ArticlePDF Available

Consequential life cycle air emissions externalities for plug-in electric vehicles in the PJM interconnection


Abstract and Figures

We perform a consequential life cycle analysis of plug-in electric vehicles (PEVs), hybrid electric vehicles (HEVs), and conventional gasoline vehicles in the PJM interconnection using a detailed, normative optimization model of the PJM electricity grid that captures the change in power plant operations and related emissions due to vehicle charging. We estimate and monetize the resulting human health and environmental damages from life cycle air emissions for each vehicle technology. We model PJM using the most recent data available (2010) as well as projections of the PJM grid in 2018 and a hypothetical scenario with increased wind penetration. We assess a range of sensitivity cases to verify the robustness of our results. We find that PEVs have higher life cycle air emissions damages than gasoline HEVs in the recent grid scenario, which has a high percentage of coal generation on the margin. In particular, battery electric vehicles with large battery capacity can produce two to three times as much air emissions damage as gasoline HEVs, depending on charge timing. In our future 2018 grid scenarios that account for predicted coal plant retirements, PEVs would produce air emissions damages comparable to or slightly lower than HEVs.
This content is subject to copyright. Terms and conditions apply.
Environ. Res. Lett. 11 (2016)024009 doi:10.1088/1748-9326/11/2/024009
Consequential life cycle air emissions externalities for plug-in electric
vehicles in the PJM interconnection
Allison Weis
, Paulina Jaramillo
and Jeremy Michalek
Engineering and Public Policy, Carnegie Mellon University, 5000 Forbes Ave, PA 15213, USA
Mechanical Engineering, Carnegie Mellon University, 5000 Forbes Ave, PA 15213, USA
Keywords: electric vehicles, LCA, social costs
Supplementary material for this article is available online
We perform a consequential life cycle analysis of plug-in electric vehicles (PEVs), hybrid electric
vehicles (HEVs), and conventional gasoline vehicles in the PJM interconnection using a detailed,
normative optimization model of the PJM electricity grid that captures the change in power plant
operations and related emissions due to vehicle charging. We estimate and monetize the resulting
human health and environmental damages from life cycle air emissions for each vehicle technology.
We model PJM using the most recent data available (2010)as well as projections of the PJM grid in
2018 and a hypothetical scenario with increased wind penetration. We assess a range of sensitivity
cases to verify the robustness of our results. We nd that PEVs have higher life cycle air emissions
damages than gasoline HEVs in the recent grid scenario, which has a high percentage of coal
generation on the margin. In particular, battery electric vehicles with large battery capacity can
produce two to three times as much air emissions damage as gasoline HEVs, depending on charge
timing. In our future 2018 grid scenarios that account for predicted coal plant retirements, PEVs
would produce air emissions damages comparable to or slightly lower than HEVs.
1. Introduction
Plug-in electric vehicle (PEV)technologies, including
plug-in hybrid electric vehicles (PHEVs)and battery
electric vehicles (BEVs), have the potential to reduce
environmental impacts from the transportation sys-
tem by reducing or eliminating tailpipe emissions.
However, the emissions associated with producing
PEVs and generating the electricity to charge PEVs
affect whether these vehicles have higher or lower life
cycle environmental and health impacts compared to
efcient gasoline vehicles (Michalek et al [1], Tessum
et al [2]). Evaluating the sustainability of different
transportation choices thus requires both a considera-
tion of the full life cycle of the technology as well as an
analysis of the impacts of the technology choice [3].
While there has been signicant research to under-
stand the life cycle environmental impacts of PEVs,
most of this prior work has followed an attributional
life cycle assessment (LCA)approach to answer the
question what air emissions are PEV charging respon-
sible for?Such approaches have assumed that PEV
charging produces emissions proportional to the
average emissions rate for electricity generation in the
political boundary (country, state, etc)or grid region
(NERC region, eGRID sub-region, interconnect, bal-
ancing area, etc)where the vehicle is charged. Alter-
natively, some studies have used hypothetical
emissions factors to evaluate a broader set of scenarios
of the effects of PEV charging [1,2,4,5]. The results
from previous work thus vary depending on the
researchers value judgment related to the emissions
that a PEV should be responsible for when charging in
a particular location.
In contrast, a consequential approach answers
what are the air emissions implications of PEV
adoption in a regionby assessing how grid operations
change in response to new charging demand. These
24 August 2015
24 December 2015
5 January 2016
9 February 2016
Original content from this
work may be used under
the terms of the Creative
Commons Attribution 3.0
Any further distribution of
this work must maintain
attribution to the
author(s)and the title of
the work, journal citation
and DOI.
© 2016 IOP Publishing Ltd
consequential effects of PEV charging on grid emis-
sions have been examined using empirical top-down
methods and normative bottom-upmethods [6].
Top-down empirical models of the power system, like
those developed by Siler-Evans et al [7]and Graff Zivin
et al [8], use regressions on historical data to estimate
marginal emissions rates. Such analysis is grounded in
the actual operations of the power system. However,
the approach is limited to historical scenarios and is
only appropriate for the analysis of small changes in
generation or load. It also suffers from error in coun-
terfactual analysis because correlations in past data do
not necessarily imply causality. The gap between cor-
relation and causality is particularly evident for the
dispatch of hydroelectric plants, which may change
generation timing in response to new load but typi-
cally will not change total energy generated in response
to new load. Alternatively, bottom-up normative
models of the power system, such as those used by
Sioshansi et al [9], Peterson et al [10], Choi et al [11]
and Weis et al [12], use optimization models to esti-
mate how a power system should operate to minimize
costs subject to a variety of constraints. These models
can assess changes of grid operation in response to
new PEV load. Such analysis can model future power
plant scenarios and large load changes. While this
approach has limited scalability for modeling large
systems, and it is typically not possible to model all
possible considerations that affect grid operations in
practice, there is growing interest in using bottom-up
models of the power system in consequential LCA.
Table 1summarizes prior consequential LCA stu-
dies of PEV air emissions. Tamayao et al [6]apply two
different top-down regression models [4,5]to assess
consequential greenhouse gas (GHG)emissions in the
US, while Ma et al [13]perform their own regression
on the operation of the UK grid. Sandy [14]and Onat
et al [15]average results from ORNLs analysis of mar-
ginal emissions for electric vehicles using a dispatch
model and different vehicle charging patterns [16].
Finally, Choi et al [11]construct a bottom-up capacity
expansion and unit commitment model to assess con-
sequential GHG emissions. In this paper, we adopt the
bottom-up normative approach to model the power
system under different scenarios in the PJM system in
order to inform a consequential LCA of PEVs. Unlike
previous work, our analysis includes the valuation of
social damages associated with emissions of criteria air
pollutants as well as GHG emissions. PJM (an inde-
pendent system operator in Pennsylvania, New Jersey,
Maryland, Ohio, and several other states)is an inter-
esting power system to examine, as it is the largest
independent system operator in the United States by
population and has a large installed coal capacity. The
supplemental information (SI)includes an expanded
comparison of this study with previous life cycle stu-
dies of PEVs in the United States.
2. Methods
2.1. Life cycle boundary
We estimate the life cycle emissions of CO
, CO, SO
, and VOCs for conventional, hybrid, and
PEVs, including the emissions from vehicle manufac-
turing, fuel production, and use. The analysis focuses
on the region covered by the PJM interconnection,
which has the largest electricity market in the US and
serves 13 states in the Mid-Atlantic region [17]. For
CVs and HEVs, the emissions from fuel production
and use include upstream emissions from petroleum
drilling and rening as well as the tailpipe emissions
during vehicle operations. For BEVs, fuel production
emissions include power plant emissions and
upstream emissions from coal and natural gas produc-
tion, but these vehicles do not have tailpipe emissions.
Finally, PHEVs have emissions associated with both
gasoline and electricity. Figure 1shows the scope of the
life cycle inventory. We do not consider end of life
emissions. Further, we assume a total vehicle life of
160 000 miles for all vehicle types, following Michalek
et al [1].
2.2. Vehicle and power grid scenarios
For this analysis we rely on a unit commitment and
economic dispatch (UCED)model previously devel-
oped and used to evaluate the costs and benets of
controlled charging of electric vehicles in the PJM
system [12,18]. This model uses a mixed integer linear
optimization program to minimize the costs of operat-
ing the power system to meet load given constraints,
including power plant operating constraints (mini-
mum and maximum load, ramping rates, and mini-
mum up and down time), as well as transmission and
operating reserves constraints. Furthermore, the
model incorporates vehicle charging by including
charging requirements and battery constraints.
In order to account for changes in the composition
of the PJM power plant eet that may result from
environmental regulations, renewable energy man-
dates, and changes in energy prices, we develop three
scenarios of the power system. For the rst scenario,
we use the Environmental Protection Agency (EPA)s
NEEDS database in order to represent the recent PJM
system [19]. Our second scenario, meant to represent a
future grid, includes retirement of power plants pre-
dicted by the EPA [20]and a 3% wind penetration, as
described in Weis et al [12]. Finally, in the third sce-
nario we include the power plant retirements EPA pre-
dicts, and we add 20% wind. The additional wind
generation for the future grid comes from NRELs
Eastern Wind Integration and Transmission Study
dataset [21]. For each scenario, we add wind sites in
order of capacity factor to reach the required wind
penetration level (3% for the second scenario and 20%
for the third scenario).
Environ. Res. Lett. 11 (2016)024009
Table 1. Summary of literature estimating consequential life cycle air emissions from PEVs.
Study Location Air emissions Model type Model detail
Tamayao et al [6]US GHG Empirical Applies regression results from [4,5]
Choi et al [11]Eastern Interconnect GHG Normative Capacity expansion and unit commitment
Ma et al [13]CA, UK GHG Empirical Regression on plant operation to nd average marginal emission factor
Sandy [14]US GHG Normative Averages resulting emissions from power plant dispatch in [16]
Onat et al [15]US GHG Normative Averages resulting emissions from power plant dispatch in [16]
This study US PJM GHG, SO
, CO, VOC Normative Optimal dispatch
Environ. Res. Lett. 11 (2016)024009
We also account for differences in vehicle technol-
ogy using Argonne National Laboratorys 2013
GREET 1 and 2 models [22,23]. The base case PEV for
the recent grid scenario is a 2010 PHEV-35, sized to
represent the Chevy Volt. We also include a smaller
battery size based on the Toyota Plug-in Prius and a
larger battery size based on the Tesla Model S. The ef-
ciency of the long-range BEV is that of the 2012 Tesla
Model S, as measured by the EPA [24](we examine
more efcient BEVs in sensitivity analysis). For the
future grid scenarios we use GREETs 2015 PHEV-35
vehicle specications. Table 2summarizes the differ-
ences across scenarios.
2.3. Life cycle inventory
We determine the life cycle emissions for each stage
shown in gure 1for each vehicle type based on a
160 000 mile vehicle life. Table 3provides a summary
of the data used for each stage of the life cycle.
2.3.1. Emissions from power plant operations
In order to estimate the change in emissions from the
power system resulting from vehicle charging, we solve
the UCED model both with and without electric
vehicles and compare the difference in the operating
schedules of the power plants. We add the charging
load to the existing non-vehicle electricity load by
assuming electric vehicles make up 10% of the vehicle
eet in PJM. Furthermore, we assume that electric
vehicles are distributed throughout the PJM system
proportionally to population (we test alternative
assumptions in sensitivity analysis).
Using data from the National Household Travel
Survey (NHTS)[25]we estimate the vehicle charging
load, following Weis et al [12,26]. The NHTS data
provide the distance driven during each trip through-
out the day surveyed as well as the time of each trip for
approximately 100 000 passenger cars across the US.
We use the distance driven in a day, the vehicle ef-
ciency, and the electric range of the vehicle to calculate
both the distance driven in charge-depleting versus
charge-sustaining mode and the total charging load
per day. We assume that all PHEVs drive as far as pos-
sible in charge-depleting mode before switching to
charge-sustaining mode. We also assume that vehicle
charging occurs at home after the last trip of the day
and that each vehicle is fully charged by the rst trip of
the next day. For each eet/vehicle type scenario, we
calculate the hourly charging load for a scenario where
charging begins at full power following the last trip of
the day (uncontrolled charging)and a scenario where
the utility dynamically controls the rate at which each
vehicle is charged in order to minimize generation cost
(controlled charging). Controlled charging of electric
vehicles may provide operational benets to the grid,
but it also changes the emissions associated with elec-
tric vehicle charging [12], so it represents an additional
scenario for analysis. For the controlled charging sce-
narios, we use 20 representative vehicle proles to
model the load from the electric vehicles as described
in Weis et al [12]and assume that the vehicle is con-
strained to be fully charged by the rst trip of the fol-
lowing day while the charging rate is limited to Level 2
charging (7.2 kW).
2.3.2. Tailpipe emissions
The distance driven using the gasoline engine (all times
for the CV and HEV and during charge-sustaining
mode for PHEVs)and the vehicles emission rate
determine its tailpipe emissions. We assume that each
vehicle travels 160 000 miles over its lifetime. Table 4
shows tailpipe emission rates for GHG, CO, SO
, and VOCs from the GREET 1 model [22].
While the two PHEVs burn some gasoline in charge-
depleting mode, the BEV-265 operates without any
tailpipe emissions.
2.3.3. Upstream emissions
Our life cycle emissions include the emissions from
the production, processing, and delivery of fossil fuels
(either to fuel the vehicles or to generate electricity),as
well as the emissions from vehicle and battery
Figure 1. Life cycle inventory for plug-in electric, hybrid electric, and conventional vehicles.
Environ. Res. Lett. 11 (2016)024009
Table 2. Summary of scenarios.
Scenario Power system data Conventional vehicle Hybrid vehicle Plug-in electric vehicles
2010 GREET PHEV-10 (Plug-in Prius-sized)
Recent (2010)2010 PJM 2010 GREET ICEV 2010 GREET HEV 2010 GREET PHEV-35 (Volt-sized)
2012 BEV-265 (Tesla-sized)
Future (~2018)EPA forecasted 2018 PJM with 3% wind penetration 2015 GREET ICEV 2015 GREET HEV 2015 GREET PHEV-35 (Volt-sized)
Hypothetical high wind future EPA forecasted 2018 PJM with 20% wind penetration 2015 GREET ICEV 2015 GREET HEV 2015 GREET PHEV-35 (Volt-sized)
Environ. Res. Lett. 11 (2016)024009
manufacturing. Argonne National Laboratorys
GREET 1 model [22]provides the emission rates for
oil drilling and rening, which are the basis of the
upstream emissions estimate for gasoline. Upstream
emissions for vehicle and battery manufacturing come
from the GREET 2 model [23].
To account for the emissions from the production,
processing, and delivery of coal and natural gas for
power plants used to charge the electric vehicles, we
rely on the GREET 1 model, which provides these
emissions on a per MWh basis (as shown in the sup-
porting information). We then apply these emissions
factors to the amount of electricity generated from
coal and natural gas plants in the UCED model results.
We include only the upstream emissions from coal-
and natural gas-based electricity as these fuels account
for the majority of generation response to additional
vehicle charging load [12]. In the high wind scenario,
wind power also contributes to vehicle charging, but
wind requires no fuel and thus has no upstream emis-
sions (our model does not include the emissions from
building physical infrastructure).
2.4. Life cycle damages
Air emissions cause environmental degradation and
affect human health. For CO
, we use estimates of the
social cost of carbon that the US EPA uses for
regulatory impact assessment [27]. EPA reports this
social cost of carbon for three different discount rates:
2.5%, 3%, and 5%. We use the 3% discount rate
average value for 2010 as our base value for all
In order to estimate the damages from SO
, and VOCs we use the values from the AP2 model
[28], which estimates the marginal health and envir-
onmental damages for emissions of each criteria air
pollutant in each county in the United States. This
model has many uncertain parameters, including the
value of statistical life, which is used to monetize mor-
bidity and mortality from air pollution. The AP2
model includes results from a Monte Carlo analysis
of the damages in each county for the baseline
year (2005). As a base case, we assume that these
2005 marginal damages per unit of emission in each
location apply also to the recent grid and future grid
scenarios. We use the distribution of the results from
the Monte Carlo analysis to characterize the uncer-
tainty within the AP2 model. Damages from vehicle
charging are based on the change in the annual genera-
tion and emissions from each power plant in the
UCED model that results from increased charging
load. Since the AP2 values are specic to individual
counties where emissions take place, we also need to
incorporate the location of the vehicle tailpipe emis-
sions. To do so, we allocate vehicles to counties
within PJM proportionally to population. We further
assume that each vehicle is driven within its respective
Table 3. Data for the life cycle emissions for each stage.
Stage Emission rate Source Other assumptions Source
Power plant operation Short Ton/year Unit commitment
Driving patterns and
vehicle efciency
NHTS GREET 1, fuelec-
Tailpipe lb/mile GREET 1 Driving patterns NHTS
Vehicle manufacturing Short Ton/lifetime GREET 2
Battery manufacturing Short Ton/lifetime GREET 2
Oil production Short Ton/mile GREET 1
Gasoline rening Short Ton/mile GREET 1
Coal production Short Ton/MWh GREET 1 MWh produced using coal Unit commitment model
Natural gas production Short Ton/MWh GREET 1 MWhproduced using nat-
ural gas
Unit commitment model
Table 4. Tailpipe emissions in pounds per 1000 miles from GREET 1. The recent grid scenario relies on the
data for the 2010 vehicles, while the future scenarios are based on the characteristics of the 2015 vehicles. CS:
charge-sustaining mode. CD: charge-depleting mode.
Vehicle CO
2010 CV 772 0.375 6.393 0.265 0.026 0.011
2010 HEV 551 0.265 6.393 0.220 0.026 0.008
2010 PHEV-10 CS (Plug-in Prius)529 0.265 6.393 0.220 0.026 0.008
2010 PHEV-10 CD (Plug-in Prius)265 0.088 2.205 0.079 0.006 0.004
2010 PHEV-35 CS (Volt)683 0.265 6.393 0.220 0.026 0.010
2010 PHEV-35 CD (Volt)44.1 0.022 0.375 0.013 0.001 0.001
2010 BEV-265 (Tesla Model S)0.00 0.00 0.00 0.00 0.00 0.00
2015 CV 705 0.375 6.393 0.265 0.026 0.011
2015 HEV 507 0.265 6.393 0.220 0.026 0.007
2015 PHEV CS (Volt)573 0.265 6.393 0.220 0.026 0.009
2015 PHEV CD (Volt)41.9 0.016 0.397 0.014 0.001 0.001
Environ. Res. Lett. 11 (2016)024009
We allocate the emissions from vehicle and battery
manufacturing to US counties identied by the US
census as having automobile and parts manufacturing
activity, weighted by the number of automotive manu-
facturing workers, following Michalek et al [1].We
also allocate coal, oil, and natural gas upstream emis-
sions to US counties where mines, oil and gas elds,
and reneries are located, weighted by the production
in each county. Figure 2shows the resulting cumula-
tive probability distribution of damages from manu-
facturing, coal, oil and gas production, and oil
rening. The damage calculations assume that all
emissions and damages occur in the United States,
though in practice some of these processes occur
outside of US borders. Marginal emissions in other
countries could incur higher or lower damages than in
the US, depending on the existing pollutant con-
centrations and populations in those areas. At this
time, however, we are unable to include these dama-
ges. We allocate the damages for all life cycle stages
except for vehicle and battery manufacturing across
the years of the vehicle lifetime, ignoring changes in
the electricity grid over the vehicle life. We then used a
3% discount rate to nd the present value of these
damages, consistent with the calculation of CO
3. Results
3.1. Life cycle emissions
Figure 3shows the breakdown of estimated emissions
by life cycle stage for each scenario. In the recent grid
scenario, the PEVs have higher GHG, SO
, and
emissions and lower CO and VOC emissions
than the HEV. Compared to the CV, PEVs have higher
emissions and lower CO and VOC emissions,
while GHG, NO
and PM
emissions may be higher
or lower, depending on the PEV characteristics and
the charging scenario. Controlled charging of PEVs
increases emissions of GHGs, SO
, and PM
while reducing emissions of VOCs compared to
uncontrolled charging, due in part to the increased use
of coal-red power plants available at night. Use of
coal generation increases with controlled charging in
all scenarios because these plants have lower marginal
cost than gas plants and have more excess capacity
available in off-peak hours than during uncontrolled
charging ours (the hours immediately following the
vehiclesrst home arrival at the end of the day). For a
detailed description of the analysis and assumptions
regarding the response of power plant dispatch to
additional electric vehicle charging load see [12].
In the future grid scenarios, compared to the HEV,
the PHEV-35 has higher SO
emissions and lower
, VOC, and CO emissions, while GHG and NO
emissions may be higher or lower, depending on the
charging scenario and wind power scenario. Com-
pared to the CV, the PHEV-35 has higher SO
lower GHG, NO
, VOC, and CO emissions.
Controlled charging of the PHEV-35 increases SO
and NO
while decreasing VOC emissions, but the
effect on GHG and PM
emissions depends on the
presence of wind, and the effect on CO emissions is
Figure 2. Cumulative probability distribution of damages per ton of pollutant released for assumed locations of upstream production
emissions by pollutant type (using AP2 data for 2005).
Environ. Res. Lett. 11 (2016)024009
3.2. Life cycle damages
3.2.1. Expected values
PEVs have higher expected life cycle damages than
hybrid vehicles in the recent PJM scenario in all cases
examined, as shown in gure 4. Their expected
damages are also higher than those of conventional
vehicles, except for the case of the PHEV-10 with
uncontrolled charging. Long-range BEVs cause two to
three times as much air emissions damage as HEVs.
The electricity generation damages come largely as a
result of the SO
emissions from the coal plants used to
charge the vehicles in off-peak hours. Controlled
charging increases life cycle damages relative to
uncontrolled charging because of the increases in
emissions associated with higher levels of coal genera-
tion. Uncertainty is not presented here because
common sources of uncertainty create correlated
uncertainty across scenarios, so error bars could be
misleading in comparing across cases. Instead,
section 3.2.2. includes an analysis of uncertainty and
robustness in these results.
In the future scenarios, shown in gure 5, the
PHEV-35 is able to reduce life cycle damages by a few
hundred dollars over its lifetime compared to the CV
and the HEV. Again, PEVs tend to produce larger
damages under controlled charging than under
uncontrolled charging, but in the future scenarios the
PHEV-35 provides benets compared to the CV and
the HEV regardless of the charging scheme. The high-
wind future scenarios do not necessarily imply lower
consequential damages than the low-wind future sce-
narios for PEVs because added wind displaces fossil
fuel plants, which can increase the availability of coal
on the margin when PEVs charge.
Figure 3. Life cycle emissions by pollutant and life cycle stage for each vehicle type in the recent (a)and future (b)PJM grid. UC stands
for uncontrolled charging and CC stands for controlled charging.
Environ. Res. Lett. 11 (2016)024009
3.2.2. Uncertainty and robustness
Qualitatively, our key ndings are that (1)the PEVs
cause more damage than the CV and HEV in the recent
PJM grid, (2)the PHEV-35 causes less damage than
the CV in the future PJM grid scenarios, but these
damages are not much lower than those for the HEV,
and (3)controlled charging tends to increase damages
compared to uncontrolled charging, though the PEVs
provide benets compared to the CV and HEV in the
future grid scenarios regardless of charging scheme.
To characterize uncertainty and robustness of these
ndings, we use the Monte Carlo analysis results from
the AP2 model and assess the probability that each
vehicle technology has higher life cycle air emissions
damages than the HEV. Table 5reveals that the
conclusions above are robust, especially in the recent
grid. The uncertainty in the AP2 models does not
signicantly affect the comparison across vehicle types
because most of this uncertainty is a result of
uncertainty in the value of a statistical life, which
affects damage estimates across all cases. As a result,
this uncertainty typically changes only the magnitude
of the difference between hybrids and other vehicles,
not the sign.
The results in table 5incorporate only the
uncertainty quantied in the AP2 model due to uncer-
tainty in input parameters, such as the value of a statis-
tical life. We examine the effect of some of the
other important input parameters and assumptions
through sensitivity cases summarized in table 6
Figure 4. Expected value of life cycle air emission damages for each vehicle type in the recent PJM grid. UC stands for uncontrolled
charging and CC stands for controlled charging for the electric vehicles.
Figure 5. Expected value of life cycle damages in the future PJM grid. The high wind scenario has 20% of demand met by wind. UC
stands for uncontrolled charging and CC stands for controlled charging.
Environ. Res. Lett. 11 (2016)024009
and detailed in the supporting information. These
include alternative assumptions for (1)emissions
damage models, (2)BEV efciency, (3)power grid
characteristics that inuence dispatch decisions, (4)
PEV adoption patterns, and (5)policy effects on con-
sequential life cycle implications. Our key ndings
are generally robust, though future fuel prices could
affect the consequential emissions benets of PEVs
versus HEVs; binding SO
caps would likely make
PEVs more competitive; and assessing a larger con-
sequential life cycle scope that accounts for the effect
of PEV adoption on each manufacturers vehicle
eet emissions due to US corporate average fuel
economy and GHG emission policy results in the
conclusion that PEV adoption increases damages
in all scenarios modeled (the policy extends through
4. Discussion and conclusions
Using a consequential LCA approach for the vehicles
and scenarios modeled, we nd that (1)PEVs cause
more damage than HEVs in the recent PJM grid, (2)
PEVs cause comparable or slightly lower emissions
than HEVs in the future PJM grid scenarios, and (3)
utility-controlled (mostly nighttime)PEV charging
tends to increase life cycle emissions compared to
uncontrolled charging. However, (1)changes in future
fuel prices could affect whether PEVs have higher or
Table 5. Robustness of results for the damage difference between hybrid vehicles and each other vehicle type. CV=conventional vehicle.
Scenario Charging
Probability damages are larger than
for HEVs
Mean change in life cycle damages com-
pared to HEVs
CVrecent grid 100% $650
PHEV-10recent grid Uncontrolled 81% $210
Controlled 98% $1100
PHEV-35recent grid Uncontrolled 95% $1200
Controlled 99% $4200
BEV-265recent grid Uncontrolled 99% $4800
Controlled 99% $8400
CVfuture grid 99% $580
PHEV-35future grid Uncontrolled 4% $420
Controlled 28% $60
PHEV-35future grid with
high wind
Uncontrolled 17% $150
Controlled 18% $150
Table 6. Summary of ndings from sensitivity analysis.
Sensitivity scenario Purpose of scenario Finding
AP2 2011 damage estimates Test more recent estimates of damages per
ton emissions from each location
Higher estimated damages, but key ndings are robust.
Increased potential for PEVs to lower damages ver-
sus HEVs in future grid
EASIUR damage estimates Test an alternative method for estimating
Higher estimated damages, but key ndings are robust.
PEVs have comparable damages versus HEVs in
future grid
High efciency long-
range BEV
Test the life cycle damages of a BEV with
comparable performance to the other
Lower estimated damages but key ndings are robust.
BEVs have higher damages than CVs, HEVs, and
PHEVs in the recent grid
Future with expensive coal,
cheap natural gas
Test effect of possible future fuel prices PEVs slightly increase damages versus HEVs when
uncontrolled, reduce damages when controlled
Future with cheap coal, expen-
sive natural gas
Test effect of possible future fuel prices PEVs slightly reduce damages versus HEVs whether
uncontrolled or controlled
Additional spinning reserves
required for high wind case
Test effect of higher reserve requirements for
intermittent wind
Key ndings are robust
No transmission constraints Test effect of transmission constraints and
location of loads and responding
Key ndings are robust
PEVs in urban areas only Test effect of location of vehicles Key ndings are robust
cap Test effect of a binding SO
cap SO
damages likely reduced, but not necessarily to zero
since PEV charging may affect location of SO
sions. PEVs likely reduce damages
CAFE leakage effect Test effect of vehicle eet-wide emissions
enabled by PEV sales due to CAFE incen-
tives for AFV adoption
PEVs increase overall US damages versus HEVs even in
future grid and even with high wind penetration
Environ. Res. Lett. 11 (2016)024009
lower damages than HEVs in a future grid and whether
controlled charging increases or decreases emissions,
and (2)when the effect of US PEV adoption on
automaker eet emissions is accounted for, PEV
adoption could increase emissions in all scenarios due
to leakage effects in federal fuel economy standards
[29,3436]. The SI includes a discussion about the
limitations of the model beyond the scope of the
sensitivity analysis.
Our results for the recent PJM grid are consistent
with those from Tessum et al using the 2007 electricity
mix [2]: PEVs have higher damages than gasoline vehi-
cles in the recent grid. Michalek et al [1]found that
PEVs with larger batteries cause more damage, which
we also observe in our recent grid scenarios. None of
the PEVs in our study, regardless of battery size, have
lower damages than the HEV in the recent grid due to
the large amounts of coal on the margin in PJM com-
pared to the average mix used in Michalek et al [1]. The
future of the grid past 2018, which is relevant for
future PEV adoption, is expected to be lower-emitting
than the recent grid, but consequential emissions from
a grid far into the future are difcult to meaningfully
project. Both Michalek et al [1]and Tessum et al [2]
nd that PEVs can reduce damages if charged with
zero-emission electricity. This is a useful bounding
case but not a scenario likely to occur soon, since even
if wind, solar, nuclear, and hydroelectric power make
up a much larger portion of the grid mix in the future,
the consequential effect of PEV charging (the differ-
ence between grid operations when PEVs are present
versus absent)is still primarily to increase generation
from fossil fuel plants. Since most wind, solar, nuclear,
and hydroelectric power is fully used in the absence of
PEV load, PEV adoption will not cause an increase in
generation from these plants. Only when low-emis-
sion plants would have been curtailed in the absence of
PEVs can PEV adoption result in increased use of these
plants. Thus, the consequential emissions of PEV
charging are affected more by the mix of coal and nat-
ural gas plants in a region than by the amount of
renewable or low-emission generation capacity.
The difference between coal and natural gas gen-
eration is signicant. We show that even in one of the
power systems in the country with the highest coal
generation, PEVs could reduce transportation health
and environmental damages in the near future, long
before a zero-carbon electricity mix is achieved, due
primarily to substitution of natural gas for coal on the
While PEVs can double or triple air emission
damages in the recent grid relative to HEVs, they could
reduce damages in a future grid. However, we estimate
that near future (2018)potential air emissions bene-
ts from PEV adoption in PJM are small relative to
HEVs (or even negative when considering the net
effect on the automakerseet under federal fuel econ-
omy policy). Nevertheless, electrication may offer a
promising long term option to signicantly reduce air
emissions from the transportation sector compared to
some other alternative transportation fuels, including
biofuels and natural gas, that have been shown to offer
small-to-no reductions in GHG emissions and could
have unintended consequences like higher global food
prices [30,31]. Indeed, the logistics of regulating emis-
sions from individual vehicles over their functional
lives are more difcult than regulation of power plant
emissions [37].
Continued regulation of the electricity system can
increase the benets of vehicle electrication, and con-
sequential air emissions implications of PEV charging
are already lower in many regions than in PJM [6].
While near-term benets of PEV adoption in PJM are
estimated to be small or negative, a transition of the
transportation system could lead to long-term benets
outside the scope of this analysis, including greater
benets in other regions and future emissions savings
enabled by a transition to electric vehicles as the elec-
tricity grid becomes cleaner and as public policy
adjusts [32,33].
We would like to thank to Nick Muller for his support
with the APEEP and AP2 models as well as Peter
Adams and Jinhyok Heo for their help with the
EASIUR model. The Doris Duke Charitable Founda-
tion, the Richard King Mellon Foundation, the Electric
Power Research Institute, and the Heinz Endowment
provided support for this research through the Renew-
Elec project ( The National
Science Foundation also provided nancial support
through a Graduate Research Fellowship, and Toyota
Motor Corporation provided support through a
nancial gift. Findings and conclusions are the sole
responsibility of the authors and may not represent the
views of the funding entities.
[1]Michalek J et al 2011 Valuation of plug-in vehicle life-cycle air
emissions and oil displacement benets Proc. Natl Acad. Sci.
USA 108 165548
[2]Tessum C, Hill J and Marshall J 2014 Life cycle air quality
impacts of conventional and alternative light-duty
transportation in the United States Proc. Natl Acad. Sci. USA
111 184905
[3]Eckelman M 2013 Life cycle assessment in support of
sustainable transportation Environ. Res. Lett. 8021004
[4]Samaras C and Meisterling K 2008 Life cycle assessment of
greenhouse gas emissions from plug-in hybrid vehicles:
implications for policy Environ. Sci. Technol. 42 31706
[5]Hawkins T, Singh B, Majeau-Bettez G and Strømman A H
2013 Comparative environmental life cycle assessment of
conventional and electric vehicles J. Ind. Ecology 17 5364
[6]Tamayao M, Michalek J J, Hendrickson C and Azevedo I 2015
Regional variability and uncertainty of electric vehicle life cycle
emissions across the United States Environ. Sci. Technol.
49 884455
[7]Silas-Evans K, Azevedo I and Morgan M G 2012 Marginal
emissions factors for the US electricity system Environ. Sci.
Technol. 46 47428
Environ. Res. Lett. 11 (2016)024009
[8]Graff Zivin J, Kotchen M and Mansur E 2014 Spatial and
temporal heterogeneity of marginal emissions: implications
for electric cars and other electricity-shifting policies J. Econ.
Behav. Organ. 107 24868
[9]Sioshansi R and Denholm P 2010 The value of plug-in hybrid
electric vehicles as grid resources Energy J. 31 110
[10]Peterson S B, Whitacre J F and Apt J 2011 Net air emissions
from electric vehicles: the effect of carbon price and charging
strategies Environ. Sci. Technol. 45 17927
[11]Choi D G, Kreikebaum F, Thomas V and Divan D 2013
Coordinated EV adoption: double-digit reductions in
emissions and fuel use for $40/vehicle-year Environ. Sci.
Technol. 47 1070310707
[12]Weis A, Michalek J, Jaramillo P and Lueken R 2015 Emissions
and Cost implications of controlled electric vehicle charging
in the US PJM interconnection Environ. Sci. Technol. 49
[13]Ma H, Balthasar F, Tait N, Riera-Palou X and Harrison A 2012
A new comparison between the life cycle greenhouse gas
emissions of battery electric vehicles and internal combustion
vehicles Energy Policy 44 16073
[14]Thomas S 2012 US marginal electricity grid mixes and EV
greenhouse gas emissions Int. J. Hydrog. Energy 37 1923140
[15]Onat N C, Kucukvar M and Tatari O 2015 Conventional,
hybrid, plug-in hybrid or electric vehicles? State-based
comparative carbon and energy footprint analysis in the
United States Appl. Energy. 150 3649
[16]Hadley W S and Tsvetkova A 2008 Potential Impacts of Plug-in
Hybrid Electric Vehicles on Regional Power Generation Oak
Ridge, Tennessee (
[17]FERC (
[18]Lueken R and Apt J 2014 The effects of bulk electricity storage
on the PJM market Energy Syst. 5677704
[19]EPA 2010 National Electric Energy Data System (NEEDS)
v.4.10 (
[20]Environmental Protection Agency 2018 Parsed Results
Detailed Output Files (
[21]NREL Eastern Wind Dataset(
transmission/eastern_wind_dataset.html)(accessed 7
July 2011)
[22]Argonne National Laboratory 2013 GREET 1 (
[23]Argonne National Laboratory 2013 GREET 2 (
[24]EPA 2012 Tesla Model S,
[25]US Department of Transportation, Federal Highway
Administration 2009 National Household Travel Survey 2009
( 6 October 2011)
[26]Weis A, Jaramillo P and Michalek J 2014 Estimating the
potential of controlled plug-in hybrid electric vehicle charging
to reduce operational and capacity expansion costs for electric
power systems with high wind penetration Appl. Energy 115
[27]Interagency Working Group on the Social Cost of Carbon
Technical Support Document:Technical Updated of the
Social Cost of Carbon for Regulatory Impact AnalysisUnder
Executive Order 12866 (
[28]Muller N AP2 (APEEP)Model (
[29]Jenn A, Michalek J J and Azevedo I L Unintended
consequences: Why US alternative fuel vehicle adoption
increases gasoline consumption and greenhouse gas emissions
(in review)
[30]Mullins K A, Grifn W M and Matthews H S 2011
Policy Implications of uncertainty in modeled life-cycle
greenhouse gas emissions of biofuels Environ. Sci. Technol. 45
[31]Tong F, Jaramillo P and Azevedo I M L 2015 Comparison of life
cycle greenhouse gases from natural gas pathways for light-
duty vehicles Energy Fuels 29 600818
[32]Greene D, Park S and Liu C 2013 Analyzing the Transition to
Electric Drive in California The Howard H. Baker Jr Center for
Public Policy (
[33]National Research Council of the National Academies 2013
Transitions to Alternative Vehicles and Fuels (
[34]Gecan R, Kile J and Beider P 2012 Effects of Federal Tax Credits
for the Purchase of Electric Vehicles US Congressional Budget
Ofce pp 136 (
[35]Goulder L, Jacobsen M and Van Benthem A 2012 Unintended
consequences from nested state and federal regulations: the
case of the Pavley greenhouse-gas-per-mile limits J. Environ.
Econ. Manage. 63 187207
[36]Jenn A 2014 Advanced and alternative fuel vehicle policies:
regulations and incentives in the United States PhD
Dissertation Department of Engineering and Public Policy,
Carnegie Mellon University, Pittsburgh, PA
[37]Beaton S P et al 1995 On-road vehicle emissionsregulations,
costs, and benets Science 268 9913
Environ. Res. Lett. 11 (2016)024009
... Using reported annual on-road emissions from the U.S. EPA National Emissions Inventory (NEI) and the California Air Resources Board, we estimated associated social costs for the years 2008, 2011, 2014, and 2017, corresponding to the years with published data in the NEI. We compared changes in emissions across these years, and the corresponding social costs were calculated by multiplying emissions by the RCM estimates produced by InMAP, EASIUR, and AP2 as well as the social costs of carbon estimates published by the US EPA (Weis et al 2016, Clark et al 2017, Heo et al 2017, Muller 2017, Muller and Jha 2017, Tessum et al 2017, Zhao et al, Zimmerman et al 2017. Then the social costs were categorized by rural/suburban/urban classification, fuel types, and vehicle types in order to determine the largest sources of social costs from on-road emissions. ...
... Here, we assess health and climate consequences that are incurred due to on-road emissions, in monetized terms (herein referred to as 'social costs'). For the health pollutants, we estimated the social costs using three RCMs-AP2, EASIUR, and InMAPwhich provide county-specific marginal social costs of each pollutant (in $ ton −1 ) (Weis et al 2016, Clark et al 2017, Heo et al 2017, Muller 2017, Muller and Jha 2017, Tessum et al 2017, Zhao et al 2017, Zimmerman et al 2017, CACES 2019. We used a concentration response function (CRF-function defining the relationship between an increase in annual exposure to PM 2.5 and mortality risk) based on the American Cancer Society's CRFs and the EPA's 2017 statistical value of a human life (VSL) (∼9.0 million $2017 USD). ...
Full-text available
Local and state policymakers have become increasingly interested in developing policies that both reduce greenhouse gas (GHG) emissions and improve local air quality, along with public health. Interest in developing transportation-related policies has grown as transportation became the largest contributing sector to GHG emissions in the United States in 2017. Information on current emissions and health impacts, along with trends over time, is helpful to policymakers who are developing strategies to reduce emissions and improve public health, especially in areas with high levels of transportation-related emissions. Here, we provide a comprehensive assessment of the public health and climate social costs of on-road emissions by linking emissions data generated by the U.S. Environmental Protection Agency to reduced complexity models that provide impacts per ton emitted for pollutants which contribute to ambient fine particulate matter, and the social costs of GHG emissions from on-road transportation. For 2017, social costs totaled $184 billion (min: $78 billion; max: $280 billion) for all on-road emissions from the eight health and climate pollutants that we assessed in the continental U.S. (in $2017 USD). Within this total social cost estimate, health pollutants constituted $93 billion of the social costs (min: $52 billion; max: $146 billion), and climate pollutants constituted $91 billion (min: $26 billion; max: $134 billion). The majority of these social costs came from CO2 followed by NO x emissions from privately owned individual vehicles in urban counties (CO2 contributed $51 billion and NO x contributed $16 billion in social costs from individual vehicles in urban counties). However, it is important to note that not all the attention should be placed solely on individual vehicles. Although the climate social costs of individual vehicle emissions are higher than those from commercial vehicles in urban counties (by two to eight times depending on the climate pollutant), the health social costs of individual vehicle emissions are roughly equal to those from commercial vehicles in urban counties. Regardless of each pollutant’s contributions to the social costs, the highest social benefits from reducing 1 ton of CO2 and its co-pollutants would occur in urban counties, given their high population density.
... However, such an option may not explicitly consider the impact of EV integration on the future electricity generation investment and could not consider the time-dependent character of electricity mix or EV charging patterns. In response to these deficiencies, an emerging trend is to combine LCA for EV with energy system modeling, as suggested by Arvesen et al. (2021), Jochem et al. (2015), and Weis et al. (2016). This calls for the development of novel modeling methods, which dynamically combines LCA approaches with energy system modeling and considers sophisticated and empirical charging demand of EV including their flexibilities for providing V2G. ...
The electrification of the transport sector plays a key role in the global energy transition and it is of great necessity to assess emissions induced by electric vehicles in the long term for effective policy‐making. Typical life cycle assessment may not consider the impact of electric vehicle integration in future electricity systems adequately, or the time‐dependent characteristics of electricity generation mix and EV charging patterns. The solution requires modeling methods to integrate electric vehicle into energy system models, especially with vehicle‐to‐grid option. However, relevant methods have not been evaluated, yet. This integration is mathematically ambitious especially for huge and heterogeneous fleets of electric vehicles and brings energy system models to their computational limits. So far, current studies have proposed several aggregation methods for the load from electric vehicle charging, which simplify the original problem but may provoke bias. In our contribution, we propose a novel method of integrating vehicle‐to‐grid compliant electric vehicles into energy system models and demonstrate its feasibility by comparing it with two recent others from the literature. Taking the performance of the individual modeling method as the benchmark, we improve one of the two methods from the literature with updated parameters and additional constraints. We apply all three aggregation methods in a simple energy system model for comparing and analyzing their performances from multiple aspects, that is, solution accuracy, computational complexity, parameter requirement, and their impact on greenhouse gas emissions. Finally, we discuss the reasons behind the differences and give recommendations for further research.
... Also, as new consumers are increasingly looking for lawn equipment that is easier to use and causes less environmental impact, original equipment manufacturers are extending their portfolio of electricpowered models. Interestingly, renewable sources (solar, wind, hydro) used for electric generation are projected to be increasingly exploited (Weis et al., 2016;UNEP 2019;Visual Capitalist 2019). From an environmental standpoint, as more renewable electricity enters the grid, the environmental impacts of electric-powered solutions will further diminish. ...
Full-text available
Purpose Gasoline-powered lawn mowers and garden equipment are emitting 30 million tons of pollutants yearly in the USA, accounting for a quarter of all non-road gasoline emissions. While the US market is dominated by gasoline-powered lawn mowers, this study offers an assessment of the environmental implication and cost of electrifying the lawn mower industry. Methods First, the lifecycle environmental footprint and total cost of ownership of electric-powered mowers are calculated and compared to those of conventional gasoline-powered counterparts, using life cycle assessment (LCA) and life cycle costing (LCC) methodologies. A multi-indicator impact assessment is notably conducted, using the SimaPro software (v8.5), the ReCiPe methodology (H), and the ecoinvent database (v3.4) completed with data from the GREET model for the use phase. Second, an extrapolation model is computed to interpret the results at a national and regional scale, considering the proper energy mix in each US state. The combination of LCA and LCC results, mapped out in a two-dimensional chart, allows a clear visual representation of the environmental and economic trade-offs between the gasoline and electric solutions. Results and discussion The findings indicate a reduction of 49.9% and 32.3% of CO2 emissions, respectively, for push and riding mowers, by using the electric solution instead of the conventional one over their lifecycle. Yet, the total cost of ownership is slightly higher (4.7–10.6%) for the electric solutions, even if the operating costs are lower. And as the initial buying price of the electric solution is more expensive than the gasoline solution of the same category, this could be a real hindrance for consumers who do not systematically consider the overall lifecycle cost when comparing mowers. In this line, the quantification of a suitable financial incentive to support the electrification of the lawn mower market is of utmost importance and appears as a promising line for future work. Conclusions The present results are significant in at least two major respects for the potential electrification of lawn mowing equipment. First, they show how an increased market share of electric mowers can contribute to cutting down greenhouse gas emissions. Second, such quantitative results can be useful for decision-makers in businesses and state governments to take appropriate ecological actions, e.g., in the development of adequate financial incentives or green policy to support the energy transition in this sector, and thus tackle global warming.
... Papers were excluded that did not match these criteria. There are a number of closely related fields of study that we exclude from this review scope, including research that: [1] evaluates the climate forcing potential of fuel switching (33,34); [2] omits an explicit calculation of emissions associated with the policy or technology (35,36); [3] calculates (25,37) or applies (11,12,17,38,39) scaling factors to approximate air quality impacts; [4] uses optimization methods to evaluate the most cost-effective way to achieve air quality and greenhouse gas reduction targets (11,17,(40)(41)(42)(43); [5] assumes a fixed criteria air pollutant or greenhouse gas emissions reduction (44)(45)(46)(47)(48), including zeroing out the emissions of a particular sector (49) and global warming temperature targets (50-55); [6] evaluates climate change's impact on air pollution without the context of a climate mitigation action (56, 57); [7] conducts life cycle assessments (58); and [8] has not been published in peerreviewed journals (15,59). Beyond the papers from the WoS keyword searches, we included research cited by retrieved papers or that cited retrieved papers, as long as papers met our inclusion criteria. ...
Full-text available
Research on air quality and human health “co-benefits” from climate mitigation strategies represents a growing area of policy-relevant scholarship. Compared to other aspects of climate and energy policy evaluation, however, there are still relatively few of these co-benefits analyses. This sparsity reflects a historical disconnect between research quantifying energy and climate, and research dealing with air quality and health. The air quality co-benefits of climate, clean energy, and transportation electrification policies are typically assessed with models spanning social, physical, chemical, and biological systems. This review article summarizes studies to date and presents methods used for these interdisciplinary analyses. Studies in the peer-reviewed literature (n = 26) have evaluated carbon pricing, renewable portfolio standards, energy efficiency, renewable energy deployment, and clean transportation. A number of major findings have emerged from these studies: [1] decarbonization strategies can reduce air pollution disproportionally on the most polluted days; [2] renewable energy deployment and climate policies offer the highest health and economic benefits in regions with greater reliance on coal generation; [3] monetized air quality health co-benefits can offset costs of climate policy implementation; [4] monetized co-benefits typically exceed the levelized cost of electricity (LCOE) of renewable energies; [5] Electric vehicle (EV) adoption generally improves air quality on peak pollution days, but can result in ozone dis-benefits in urban centers due to the titration of ozone with nitrogen oxides. Drawing from these published studies, we review the state of knowledge on climate co-benefits to air quality and health, identifying opportunities for policy action and further research.
... In the life cycle of a BEV that includes its manufacturing and disposal, other important sources of emissions are the production of the battery and the vehicle itself (Tessum, Hill, & Marshall, 2014). Weis, Jaramillo, and Michalek (2016) conducted a study comparing BEVs and conventional gasoline vehicles and found that BEVs have higher sulfur dioxide but lower CO and volatile organic compound life-cycle emissions. Based on a similar comparative analysis, Nealer et al. (2015) found that, on average, BEVs produce less than half the GHG emissions of comparable gasoline-powered vehicles. ...
The transportation sector is experiencing a paradigm shift. Three emerging technologies—connected and autonomous vehicles, on-demand mobility services, and zero-emission vehicles are changing the way people and goods are moved. In this chapter, potential benefits and unintended consequences due to the adoption of these technologies are analyzed from emissions, exposure, and health standpoints. The impacts of these technologies are classified according to their level of immediacy to emissions-related health concerns. From the most immediate to the least, the impacts are changes in exposure time, vehicle and fuel technologies, short-term travel activity, and long-term travel behavior. In addition, the likelihood, directionality, and level of impacts of the three technologies are qualitatively assessed based on evidence in the existing literature. Finally, implementation considerations specific to each of the three technologies for lowering emissions and consequent health impacts are discussed.
Full-text available
New energy vehicles (NEVs), especially electric vehicles (EVs), address the important task of reducing the greenhouse effect. It is particularly important to measure the environmental efficiency of new energy vehicles, and the life cycle analysis (LCA) model provides a comprehensive evaluation method of environmental efficiency. To provide researchers with knowledge regarding the research trends of LCA in NEVs, a total of 282 related studies were counted from the Web of Science database and analyzed regarding their research contents, research preferences, and research trends. The conclusion drawn from this research is that the stages of energy resource extraction and collection, carrier production and energy transportation, maintenance, and replacement are not considered to be research links. The stages of material, equipment, and car transportation and operation equipment settling, and forms of use need to be considered in future research. Hydrogen fuel cell electric vehicles (HFCEVs), vehicle type classification, the water footprint, battery recovery and reuse, and battery aging are the focus of further research, and comprehensive evaluation combined with more evaluation methods is the direction needed for the optimization of LCA. According to the results of this study regarding EV and hybrid power vehicles (including plug-in hybrid electric vehicles (PHEV), fuel-cell electric vehicles (FCEV), hybrid electric vehicles (HEV), and extended range electric vehicles (EREV)), well-to-wheel (WTW) average carbon dioxide (CO2) emissions have been less than those in the same period of gasoline internal combustion engine vehicles (GICEV). However, EV and hybrid electric vehicle production CO2 emissions have been greater than those during the same period of GICEV and the total CO2 emissions of EV have been less than during the same period of GICEV.
Full-text available
Light weighting by material substitution is a key to reducing GHG emissions during vehicle operation. The GHG benefits are a salient factor in selecting lightweight materials for vehicles. Although the literature has performed lightweight material selections using GHG benefits under product- and fleet-based life-cycle inventory (LCI) analyses, recycling effects have therein been accounted for by arbitrarily selecting allocation methods for recycling, as the consensus on their selection is absent. Furthermore, studies have mistreated the temporal variations of the LCI parameters (the dynamic inventory (DI)), though that could be an important factor affecting the overall LCI results when allocation methods for recycling are in place. Therefore, to investigate their influence on greenhouse gas (GHG) benefit evaluations, an LCI case study was conducted, centered on aluminum- and magnesium-substituted internal combustion engine vehicles (ICEVs) at the product- and fleet- levels. “CO2 savings” and the “CO2 payback time”, as well as four allocation methods for recycling, were considered to represent the GHG benefits and address the recycling effects, respectively. The dynamic inventory was based on the world average electricity grid mix change. The results indicate that changing the conditions of the DI and the allocation methods for recycling could alter the better performing material under fleet-based analyses. Therefore, we ascertained that the choice of the allocation method for recycling and conducting fleet-scale dynamic LCI analyses in the presence of the DI is pivotal for material selections.
The past century has seen a remarkable rise in personal mobility and heavy goods transport. The development of the internal combustion engine has played a pivotal role in this development. Significant progress has been made in improving engine efficiency and reducing emissions. However, further improvements are necessary in order to meet local zero emission regulation as well as global climate goals. A rapid transition to renewable energy sources is key, enabling clean electricity generation and widespread deployment of sustainable fuels. Every country has a role to play. Developing nations must learn to become less dependent on fossil fuels as they grow their economies and industrialized nations must continue their sustainability journey and quickly transfer critical knowledge and lessons learned. Technologies should be assessed in terms of their life cycle impact and not simply their tailpipe emissions. As we consider the wide range of disparate applications across the transportation sector, we would be wise to embrace a fact-driven approach, keeping multiple options open and to build on past successes. Rather that betting it all on a single technology, a diverse mix of low-carbon technologies should be pursued.
Ridesourcing services from transportation network companies, like Uber and Lyft, serve the fastest growing share of U.S. passenger travel demand.1 Ridesourcing vehicles' high use intensity is economically attractive for electric vehicles, which typically have lower operating costs and higher capital costs than conventional vehicles. We optimize fleet composition (mix of conventional vehicles (CVs), hybrid electric vehicles (HEVs), and battery electric vehicles (BEVs)) and operations to satisfy demand at minimum cost and compare findings across a wide range of present-day and future scenarios for three cities. In nearly all cases, the optimal fleet includes a mix of technologies, HEVs and BEVs make up the majority of distance traveled, and CVs are used primarily for periods of peak demand (if at all). When life cycle air pollution and greenhouse gas emission externalities are internalized via a Pigovian tax, fleet electrification increases and externalities decrease, suggesting a role for policy. Externality reductions vary from 10% in New York (where externality costs for both gasoline and electricity consumption are relatively high and a Pigovian tax induces a partial shift to BEVs), to 22% in Los Angeles (where high gasoline and low electric grid externalities lead a Pigovian tax to induce a near-complete shift to BEVs).
Full-text available
Many researchers use life cycle assessment methodology to investigate the energy and environmental impacts of energy-saving and new energy vehicles. However, in the context of China, the life cycle energy-saving and emission-reduction effects of extended-range electric vehicles (EREVs), and the optimal applicable vehicle size and driving conditions for EREVs have been rarely studied. In this study, based on the life cycle assessment theory, the resource consumption, energy exhaustion, and environmental impact of EREVs were comprehensively analyzed. In addition, a differential evaluation model of ecological benefits was established for comparing EREVs with other vehicles with different power sources. Finally, scenario analysis was performed in terms of different vehicle sizes and driving conditions. The results have shown that EREV has great advantages in reducing mineral resource consumption and fossil energy consumption. The consumption of mineral resources of EREV is 14.68% lower than that of HEV, and the consumption of fossil energy is 34.72% lower than that of ICEV. In terms of environmental impact, EREV lies in the middle position. The scenario analysis has revealed that, for EREV in China, the optimal vehicle size is the passenger car and the optimal driving condition is the suburban condition. This work helps to understand the environmental performance of EREVs in China and may provide a decision-making reference for the government.
Full-text available
Electric vehicles (EVs), plug-in hybrid electric vehicles (PHEVs), and hybrid electric vehicles (HEVs) are often considered as better options in terms of greenhouse gas emissions and energy consumption compared to internal combustion vehicles. However, making any decision among these vehicle options is not a straightforward process due to temporal and spatial variations, such as the sources of the electricity used and regional driving patterns. In this study, we compared these vehicle options across 50 states, taking into account state-specific average and marginal electricity generation mixes, regional driving patterns , and vehicle and battery manufacturing impacts. Furthermore, a policy scenario proposing the widespread use of solar energy to charge EVs and PHEVs is evaluated. Based on the average electricity generation mix scenario, EVs are found to be least carbon-intensive vehicle option in 24 states, while HEVs are found to be the most energy-efficient option in 45 states. In the marginal electricity mix scenario , widespread adoption of EVs is found to be an unwise strategy given the existing and near-future marginal electricity generation mix. On the other hand, EVs can be superior to other alternatives in terms of energy-consumption, if the required energy to generate 1 kW h of electricity is below 1.25 kW h.
Full-text available
Commonly considered strategies for reducing the environmental impact of light-duty transportation include using alternative fuels and improving vehicle fuel economy. We evaluate the air quality-related human health impacts of 10 such options, including the use of liquid biofuels, diesel, and compressed natural gas (CNG) in internal combustion engines; the use of electricity from a range of conventional and renewable sources to power electric vehicles (EVs); and the use of hybrid EV technology. Our approach combines spatially, temporally, and chemically detailed life cycle emission inventories; comprehensive, fine-scale state-of-the-science chemical transport modeling; and exposure, concentration-response, and economic health impact modeling for ozone (O3) and fine particulate matter (PM2.5). We find that powering vehicles with corn ethanol or with coal-based or "grid average" electricity increases monetized environmental health impacts by 80% or more relative to using conventional gasoline. Conversely, EVs powered by low-emitting electricity from natural gas, wind, water, or solar power reduce environmental health impacts by 50% or more. Consideration of potential climate change impacts alongside the human health outcomes described here further reinforces the environmental preferability of EVs powered by low-emitting electricity relative to gasoline vehicles.
Full-text available
Simulations predict that the introduction of PHEVs could impact demand peaks, reduce reserve margins, and increase prices. The type of power generation used to recharge the PHEVs and associated emissions will depend upon the region and the timing of the recharge.
Low prices and abundant resources open new opportunities for using natural gas, one of which is the production of transportation fuels. In this study, we use a Monte Carlo analysis combined with a life cycle analysis framework to assess the greenhouse gas (GHG) implications of a transition to natural gas-powered vehicles. We consider six different natural gas fuel pathways in two representative light-duty vehicles: a passenger vehicle and a sport utility vehicle. We find that a battery electric vehicle (BEV) powered with natural gas-based electricity achieves around 40% life cycle emissions reductions when compared to conventional gasoline. Gaseous hydrogen fuel cell electric vehicles (FCEVs) and compressed natural gas (CNG) vehicles have comparable life cycle emissions with conventional gasoline, offering limited reductions with 100-year global warming potential (GWP) yet leading to increases with 20-year GWP. Other liquid fuel pathways (methanol, ethanol, and Fischer-Tropsch liquids) have larger GHG emissions than conventional gasoline even when carbon capture and storage technologies are available. Life cycle GHG emissions of natural gas pathways are sensitive to the vehicle fuel efficiency, to the methane leakage rates of natural gas systems, and to the GWP assumed. With the current vehicle technologies, the break-even methane leakage rates of CNG, gaseous hydrogen FCEV, and BEV are 0.9%/2.3%, 1.2%/2.8%, and 4.5%/10.8% (20-year GWP/100-year GWP). If the actual methane leakage rate is lower than the break-even rate of a specific natural gas pathway, that natural gas pathway reduces GHG emissions compared to conventional gasoline; otherwise, it leads to an increase in emissions.
Battery electric vehicles (BEVs) are often portrayed as "green," implying negligible greenhouse gas (GHG) emissions. While BEVs are zero emission vehicles, the electrical power generators used to recharge vehicle batteries do emit copious GHGs. Some analysts have estimated the power plant GHG emissions due to charging EV batteries using the average electrical generator grid mix for a given region. However, the GHG protocol specifies that analysts should use the marginal grid mixes to accurately calculate GHG emissions from adding EVs to the vehicle fleet. This paper utilizes the marginal grid mixes for each electrical power region in the US, and calculates the vehicle-weighted average GHG emissions for the entire country. These calculations demonstrate that, on the average, each BEV that displaces a gasoline hybrid electric vehicle (HEV) will increase GHGs by more than 7% and each PHEV put in service will increase GHGs by an average of 10% compared to a gasoline HEV. Copyright
We characterize regionally specific life cycle CO2 emissions per mile traveled for plug-in hybrid electric vehicles (PHEVs) and battery electric vehicles (BEVs) across the United States under alternative assumptions for regional electricity emission factors, regional boundaries, and charging schemes. We find that estimates based on marginal vs average grid emission factors differ by as much as 50% (using National Electricity Reliability Commission (NERC) regional boundaries). Use of state boundaries versus NERC region boundaries results in estimates that differ by as much as 120% for the same location (using average emission factors). We argue that consumption-based marginal emission factors are conceptually appropriate for evaluating the emissions implications of policies that increase electric vehicle sales or use in a region. We also examine generation-based marginal emission factors to assess robustness. Using these two estimates of NERC region marginal emission factors, we find the following: (1) delayed charging (i.e., starting at midnight) leads to higher emissions in most cases due largely to increased coal in the marginal generation mix at night; (2) the Chevrolet Volt has higher expected life cycle emissions than the Toyota Prius hybrid electric vehicle (the most efficient U.S. gasoline vehicle) across the U.S. in nearly all scenarios; (3) the Nissan Leaf BEV has lower life cycle emissions than the Prius in the western U.S. and in Texas, but the Prius has lower emissions in the northern Midwest regardless of assumed charging scheme and marginal emissions estimation method; (4) in other regions the lowest emitting vehicle depends on charge timing and emission factor estimation assumptions.
We develop a unit commitment and economic dispatch model to estimate the operation costs and the air emissions externality costs attributable to new electric vehicle electricity demand under controlled vs. uncontrolled charging schemes. We focus our analysis on the PJM Interconnection and use scenarios that characterize (1) the most recent power plant fleet for which sufficient data are available, (2) a hypothetical 2018 power plant fleet that reflects upcoming plant retirements, and (3) the 2018 fleet with increased wind capacity. We find that controlled electric vehicle charging can reduce associated generation costs by 23%-34% in part by shifting loads to lower-cost, higher-emitting coal plants. This shift results in increased externality costs of health and environmental damages from increased air pollution. On balance, we find that controlled charging of electric vehicles produces negative net social benefits in the recent PJM grid but could have positive net social benefits in a future grid with sufficient coal retirements and wind penetration.
Plug-in hybrid electric vehicles (PHEVs) can become valuable resources for an electric power system by providing vehicle to grid (V2G) services, such as energy storage and ancillary services. We use a unit commitment model of the Texas power system to simulate system operations with different-sized PHEV fleets that do and do not provide V2G services, to estimate the value of those services. We demonstrate that a PHEV fleet can provide benefits to the system, mainly through the provision of ancillary services, reducing the need to reserve conventional generator capacity. Moreover, our analysis shows that PHEV owners are made better off by providing V2G services and we demonstrate that these benefits can reduce the time it takes to recover the higher upfront capital cost of a PHEV when compared to other vehicle types.
Recent advancements in battery technologies may make bulk electricity storage economically feasible. We analyze the value of two electrochemical storage technologies and traditional pumped hydropower storage in the 2010 PJM day-ahead energy market, using a reduced-form unit commitment model. We find that large-scale storage would increase overall social welfare in PJM. However, the annualized capital costs of storage would exceed social welfare gains. Consumers would save up to $4 billion annually due to reduced peak prices and reduced reliance on expensive peaking generators. These savings are equivalent to \(\sim \) 10% of sales in the PJM day-ahead energy market. Savings come largely at the expense of generator surplus. Existing market mechanisms are insufficient to encourage the socially optimal quantity of storage. Storage reduces the profitability of generators and the need for peaking generation capacity. Storage modestly increases emissions of CO \(_{2}\) and other pollutants in a system with 2010 PJM characteristics.
Electric power systems with substantial wind capacity require additional flexibility to react to rapid changes in wind farm output and mismatches in the timing of generation and demand. Controlled variable-rate charging of plug-in electric vehicles allows demand to be rapidly modulated, providing an alternative to using fast-responding natural gas plants for balancing supply with demand and potentially reducing costs of operation and new plant construction. We investigate the cost savings from controlled charging of electric vehicles, the extent to which these benefits increase in high wind penetration scenarios, and the trade-off between establishing a controlled charging program vs. increasing the capacity of generators in the power system. We construct a mixed integer linear programming model for capacity expansion, plant dispatch, and plug-in hybrid electric vehicle (PHEV) charging based on the NYISO system. We find that controlled charging cuts the cost of integrating PHEVs in half. The magnitude of these savings is ∼5% to 15% higher in a system with 20% wind penetration compared to a system with no wind power, and the savings are 50–60% higher in a system that requires capacity expansion.