automotive industry. Hwang (2008), the two-stage DEA model can be formulated
The performance of vehicle designs was evaluated in the first as follows:
stage, whereas the environmental impact of vehicles was eval-
P
q
uated in the second stage. By combining the results of the two wq Zpk
stages, the overall efficiency scores of each vehicle model were p1
E1k max m ; (1)
calculated. P
vi Xik
i1

subject to
Methods
P
q
A two-stage model wq Zpk
p1
Figure 1 shows a two-stage model using DEA to evaluate the  1; (2)
P
m
performance of vehicles. In the first stage, an engineering vi Xik
i1
specification can be defined as a precise description of an
engineering characteristic incorporated in a product design, wp ;vi  e; p 1; 2; . . . ; q; i 1; 2; . . . ; m; (3)
and a product attribute can be defined as one of the main P s
physical features of a product as the combination of a number ur Yrk
of engineering specifications (Chen et al. 2012). Thus, pro- E2k max r q 1 ; (4)
P
duct attributes, such as passenger volume and fuel economy, wp Zpk
are strongly affected by the engineering characteristics of p1

vehicles. In the second stage, the environmental performance subject to


of vehicles was calculated by taking product attributes from
the first stage as inputs and environmental performances as P
s
ur Yrk
outputs. When automotive manufacturers attempt to reduce r1
 1; (5)
the environmental impact of vehicles by decreasing CO2 and Pq
wp Zpk
GHG emissions, they must improve vehicles performance. p1
Thus, in order to improve the environmental impact of
green vehicles, for an automotive manufacturer, trade-offs ur ;wp  e; r 1; 2; . . . ; s; p 1; 2; . . . ; q: (6)
should be made between improving environmental impact
The overall efficiency score can be obtained as follows:
and product attributes. Moreover, the engineering design of
vehicle should also be re-considered to make sure that the Ek E1k  E2k ; (7)
vehicle works properly.
Following the two-stage model proposed by Chen et al. where
(2012), each line of vehicle is treated as one DMU. Let k be P
s
the number of DMUs. In this first stage, m and q are the ur Yrk
r1
numbers of input variables and output variables, respec- Ek  1:
P
m
tively. In the second stage, q and s are the numbers of vi Xik
i1
input variables and output variables, respectively. Assume
that for each green vehicle, which is denoted by DMUj In this study, the vehicle evaluation process was divided
(j = 1,2, . . ., k), there are m engineering specifications as into two separate stages. In the first stage, some engineering
inputs, which are denoted by Xik (i = 1,2, . . ., m), and q specifications, such as the displacement (X1), rated horse-
product attributes as outputs, which are denoted by Zpk power (X2), axel ratio (X3), and speed ratio (X4), were selected
(p = 1,2, . . ., q). These q outputs are used as inputs for as inputs. The traditional performance of vehicles, such as
the second stage and will be referred to as intermediate passenger volume (Z1) and combined fuel economy (Z2), was
measures. The s outputs from the second stage, which are measured as outputs. In the second stage, the environmental
denoted by Yrk (r = 1,2, . . ., s), are the environmental performance of vehicles, which includes the GHG emission
score (Y1) and air pollution score (Y2), was evaluated. The
product attributes from the first stage were used as inputs,
and the environmental performance data were selected as
outputs. Finally, the overall efficiency was evaluated by com-
bining the results from first and second stages. More details of
all inputs and outputs are provided as follows:
Displacement (X1): The volume that is swept by all pistons
inside the cylinders of a reciprocating engine in a single
movement from the top dead center to the bottom dead
center, measured in liters.
Rated horsepower (X2): The rated power of the output of
Figure 1. A two-stage DEA model for evaluating green vehicles. engine.
1166 F.-K. WANG AND M. SAITO

Axel ratio (X3): The ratio of the drive shaft speed, which is selected for analysis. For example, Volkswagen-Jetta (diesel
measured as revolutions per minute to drive the wheel speed. vehicle) has X1 = 2, X2 = 155, X3 = 32.2, X4 = 3.65, Z1 = 91,
A higher ratio indicates a faster engine (in revolutions) at a Z2 = 33, Y1 = 8, and Y2 = 6.
certain speed. The data were collected based on the US EPA database (US
Speed ratio (X4): The ratio of speed between engine and Environmental Protection Agency (USEPA) 2013). When
driving wheel. there were duplicate data of identical car lines, the data were
Passenger volume (Z1): The sum of front- and rear-seat averaged. Moreover, identical car models with different
passenger volumes in cubic feet. options (i.e. automatic vs. manual, two-wheel drive vs. four-
Combined fuel economy (Z2): Fuel economy from driving a wheel drive) were identified as different carlines. When some
combination of 55% city miles and 45% highway miles in mpg. of the data were missing, the carlines were removed from the
It is calculated as 1/[(0.55/city mpg) + (0.45/highway mpg)]. study. However, since electric vehicles do not have engine
Greenhouse gas emission score (Y1): The score that reflects displacement, the value of 0.01 was used to make results
the fuel emissions of CO2 and other greenhouse gases. The comparable with other types of green vehicles because the
score is given on a scale of 110. A higher score indicates that value that is used for DEA should be greater than 0.
the vehicle has lower GHG emissions.
Air pollution score (Y2): Based on government emission
Test hypotheses
standards, with which the vehicle was certified to comply and
which reflects the vehicles tailpipe emissions that contribute to The data collected using this procedure were tested with the
local and regional air pollution and create problems such as haze two-stage DEA to evaluate each carlines efficiency. In addition,
and health issues. The score is given on a scale of 110. A higher since the efficiencies of vehicles may be affected by indepen-
score indicates that the vehicle has a lower air pollution impact. dent variables, four hypotheses were made to test relationship
between the independent variables of vehicles and the effi-
ciency using truncated regression model as follows:
A truncated regression model
According to ek (2012), a truncated regression model can Hypothesis 1 (Type of vehicle): A green vehicle that runs on
be modeled using the maximum likelihood when the latent either fully or partially electricity offers higher
linear regression model is considered. Assume the liner efficiency than a green vehicle that runs on
regression model as follows: alternative fuels. For example, electric vehicles
may have better efficiency than alternative fuel
yi xiT b0 ei ; (8) vehicles because the vehicles run on electric
where yi 2 0; 1 is the latent response variable, xi 2 Rp is the power without gas emissions.
vector of explanatory variables, and b0 is the true value of the
Hypothesis 2 (Vehicle manufacturer): A green vehicle made by
parameter vector b0 2 b 2 Rp and ei is the latent error term
p western manufacturer shows higher efficiency
with standard deviation, s varei . Without loss of gen- than that made by Asian manufacturer.
erality, by assuming the truncation occurs at zero from below, Different manufacturers produce green vehicles
only the points where yi ; xi yi ; xi are observed: that are sold in the US market. The origin coun-
yi xTi b0 ei if xTi b0 ei > 0: (9) try of the manufacturer from both West (i.e. US
and Europe) and Asia (i.e. Japan and South
Denoting di I yi > 0, the parameter vector b0 can be Korea) may affect the efficiency of green vehicles.
estimated using the maximum likelihood method that max-
imizes the truncated regression (9): Hypothesis 3 (Vehicle size): Green vehicles with smaller sizes
X
n show higher efficiency than those with larger size.
1nLn b; s f1nf s yi  xTi b0  1n1  Fs 0  xTi b0 g:
i1 Hypothesis 4 (Company type): Larger automotive companies
(10) show higher efficiency than smaller companies
Empirical study such as the subsidiaries and divisions of parent
firms. Since larger automotive companies can
Data collection
access more resources than smaller companies,
We used the categorization of green vehicles used by the US green vehicles developed by larger companies
EPA. Four types of green vehicles were selected for the study: can be more efficient than those developed by
HEV, FFV, EV, and PHEV. In order to compare the efficiency smaller companies.
of green and conventional vehicles, we included diesel vehicle
in this study. The US EPA publishes emission data and basic
Analysis results
engineering specifications of all new car models that are sold
each year for the public (US Environmental Protection According to the data of vehicles from 2008 to 2012, a two-stage
Agency (USEPA) 2013). Data of five years (20082012) were DEA was conducted based on the procedure discussed in the
obtained from the US EPA database. Four inputs (X1X4), two previous section. The number of each type of green and diesel
intermediates (Z1 and Z2), and two outputs (Y1 and Y2) were vehicles that were tested for each year is provided in Table 2.
INTERNATIONAL JOURNAL OF GREEN ENERGY 1167

Table 2. The number of each type of vehicle tested from 2008 to 2012. (2) The difference in efficiency scores increased in 2010. The
Types of vehicles 2008 2009 2010 2011 2012 efficiency scores range from 0.1948 to 1.0000. The most
Diesel 4 6 9 10 12 inefficient green vehicle of the year was carline #40. The
HEV 8 12 19 22 26
FFV 11 9 22 26 33
first-stage efficiency score of the vehicles was 0.2023,
EV 0 0 3 4 7 whereas the second-stage score was 0.9630 because the
PHEV 0 0 0 3 4 engine power of the vehicle was high but the fuel econ-
Total 23 27 53 65 82
omy was not high and the passenger volume was not
large. One of the reasons is the appearance of electric
In Table 2, the number of tested vehicles gradually increased. vehicles. This year was the first year in which electric
In 2009, three types of vehicles, including diesel, HEV, and FFV, vehicles were widely introduced into the market. There
were available in the market, whereas EV and PHEV were were three electric vehicles in the results for the year, two
introduced later. For each car line, the first-stage, second-stage, of which were identified as efficient. None of the overall
and overall efficiency scores were calculated by using R pro- efficiency scores of diesel vehicles reached 0.8000 in this
gramming language with the benchmark package. All first-stage, year, whereas two electric vehicles reached efficiency
second-stage, and overall efficiency scores for each green vehicle scores of 1.0000, possibly because diesel vehicles could
from 2008 to 2012 are provided in the Appendix. not catch up with the improvements in other types of
Table 3 shows the average efficiency scores of each type of green vehicles in terms of environmental performance. In
vehicle from 2008 to 2012. The results indicate that all types addition, one automotive manufacturer in Europe domi-
of vehicles had notably similar average efficiency scores in nated the diesel vehicle market in the United States.
2008 (0.940.97). However, the difference increases in 2012 (3) In 2011, PHEVs became available in the US market. In
(0.800.94), which may have been caused by the introduc- this year, the efficiency scores ranged from 0.5818 to
tion of electric vehicles into the market and the fact that 1.0000, and five of the 65 vehicles were identified as
automotive manufactures put more resources into certain efficient: carlines #19, 29, 49, 50, and #51. The data
types of green vehicles for further development. The results show that no EVs and PHEVs were identified as effi-
also indicate that green vehicles performed better than diesel cient, which indicates that American and Asian auto-
vehicle. For HEV and FFV in 20082010, the second stage makers were eager to pursue the development of green
has higher average efficiency scores than the first stage, vehicles with less environmental impact.
which indicates that green vehicles have relatively higher (4) According to the results from 2012, six vehicles were
efficiency in environmental performance in the second identified as efficient: carlines #32, 48, 61, 72, 74, and
stage than the traditional performance in the first stage. #79. Only two green vehicles that were identified as effi-
This result is consistent with the original purpose for devel- cient in 2011 retained their definitions as efficient green
oping green vehicles. vehicles in 2012. For example, the efficiency score of car-
Based on these results and the results given in the line #18 changed from 1.00 to 0.94, and that of carline #64
Appendix, the trends of green vehicles are obtained as follows: changed from 1.00 to 0.94. The efficiency scores decreased
because of the second-stage result; the first-stage efficiency
(1) In 2008 and 2009, the efficiency scores of green vehi- scores of these vehicles remained 1.00. This result indicates
cles were 0.841.00 and 0.811.00, respectively. The that although those vehicles were relatively efficient as
results indicate that there is little difference among before in terms of traditional performance, they were no
green vehicles. In 2008, carlines #5, 9, 11, 13, 14, 19, longer efficient vehicles because their environmental per-
22. and #23 were classified as efficient DMUs. In 2009, formance was not as good as that of other vehicles.
carlines #7, 8, 11, 13, 19, 26, and #27 were classified as (5) Although no Asian automotive company produces
efficient DMUs. For example, one of the oldest hybrid FFVs, most Asian automotive manufacturers focused
vehicles that were made in Japan and brought in the on developing hybrid, electric, and plug-in hybrid
US market was considered efficient. vehicles, which may reflect consumer preferences or
market situations in their home markets. Hence,
although most makers sell their vehicles worldwide,
Table 3. The average efficiency score of each type of vehicle from 2008 to 2012. their home market affects their decisions on the type
Year Diesel HEV FFV EV PHEV Total of green vehicles that they develop.
2008 First stage 0.96 0.98 0.94 0.96
Second stage 0.98 0.98 0.99 0.99
Overall 0.94 0.97 0.94 0.95 To test the difference of each input and output between
2009 First stage 0.95 0.98 0.94 0.95 efficient and inefficient DMUs, MannWhitney U test was
Second stage 0.94 0.99 1.00 0.94
Overall 0.88 0.96 0.94 0.88
applied. Our hypotheses were formulated for each input and
2010 First stage 0.84 0.92 0.75 1.00 0.84 output as follows:
Second stage 0.82 0.95 0.98 0.95 0.94 H0: There is no significant difference in the input/output
Overall 0.69 0.88 0.73 0.95 0.79
2011 First stage 0.91 0.93 0.87 0.84 1.00 0.90 value between efficient and inefficient vehicles; H1: There is a
Second stage 0.74 0.91 0.98 0.87 0.84 0.90 significant difference in the input/output value between effi-
Overall 0.67 0.84 0.85 0.72 0.84 0.81
2012 First stage 0.92 0.94 0.87 1.00 1.00 0.92 cient and inefficient vehicles.
Second stage 0.86 0.92 0.96 0.84 0.94 0.92 Table 4 shows that all inputs and outputs in 2008 and 2009
Overall 0.80 0.86 0.84 0.84 0.94 0.84 are not significantly different for efficient and inefficient
1168 F.-K. WANG AND M. SAITO

Table 4. Results of MannWhitney U test for inefficient and efficient DMUs from Table 5. Results of truncated regression analysis from 2008 to 2012.
2008 to 2012. Independent variables Coefficient Standard error t-value p-value
2008 2009 2008
Median Median Intercept 0.9189 0.0336 27.324 2.2E-16
Vehicle type 0.0128 0.0375 0.343 0.732
Inputs/ Inefficient Efficient P- Inputs/ Inefficient Efficient p- Vehicle maker 0.0209 0.0410 0.509 0.611
outputs DMU DMU value outputs DMU DMU value Vehicle size 0.0068 0.0175 0.386 0.700
X1 3 2.7 0.796 X1 2.45 2.7 0.525 Company type 0.0054 0.0139 0.388 0.698
X2 224 192.5 0.349 X2 171.5 234 0.543 Sigma 0.0492 0.0073 6.782 1.1E-11
X3 31.3 27.35 0.186 X3 29 29.1 0.761 2009
X4 3.65 3.065 0.071 X4 3.07 3.41 0.543 Intercept 0.9625 0.0315 30.585 2.20E-16
Z1 96 101 0.146 Z1 94.5 101 0.677 Vehicle type 0.0529 0.0274 1.930 0.0536
Z2 23 19.5 0.628 Z2 31.5 21 0.933 Vehicle maker 0.0102 0.0301 0.337 0.7363
Y1 6 7 0.439 Y1 7 8.5 0.219 Vehicle size 0.0133 0.0155 0.861 0.3893
Y2 6 6.25 0.949 Y2 8 6 0.960 Company type 0.0172 0.0109 1.575 0.1153
2010 2011 Sigma 0.0473 0.0064 7.349 2.01E-13
Median Median 2010
Intercept 0.7081 0.0567 12.478 2.20E-16
Inputs/ Inefficient Efficient P- Inputs/ Inefficient Efficient p- Vehicle type 0.1863 0.0603 3.088 0.0020*
outputs DMU DMU value outputs DMU DMU value Vehicle maker 0.0077 0.0662 0.116 0.9079
X1 2.5 2.1 0.260 X1 2 3.6 0.141 Vehicle size 0.0153 0.0276 0.555 0.5791
X2 220 128.5 0.027a X2 182 173 0.693 Company type 0.0167 0.0228 0.734 0.4628
X3 28.6 27.95 0.578 X3 3.185 2.44 0.060 Sigma 0.1426 0.0139 10.295 2.20E-16
X4 3.07 3.34 0.921 X4 29.35 26.7 0.076 2011
Z1 98 97.5 0.980 Z1 94.5 100 0.208 Intercept 0.8475 0.0405 20.925 2.20E-16
Z2 21.5 40 0.106 Z2 27.5 16 0.459 Vehicle type 0.0554 0.0367 1.507 0.1317
Y1 6 8 0.005a Y1 6 6 0.338 Vehicle maker 0.1501 0.0408 3.677 0.0002a
Y2 5 9 0.496 Y2 7 5 0.980 Vehicle size 0.0031 0.0179 0.174 0.8620
2012 Company type 0.0424 0.0154 2.747 0.0060a
Sigma 0.1026 0.0090 11.402 2.20E-16
Median
2012
Inputs/outputs Inefficient DMU Efficient DMU p-value Intercept 0.9131 0.0288 31.657 2.00E-16
X1 2.4 1.005 0.152 Vehicle type 0.0272 0.0213 1.280 0.2007
X2 215.5 144 0.014a Vehicle maker 0.0481 0.0259 1.853 0.0639
X3 3.165 3.005 0.593 Vehicle size 0.0374 0.0147 2.545 0.0109a
X4 29 41.3 0.105 Company type 0.0248 0.0103 2.414 0.0158a
Z1 97.5 89 0.095 Sigma 0.0783 0.0061 12.806 2.00E-16
Z2 26 38.5 0.224 Notes: Vehicle type (alternative fuels = 0, Electric = 1); vehicle maker (wes-
Y1 6 7.8 0.017a tern = 0, Asian = 1); vehicle size (small sedan = 0, middle sedan = 1, large
Y2 7 10 0.075 sedan = 2, station wagon = 3); company type (small = 0, middle = 1, large = 2);
Note: aSignificant at 95% Confidence Level. a
significant level at 0.05.

DMUs at a 95% confidence level, which is consistent with the Thus, in 2011, green vehicles made by Asian companies and
results from the previous section. The efficiency scores of smaller companies showed higher efficiency.
vehicles do not vary significantly. The results from 2010 Finally, in 2012, the vehicle and company size are signifi-
indicate that an efficient vehicle in this year has relatively cant for the efficiency of vehicles. Smaller vehicles show
higher rated horsepower and air pollution scores than ineffi- higher efficiency, which indicates that although smaller vehi-
cient vehicles. In 2011, no inputs and outputs were considered cles tend to have lower engine power and smaller passenger
as significantly different. However, the rated horsepower and volume than large vehicles, smaller vehicles perform relatively
air pollution rate were again significantly different between better than large vehicles in terms of efficiency. In addition,
efficient and inefficient DMUs. Therefore, efficient DMUs bigger automotive companies indicate higher levels of effi-
have higher rated horsepower and air pollution scores than ciency because bigger companies have more resources and
inefficient DMUs. capital to develop more efficient green vehicle, whereas smal-
A truncated regression model was applied to test the rela- ler companies have limitations on their resources.
tionship among four independent variables: vehicle type, ori-
gin of maker, vehicle size, and type of company. Table 5
shows the results of truncated regression model, which indi- Discussion
cate that all independent variables are not significantly related From the analysis results of green vehicles and diesel vehicle
to efficiency in 2008 and 2009 at a 95% confidence level. The from 2008 to 2012, four major findings are discussed below:
DEA results show a small gap between efficient and inefficient
green vehicles. Therefore, no variable was identified as sig- (1) Vehicles that either partially or fully run on electricity
nificant. In 2010, the vehicle type is significantly related to the were more efficient than those that run on alternative
efficiency of green vehicles, which proves that green vehicles fuels in 2010. However, the vehicle type was not a
that fully or partially run on electricity, exhibit higher effi- significant variable because gap in efficiency between
ciency than the vehicles run on alternative fuels for this year. these two types had been narrowed.
However, the vehicle type was no longer significant in 2011 (2) In 2011, green vehicles made by Asian automotive
and 2012, whereas in 2011, the origin of vehicle maker and companies had higher efficiency scores than those
company size were significantly correlated with efficiency. made by Western automakers. Asian automakers
INTERNATIONAL JOURNAL OF GREEN ENERGY 1169

used to have advantage in developing more efficient References


green vehicles than Western automakers. However, the
Banker, R.D., A. Charnes, and W.W. Cooper. 1984. Some models for
gap is shrinking. estimating technical and scale inefficiencies in data envelopment ana-
(3) Smaller green vehicles were more efficient than larger lysis. Management Science 30:107892.
green vehicles that were produced in 2012. This result Beltramello, A. 2012. Market development for green cars.
will be a good indication for customers who are look- OECD Green Growth Papers, No. 201203. Paris: OECD
ing for efficient green vehicles. In addition, automotive Publishing.
Browne, D., M. OMahony, and B. Caulfield. 2012. How should barriers
firms with limited budgets may focus on developing to alternative fuels and vehicles be classified and potential policies to
smaller green vehicles. promote innovative technologies be evaluated? Journal of Cleaner
(4) Smaller automotive firms, including subsidiaries and Production 35:40151.
divisions of parent firms, produced more efficient Chang, T.Y., P.H. Chung, and S.S. Hsu. 2012. Two-stage performance
green vehicles than larger automotive firms, possibly model for evaluating the managerial efficiency of higher education:
Application by the Taiwanese tourism and leisure department.
because it is easier for smaller manufacturers to focus Journal of Hospitality, Leisure, Sport & Tourism Education
on developing a few green vehicle lines, whereas 11:16871.
larger companies usually have a wide range of vehicle Charnes, A., W.W. Cooper, and E. Rhodes. 1978. Measuring the effi-
lines. ciency of decision-making units. European Journal of Operational
Research 2:429359.
Chen, Z., J. Liu, Z. Han, B. Du, Y. Liu, and C. Lee. 2013. Study on performance
and emissions of a passenger-car diesel engine fueled with butanol-diesel
Conclusions blends. Energy 55:63846.
Chen, C., J. Zhu, J. Yu, and H. Noori. 2012. A new methodology for
In this study, the efficiency of green and diesel vehicles was evaluating sustainable product design performance with two-stage
evaluated by combining traditional and environmental per- network data envelopment analysis. European Journal of Operational
formances using a two-stage DEA model. For a two-stage Research 221:34859.
ek, P. 2012. Semiparametric robust estimation of truncated
DEA model, four inputs, two intermediates, and two outputs
and censored regression models. Journal of Econometrics
are identified. The results show that there was little difference 168:34766.
in the efficiency of vehicles in 2008 and 2009, but the gap Choi, H. and I. Oh. 2010. Analysis of product efficiency of hybrid
between green and diesel vehicles became significant in later vehicles and promotion policies. Energy Policy 38:226271.
years. Furthermore, a truncated regression analysis was con- Consumer Report Org. 2013. Car brand perception survey. Available at
http://www.consumerreports.org/cro/index.htm (accessed April 11,
ducted to analyze the relationship between efficiency and
2013).
variable of interest. The results indicate that larger automotive Correia, G.N., T.P. Batista, S.S. Marques, and C.M. Silva. 2014.
companies and smaller green vehicles show higher efficiency How car material life-cycle emissions are considered in environ-
in recent years, whereas type of green vehicles and the origin mental rating methodologies? Suggestion of expedite models
of automotive companies are not significantly related to the and discussion. Renewable and Sustainable Energy Reviews
38:2035.
efficiency scores of green vehicles.
Emrouznejada, A., B.R. Parkerb, and G. Tavaresc. 2008. Evaluation of
Since other types of performance measurements were not research in efficiency and productivity: A survey and analysis of the
considered in this study, the results should be compared first 30 years of scholarly literature in DEA. Socio-Economic Planning
with other environmental ranking methodologies based on Sciences 42:15157.
LCA principles. Other variables, such as the ratio of power/ Fiksel, J.R. 2009. Design for environment: A guide to sustainable product
development, 2nd ed. New York: McGraw-Hill.
weight, the number and range of speed ratios available, life-
Flamm, B.J. and A.W. Agrawal. 2012. Constraints to green vehicle own-
time of vehicle, fractions of recycled materials, engine noise, ership: A focus group study. Transportation Research Part D:
and energy consumption, can be added into the proposed Transport and Environment 17:10815.
model. Some companies have implemented eco-efficiency in Hwang, S., C. Chen, Y. Chen, H. Lee, and P. Shen. 2013. Sustainable
their production and operations to communicate the results design performance evaluation with applications in the automobile
industry: Focusing on inefficiency by undesirable factors. Omega
to the public. Furthermore, since this study only included
41:55358.
the data provided by the US EPA, which is a government Kang, M.J. and H. Park. 2011. Impact of experience on government
agency, the views of green vehicle users, such as customer policy toward acceptance of hydrogen fuel cell vehicles in Korea.
satisfaction of green vehicles, can be included to further Energy Policy 39:346575.
evaluate green vehicles. A limitation of this study stems Kao, C. and S.N. Hwang. 2008. Efficiency decomposition in two-stage
data envelopment analysis: An application to non-life insurance
from small sample size. A full-scale empirical eco-efficiency
companies in Taiwan. European Journal of Operational Research
analysis of green vehicles would provide an interesting area 185:41829.
for future research. Lane, B. and S. Potter. 2007. The adoption of cleaner vehicles in the UK:
Exploring the consumer attitudeaction gap. Journal of Cleaner
Production 15:108592.
Nakata, T. 2003. Energy modeling on cleaner vehicles for reducing CO2
Acknowledgments emissions in Japan. Journal of Cleaner Production 11:38996.
Samaras, C. and K. Meisterling. 2008. Life cycle assessment of
The authors thank editor-in-chief, Dr. Li, and two reviewers for their greenhouse gas emissions from plug-in hybrid vehicles:
valuable comments and suggestions to improve the quality of this paper.
1170 F.-K. WANG AND M. SAITO

Implications for policy. Environmental Science & Technology Sushandoyo, D. and T. Magnusson. 2014. Strategic niche
42:317076. management from a business perspective: Taking cleaner
Sanitthangkul, J., A. Ratsamewongjan, W. Charoenwongmitr, and J. vehicle technologies from prototype to series production.
Wongkantarakorn. 2012. Factors affecting consumer attitude toward Journal of Cleaner Production 74:1726.
the use of eco-car vehicles. Procedia Social and Behavioral Sciences The Society of Motor Manufacturers and Traders. (2013). Automotive
40:46166. sustainability report, 14th ed. Available at http://www.smmt.co.uk/
Seiford, L.M. and J. Zhu. 1999. Profitability and marketability sustainability/ (accessed May 19, 2014).
of the top 55 US commercial banks. Management Science 45:1270 Wang, X. and M. Chen. 2011. Implementing extended producer respon-
88. sibility: vehicle remanufacturing in China. Journal of Cleaner
Shimshak, D.G., M.L. Lenard, and R.K. Klimberg. 2009. Incorporating Production 19:68086.
quality into data envelopment analysis of nursing home performance: US Environmental Protection Agency (USEPA). (2013). Database.
A case study. Omega 37:67285. Available at http://www.epa.gov/ (accessed April 15, 2014).
INTERNATIONAL JOURNAL OF GREEN ENERGY 1171

Appendix DEA results in 2010

Efficiency score
DEA results in 2008
Carline Vehicle First Second
Type (DMU) maker Transition Engine stage stage Overall
Efficiency score Diesel 1 0 A-S6 3.0/6 0.69 0.81 0.56
Carline Vehicle First Second 2 0 A-S6 2.0/4 0.90 0.81 0.73
Type (DMU) maker Transition Engine stage stage Overall 3 0 M-6 2.0/4 0.93 0.81 0.76
4 0 A-S6 2.0/4 0.90 0.81 0.73
Diesel 1 0 A-S6 2.0/4 0.96 0.99 0.95 5 0 M-6 2.0/4 0.93 0.81 0.76
2 0 M-6 2.0/4 0.96 0.99 0.94 6 0 A-7 3.0/6 0.71 0.81 0.58
3 0 A-S6 2.0/4 0.95 0.98 0.93 7 0 A-S6 2.0/4 0.73 0.82 0.60
4 0 M-6 2.0/4 0.96 0.98 0.93 8 0 A-S6 2.0/4 0.88 0.82 0.72
HEV 5 1 AV 1.3/4 1.00 1.00 1.00 9 0 M-6 2.0/4 0.91 0.84 0.77
6 1 A-S6 3.5/6 0.95 0.96 0.91 HEV 10 1 AV-S7 1.5/4 1.00 0.99 0.99
7 1 A-S8 5.0/8 1.00 0.96 0.96 11 1 M-6 1.5/4 0.83 1.00 0.83
8 0 A-4 2.4/4 0.90 1.00 0.90 12 1 AV 1.3/4 1.00 1.00 1.00
9 1 AV 2.5/4 1.00 1.00 1.00 13 1 AV-S7 1.3/4 1.00 0.99 0.99
10 0 A-4 2.4/4 1.00 1.00 1.00 14 1 AV 1.3/4 1.00 0.99 0.99
11 1 AV 2.4/4 1.00 1.00 1.00 15 1 AV 1.8//4 1.00 0.97 0.97
12 0 AV 1.5/4 1.00 0.95 0.95 16 1 AV-S6 3.5/6 0.67 0.98 0.66
FFV 13 0 A-4 2.7/6 1.00 1.00 1.00 17 1 AV 2.4/4 0.81 0.95 0.77
14 0 A-4 2.7/6 1.00 1.00 1.00 18 1 AV-S8 5.0/8 0.73 1.00 0.73
15 0 A-7 3.0/6 0.92 1.00 0.92 19 0 A-S6 4.4/8 0.73 0.85 0.62
16 0 A-7 3.0/6 0.92 1.00 0.92 20 0 A-S8 4.4/8 1.00 0.85 0.85
17 0 A-4 3.5/6 0.89 0.94 0.84 21 0 AV 2.5/4 0.88 1.00 0.88
18 0 A-4 3.5/6 0.84 1.00 0.84 22 1 A-6 2.4/4 1.00 0.80 0.80
19 0 A-4 3.5/6 1.00 1.00 1.00 23 1 A-6 2.4/4 0.94 0.80 0.75
20 0 A-4 3.9/6 0.90 0.97 0.87 24 0 AV 2.5/4 1.00 1.00 1.00
21 0 A-4 4.6/8 0.91 1.00 0.91 25 1 AV 2.5/4 0.89 1.00 0.89
22 0 A-4 4.6/8 1.00 1.00 1.00 26 1 AV 2.4/4 1.00 1.00 1.00
23 0 A-4 4.6/8 1.00 1.00 1.00 27 1 AV 1.8/4 1.00 1.00 1.00
28 0 A-7 3.5/6 1.00 0.95 0.95
FFV 29 0 A-S6 6.0/12 0.44 1.00 0.44
30 0 A-6 3.6/6 0.60 1.00 0.60
DEA results in 2009 31 0 A-5 3.6/6 0.59 0.96 0.56
32 0 A-7 3.0/6 0.66 0.99 0.65
33 0 A-7 3.0/6 0.66 0.99 0.65
34 0 A-S6 2.0/4 0.83 1.00 0.83
Efficiency score 35 0 M-6 2.0/4 0.82 0.93 0.76
Carline Vehicle First Second 36 0 A-4 3.9/6 0.79 0.96 0.76
Type (DMU) maker Transition Engine stage stage Overall 37 0 A-S6 2.4/4 0.80 1.00 0.80
38 0 A-4 3.5/6 0.79 0.95 0.75
Diesel 1 0 A-S6 3.0/6 0.97 0.94 0.91 39 0 A-4 3.9/6 0.83 0.96 0.80
2 0 A-S6 2.0/4 0.97 0.92 0.89 40 0 A-5 3.6/6 0.20 0.96 0.19
3 0 M-6 2.0/4 1.00 0.92 0.92 41 0 A-6 3.6/6 0.64 1.00 0.64
4 0 M-6 2.0/4 1.00 0.95 0.95 42 0 A-5 3.6/6 0.74 0.94 0.70
5 0 A-S6 2.0/4 0.84 0.97 0.81 43 0 A-S6 3.0/6 0.86 0.97 0.84
6 0 M-6 2.0/4 0.89 0.94 0.83 44 0 A-4 4.6/8 0.86 1.00 0.86
HEV 7 1 AV-S7 1.3/4 1.00 1.00 1.00 45 0 A-S6 3.0/6 0.69 0.97 0.67
8 1 AV 1.3/4 1.00 1.00 1.00 46 0 A-S6 3.0/6 1.00 0.97 0.97
9 1 AV-S6 3.5/6 0.95 1.00 0.95 47 0 A-4 4.6/8 1.00 1.00 1.00
10 1 AV 2.4/4 0.90 1.00 0.90 48 0 A-S6 2.0/4 1.00 1.00 1.00
11 1 AV-S8 5.0/8 1.00 1.00 1.00 49 0 M-6 2.0/4 0.81 1.00 0.81
12 0 A-4 2.4/4 0.95 0.97 0.92 50 0 A-4 4.6/8 1.00 1.00 1.00
13 0 AV 2.5/4 1.00 1.00 1.00 EV 51 0 N/A N/A 1.00 1.00 1.00
14 0 AV 2.5/4 0.99 1.00 0.99 52 0 N/A N/A 1.00 1.00 1.00
15 1 AV 2.5/4 0.97 1.00 0.97 53 1 N/A N/A 1.00 0.86 0.86
16 1 AV 2.4/4 1.00 0.99 0.99
17 1 AV 1.8/4 1.00 0.90 0.90
18 0 A-7 3.5/6 1.00 0.97 0.97
FFV 19 0 A-4 2.7/6 1.00 1.00 1.00
20 0 A-7 3.0/6 0.86 1.00 0.86
21 0 A-7 3.0/6 0.86 1.00 0.86
22 0 A-S6 2.4/4 0.95 1.00 0.95
23 0 A-S6 3.0/6 0.91 1.00 0.90
24 0 A-4 4.6/8 0.92 1.00 0.92
25 0 A-S6 3.0/6 0.97 1.00 0.97
26 0 A-4 4.6/8 1.00 1.00 1.00
27 0 A-4 4.6/8 1.00 1.00 1.00
1172 F.-K. WANG AND M. SAITO

DEA results in 2011

Efficiency score
Type Carline (DMU) Vehicle maker Transition Engine First stage Second stage Overall
Diesel 1 0 A-S6 2.0/4 0.90 0.74 0.67
2 0 M-6 2.0/4 0.99 0.74 0.73
3 0 A-S6 2.0/4 0.90 0.74 0.67
4 0 M-6 2.0/4 0.94 0.74 0.70
5 0 A-S6 2.0/4 0.86 0.72 0.62
6 0 M-6 2.0/4 1.00 0.70 0.70
7 0 A-S6 2.0/4 0.90 0.76 0.68
8 0 A-7 3.0/6 0.96 0.77 0.74
9 0 A-7 3.0/6 1.00 0.71 0.71
10 0 A-S6 2.0/4 0.67 0.76 0.51
HEV 11 1 AV-S7 1.5/4 0.95 1.00 0.95
12 1 M-6 1.5/4 0.74 1.00 0.74
13 1 AV 1.5/4 0.95 0.95 0.90
14 1 AV-S7 1.3/4 0.84 0.96 0.81
15 1 AV 1.3/4 0.84 0.96 0.81
16 1 AV 1.8/4 0.98 1.00 0.98
17 1 AV 2.4/4 0.88 0.85 0.75
18 1 AV 5.0/8 0.92 1.00 0.92
19 1 AV 1.5/4 1.00 1.00 1.00
20 0 A-S8 3.0/6 0.89 0.76 0.68
21 0 A-S8 4.4/8 0.95 0.78 0.74
22 0 AV-S6 2.4/4 1.00 0.79 0.79
23 0 AV-S6 2.4/4 0.85 0.80 0.67
24 1 A-6 2.4/4 1.00 0.94 0.94
25 1 A-S7 3.5/6 0.97 0.77 0.75
26 1 A-6 2.4/4 0.91 0.94 0.86
27 1 AV 2.5/4 1.00 0.98 0.98
28 1 AV 2.5/4 1.00 1.00 1.00
29 1 AV 1.8/4 1.00 1.00 1.00
30 0 A-S8 4.4/8 1.00 0.75 0.75
31 0 A-8 3.0/6 0.73 0.79 0.58
32 1 AV 1.8/4 1.00 0.91 0.91
FFV 33 0 A-S6 6.0/12 0.73 1.00 0.73
34 0 A-7 3.5/6 0.84 1.00 0.84
35 0 A-7 3.5/6 0.86 1.00 0.86
36 0 A-7 3.5/6 0.86 0.97 0.83
37 0 AM-6 2.0/4 0.87 0.93 0.81
38 0 M-5 2.0/4 0.87 0.93 0.81
39 0 AM-6 2.0/4 0.74 1.00 0.74
40 0 A-7 3.0/6 0.83 0.96 0.79
41 0 A-7 3.5/6 0.85 0.94 0.80
42 0 A-S6 2.4/4 0.84 1.00 0.84
43 0 A-S6 3.6/6 1.00 0.97 0.97
44 0 A-S6 3.6/6 0.92 1.00 0.92
45 0 A-S6 2.0/4 0.81 1.00 0.81
46 0 M-6 2.0/4 0.76 0.96 0.73
47 0 A-S6 2.4/4 0.98 0.96 0.94
48 0 A-S6 2.4/4 0.92 1.00 0.92
49 0 A-6 3.6/6 1.00 1.00 1.00
50 0 A-6 3.6/6 1.00 1.00 1.00
51 0 A-6 3.6/6 1.00 1.00 1.00
52 0 A-8 3.6/6 1.00 0.85 0.85
53 0 A-8 3.6/6 0.79 1.00 0.79
54 0 A-8 3.6/6 0.81 1.00 0.81
55 0 A-S6 2.0/4 0.80 1.00 0.80
56 0 M-6 2.0/4 0.90 1.00 0.90
57 0 A-S6 2.0/4 0.82 1.00 0.82
58 0 M-6 2.0/4 0.80 1.00 0.80
EV 59 0 N/A N/A 0.59 1.00 0.59
60 1 N/A N/A 1.00 0.85 0.85
61 1 N/A N/A 0.78 0.80 0.62
62 0 N/A N/A 1.00 0.82 0.82
PHEV 63 0 N/A N/A 1.00 0.96 0.96
64 0 N/A N/A 1.00 0.80 0.80
65 1 N/A N/A 1.00 0.77 0.77
INTERNATIONAL JOURNAL OF GREEN ENERGY 1173

DEA results in 2012

Efficiency score
Type Carline (DMU) Vehicle maker Transition Engine First stage Second stage Overall
Diesel 1 0 AM-S6 2.0/4 0.92 0.92 0.85
2 0 M-6 2.0/4 0.90 0.92 0.83
3 0 AM-S6 2.0/4 0.91 0.88 0.80
4 0 M-6 2.0/4 1.00 0.88 0.88
5 0 AM-S6 2.0/4 0.92 0.88 0.81
6 0 M-6 2.0/4 0.97 0.88 0.85
7 0 AM-S6 2.0/4 0.90 0.83 0.74
8 0 M-6 2.0/4 1.00 0.82 0.82
9 0 AM-S6 2.0/4 0.91 0.86 0.78
10 0 A-7 3.0/6 0.97 0.83 0.81
11 0 A-7 3.0/6 1.00 0.75 0.75
12 0 AM-S6 2.0/4 0.69 0.92 0.64
HEV 13 1 AV-S7 1.5/4 0.85 1.00 0.85
14 0 A-S8 3.0/6 0.96 0.85 0.81
15 1 AV 1.5/4 0.95 1.00 0.95
16 1 AV-S7 1.3/4 0.86 1.00 0.86
17 1 AV 1.3/4 0.86 1.00 0.86
18 1 AV 1.5/4 1.00 0.94 0.94
19 0 AM-S7 1.4/4 0.94 0.92 0.87
20 0 A-S8 3.0/6 0.93 0.82 0.76
21 0 A-S6 2.4/4 1.00 0.87 0.87
22 0 A-S6 2.4/4 0.84 0.89 0.74
23 0 A-S6 2.4/4 1.00 0.91 0.91
24 0 AV 2.0/4 0.98 0.86 0.84
25 0 A-S7 3.5/6 1.00 0.85 0.85
26 0 AV-S6 2.5/4 0.98 0.93 0.91
27 1 AV-S6 3.5/6 1.00 0.96 0.96
28 1 AV-S8 5.0/8 0.97 1.00 0.97
29 1 AV 2.0/4 0.85 0.88 0.75
30 0 A-7 3.5/6 0.89 1.00 0.89
31 0 AV 2.5/4 1.00 0.98 0.98
32 1 AV 2.5/4 1.00 1.00 1.00
33 1 AV 1.8/4 1.00 0.89 0.89
34 0 A-S8 3.0/6 1.00 0.72 0.72
35 0 AV 2.0/4 0.88 0.86 0.76
36 0 A-7 3.5/6 0.87 0.95 0.83
37 0 A-8 3.0/6 0.80 0.83 0.66
38 1 AV 1.8/4 1.00 0.91 0.91
FFV 39 0 A-S8 2.0/4 0.96 0.99 0.95
40 0 A-S6 6.0/12 0.74 1.00 0.74
41 0 A-7 3.5/6 0.86 1.00 0.86
42 0 A-7 3.5/6 0.87 1.00 0.87
43 0 A-S8 2.0/4 0.90 0.99 0.89
44 0 A-S8 2.4/4 0.74 1.00 0.74
45 0 A-S6 2.4/4 0.82 0.96 0.78
46 0 A-S6 3.6/6 0.79 0.99 0.78
47 0 A-S6 3.6/6 0.79 0.98 0.78
48 0 A-6 3.6/6 1.00 1.00 1.00
49 0 A-S6 2.0/4 0.75 1.00 0.75
50 0 AM-6 2.0/4 0.74 1.00 0.74
51 0 AM-6 2.0/4 0.75 1.00 0.75
52 0 A-7 3.5/6 0.88 1.00 0.88
53 0 A-7 3.5/6 0.87 0.97 0.84
54 0 A-S6 3.6/6 0.92 0.97 0.89
55 0 A-S6 3.6/6 0.92 1.00 0.92
56 0 A-S6 2.0/4 0.85 0.98 0.83
57 0 M-6 2.0/4 0.81 1.00 0.81
58 0 A-S6 2.4/4 0.86 0.93 0.79
59 0 A-6 3.6/6 1.00 0.95 0.95
60 0 A-6 3.6/6 1.00 0.95 0.95
61 0 A-6 2.0/4 1.00 1.00 1.00
62 0 A-7 3.5/6 0.87 0.93 0.81
63 0 A-7 3.5/6 0.87 0.96 0.84
64 0 A-6 3.6/6 1.00 0.94 0.94
65 0 A-8 3.6/6 0.93 0.90 0.83
66 0 A-8 3.6/6 0.82 0.87 0.71
67 0 A-5 3.6/6 0.92 0.91 0.84
68 0 A-8 3.6/6 1.00 0.90 0.90
69 0 A-8 3.6/6 0.81 0.87 0.70
70 0 A-S6 3.5/6 0.88 0.88 0.78
71 0 A-S6 3.5/6 0.79 0.88 0.70
(Continued )
1174 F.-K. WANG AND M. SAITO

(Continued).
Efficiency score
Type Carline (DMU) Vehicle maker Transition Engine First stage Second stage Overall
EV 72 0 N/A N/A 1.00 1.00 1.00
73 0 N/A N/A 1.00 0.81 0.81
74 0 N/A N/A 1.00 1.00 1.00
75 1 N/A N/A 1.00 0.79 0.79
76 1 N/A N/A 1.00 0.73 0.73
77 0 N/A N/A 1.00 0.83 0.83
78 1 N/A N/A 1.00 0.74 0.74
PHEV 79 0 N/A N/A 1.00 1.00 1.00
80 0 N/A N/A 1.00 0.93 0.93
81 0 N/A N/A 1.00 0.93 0.93
82 1 N/A N/A 1.00 0.89 0.89
Note: Vehicle maker (West = 0 and Asia = 1).

Vous aimerez peut-être aussi