-------
1.1.3.2 Development of truck curve
Historically, there has been a significant increase (offset) between the car and light truck
footprint-based curves to reflect the additional utility of trucks. The large shift in sales from car
crossovers to truck crossovers might suggest that the size of this offset was not appropriate for
vehicles with similar towing and hauling capability - for example, crossover vehicle models
(trucks) equipped with AWD compared to those same models with 2WD (cars). Most of these
vehicles available with both driveline options exhibit the same tow rating and nearly identical
GCWR.
In redesigning the truck curve, EPA considered the "base utility" of moving people for
passenger vehicles and light trucks to be similar (this is especially true for crossover vehicles and
wagons, for example). However, larger trucks which are designed for more towing and hauling
capability do require design changes to allow for handling of these larger loads and this is
reflected in increased engine capability, body-on-frame design, and greater structural mass. EPA
analyzed empirical fleet data to quantify the additional tailpipe CO2 resulting from these required
design changes and use it as a basis for a "utility offset" that is built into the slope of the final
truck curves.
The truck curve is based on the car curve, but with additional allowances for 1) AWD and 2)
towing and hauling utility. The analysis that went into the determination of each offset, and the
resulting truck slope, is detailed below.
3) AWD Offset
EPA analyzed certification data (Ellies 2023) from MY 2019 (the latest at the time the
analysis was completed) to compare the tailpipe CO2 emissions of crossover vehicle models with
2WD and AWD driveline configurations and identical engines. In total, 32 vehicle models were
offered in both a 2WD and an AWD version and were subject to passenger vehicle and light
truck CO2 compliance targets, respectively.
Figure 1-6. Increase in Tailpipe CO2 Emissions: MY 2019 AWD vs. 2WD Crossovers.
1-9
-------
Figure 1-6 shows the distribution of tailpipe increase between unique 2WD and AWD vehicle
models. The median increase in tailpipe CO2 is 12.5 g/mile for these models, although several
models showed increases below 10 g/mi. As this characteristic is the only attribute distinguishing
a "truck" crossover from a "car" crossover that should produce measurable tailpipe CO2
differences, it forms the basis for the offset between the car and truck curves for vehicles of
equivalent towing capacity. Based on this analysis, EPA's final footprint curves reflect an offset
between the car and truck curves of 10 g/mile for ICE vehicles equipped with AWD.
4) Towing and Hauling Utility Offset
In determining an offset for truck utility, EPA reviewed vehicle specifications available in the
MY 2019 fleet data. One way to quantify a vehicle's utility (or maximum output) is by its gross
combined weight rating (GCWR).8 GCWR is the value specified by the vehicle manufacturer as
the maximum weight of a loaded vehicle and trailer (40 CFR § 86.1803-01 2023).
In its simplest form,
GCWR = GVWR + maximum loaded trailer weight,
where:
GVWR (gross vehicle weight rating) is the value specified by the manufacturer as the
maximum design loaded weight of a single vehicle (40 CFR § 86.1803-01 2023).
EPA first reviewed MY 2019 vehicle models and plotted GCWR vs engine performance. Of
horsepower or engine torque, engine torque correlated best with a truck's utility. As shown in
Figure 1-7, there is a positive correlation between a vehicle's GCWR and its rated engine torque.
8 GVWR describes the maximum load that can be carried by a vehicle, including the weight of the vehicle itself.
GCWR describes the maximum load that the vehicle can haul, including the weight of a loaded trailer and the
vehicle itself. For more information, please refer to the Medium and Heavy-duty GHG Phase 2 FRM (81 FR 73478
2016).
1-10
-------
20000
18000
16000
14000
_n
12000
cc
&
10000
>
O
15
8000
6000
4000
2000
GCWR vs Torque: All
y = 37,767k
k2 - B.9M9
1
..-•i i
•
~
1
\ •
|V»
• • >- 1
• •
•
•V
•
..w* ' I
1 :
%
iP* • 1
50 100 ISO 200 250 300
Rated Torque (ft-lb)
350
400
450
500
Figure 1-7. GCWR-Torque Relationship, MY 2019 Light Truck Data.
As seen in the fleet data, vehicle models which are offered at a higher tow rating than the base
model will be equipped with a more powerful engine (and accompanying transmission, driveline,
and chassis improvements). From a modeling perspective EPA focused on the increase in engine
torque based on the relationship observed above.
EPA then evaluated the increase in tailpipe CO2 for additional towing capacity using response
surface equations (RSEs) from ALPHA model results as follows:
• First, we estimated the required nominal engine torque for three vehicle models with
different body styles (small pickup, SUV, and full-size pickup) at various tow rating
levels by calculating the GCWR and applying the relationship seen in Figure 1-7.
• Then we scaled each engine model to an appropriate displacement (to match required
torque) for various modeled engine architectures9 based on each modeled engine's
BMEP. Test weight (curb + 300 pounds) was increased slightly to account for heavier
powertrain, driveline, suspension, and brakes that are required for greater towing
capacity. Road loads were modified slightly based on this increased weight. We were
then able to predict CO2 based on the RSE results for a downsized turbocharged
engine and various gasoline GDI engine models.
9 ALPHA modeled engines include GDI with and without cylinder deactivation and Turbo Gas for pickups, and GDI
and Atkinson for CUVs.
1-11
-------
• The modeling results show the increase in CO2 as a function of an increase in towing
capacity in Figure 1-8. The data suggests that the average increase in CO2 for a given
vehicle is about 9 g/mile per addi tional 1000 pounds of tow capability
en
Incremental C02 vs Incremental Towing
V = G.0087X
R2 =0.9728
50
40
30
20
10
0 <
9
•
,.'Jh
•
w
•
k-*"
00
3 1000 2000 3000 4000 5000 6000 70
Figure 1-8. Incremental CO2 as a Function of Increased Towing Capacity.
Finally, MY 2019 data shown in Figure 1-9 indicates that tow rating is directionally
proportional with footprint (as longer wheelbases are required for stability during increased
towing demands). The difference in towing capacity between a 70 square foot truck (at a sales-
weight average tow rating slightly over 9000 pound) and that of a 45 square foot truck (with
average tow rating just over 2000 pound) is 7000 pounds. Based on the relationship derived
above for CO2 vs. towing capacity, this would correspond to an additional 63 g/mile of tailpipe
CO2 between 45 and 70 square feet.10 EPA combined these relationships to establish an
appropriate footprint-based truck slope that is based on the additional utility that trucks are
designed for. This represents the full utility-based offset of the truck curve for a 100 percent ICE
vehicle fleet.
10 EPA is not considering towing differences for trucks greater than 70 square feet or smaller than 45 square feet.
1-12
-------
Tow Rating vs Footprint
14000
y = 278.59x-10134
12000 R2« 0.6417
• * *
— »' it'l —
| 8000 I— . • » * . *,'•••
•rz m • ••# 1** ® ••
| ? iV. • *
j 6000 * • r - • t •
tS • • •
• • •
4000 •• • m
• { .•# • • • • ••• ••
•
J •
2000 •• • • • •
• • • • •
• • • • •
o -
40 45 50 55 60 65 70 75
Footprint (sq ft)
Figure 1-9. Tow Rating-Footprint Relationship, MY 2019 Trucks.
For a strictly ICE vehicle fleet, the AWD and utility offset would look as shown in Figure
1-10.
y = 278.59X -1013/
•
•
• • •
*
# R2# 0.6417
•
: •
1
I
V
•
!•
# •
m m 9 j
••• ¦
• ••
• — •
••
•• %!
•
• m
• • •
• •
•
•
•
•
•
••
•
•
• •% ••
• • «
• •
> ••• ••
~
• • •
•
1-13
-------
Buildup of Truck Curve from
"Base" Car Cum
3
\
\
\
\
%
S
s
s
s
s
~
CUD
QJ
ClO
~
s
s
s
s
£
I
CD
s
s
*
_ — —
35 40 4
— — Car n
0
|
5 50 55 60 65 70 75 8
Footprint (sq ft)
o CP w/AWD — — Truck no CP w/AWD and Utility Offsets
Figure 1-10. AWD and Utility Offset Applied to Establish Truck Curve (100 percent
ICE).11
However, as described in Section III.C.2.ii of the preamble, we are scaling the car and truck
curves as appropriate to reflect expected increases in technologies with increasingly lower
tailpipe GHG emissions. For the 2030 fleet we are applying to these offsets a 50 percent factor
(associated with an average tailpipe target that is nominally 50 percent of the base year target), as
well as a 50 percent factor to the base car slope. We recognize that across our sensitivity analyses
there is a wide range of technology penetrations, and believe this approach is appropriate to
capture the range of low-GHG emitting technologies represented in our future projections. This
reduces the AWD offset to 5 g/mile and the full-size truck utility offset to 31.5 g/mile as shown
in Figure 1-1 L
11 For this figure and the subsequent figures, "no CP" indicates that no outpoints were reflected in these plots.
1-14
-------
Buildup of Truck Curve from
"Base" Car Curv<
3
00
CD
ClO
**
**
£
CD
I
O
**
3
0
i
5 40 45 50 55 60 65 70 75 8
Footprint (sq ft)
— — Car no CP w/AWD — — Truck no CP w/AWD and Utility Offsets
Figure 1-11. A WD and Utility Offset Applied to Establish Truck Curve (Scaled).
1.1.3.3 Analysis of Footprint Response to Standards
To confirm that the slopes for car and truck curves would not incentivize a shift in vehicle
size, we analyzed the projected trend in vehicle footprint for the standards to confirm a minimal
overall change in vehicle size for the combined fleet. Figure 1-12 shows a comparison of 2020
base year footprint (blue) compared to the MY 2032 average projected footprint (orange) for the
standards, for BEV and ICE vehicles, by body style. As can be seen, the BEVs increase slightly
in size, while the ICE vehicles decrease slightly. These two tendencies offset each other to
minimize the overall change in fleet size.
1-15
-------
80.0
70.0
60,0
4->
4—
cr
i/>
50.0
C
40.0
u
a.
o
30.0
o
Ll_
20.0
10.0
0.0
Values
¦ Sum of BYFP_SW
¦ SumofFP SW
BEV
ICE
BEV
ICE
BEV
ICE
sedan cuv_suv pickup
Figure 1-12. Comparison of Average Footprint to Base Year Footprint for Final Standards
Table 1-2 shows the numerical MY 2032 average footprint (FP) for the various body styles
for BEVs and ICE vehicles, and the fleet averages, compared to base year (MY 2020) footprint
for the final standards. The overall change in average footprint (51.3 square feet) compared to
the base year footprint (50.6) is minimal (an increase of 1 percent).
Table 1-2: Comparison of MY 2032 Footprint to Base Year Footprint, Final Standards
BEV
ICE
Combined
Base FP
MY 2032 FP
Base FP
MY 2032 FP
Base FP
MY 2032 FP
Sedan
46.5
48.1
46.0
43.7
46.4
47.1
CUV/SUV
49.0
50.9
49.0
46.7
49.0
49.7
Pickup
65.8
69.1
65.5
63.2
65.7
65.7
Total
49.7
51.7
52.7
50.3
50.6
51.3
1-16
-------
1.1.3.4 Cut points
EPA evaluated the sales weight-average footprint for full size pickups in determining the
appropriate upper truck cutpoint for this rule. Figure 1-13 shows that the average footprint has
increased for full size pickups from 67 square feet to over 69 square feet in 2021. The upper
cutpoint has increased from 66 square feet in MY 2018 to 69 square feet in 2021. To avoid any
incentive to further upsize the full-size pickups, EPA is finalizing a provision that phases down
the long-term upper truck cutpoint to 70 square feet.12 The upper cutpoint for cars is unchanged
at 56 square feet.
70
2012 2013 2014 2015 2016 2017 2018 2019 2020 2021
Model Year
^¦Full Size Truck - SWFP LT Upper Cutpt
Figure 1-13. Sales-weighted Footprint of Full-Size Pickups, 2012-2021 MY
EPA is finalizing that vehicles smaller than 45 square feet should not be subject to more
stringent standards based on an extrapolation of the utility offset approach described above.
Many vehicle models smaller than 45 square feet, both cars and trucks, are offered and EPA does
not want to discourage vehicles in this segment for equity and affordability concerns. These
12 In the 2021 rale, for MYs 2023 and beyond the upper track cutpoint was restored to the original 74 square foot
value first finalized in 2012. EPA's final rale reduces the upper cutpoint beginning in MY 2027, with Ml phase
down (from 74 in 2026) to 70 square feet by 2030.
1-17
-------
include popular vehicles such as the Subaru Crosstrek, Nissan Kicks, the Chevy Trax, and the
Honda HR-V.
Applying the cutpoints to the preceding methodology yields the final curve shape that is
shown in Figure 1-14.
Buildup of Truck Curve from "Base" Curve
Footprint (sq ft)
— —Car no CP w/AWD — — w/tow Car w/CP Trkw/CP
Figure 1-14. Car and Truck Curves, Scaled, with Cutpoints
1.2 Development of the final GHG standards for Medium-Duty Vehicles
1.2.1 History of GHG standards for Medium-Duty Vehicles
In the Phase 1 Heavy-duty rule, EPA established a GHG standards program structure for
complete Class 2b and 3 heavy-duty vehicles (referred to in this rule as 'medium-duty pickups
and vans") as part of a joint GHG and CAFE program with NHTSA (76 FR 57106 2011). The
Phase 1 standards began to be phased-in for MY 2014 with the final Phase 1 stringency levels
stabilizing in MY 2018. The Phase 1 program worked well to establish a first time GHG
standards program for these work-oriented vehicles. The Phase 2 program established more
stringent standards for MY 2027, phased in over MYs 2021-2027, requiring additional GHG
reductions (81 FR 73478 2016). The MY 2027 standards would remain in place unless and until
amended by the agency. Medium-duty vehicles (previously described as heavy-duty vehicles in
the Phase 1 and Phase 2 FID GHG rules) with a gross vehicle weight rating (GVWR) between
1-18
-------
8,501 and 10,000 pounds are classified in the industry as Class 2b motor vehicles while vehicles
with GVWR between 10,001 and 14,000 pounds are classified as Class 3 motor vehicles.
Class 2b includes vehicles classified as medium-duty passenger vehicles (MDPVs) such as
very large SUVs (40 CFR § 86.1803-01 2023). We are finalizing changes in the definition of
MDPV in 40 CFR § 86.1803-01, as we described in Section III.E of the preamble. EPA
performed a study to assess the GHG increases of a medium-duty pickup compared to a similar
sized light-duty pickup when they are operated similarly as primarily unloaded vehicles
transporting just the operator and also if they are lightly loaded with 1/2 the payload capacity.
Results of this study are summarized in RIA Chapter 3.5. This comparison reflects the issue that
medium-duty pickups have certain heavier duty design aspects (frames, axles, brakes,
transmissions, etc.) intended for trailer towing work that negatively impact GHG emissions when
they are only operated with lighter loads similar to the expected operation from a light-duty
pickup.
Table 1-3 summarizes the chassis test data for the F150 and the F250, each tested in its
original configuration and alternative configuration (as a 2b for the F150, and as an light-duty
LDT4 for the F250). The F250 with the 7.3L engine, tested at a light-duty ETW of curb+300
pounds, emitted 172 g/mile more than the F150. Similarly, the F250 emitted 170 g/mile more
than the F150 with both tested at ALVW.
Table 1-3: GHG Emissions Comparison of LD and MD pickup
Targets
Test vehicle #
Model Test config
A B
C
RLHP
50
ETW
Dyno
cfg
FTP
IIWI
E
US06
55/45
Notes
KFA20095
2019
^ LDT4 @ ETW
46.347 0.2527
0.03984
21.1
5142
4WD
476
322
515
407
native config / 3
test avg.
tested as 2b @ ALVW
60.03 0.2527
0.03984
23.0
6698
4WD
529
359
591
452
3-test avg
Delta
9%
30%
11%
11%
15%
11%
MDF250 RLD1
tested as LDT4 @ ETW
2022
P25Q 2b @ ALVW
48.87 0.12
59.33 0.12
0.050665
0.050665
Delta
24.2 6896
25.6 8373
6% 21%
4WD 682
4WD 736
8%
454 707
483 808
6% 14%
-5?8-
622
7%
3-test avg
native config / 3
test avg.
The GHG emission difference observed in the data indicates that light- to medium-load
operation results in much higher CO2 emissions in the medium-duty pickup under similar
passenger or payload conditions. The medium-duty pickup is designed primarily for regular
towing and therefore may have higher emissions under other operating conditions compared to
light-duty pickups designed more for transportation of passengers or cargo in the bed.
Because MDPVs are designed primarily to be used as light-duty passenger vehicles, they are
regulated under the light-duty vehicle rules. Thus, the requirements for MDPVs in this
rulemaking are the same as the light-duty pickups with respect to both GHG and criteria
emission standards.
1-19
-------
Historically, about 90 percent of medium-duty pickups and vans have been what are often
referred to as "3/4-ton" and "1-ton" pickup trucks,13 12- and 15-passenger vans, and large work
vans that are sold by vehicle manufacturers as complete vehicles, with no secondary
manufacturer making substantial modifications prior to registration and use. Most of these
vehicles are produced by companies with major light-duty markets in the United States,
primarily Ford, General Motors, and Stellantis.14 Often, the technologies available to reduce
GHG emissions from this segment are similar to the technologies used for the same purpose on
light-duty pickup trucks and vans, including both engine efficiency improvements (for gasoline
and diesel engines) and vehicle efficiency improvements. In the Heavy-Duty Phase 1 (76 FR
57106 2011) and Phase 2 (81 FR 73478 2016) rules, EPA adopted GHG standards for medium-
duty pickups and vans based on the whole vehicle (including the engine), expressed as grams of
CO2 per mile, consistent with the way these vehicles are regulated by EPA today for criteria
pollutants.
Vehicle testing for both the medium-duty and light-duty vehicle programs is conducted on
chassis dynamometers using the drive cycles from the EPA Federal Test Procedure (Light-duty
FTP or "city" test) and Highway Fuel Economy Test (HFET or "highway" test) (40 CFR §
1066.801 Subpart 12023). For the light-duty GHG standards, EPA factored vehicle attributes
into the standards by basing the GHG standards on vehicle footprint (the wheelbase times the
average track width). For those standards, passenger cars and light trucks with larger footprints
are assigned higher GHG targets (see Chapter 1.1.1.1). For HD pickups and vans, the agencies
also set GHG standards based on vehicle attributes but used a work-based metric as the attribute
rather than the footprint attribute utilized in the light-duty vehicle rulemaking. Work-based
measures such as payload and towing capability are key among the parameters that characterize
differences in the design of these vehicles, as well as differences in how the vehicles will be
utilized. Buyers consider these utility-based attributes when purchasing a HD pickup or van.
EPA therefore finalized Phase 1 and 2 standards for medium-duty pickups and vans based on a
"work factor" attribute that combines the vehicle's payload and towing capabilities, with an
added adjustment for 4-wheel drive vehicles.
For Phase 1 and 2, the agencies adopted provisions such that each manufacturer's fleet
average standard is based on production volume-weighting of target standards for all vehicles
that in turn are based on each vehicle's work factor (76 FR 57106 2011) (81 FR 73478 2016).
These target standards are taken from a set of curves (mathematical functions). The Phase 2 work
factor GHG standards are shown in Figure 1-15 for reference. The agencies established separate
standards for diesel and gasoline medium-duty pickups and vans. Note that this approach does
not create an incentive to reduce the capabilities of these vehicles because less capable vehicles
are required to have proportionally lower emissions and fuel consumption targets.
13 "3/4-ton" and "1-ton" are common industry terms, not regulatory definitions. These terms typically refer to Class
2b and Class 3 trucks, respectively. For specific regulatory definitions for Class 2b and Class 3, please refer to 40
CFR § 86.1803-01.
14 Formerly Fiat-Chrysler during the period when the Heavy-duty Phase 1 and 2 standards were developed.
1-20
-------
Q)
I
o
u
5000
Work Factor
O
O
_o
"ra
3B
c
_o
Q.
c
o
u
Diesef Standards
O
O
O
u
Work Factor
Figure 1-15: Heavy-duty Phase 2 work factor-based GHG standards for medium-duty
pickups and vans (81 FR 73478 2016).
1-21
-------
1.2.2 Development of the final standards for Medium-Duty Vehicles
Medium-duty-vehicles (MDV)15 are similar to the light-duty trucks addressed in this program
with respect to both technological opportunity for electrification as well as in terms of how they
are manufactured. Several light-duty manufacturers are also the primary manufacturers of the
majority of medium-duty pickups and vans. Medium-duty pickups and vans share close parallels
to the light-duty program regarding how EPA has developed our medium-duty standards and
compliance structures with the penetration of technologies such as electrification. The primary
difference between the light-duty and the MDV standards is that MDV standards continue to be
based on work attributes rather than vehicle footprint. MDV pickups and vans are true work
vehicles that are designed for much higher towing and payload capabilities than are light-duty
vehicles. The technologies applied to light-duty vehicles are not all applicable to MDVs at the
same adoption rates, and the internal combustion engine technologies often produce a lower
percent reduction in CO2 emissions when used in many medium-duty vehicles. For example,
electrification of an MDV pick-up designed and used solely for high towing capacity may be
more challenging in the earlier years of the program. Conversely, delivery vans or payload-
oriented pick-ups that operate over limited distances and daily routes present a significant
opportunity for electrification, as evidenced by current last mile delivery companies'
announcements and purchase orders for EVs in this segment. Due to this expected usage
difference of MDVs, there are fewer parallels with the structure of the light-duty program. In
addition, the phase-in provisions in the MDV program, although structurally different from those
of the light-duty program due to CAA requirements, serve the same purpose, which is to allow
manufacturers to achieve large reductions in emissions while providing a broad mix of products
to their customers.
The form and stringency of the original Phase 1 and 2 standards curves were based on the
performance of a set of vehicle, engine, and transmission technologies expected (although not
required) to be used to meet the GHG emissions standards with full consideration of how these
technologies were likely to perform in medium-duty vehicle specific testing and use. The
technologies included:
• Advanced engine improvements for friction reduction and low friction lubricants
• Improved engine parasitics, including fuel pumps, oil pumps, and coolant pumps
• Valvetrain variable lift and timing
• Cylinder deactivation
• Direct gasoline injection
• Cooled exhaust gas recirculation
• Turbo downsizing of gasoline engines
15 In our proposal we are defining a new MDV category that combines Class 2b and Class 3 and that excludes
MDPV. For the full definition, please refer to Section III. A. 1 of the Preamble to this rule.
1-22
-------
Diesel engine efficiency improvements
• Downsizing of diesel engines
• Electric power steering
• High efficiency transmission gear boxes and driveline
• Further improvements in accessory loads
• Additional improvements in aerodynamics and tire rolling resistance
• Low drag brakes
• Mass reduction
• Mild hybridization
• Strong hybridization
• Advanced 8 and higher speed automatic transmissions
• Diesel aftertreatment optimization
• BEV
Substantial opportunity still exists to further implement and make improvements to most of
these technologies to achieve further reductions in GHG emissions beyond those achieved in the
initial implementation of the Heavy-duty Phase 2 program as it applies to Class 2b and Class 3
vehicles (81 FR 73478 2016). Many of these technologies have not yet been implemented since
the Phase 2 standards are still within a phase-in period continuing through MY 2027. The agency
still expects to see additional penetration of many of these technologies.
The electrification of MDVs in the form of BEVs, particularly in delivery vans and some
pick-ups, has the highest potential for GHG reductions of all technologies investigated by the
agency. However, mild and strong hybridization and targeted PHEV implementation,
particularly PHEV Class 2b pickup trucks, may also provide substantial GHG emission
reductions as well as potential improvements in internal combustion engines, transmissions, and
vehicle technologies.
1.2.2.1 Final MDV GHG Standards
Our final GHG standards for all MDVs16 are entirely chassis-dynamometer based and
continue to be work-factor-based as with the previous heavy-duty Phase 2 standards. The
standards also continue to use the same work factor (WF) and GHG target definitions (81 FR
73478 2016). The chassis dynamometer testing methodology for MDVs does not directly
incorporate any GCWR related direct load or weight increases (e.g., trailer towing), however
16 Pickup trucks, vans, incomplete vehicles, and other vehicles having GVWR between 8,501 and 14,000 pounds,
excluding MDPVs. See section III. A. 1 of the Preamble to this rule.
1-23
-------
they would be reflected in the higher target standards when calculating the GHG targets using
GCWR values above approximately 22,000 pounds, which approximately corresponds to work
factors above 5,500 pounds. Without some limiting "cap" on GHG emissions, the resulting high
target standards relative to actual measured performance would be unsupported within the test
data used to demonstrate compliance and would generate windfall compliance credits for higher
GCWR ratings. For the final rule, the GHG standards were thus flattened to a constant value for
work factors above 5,500 pounds and for MY 2028 and later MDV. For further discussion of the
rationale behind the final MDV GHG standards, please refer to Section III.C.3 of the Preamble
for this final rule. The equations for MDV compliance with the final GHG standards are:
CC>2e Target (g/mi) = [a x WF] + b
WF = Work Factor = [0.75 x [Payload Capacity + xwd] + [0.25 x Towing Capacity]
Payload Capacity = GVWR (pounds) - Curb Weight (pounds)
xwd = 500 pounds if equipped with 4-wheel-drive, otherwise 0 pounds
Towing Capacity = GCWR (pounds) - GVWR (pounds);
and with coefficients:
Table 1-4: Final coefficients for MDV GHG standards for WF < 5,500 pounds
Model Year
a
b
2027
0.0348
268
2028
0.0339
270
2029
0.0310
246
2030
0.0280
220
203 1
0.0251
195
2032
0.0221
170
Table 1-5: Final coefficients for MDV GHG standards for WF > 5,500 pounds
Model Year a b
2027 0.0.148 268
2028 0 456
2029 0 417
2030 0 374
2031 0 333
2032 0 292
The feasibility of the MYs 2027 - 2032 GHG standards is based primarily upon an assessment
of the potential for a steady increase in MDV electrification, primarily within the van segment.
Relative to all of MDV, OMEGA compliance results had approximately 20 percent of MDV
sales as BEV in 2030, approximately 25 percent in 2031, and approximately 37 percent in 2032.
The feasibility of the initial year of compliance (2027) is from continued introduction of
technologies phasing into use for compliance with HD GHG Phase 2 as described in RIA
Chapter 1.2.2. Note that the fuel neutral standard in 2027 is a revision that would replace the last
1-24
-------
year of phase-in into the HD Phase 2 GHG program and applies solely to MDVs within that
program.
Beginning in MY 2027, the MDV GHG program moves gasoline, diesel, and PEV MDVs to
fuel-neutral standards, i.e., identical standards regardless of the fuel or energy source used.
Beginning in MY 2028, the GHG standards flatten above a work factor of 5,500 pounds.
650
600
550
500
450
J2D
400
r\i
o
u
350
300
250
200
MY 2028: Flatten
at WF>8
000 lbs.
3,000 3,500 4,000 4,500 5,000 5,500 6,000 6,500 7,000 7,500 8,000 8,500 9,000 9,500 10,000
Work Factor (lbs.)
Figure 1-16: Final MDV GHG target standards
MY 2030+: Flatte
n at WF>
,500 lbs
MY 2029: Fl
atten at WF>6,800 lbs
1.3 Development of the final battery durability standards
As described in sections III.G.2 and III.G.3 of the Preamble, EPA is establishing new battery
durability and warranty standards for PEVs.
In developing the standards, EPA took into consideration the provisions established in United
Nations Global Technical Regulation No. 22, as well as the California Air Resources Board
battery durability and warranty requirements under the Advanced Clean Cars II program.
Although EPA's durability and warranty provisions are not identical to either program, we
recognize the fact that automakers may be subject to GTR No. 22 in markets outside the U.S.,
and that many may also be subject to the durability and warranty requirements under the State of
California ACC II program. In considering the design and feasibility of the standards, EPA has
considered the specific features and purposes of both programs and has considered opportunities
for harmonization.
1-25
-------
The following discussion provides background on GTR No. 22, and on the California Air
Resources Board ACC II durability and warranty requirements. For a complete discussion of the
requirements and their relation to these other programs, please refer to Preamble III.G.2 and
III.G.3.
1.3.1 United Nations Global Technical Regulation No. 22 on In-Vehicle Battery
Durability
For several years, EPA has worked closely with the United Nations Economic Commission
for Europe (UNECE) Working Party on Pollution and Energy (GRPE) to develop a world
harmonized Global Technical Regulation (GTR) for In-vehicle Battery Durability for Electrified
Vehicles, or GTR No. 22 (UN ECE 2022). This GTR was created within a GRPE Informal
Working Group (IWG) known as Electric Vehicles and the Environment (EVE).
The EPA battery durability program is described primarily in Section III.G.2 of the Preamble.
The program largely adopts the general framework and requirements described in GTR No. 22,
with minor adaptations to incorporate established EPA test procedures and to achieve specific
program objectives. In addition to the published GTR, the EVE also produced a document which
outlines the technical justification and the development process of the GTR requirements (UN
ECE 2021).
In 2015 the UNECE began studying the need for a GTR governing battery durability in light-
duty vehicles. In 2021 it finalized GTR No. 22, which provides a regulatory structure for
contracting parties to set standards for battery durability in light-duty BEVs and PHEVs. The
European Commission and other contracting parties are currently working to adopt this standard
in their local regulatory structures. EPA representatives chaired the informal working group that
developed this GTR and worked closely with global regulatory agencies and industry partners to
complete its development in a form that could be adopted in various regions of the world,
including potentially the United States.
GTR No. 22 establishes a framework for regulating battery durability of BEVs and PHEVs by
establishing durability metrics, durability performance monitoring requirements, minimum
performance requirements, and procedures for determining monitor accuracy and determining
compliance. It does not include battery warranty requirements. To monitor durability
performance, it requires that manufacturers implement two ways of monitoring battery state-of-
health (SOH): State of Certified Energy (SOCE) and State of Certified Range (SOCR). SOCE
(and potentially in the future, SOCR) is then used to determine compliance with a Minimum
Performance Requirement (MPR) at two points during the vehicle's life, as described below. In
the current version of the GTR, the monitor requirements apply to Category 1-1, 1-2, and
Category 2 vehicles. The MPR applies only to Category 1-1 and Category 1-2 vehicles. The
IWG chose not to set an MPR for Category 2 vehicles at this time, largely because the early
stage of adoption of these vehicles meant that in-use data regarding battery performance of these
vehicles was difficult to obtain, and because these vehicles are more likely to have auxiliary
work-related features that use power from the battery for non-propulsion purposes, and the
impact of these features on battery life was not currently well characterized. MPR requirements
for category 2 vehicles were therefore reserved for possible inclusion in a future amendment to
the GTR.
1-26
-------
SOCE is an estimate of remaining usable battery energy (UBE) capacity at a point in the
vehicle's life, expressed as a percentage of the original UBE capacity when the vehicle was new.
In most jurisdictions, including the U.S. and those that have adopted the WLTP, original UBE is
already measured as part of the vehicle certification or range labeling process when the vehicle is
new. The GTR requires the SOCE monitor estimate of remaining UBE capacity to be readable
by the customer and by regulatory authorities. The algorithm for estimating and updating SOCE
during the lifetime of the vehicle is left to the manufacturer. The SOCE monitor value is required
to be on average no more than 5 percent higher than the actual value that would be obtained if
the true remaining UBE capacity were to instead be determined by the test procedure that was
used at certification. Accuracy is determined by a test program in which a statistical test is
applied to test results from a sample of test vehicles within a defined test group.
SOCR is an estimate of the total electric driving range that the vehicle battery remains capable
of providing at a point in the vehicle's life, expressed as a percentage of the original electric
driving range when the vehicle was new. As with UBE, electric driving range is already
measured and collected under applicable regional certification or type approval procedures when
the vehicle is new. The GTR requires SOCR to be readable by regulatory authorities but not
necessarily by the consumer. The SOCR monitor is also subject to the requirements for
determination and reporting of monitor accuracy but is not currently subject to the accuracy
requirement.
The GTR establishes a Minimum Performance Requirement (MPR) that specifies a minimum
percentage retention of SOCE and SOCR at two points in the vehicle's life. During the first phase
of implementation of the GTR, only the SOCE MPR will be enforced, although SOCR will be
collected for information purposes. As shown in Table 1-6, the MPRs established by GTR No.
22 require retention of at least 80 percent SOCE at 5 years or 100,000 km (about 62,000 mi), and
70 percent SOCE at 8 years or 160,000 km (about 100,000 miles).
Table 1-6. Battery durability performance requirements of UN GTR No. 22
Percent retention of at Mileage Percent of sample
must pass
80% SOH (UBE) 5 years 1 ()().()()() km 90%
70% 8 years 160.000 km
In the GTR, compliance with the SOCE MPR is determined for the vehicles within a given
durability test group by collecting a large sample of SOCE monitor values from in-use vehicles
at appropriate points in their life. The test group is compliant if at least 90 percent of the vehicles
monitored meet the applicable SOCE MPR.
This Section has outlined the provisions of GTR No. 22. For a description of the specifics of
the EPA battery durability program and how they compare to the provisions of the GTR, please
refer to Section III.G.2 of the Preamble and to the regulatory text.
1.3.2 California Air Resources Board battery durability and warranty provisions
under the ACC II program
In 2022, the California Air Resources Board (CARB), as part of its Advanced Clean Cars II
(ACC II) program, established battery durability and battery warranty requirements as part of a
1-27
-------
suite of customer assurance provisions designed to ensure that zero-emission vehicles maintain
similar standards for usability, useful life, and maintenance as conventional vehicles. The
performance requirements under the initial proposed version of the CARB durability standard
were significantly more stringent than those of UN GTR No. 22. After taking public comment
and consulting with the Board, the performance requirements were modified to a level closer to
that of GTR No. 22, while certain aspects of the program remain more stringent than those of the
GTR.
In contrast to GTR No. 22, the CARB battery durability requirement applies to electric
driving range instead of capacity, and phases in according to model year (MY). As shown in
Table 1-7, for MYs 2026 through 2029, a vehicle test group is compliant if at least 70 percent
of the vehicles in the group maintain 70 percent of certified range after 10 years or 150,000 miles
(240,000 km). For MYs 2030 and later, a test group is compliant if, on average, the vehicles in
the group maintain 80 percent of certified range after 10 years or 150,000 miles (240,000 km).
Details on monitor accuracy requirements, thresholds for determination of non-conformance, and
specific data reporting requirements are outlined in the regulations (State of California 2022a),
(State of California 2022b).
The CARB warranty requirement also phases in by model year, but instead of range it refers
to a state of health as expressed by usable battery energy (UBE). As shown in Table 1-8, for
MYs 2026 to 2030, the battery must maintain 70 percent state of health after 8 years or 100,000
miles (160,000 km). For MYs 2031 and later, it increases to 75 percent state of health. The
warranty requirement applies to the first purchaser and each subsequent purchaser. The warranty
requirements are further outlined in the regulation (Title 13, California Code of Regulations
2022).
Table 1-7. CARB ACC II battery durability requirements
Model vears
2026-2029
2030+
Percent
relent ion
70%
80%
of
Range
at
10 vears
Mileage
150.000 mi
Percent of
sample must
pass
70%
On average
Table 1-8. CARB battery warranty requirements
Model years Percent retention of at
2026-2030 70% SOH (UBE) 8 years
2031+ 75%
Mileage
100.000 mi
As described in the Preamble Sections III.G.2 and III.G.3, EPA is establishing battery
durability and warranty standards that differ to some degree from those of the CARB program,
but we have taken California's approach into consideration because we recognize that a
substantial number of vehicles sold in the United States will be subject to California's
requirements. The battery warranty requirements are implemented under the existing regulatory
structure that establishes a minimum warranty for major emission control components, and thus
retains similarities to the requirements under that program. The durability requirements are less
stringent than the CARB program and have a greater similarity to those of GTR No. 22. For a
1-28
-------
complete discussion of the requirements and their relation to these other programs, please refer
to Sections III.G.2 and III.G.3 of the preamble.
1-29
-------
Chapter 1 References
40 CFR § 1066.801 Subpart! 2023. (March 6). https://www.ecfr.gov/current/title-40/chapter-
I/subchapter-U/part-1066/subpart-I.
40 CFR § 86.1803-01. 2023.
49 CFR § 523.5. 2022. https://www.ecfr.gov/current/title-49/subtitle-B/chapter-V/part-
523/section-523.5.
76 FR 57106. 2011. (September 15). https://www.govinfo.gov/content/pkg/FR-2011-09-
15/pdf/20 ll-20740.pdf.
77 FR 62624. 2012.
81 FR 73478. 2016. (October 25). https://www.govinfo.gov/content/pkg/FR-2016-10-
25/pdf/2016-21203 .pdf.
Ellies, B. 2023. "Docket CUV Compliance Data Memo (Docket EPA-HQ-OAR-2022-0829)."
Greene, et al. 2018. "Consumer willingness to pay for vehicle attributes: What do we know?"
Transportation Research Part A.
Public Law 94-163. 1975. (December 22). Accessed 1 8, 2023.
https://www.govinfo.gov/content/pkg/STATUTE-89/pdf/STATUTE-89-Pg871.pdf.
State of California. 2022a. "California Code of Regulations, title 13, section 1962.4."
—. 2022b. "California Code of Regulations, title 13, section 1962.7."
Title 13, California Code of Regulations. 2022. Section 1962.7
U.S. EPA. 2011. "Draft Joint Technical Support Document: Proposed Rulemaking for 2017-
2025 Light-Duty Vehicle Greenhouse Gas Emission Standards and Corporate Average Fuel
Economy Standards."
U.S. EPA. 2022. "The EPA Automotive Trends Report: Greenhouse Gas Emissions, Fuel
Economy, and Technology since 1975."
UN ECE. 2021. "Proposal for a technical report on the development of a new UN Global
Technical Regulation on In-Vehicle Battery Durability for Electrified Vehicles." November.
https://unece.org/sites/default/files/2021-10/GRPE-84-02e_0.pdf.
—. 2022. "United Nations Global Technical Regulation on In-vehicle Battery Durability for
Electrified Vehicles." April, https://unece.org/transport/documents/2022/04/standards/un-gtr-
no22-vehicle-battery-durability-electrified-vehicles.
1-30
-------
Chapter 2: Tools and Inputs Used for Modeling Technologies and Adoption
Towards Compliance
This chapter summarizes the tools and inputs used for modeling technologies, adoption of
technologies, and vehicle compliance with the final standards. This includes details regarding the
OMEGA model, ALPHA vehicle simulation tools, and the Agency's approach to analyzing
vehicle manufacturing costs, consumer demand, and vehicle operational costs. The chapter also
includes a summary of modeling inputs that reflect an assessment of impacts due to the
implementation of the Inflation Reduction Act of 2022. We also discuss the rigorous technical
evaluations and peer-review processes that were used in developing the modeling tools and
technology inputs that were used in developing the standards. We describe the peer-review
process for the OMEGA and ALPHA models in Chapters 2.3 and 2.4.9 respectively. We also
discuss our detailed technical evaluation for battery cost estimates done in collaboration with
national experts at the Argonne National Laboratory in Chapter 2.5.2.1, and extensive cost tear-
down studies commissioned with an automotive consulting firm to develop estimates for non-
battery costs in Chapter 2.5.2.2.
Much of the material in this and other chapters of this RIA reflects EPA's long-standing
expertise in the area of mobile source emissions and regulatory standards development. EPA's
Office of Transportation and Air Quality (OTAQ) has more than fifty years of experience in
developing standards to reduce air pollution and greenhouse gas emissions from mobile sources.
This work has historically involved not only broad stakeholder engagement and foundational
work in regulatory design but also the development of deep scientific and technical expertise in
the engineering and science surrounding the measurement, modeling, and control of mobile
source emissions. This has included the development of sophisticated modeling tools to assess
mobile source-related air quality problems; establishing national and international standards to
reduce emissions; implementing standards through certification processes and in-use monitoring
strategies; developing fuel efficiency programs and technologies; and researching, evaluating,
and developing advanced technologies and new strategies for controlling emissions. Staff have a
variety of technical, legal, policy, and communications backgrounds to work effectively with
diverse stakeholders throughout this process. This includes employing well over a hundred staff
with undergraduate and graduate degrees in mechanical engineering, electrical engineering,
automotive engineering, computer science and engineering, chemical engineering, material
science, physics, chemistry, and other engineering, science, and related fields, including
economics. OTAQ also staffs and operates the National Vehicle and Fuel Emissions Laboratory
(NVFEL) in Ann Arbor, Michigan. For nearly 50 years, NVFEL has been a world-class, state-
of-the-art testing facility that provides emission testing support for EPA programs related to
light- and heavy-duty vehicles, heavy-duty engines, and nonroad engines, including testing of
gasoline and diesel engines and vehicles, HEVs, PHEVs, BEVs, electric machines, and high-
voltage batteries. EPA staff each year conduct hundreds of tests of vehicles and engines to
measure emissions, fuel economy, and performance. EPA also represents the U.S. at the United
Nations World Forum for the Harmonization of Vehicle Regulations, where EPA OTAQ
employees have chaired several working groups that have developed Global Technical
Regulations to establish international test procedures and emission standards for light-duty
vehicles, motorcycles, heavy-duty engines and vehicles, and electric vehicles. EPA OTAQ staff
also routinely work with major independent technical automotive laboratories and engineering
contractors - the same firms that are utilized by many of the light-, medium- and heavy-duty
2-1
-------
engine and vehicle manufacturers. These include multi-year contracts with Southwest Research
Institute and FEV North America. EPA utilizes these contracts to expand its access to additional
laboratory testing capabilities and expertise, including expertise in light-, medium- and heavy-
duty vehicle technology assessments. OTAQ has established Cooperative Agreements with
major transportation research universities, including the University of Michigan, the University
of California - Davis, and Michigan State University. EPA OTAQ has utilized interagency
agreements with several of the Department of Energy and the Department of Transportation
National Laboratories to collaborate on transportation sources research investigations, and the
National Vehicle and Fuel Emissions Laboratory has a long-standing, multi-decade Cooperative
Research and Development Agreement with the major U.S. car manufacturers and the California
Air Resources Board to "identify, encourage, evaluate and develop the instrumentation and
techniques to accurately and efficiently measure emissions from motor vehicles." EPA OTAQ
staff have authored and co-authored hundreds of peer reviewed articles in the engineering,
scientific, and economic literature, including publications by the Society of Automotive
Engineers, the American Society of Mechanical Engineers, the Energy Policy Journal, the
International Review of Environmental and Resource Economics, the World Electric Vehicle
Journal, Transportation Research, the International Journal of Environmental Research and
Public Health, and many others. EPA publications in the literature cover a wide range of topics,
including the development of emission reduction technologies, new test vehicle and engine
testing procedures, technology cost projections based on vehicle and sub-system tear-down
assessments, vehicle and engine performance and emissions benchmarking, emission
measurement programs, vehicle modeling techniques, vehicle fuel testing programs, and public
health assessments of transportation emissions. EPA OTAQ employees have also frequently
been asked to serve as peer reviewers for a number of these journals. EPA OTAQ employees
working at the National Vehicle and Fuel Emissions Laboratory have also been granted over 100
U.S. patents covering a wide range of engine, and vehicle related technologies, including
technologies for reducing criteria pollutant and GHG emissions, improving fuel efficiency, and
technologies for the measurement of mobile source emissions.
2.1 Overview of EPA's Compliance Modeling Approach
EPA's technical analysis supporting emissions standards, at its highest level, is based on the
following major tools that are used in the assessment of emissions reduction technologies and
costs. These are, in order of execution: ALPHA, response surface modeling, and OMEGA. They
are used in an integrated manner as follows:
• EPA's ALPHA model is a vehicle simulation tool used to predict tailpipe CO2
emissions and energy consumption for advanced technologies. ALPHA is detailed in
Chapter 2.4.
• Response surface methodology (RSM) incorporates ALPHA results for various
vehicle technologies over thousands of vehicle combinations into response surface
equations (RSE) which can be quickly referenced to characterize any future vehicle's
GHG emissions based on its size, weight, power, and road loads. This approach is
described in Chapter 2.4.10.
2-2
-------
• EPA's manufacturer compliance model, OMEGA, incorporates RSEs, technology
costs, and other inputs into its algorithms for finding cost-efficient pathways for
manufacturers to achieve compliance with desired emissions standards. The
compliance modeling produces a fleet of new light- and medium-duty vehicles for
each analyzed model year, which OMEGA integrates into projections of the on-road
vehicle stock and VMT. Finally, OMEGA tabulates the costs and benefits, emissions
inventories, and physical effects that arise from the usage of vehicles over their
lifetimes. A schematic of the overall analytical workflow is provided in Figure 2-1.
r
ALPHA & OMEGA Models
Inputs
V
ALPHA
Vehicle Modeling
Other
Inputs
V
Response Surface
Methodology
(RSM)
OMEGA
Create Response Surface Equations Compliance Modeling
Figure 2-1: Compliance modeling workflow.
Finally, the results from OMEGA are used to inform its fleet onroad vehicle emissions model
(MOVES) to generate fleet vehicle emissions and project benefits due to updated standards. A
discussion of MOVES is provided in section 8.2.1.
2.1.1 OMEGA Compliance and Model Overview
The OMEGA model has been developed by EPA to evaluate policies for reducing greenhouse
gas (GHG) emissions from light-duty vehicles. Like the prior releases, this latest version is
intended primarily to be used as a tool to support regulatory development by providing estimates
of the effects of policy alternatives under consideration. These effects include the costs
associated with emissions-reducing technologies and the monetized effects normally included in
a societal benefit-cost analysis, as well as physical effects that include emissions quantities, fuel
consumption, and vehicle stock and usage. In developing OMEGA version 2.0 (OMEGA2), the
goal was to improve modularity, transparency, and flexibility so that stakeholders can more
easily review the model, conduct independent analyses, and potentially adapt the model to meet
their own needs.
2.1.2 OMEGA Version 2.0
EPA created OMEGA version 1.0 (OMEGA1) to analyze new GHG standards for light-duty
vehicles proposed in 2011. The 'core' model performed the function of identifying
manufacturers' cost-minimizing compliance pathways to meet a footprint-based fleet emissions
standard specified by the user. A preprocessing step involved ranking the technology packages to
2-3
-------
be considered by the model based on cost-effectiveness. Postprocessing of outputs was
performed separately using a spreadsheet tool, and later a scripted process which generated table
summaries of modeled effects. An overview of OMEGA1 is shown on the left of Figure 2-2.
In the period since the release of OMEGA1, there have been significant changes in the light-
duty vehicle market including technological advancements, a diversification of powertrain types
across a wide range of vehicles, and the introduction of new mobility services. Advancements in
battery electric vehicles (BEVs) with greater range, faster charging capability, and expanded
model availability are particularly relevant when considering pathways for greater levels of
emissions reduction in the future. GMEGA2 has been developed with these trends in mind. The
model's interaction between consumer and producer decisions allows a user to represent
consumer responses to these new vehicles and services. The model also has been designed to
have expanded capability to model a wider range of GHG program options, which is especially
important for the assessment of policies that are designed to address future GHG reduction goals.
In general, with the release of OMEGA2, the goal is to improve usability and flexibility while
retaining the primary functions of OMEG A 1. The right side of Figure 2-2 shows the overall
model flow for OMEGA2 and highlights the main areas that have been revised and updated.
Technologies
Policy Alternatives +
Base Year Fleet
New Vehicles with
Applied Technologies
Effects
Postprocessing
Societal
Costs
Environmental
Effects
OMEGA ver. 2
Analysis
Context
. Policy
Alternatives
Compliance
Iteration
Environmental
Effects
Figure 2-2: Comparison of OMEGA1 and OMEGA2.
The following updates constitute a summary of improvements in OMEGA2 from OMEGA 1.
Update #1: Expanded model boundaries. In defining the scope of this model version, EPA has
attempted to simplify the process of conducting a run by incorporating into the model some of
the pre- and post-processing steps that had previously been performed manually. At the same
time, EPA recognizes that an overly expansive model boundary can result in requirements for
inputs that are difficult to specify. To avoid this, the input boundary has been set only so large as
to capture the elements of the system it is assumed are responsive to policy. This approach helps
to ensure that model inputs such as technology costs and emissions rates can be quantified using
data for observable, real-world, characteristics and phenomena, and in that way enable
transparency by allowing the user to maintain the connection to the underlying data. For the
2-4
-------
assumptions and algorithms within the model boundary, the aim is transparency through well-
organized model code and complete documentation.
Update #2: Independent Policy Module. OMEGA1 was designed to analyze a very specific
GHG policy structure in which the vehicle attributes and regulatory classes used to determine
emissions targets were incorporated into the code throughout the model. To make it easier to
define and analyze other policy structures, the details regarding how GHG emissions targets are
determined and how compliance credits are treated over time are now included in an independent
Policy Module and associated policy inputs. This allows the user to incorporate new policy
structures without requiring revisions to other code modules. Specifically, the producer decision
module no longer contains any details specific to a GHG program structure, and instead
functions only on very general program features such as fleet averaging of absolute GHG credits
and required technology shares.
Update #3: Modeling of multi-year strategic producer decisions. As a policy analysis tool,
OMEGA2 is intended to model the effect of policies that may extend well into the future, beyond
the timeframe of individual product cycles. OMEGA2 is structured to consider producer
decisions in a coherent manner across all the years in the analysis period. Year-by-year
compliance decisions account for management of credits which can carry across years in the
context of projections for technology cost and market conditions which change over time. The
timeframe of a given analysis can be specified anywhere from near-term to long-term, with the
length limited only by inputs and assumptions provided by the user.
Update #4: Addition of a consumer response component. The light-duty vehicle market has
evolved significantly in the time since the initial release of OMEGA1. As the range of available
technologies and services has grown wider, so has the range of possible responses to policy
alternatives. The model structure for this version includes a Consumer Module that can be used
to project how the light-duty vehicle market would respond to policy-driven changes in new
vehicle prices, fuel operating costs, trip fees for ride hailing services, and other consumer-facing
elements. The Consumer Module outputs the estimated consumer responses, such as overall
vehicle sales and sales shares, as well as vehicle re-registration and use, which together
determine the stock of new and used vehicles and the associated allocation of total VMT.
Update #5: Addition of feedback loops for producer decisions. OMEGA2 is structured around
modeling the interactions between vehicle producers responding to a policy and consumers who
own and use vehicles affected by the policy. These interactions are bi-directional, in that the
producer's compliance planning and vehicle design decisions will both influence, and be
influenced by, the sales and shares of vehicles demanded and the GHG credits assigned under the
policy. Iterative feedback loops have now been incorporated between the Producer and
Consumer modules to ensure that modeled vehicles would be accepted by the market at the
quantities and prices offered by the producer, and between the Producer and Policy modules to
account for the compliance implications of each successive vehicle design and production option
considered by the producer. This update has been peer reviewed as detailed in Section 2.3.
Update #6: Use of absolute vehicle costs and emissions rates. OMEGA1 modeled the
producer application of technologies to a fleet of vehicles that was otherwise held fixed across
policy alternatives. With the addition of a consumer response component that models market
share shifts, OMEGA2 utilizes absolute costs and emissions rates to compare vehicle design and
purchase decisions across vehicle types and market classes.
2-5
-------
The table below contains a list of OMEGA topics and the RIA chapters where those topics are
discussed. This list is not comprehensive, though it may help understanding of the model,
including its structure, inputs and outputs. For a qualitative discussion on analytical updates
since the proposed rule that may have impacted OMEGA results, see Section IV.A.2 of the
preamble.
Table 2-1: OMEGA topics and their RIA Chapter locations
Topic
RIA Chapters
Model structure/overview
Model structure
2.2.1; 2.1.2; 2.2.2
Inputs and outputs
2.1.1; 2.6
Module interactions
2.1.2; 2.2.2
Convergence
2.2.3
Producer module
Overview
2.2.2
Cost methodology
2.5.1; 2.5.3; 2.6.1; 2.6.2; 2.6.4
Absolute vs. incremental cost approach
2.5.1
Learning-by-doing
2.5.3
Direct manufacturing costs
2.5.2; 2.6.1; 2.6.2; 2.6.8
Battery costs and battery sizing
2.5.2.1
Indirect costs
2.5.4
Producer decisions
2.6.4
Consumer Module
Overview
2.2.2
Purchase decision methodology
2.6.3; 4.1
Generalized cost
2.6.4; 4.1.1; 4.2.2, Table 4-1
Consumer purchase incentive
2.6.8
Market shares, incl. shareweights
4.1.2
Policy Module
Overview
2.2.2
Effects Module
Overview
8
Physical effects
8
VMT and rebound
4.2.1; 4.3.2; 8.3
Safety
8.4
Electricity and liquid fuel consumption
4.3.3; 4.3.4; 8.5
Emissions
8.6
Energy security and oil imports
8.7
New Vehicle Sales
4.4.1
VMT and rebound
4.2.1; 4.3.2
Costs
4.3
Technology costs
4.3.1
Insurance
4.3.6
Repair and maintenance
4.3.7
Noise and congestion
4.3.8
Social Benefits, Social Costs and Benefit-Cost
Analysis
9
2-6
-------
2.2 OMEGA2 Model Structure and Operation
2.2.1 Inputs and Outputs
Like other models, OMEGA relies on the user to specify appropriate inputs and assumptions.
Some of these may be provided by direct empirical observations, for example the number of
currently registered vehicles. Others might be generated by modeling tools outside of OMEGA,
such as physics-based vehicle simulation results produced by EPA's ALPHA model, or
transportation demand forecasts from DOE's NEMS model. OMEGA has adopted data elements
and structures that are generic, wherever possible, so that inputs can be provided from whichever
sources the user deems most appropriate.
The inputs and assumptions are categorized according to whether they define the policies
under consideration or define the context within which the analysis occurs. Policy alternative
inputs describe the standards themselves, including the program elements and methodologies for
determining compliance as would be defined for an EPA rule in the Federal Register and Code of
Federal Regulations. Analysis context inputs and assumptions cover the range of factors that the
user assumes are independent of the policy alternatives. The context inputs may include fuel
costs, costs and emissions rates for a particular vehicle technology package, attributes of the
existing vehicle stock, consumer demand parameters, existing GHG credit balances, producer
decision parameters, and many more. The user may project changes in the context inputs over
the analysis timeframe based on other sources, but for a given analysis year the context definition
requires that these inputs are common across the policy alternatives being compared.
The primary outputs are the environmental effects, societal costs and benefits, and producer
compliance status for a set of policy alternatives within a given analysis context. These outputs
are expressed in absolute values, so that incremental effects, costs, and benefits can be evaluated
by comparing two policy alternatives for a given analysis context. For example, comparing a No
Action scenario to an Action (or Policy) Alternative. Those same policy alternatives can also be
compared using other analysis context inputs to evaluate the sensitivity of results to uncertainty
in particular assumptions. For example, comparing the incremental effects of a new policy in
high fuel price and low fuel price analysis contexts.
2.2.2 Model Structure and Key Modules
OMEGA2 has been set up so that primary components of the model are clearly delineated in
such a way that changing one component of the model will not require code changes throughout
the model. The four main modules — Producer, Consumer, Policy, and Effects — are each
defined along the lines of their real-world analogs. Producers and consumers are represented as
distinct decision-making agents, which each exist apart from the regulations defined in the Policy
Module. Similarly, the effects, both environmental and societal, exist apart from producer and
consumer decision-making agents and the policy. This structure allows a user to analyze policy
alternatives with consistently defined producer and consumer behavior. It also provides users the
option of interchanging any of OMEGA's default modules with their own, while preserving the
consistency and functionality of the larger model.
Producer Module: This module generates potential decisions of the regulated entities
(producers) in response to policy alternatives, while accounting for technology cost and
availability, and constraints on vehicle production. The regulated entities can be specified as
2-7
-------
individual companies, or considered in aggregate as a collection of companies, depending on the
assumptions made by the user regarding how GHG credits are averaged or transferred between
entities.
Consumer Module: This module projects a consumer response to the vehicle offerings
generated by the producer module. The final projection is determined by iterating across multiple
hypothetical vehicle offering scenarios, as described in Section 2.2.3. Under each scenario, the
purchase decisions of consumers are modeled in response to vehicle price and consumers'
preconceptions of their own driving behavior and likely fuel and operating costs, in combination
with the consumer's consideration of other vehicle attributes and individual preferences.
Policy Module: This module determines the compliance status for a producer's possible fleet
of new vehicles based on the characteristics of those vehicles and the policy defined by the user.
Policies may be defined as performance-based standards using fleet averaging (for example,
determining compliance status by the accounting of fungible GHG credits), as a fixed
requirement without averaging (for example, a minimum required share of BEVs), or as a
combination of performance-based standards and fixed requirements.
Effects Module: This module projects the physical and cost effects that result from the
modeling of producers, consumers, and policy within a given analysis context. Examples of
physical effects include the stock and use of registered vehicles, electricity, and gasoline
consumption, and the GHG and criteria pollutant emissions from tailpipe and upstream sources.
Examples of cost effects include vehicle production costs, ownership and operation costs,
societal costs associated with GHG and criteria pollutants, and other societal costs associated
with vehicle use.
2.2.3 Iteration and Convergence
OMEGA2 is intended to find a solution which simultaneously satisfies producer, consumer,
and policy requirements while minimizing the producer generalized costs. OMEGA2's Producer
and Consumer modules represent distinct decision-making entities, with behaviors defined
separately by the user. Without some type of interaction between these modules, the model
would likely not arrive at an equilibrium of vehicles supplied and demanded. For example, a
compliance solution which only minimizes producer generalized costs without consideration of
consumer demand may not satisfy the market requirements at the fleet mix and level of sales
preferred by the consumer. Similarly, the interaction between Producer and Policy modules
ensures that with each subsequent iteration, the compliance status for the new vehicle fleet under
consideration is correctly accounted for by the producer. Since there is no general analytical
solution to this problem of alignment between producers, consumers, and policy which also
allows model users to independently define producer and consumer behavior and the policy
alternatives, OMEGA2 uses an iterative search approach (U.S. EPA 2024).
2.2.4 Analysis Resolution
The policy response projections generated by OMEGA2 are centered around the modeled
production, ownership, and use of light-duty vehicles. It would not be computationally feasible
(nor would it be necessary) to distinguish between the approximately 15 million light-duty
vehicles produced for sale each year in the US, and hundreds of millions of vehicles registered
for use at any given time. Therefore, OMEGA is designed to operate using 'vehicles' which are
2-8
-------
aggregate representations of individual vehicles, while still retaining sufficient detail for
modeling producer and consumer decisions, and the policy response. The resolution of vehicles
can be set for a given analysis and will depend on the user's consideration of factors such as the
availability of detailed inputs, the requirements of the analysis, and the priority of reducing
model run time.
2.3 OMEGA2 Peer Review
The peer review was conducted during the development of OMEGA2 to facilitate
implementation of peer review comments and suggestions into the completed version utilized in
the Final Rule. This process was intended to gain additional insights for the updated structure,
new modules, processing methods, and reporting methodology of OMEGA2.
2.3.1 Charge Questions for the Peer Review:
• The overall approach to the specified modeling purposes, the specific approaches
chosen for modeling individual modules, and the methodologies chosen to achieve
that purpose.
• The appropriateness and completeness of the contents of the input files.
• The types of information which can be input to the model point to both the flexibilities
and constraints of the model.
• The accuracy and appropriateness of the model's conceptual algorithms and equations
for technology application, market impacts, and calculation of compliance.
• The congruence between the conceptual methodologies and the program execution.
• Clarity, completeness, and accuracy of the model's visualization output, in which the
technology application is displayed.
• Recommendations for any functionalities beyond what EPA has described as "future
work."
2.3.2 Information Received from Peer Review
EPA's charge to the peer reviewers requested their expert opinions on the concepts and
methodologies upon which the model relies and whether the OMEGA2 model correctly executes
the associated algorithms. EPA's charge also asked the peer reviewers to comment on specific
aspects of the model's design, execution, outputs, and documentation.
All peer reviewers commented favorably that they appreciated the increased capability and
complexity of OMEGA2 over the previous OMEGA1 version as expressed in general comments:
• "A significantly enhanced EPA OMEGA2 over the earlier OMEGA1 version is an
excellent and unique producer-consumer vehicle choice model by minimizing the
resulting effects of societal costs and emissions subject to user-input emission policies.
It is intended to find a solution which simultaneously satisfies producer, consumer,
2-9
-------
and policy requirements while minimizing the producer generalized costs. The model
and documentation available directly from Github provides the necessary detailed
model scope description and modeling approach capabilities."
The peer reviewers provided numerous detailed comments and recommendations that
indicated a thorough evaluation and good understanding of the model's design. The most
common category of comments consisted of recommendations for improving the model's
documentation by adding further explanations or specifics to enhance the user's understanding
that were added in the final version of the model.
Another category of peer reviewer comments concerned the model's overall approach,
including the functions of each module. Reviewers commented on specific details, recommended
improvements, and noted inputs and results that would benefit from further explanation. Many
peer reviewer recommendations for new or additional functionality focused on specific
enhancements of the existing modules.
Certain topics were raised by multiple peer reviewers. For example, all peer reviewers
commented on some aspect of the model's handling of greenhouse gas (GHG) emissions credits,
especially as it relates to manufacturers banking these credits from one year to the next and, in
some cases, how credit banking would interact with manufacturers' multi-year model
development cycle. Reviewers also commented on the fact that the model does not currently
include other aspects of firm profit-seeking compliance behavior, including changing vehicle
attributes to shift regulatory classes. Peer reviewers also indicated that it was likely that wide-
scale implementation of the technologies available in OMEGA2 could cause a significant change
to overall fuel prices that should be considered. The peer reviewers also requested further
explanation of how the OMEGA2 model processes hauling/non-hauling vehicles, all-wheel drive
(AWD), and OMEGA2 model algorithm's treatment of iterative convergence on a final result.
As a result of the peer review, EPA implemented multiple model revisions and improvements,
and expanded the OMEGA model's documentation and instructional materials. The peer review
revisions were adopted for the model version used for the proposal, which was the basis for the
version used for this final rulemaking. EPA's responses to individual peer review comments is
provided in the full peer review report (U.S. EPA 2023). EPA has made improvements and
updates to the OMEGA model compared to the version of the model reviewed by the peer
reviewers (OMEGA version 2.0.0). Updates were made for the version used by EPA for the
notice of proposed rulemaking (version 2.1.0), and further updates have been made since the
proposal for this final rule in response to public comments (version 2.5.0). OMEGA is a state-
of-the-art vehicle emissions compliance model, and has capabilities not available in any other
vehicle emissions compliance model publicly available today. For example, OMEGA2 now
incorporates certain consumer preferences via the new consumer module. The current OMEGA
model reflects the agency's considered expertise in modeling the development and application of
vehicle pollution control technologies over many decades, and EPA believes it is the best
available model for modeling for this rulemaking. Over time, we recognize that what is
considered state-of-the-art evolves - in response to new data, new modeling techniques, and the
evolving scientific literature, and EPA anticipates future improvements to OMEGA, beyond the
scope of the capabilities reflected in OMEGA 2.5.0, in response to future data, peer review,
modeling techniques, and the literature.
2-10
-------
2.4 ALPHA Full Vehicle Simulation and Response Surface Equations
ALPHA is a physics-based, forward-looking, full vehicle computer simulation capable of
analyzing various vehicle types with different powertrain technologies, showing realistic vehicle
behavior. The software tool is a MATLAB/Simulink based simulation.
ALPHA is capable of estimating CO2 emission values for many different vehicle types and
technology packages. OMEGA needs to quickly estimate the CO2 emission values for each
future vehicle considered along with estimates for future fleets. Because operating ALPHA in
real time to conduct full vehicle simulations is time prohibitive, EPA developed a methodology
of reproducing ALPHA model CO2 values using an industry standard statistical technique known
as response surface methodology (RSM). (Kleijnen 2015) This methodology is used to
computationally access CO2 results from a complete set of ALPHA model results by generating a
collection of response surface equations (RSEs) that represent those simulation results. In 2018,
EPA commissioned RTI International to conduct an independent peer review of an earlier
version of the RSE methodology. (RTI International 2018)
ALPHA simulates a single combination of technologies (known as a technology package)
across different combinations of vehicle parameters. Each set of ALPHA simulation outputs are
processed to create the RSEs needed for each technology package addressed. These RSEs are
subsequently used within OMEGA to quickly reproduce the ALPHA model estimates for CO2 in
real time for the various vehicle technologies. ALPHA'S role in the creation of the RSEs for use
within OMEGA is shown in Figure 2-3 and described in detail in Chapter 2.4.10.
ALPHA Simulation Scripts
for Various Technology Packages1
Conventional Vehicles
Mild Hybrid Vehicles
Strong Hybrid Vehicles
Battery Electric Vehicles
ALPHA Inputs
for Various Components2
engines
eMotors
batteries
transmissions
Notes:
1-shown in Table 2-21
2- shown in Tables 2-2 through 2-5
3- discussed in section 2.4.11.3
Relationship of ALPHA, RSEs & OMEGA
Inputs
\7
ALPHA
Vehicle Modeling
Response Surface
Methodology
(RSM)
Create Response Surface Equations
Other
Inputs
V
1=>
OMEGA
Compliance Modeling
ALPHA Simulation Outputs
for each Technology Package
Response Surface Equations (RSE)1
for each Technology Package
Figure 2-3: Relationship of ALPHA, RSEs and OMEGA.
2.4.1 General Description of ALPHA
Within ALPHA, an individual vehicle is defined by specifying the appropriate vehicle road
loads (inertia weight and coast-down coefficients) and specifications of the powertrain
2-11
-------
components. Powertrain components (e.g., engines, transmissions, e-motors) are individually
parameterized and can be exchanged within the model draft.
Vehicle control strategies are also modeled, including engine accessory loading, deceleration
fuel cut off (DFCO), hybrid behavior, torque converter lockup, and transmission shift strategy.
Transmission shifting is parameterized and controlled by ALPHAshift, (Newman, Kargul and
Barba 2015b) a shifting strategy algorithm that ensures an appropriate shifting strategy when
engine size or vehicle loading changes. The control strategies used in ALPHA are modeled after
strategies observed during actual vehicle testing.
The performance of vehicle packages defined within ALPHA can be modeled over any pre-
determined vehicle drive cycle. To determine fuel consumption values used to calculate LD
GHG rule CO2 values, the FTP and HWFET cycles are simulated, separated by a HWFET prep
cycle when required such as when simulating certification testing of start-stop vehicles and
hybrids. ALPHA does not include a temperature model, so the FTP is simulated assuming warm
component efficiencies for all test phases. Additional fuel consumption due to the FTP cold start
is calculated in post-processing by applying a fuel consumption penalty to bags 1 and 2,
depending on the assumed warmup strategy (refer to Section 5.3.3.2.5 of the 2016 Draft
Technical Assessment Report (U.S. EPA; U.S. DOT-NHTSA; CARB 2016)). In addition, other
defined cycles can be used to simulate fuel economy in ALPHA. For example, the results from
the US06, NEDC, and WLTP cycles (among others) are used to tune vehicle control strategy
parameters to match simulation results to measured vehicle test results across a variety of
conditions. In addition, performance cycles have been defined, which can be used to determine
acceleration performance metrics.
2.4.2 Overview of Previous Versions of ALPHA
The Advanced Light-Duty Powertrain and Hybrid Analysis (ALPHA) tool was created by
EPA to evaluate the Greenhouse Gas (GHG) emissions of Light-Duty (LD) vehicles. In addition,
to provide additional flexibilities and transparency, EPA developed this in-house full vehicle
simulation model that could freely be released to the public. Model development, along with the
data collection and benchmarking that comes along with model calibration, is an extremely
effective means of developing expertise and deeper understanding of technologies and their
interactions. Better understanding of technologies contributes to a more robust regulatory
analysis. Having a model available in-house allows EPA to make rapid modifications as new
data is collected.
EPA began developing both light-and heavy-duty vehicle models simultaneously as these
vehicles share many of the same basic components. The light-duty vehicle model (ALPHA), and
the heavy-duty model (GEM), share much of the same basic underlying architecture.17 ALPHA
2.1 and 2.2 were developed and used previously under EPA's 2016 Draft Technical Assessment
Report (U.S. EPA; U.S. DOT-NHTSA; CARB 2016), the 2016 Proposed Determination (U.S.
EPA 2016a) (U.S. EPA 2016b), and the 2017 Final Determination (U.S. EPA 2017a) (U.S. EPA
2017b).
17 The GEM model has also been peer reviewed multiple times and was the subject of comment during the
rulemaking adopting the second phase of GHG standards for heavy-duty vehicles and engines. See 81 FR 73530-
531,538-549.
2-12
-------
As part of the Midterm Evaluation, EPA validated the ALPHA model using several sources
including vehicle benchmarking, stakeholder data, and industry literature. To further enhance
transparency, in May 2016, EPA completed an external peer review of ALPHA 2.0 (U.S. EPA
2023a). This peer review package included runnable MATLAB/Simulink source code along with
the input data provided as part of the review.
2.4.3 Current version of ALPHA
ALPHA 3.0 is the current version of the simulation tool used for this rule. The two primary
changes in ALPHA 3.0 compared to the previous version of ALPHA (ALPHA 2.2) are the
addition of electrified vehicle architectures (including hybrid, plug-in hybrid, and battery electric
vehicles) and the addition of a robust structure to allow large numbers of simulations to
characterize current and future fleets. A basic description of how ALPHA 3.0 works can be
found online (U.S. EPA 2022c).
While ALPHA 3.0 continues to be refined and calibrated, the new electrified vehicle models
of the version in use as of October 9, 2022, were externally peer-reviewed (U.S. EPA 2023a).
The concepts and methodologies upon which the model relies were examined by peer reviewers
to determine if these algorithms can deliver sufficiently accurate results. The results of the peer
review are discussed in RIA Chapter 2.4.9.
Throughout this chapter, details are provided on the major technology assumptions built into
ALPHA 3.0. EPA has also provided technical details in Chapter 3.5 which summarizes the
ALPHA inputs used for this rule. In the time since ALPHA development began, EPA has
published over twenty peer-reviewed papers describing ALPHA and the results of key testing,
validation, and analyses (U.S. EPA 2023a) (U.S. EPA 2022b).
2.4.4 ALPHA Models for Conventional and Electrified Vehicle Architectures
One of the most significant changes in ALPHA 3.0 is the addition of new electrified vehicle
architecture models. Early in the development phase of ALPHA 3.0, EPA conducted research to
determine which electrified vehicle architectures should be included in ALPHA'S suite of models
(Zhuanga, et al. 2020). Based on trends of the various hybrid and electric vehicles available for
sale in the U.S. in recent years, the conclusion was the electrified vehicle market could be
modeled in ALPHA 3.0 with the addition of five hybrid vehicle architectures and one battery
electric vehicle architecture along with the base conventional vehicle architecture, all
summarized in Figure 2-4.
2-13
-------
Convention
al Vehicle
PO Mild
Hybrid
P2
Hybrid
P2-P4
Hybrid
(REET)
SP-P4
Hybrid
(REET)
PowerSplit
Hybrid
Battery
Electric
Vehicle
HV f
Battery
Components
Architecture
oooo
Engine
electric
starter
generator
COCO
Q
electric
starter
generator
(optional)
OOOO
¦ul"
I.'> l-1--
generator
COCO
a
e-motor/
generator
e-motor/
generator
e-motor/
generator
a
e-motor/
generator
planetary
gear
t
e-motor/
generator
I I
Figure 2-4: Summary of components and architectures used in ALPHA'S modeling.
2.4.4.1 Conventional Vehicle Architecture
The CO2 performance for all conventional vehicles is modeled using the basic engine plus
transmission architecture shown in Figure 2-5. Different types of engines and transmissions
(including their many operational strategies such as cylinder deactivation, engine stop/start
control, engine deceleration fuel cut off) can be scaled to suit the different vehicle models. For
this rule, conventional vehicles are modeled using the same model described in Chapter 2.3.3.3
of the Proposed Determination on the Appropriateness of the Model Year 2022-2025 Light-Duty
Vehicle Greenhouse Gas Emissions Standards under the Midterm Evaluation: Technical Support
Document (U.S. EPA 2016b).
2-14
-------
Figure 2-5: Conventional vehicle architecture.
2.4.4.2 Hybrid Electric Vehicle (HEV) Architectures
There are a wide variety of possible hybrid-electric vehicle architectures, many of which are
or have been represented in the fleet. To assess the scope of this variety, EPA used a recent
hybrid architecture survey paper (Zhuanga, et al. 2020). Although other researchers may use a
different terminology for specific architectures, in the interest of consistency, EPA adopted the
categorization and nomenclature used by the authors in this survey paper for further discussion
of hybrid-electric vehicle architectures.
The CO2 performance of hybrid vehicles in ALPHA is modeled using one mild and four
strong hybrid architectures. The mild hybrid architecture chosen was a parallel P0 configuration
(referred to later as "P0"). The four strong hybrid architectures chosen were a light-duty vehicle
PowerSplit configuration patterned after the Toyota Prius (referred to later as "PS"), a light-duty
vehicle/truck parallel P2 configuration (referred to later as "P2"), a light-duty towing vehicle
series-parallel configuration (referred to later as "SP-P4"), and finally medium-duty towing
vehicle parallel P2-P4 plug-in configuration (referred to later as "P2-P4 PHEV").
While other mild and strong hybrid architectures also exist in the fleet, EPA's analysis in RIA
Chapters 2.4.8.5 and 2.4.8.6 shows that these hybrid variations can be adequately modeled using
the one mild hybrid and four core strong hybrid architectures chosen for incorporation into
ALPHA 3.0.
An analysis of the MY 2022 light-duty vehicle fleet revealed that 36 percent of all hybrid
vehicles in the MY 2022 fleet were mild hybrids, and the remaining 64 percent were strong
hybrids (Table 2-2). Of the strong hybrids, 60 percent were based on PowerSplit architecture, 27
percent were based on P2 hybrid technology, and the remaining 13 percent were based on
architectures such as series-parallel and pure series architectures. The following will discuss the
different hybrid models incorporated into ALPHA 3.0 to simulate these different types of hybrid
vehicles.
2-15
-------
Table 2-2: Percentage breakdown of mild and strong hybrids in the MY 2022 light-duty
vehicle fleet
ALPHA'S Mild Hybrid Model
% of Mild
Hybrids
% of all Hybrid
Vehicles
P0 Mild Hybrids
83.5%
30.1%
PI Mild Hybrids
16.5%
5.9%
ALPHA'S Strong Hybrid Model
% of Strong
Hybrids
% of all Hybrid
Vehicles
PowerSplit Strong Hybrids
59.9%
38.3%
PowerSplit PI-IEVs
P2 Strong Hybrids
26.8%
17.1%
P2 PHEVs
Other Hybrids
13.3%
8.6%
Other PHEVs
2.4.4.2.1 Mild Hybrid Architectures
Mild hybrids are modeled within ALPHA using a 48V P0 architecture, which includes a
conventional engine and transmission along with a starter/generator mounted on the front of the
engine and connected through a belt and pulley, as shown in Figure 2-6. The battery energy
capacity of a typical mid-sized mild hybrid vehicle is around 0.25 kWh. ALPHA is capable of
simulating the operation of P0 mild hybrid technology and engine start-stop technology, each
separately, or combined, depending upon the configuration of the vehicle being simulated. For
the purposes of this rulemaking, P0 mild hybrids simulations combine both P0 mild hybrid and
engine start-stop technologies.
Figure 2-6: P0 Mild hybrid-electric vehicle architecture.
Table 2-2 shows that 84 percent of the mild hybrids in the MY 2022 LD fleet are based on a
P0 design. Stellantis /Ram and Volkswagen were the two biggest producers of P0 mild hybrid
vehicles in the fleet. The other 16 percent of mild hybrids were based on a PI design, where the
2-16
-------
starter generator is directly mounted on the backside of the engine without the use of a belt.
Mercedes was the only manufacturer of PI mild hybrids in 2022.
Analysis of P0 and PI hybrids in the MY 2022 fleet presented later in Chapter 2.4.8.5,
indicates the PI variant of mild hybrids, although more efficient than the P0 architecture, can be
reasonably represented by ALPHA'S P0 mild hybrid model. Consequently, the ALPHA P0
model was chosen to simulate all the mild hybrids associated with this rule.
2.4.4.2.2 Strong Hybrid and PHEV Architectures
ALPHA 3.0 uses four distinct models to simulate strong hybrid-electric vehicles (HEVs) and
strong plug-in hybrid electric vehicles (PHEVs) in the U.S. vehicle fleet. Within each model,
both HEV and PHEV simulations use the same base supervisory controllers to determine engine
and electric motor operation. However, SOC ranges and load level SOC targets differ since the
HEVs and PHEVs operate at different SOC levels.
The Power Split strong HEV and PHEV architecture is shown in Figure 2-7. This
architecture includes a dedicated hybrid engine specifically designed to provide higher efficiency
at the more stable engine loads possible with a PowerSplit powertrain. ALPHA 3.0 models the
PowerSplit device using a planetary arrangement like that in the third-generation Prius, with the
engine mated to the planetary carrier gear, the Motor/Generator 1 (MG1) connected to the sun
gear, and the Motor/Generator 2 (and its associated planetary gear) connected to both the ring
gear and drive axle (through the final drive gear). The PowerSplit device divides the torque
between the engine, MG1, and the MG2/drive axle to provide the needed torque to the wheels
while optimizing efficiency of the powertrain components. The battery for a typical mid-sized
PowerSplit hybrid electric vehicle is around 1.5 kWh. The battery energy capacity of a similar
sized plug-in hybrid (PHEV) version of PowerSplit hybrid is around 14 kWh. In ALPHA, the
same base PowerSplit supervisory controller is used to simulate both HEVs and PHEVs, but with
different SOC ranges and load level SOC targets.Table 2-2 illustrates that 60 percent of the
strong hybrid vehicles in the MY 2022 fleet are the PowerSplit architecture (including both
strong hybrids and PHEVs). The biggest producer of PowerSplit hybrids in MY 2022 by far
(both by number of vehicle models and total sales) was Toyota. Ford, Stellantis, and Subaru also
offered a plug-in version of the PowerSplit architecture on at least one vehicle model.
2-17
-------
Figure 2-7: PowerSplit strong HEV and PHEV (& planetary gear arrangement).
e-motor/
generator 1
PowerSplit
Device
engine
e-motor/
generator 2
The PowerSplit model also delivered suitable CO2 predictions for other strong hybrid designs
(e.g., series-parallel architecture), which represent 13 percent18 of the remaining MY 2022 hybrid
fleet. In total, ALPHA'S PowerSplit strong hybrid model was used to simulate 73 percent of the
MY 2022 strong hybrid fleet.
The P2 strong HEV and PHEV architecture illustrated in Figure 2-8 is the second strong
hybrid-electric model used within ALPHA. This hybrid architecture uses a conventional or a
dedicated hybrid engine and a conventional 6 speed (or higher) automatic transmission with a
clutch and electric motor/generator in place of the standard torque converter for a conventional
vehicle. The P2 architecture has higher power and torque capability due to the full power engine
and transmission and is suitable for truck and large SUV applications with towing capability.
The battery energy capacity of a typical P2 strong hybrid vehicle is around 1.5 kWh (same as the
PowerSplit strong hybrid). The battery capacity of a similar sized plug-in hybrid (PHEV) version
of P2 hybrid is around 16 kWh. In ALPHA, the same base P2 supervisory controller is used to
simulate both HEVs and PHEVs, but with different SOC ranges and load level SOC targets.
Table 2-2 shows that 27 percent of the strong hybrids in the MY 2022 fleet are based on a P2
design (including both strong hybrids and PHEVs). Leading manufacturers of P2 hybrid and
plug-in hybrid vehicles include Hyundai/Kia, BMW, Ford, and Porsche AG.
1 ft
Nearly all these remaining vehicles are based on a series-parallel hybrid design like the Honda Accord hybrid.
While the CO2 performance of a series-parallel hybrid can be estimated using a PowerSplit hybrid architecture, EPA
developed a dedicated series-parallel model for characterizing future PHEVs using ALPHA.
2-18
-------
Figure 2-8: P2 strong HEV and PHEV architecture.
The SP-P4 strong HEV and PHEV architecture shown in Figure 2-9 was developed so
ALPHA could simulate advanced hybrid vehicles in future light-duty fleets. The SP-P4
architecture was specifically developed for future LD truck and large SUV HEV and PHEV
applications with towing capability. This model was not used to model any vehicles in the
MY2022 base-year fleet. More information about this hybrid architecture can be found in EPA's
report on the Range-Extended Electric Taick (REET) (Bhattacharjya, et al. 2023).
e-motor/
generator
P4 Front Axle
1 1
Inverter 1
a i
t
generator
Figure 2-9: SP-P4 strong HEV and PHEV architecture from the REET report
(Bhattacharjya, et al. 2023).
The P2-P4 PHEV architecture shown in Figure 2-10 was developed so ALPFIA could
simulate advanced PHEV vehicles in future medium-duty fleets. The P2-P4 PHEV architecture
was specifically developed for future MD truck PHEV applications with towing capability. This
model was not used to model any vehicles in the MD base-year fleet. More information about
this hybrid architecture can be found in EPA's report on the Range-Extended Electric Truck
(REET) (Bhattacharjya, et al. 2023).
2-19
-------
e-motor/
generator
Figure 2-10: P2-P4 PHEV architecture from the REET report (Bhattacharjya, et al. 2023).
2.4.4.3 Battery Electric Vehicle Architecture (BEV)
The energy consumption performance of battery electric vehicles (BEVs) is modeled using a
battery and an electric drive unit (EDU) consisting of inverter, motor/generator, and gearing
assembly as shown in
Figure 2-11. The battery energy capacity for a typical mid-sized vehicle with a 300-mile range
is around 80-100 kWh.
e-motor/
generator
Figure 2-11: Battery electric vehicle architecture.
2-20
-------
2.4.5 Engine, E-motor, Transmission, and Battery Components
To assess advanced vehicle powertrain technologies for regulatory feasibility, the National
Center for Advanced Technology (NCAT) within the National Vehicle and Fuels Emission
Laboratory, benchmarks advanced vehicle powertrain technologies. Using their experience and
expertise in laboratory benchmarking, EPA has been able to characterize engine and
transmission controls, energy consumption, and emissions impacts of manufacturer powertrain
technologies under operating conditions. These characterizations are then provided as inputs for
complete vehicle simulations using its ALPHA model. EPA's engine, electric motor and
transmission benchmarking methods have been peer reviewed during the publishing process for
their publicly available SAE technical papers (U.S. EPA 2022b).
ALPHA stores engine, transmission, e-motor, and battery component data in ALPHA input
files. The data included in each of the ALPHA inputs comes from various sources including EPA
and other National Laboratory benchmark testing, GT-Power combustion modeling, contracted
benchmark testing, and technical papers. Each input dataset receives extensive quality analysis
from EPA's benchmarking and engineering team to identify and remove any errors, document
primary sources of data, apply best practices when extrapolating to very low or high
speeds/torques, and ensure consistency between similar ALPHA input files.
The rest of this chapter discusses the various ALPHA input files for the internal combustion
engines, electric inverters/motors, batteries, and transmissions used for this rule. These ALPHA
inputs are listed in Table 2-3 through Table 2-6 and described in detail in RIA Chapter 3.5 Light-
Duty Engines. Table 2-3 identifies the internal combustion engines that ALPHA uses for this
rule. The details of each engine ALPHA input listed are described in the RIA Chapter 3.5.1.
Detailed information about the engines (engine efficiency map, inertia, DFCO, fuel penalties,
cylinder deactivation features, fuel used, etc.) can be found in the data package associated with
each engine (U.S. EPA 2022b) (U.S. EPA 2023c). The qualifiers in the Engine RSE Code
column are defined as: SLA- Standard Load Application, HLA- High Load Application, MDV-
medium-duty vehicle, ICE- non-hybridized internal combustion engine, P2- HEV with electric
motor in the P2 position, PS- PowerSplit HEV, P2P4 - AWD P2 type PHEV, and SPP4 - AWD
series-parallel PHEV.
2-21
-------
Table 2-3: Engine ALPHA input maps used to create ALPHA outputs for RSEs
Configuration
Engine RSE
ALPHA Component Name
Data Sou rec
Code
Port Fuel Injection
PFI
GT Power Baseline 2020 Ford 7.3L
Technical Report
Large Bore
(MDVorHLA)
Engine from Argonnc Report Tier 3 Fuel
(Argonnc/SwRI)
Gas Direct
GDI
2013 Chevrolet 2.5L Ecotec LCV Engine
Contracted Testing
Injection
Reg E10 Fuel
(FEV)
Gas Direct
GDI
2014 Chevrolet 4.3 L EcoTcc3 LV3 Engine
EPA-NCAT Testing
Injection
(HLA)
LEVI 11 Fuel
Turbo Gas
TDS11
2013 Ford 1.6L EcoBoost Engine LEV III
EPA-NCAT Testing
Fuel
Turbo Gas
TDS
2015 Ford 2.7L EcoBoost Engine Tier 3
EPA-NCAT Testing
(HLA)
Fuel
Turbo Gas
TDS
2016 Honda 1.5L L15B7 Engine Tier 3
EPA-NCAT Testing
(SLA)
Fuel
Turbo Gas Miller
MILLER (ICE)
Volvo 2.0L VEP LP Gcn3 Miller Engine
Technical Paper (2020
from 2020 Aachen Paper Octane Modified
Aachen)
for Tier 3 Fuel
Turbo Gas Miller
MILLER
Gccly 1.5L Miller GHE from 2020 Aachen
Technical Paper (2020
Dedicated Hybrid
(P2 or PS)
Paper Octane Modified for Tier 3 Fuel
Aachen)
Atkinson
ATK
2018 Toyota 2.5L A25A-FKS Engine Tier
EPA-NCAT Testing
3 Fuel
Atkinson
DHE
Toyota 2.5L TNGA Prototype Hybrid
Technical Paper (2017
Dedicated Hybrid
(P2 or PS)
Engine from 2017 Vienna Paper Octane
Vienna)
Modified for Tier 3 Fuel
Turbo Diesel
DIESEL
Duramax 3.0L
Technical Report
(Argonnc/SwRI)
High Load
MILLER
Future 3.6L HLA Hybrid Concept Engine
VW 1.5LTSI cvo
Application
(SPP4)
Tier 3 Fuel
Hybrid Concept 4 from
Dedicated Hybrid
2019 Aachen Paper
High Load
MILLER
Future 6.0L HLA Hybrid Concept Engine
VW 1.5LTSI cvo
Application
(P2P4)
Tier 3 Fuel
Hybrid Concept 4 from
Dedicated Hybrid
2019 Aachen Paper
2.4.5.1 Electric Drive Components
Table 2-4 shows the three types of electric drive components that ALPHA uses for this rule.
• BISG - Belt Integrated Starter Generator consisting of an inverter, an electric motor,
and the engine's front-end pulley/belt drive.
• EDU - Electric Drive Unit consisting of an inverter, an electric motor, and the drive
gearing.
• EMOT - Electric Motor consisting of an inverter and an electric motor (the gear
losses are accounted for elsewhere and not within this device).
The details of each electric motor ALPHA input listed are described in the RIA Chapter 3.5.2.
Detailed information about the electric component (efficiency map, losses, gear ratios, etc.) can
2-22
-------
be found in the data package associated with each component (U.S. EPA 2023a) (U.S. EPA
2023b).
Table 2-4: Electric motor/related ALPHA input maps for electrified vehicles
used to create ALPHA outputs for RSEs
Type
ALPHA Component Name
Data Sou rec
EMOT
2010 Tovola Prius 60kW 650V MG2 EMOT
ORNL
EMOT
Est 2010 Toyota Prius 60kW 650V MG1 EMOT
ORNL/NCAT
EMOT
2011 Hyundai Sonata 30kW 270V EMOT
ORNL
BISG
2012 Hyundai Sonata 8.5kW 270V BISG
ORNL
EDU
Generic IPM 150kW EDU
NCAT
EMOT
3 modern IPM electric motor/inverters used in
future LD and MD PHEVs with lowing capability
Confidential
2.4.5.2 Transmissions
Table 2-5 identifies the automatic transmissions used for this rule. These transmission models
are all traditional step automatic transmissions and are used to represent all drivetrains in
conventional and electrified vehicles (except for PowerSplit vehicles and BEVs). Transmission
losses as a function of load and gear number are built into the ALPHA input. The torque
converter efficiency and lockup logic are also programmed into each ALPHA input. The shifting
logic for each transmission is built into a function called ALPHA-shift. The TRX ECVT FWD
transmission supplies the planetary gear ratios and the gear mesh efficiency for the PowerSplit
drivetrain. EPA did not perform any additional transmission testing for this rulemaking.
For more information on most of these transmissions, please refer to the description of
ALPHA in the 2016 Final Determination (U.S. EPA 2017b).
2-23
-------
Table 2-5: Transmission ALPHA inputs used to create ALPHA outputs for RSEs
Tj |K-
A1.1*11A ( omponenl Name
5-s.pd 1\\ D AT
TRXlUlWD
5-spd RWD AT
TRX10RWD
6-spd FWD AT
TRX11FWD
6-spd RWD AT
TRX11RWD
Adv 6-spd FWD AT (no torque converter)
TRX12_F WD_P2_Hybrid
Adv 6-spd FWD AT
TRX12FWD
Adv 6-spd RWD AT
TRX12RWD
8-spd FWD AT
TRX21FWD
8-spd RWD AT
TRX21RWD
Adv 8-spd FWD AT (no torque converter)
TRX22_FWD_P2_Hybrid
Adv 8-spd FWD AT
TRX22FWD
Adv 8-spd RWD AT
TRX22RWD
PS Planetary Gear
TRXECVTFWD
Surrogate for BEV Transmission
BEV transmission
2.4.5.3 Batteries
Table 2-6 lists the drive battery packs used in electrified vehicles. EPA did not test any battery
packs for this rulemaking. We relied on battery equivalent circuit data provided by Southwest
Research Institute and other sources.
Table 2-6: Battery ALPHA inputs used to create ALPHA outputs for RSEs
Type
ALPHA Component Name
Used For
48-Voll Battery
battery _base_A 12348 V_8 Ah
PO
High-Voltage
battery _basc_Samsung_LI_Powcr_mod2
PowerSplit
Battery
High-Voltage
battery _base_9p8_kWh_NCM
P2
Battery
High-Voltage
battery _pack_NMC_58kWh
BEV
Batterv
An equivalent circuit model, as shown in Figure 2-12 is used for the battery cells in the
ALPHA. The following parameters are used to define the high voltage battery model:
• Open circuit voltage (OCV V)
• Series resistance (RS) to model ohmic effects
• Short time constant resistor and capacitor (RP ST and CP ST) to model charge
transfer dynamics
2-24
-------
• Long time constant resistor and capacitor (RPLT and CPLT) to model diffusion
dynamics
The ALPHA framework allows for these parameters to be a function of multiple variables
such as state of charge (SOC), temperature, etc. The SOC is estimated based on coulomb
counting. Additionally, the model also contains a basic thermal model that estimates battery
temperature based on the losses.
Specifically, for the HEV and BEV models validated for the program, the propulsion battery
parameters are a function of SOC (at minimum) and temperature (when data was available).
Further, it was decided that using a series resistance (RS) was sufficient based on the
performance of the model compared to the vehicle test data, so the short- and long-time constants
are disabled in such scenarios but can be enabled if data is available.
ocu u
RS RP_LT RP_ST
—'W—t—'W—t—M—*
£ CP_LT CP_ST
batt volts U
-------
Generic 150 kW EDU
Generic 200 kW EDU
200 kW
267 kW
500 kW
400 kW
300 kW
Efficiency (%)
-100
133kW
67 kW
0
-67 kW
-133 kW
-200 kW
-300 kW
^OOkW
-500 kW
-200
-300
5000 10000
Speed (RPM)
5000 10000
Speed (RPM)
Figure 2-13: Power scaling example - Electric drive unit.
100 kW
Efficiency {%)
-267 kw
-400 kW
-533 kW
-667 kW
2.4.7 Tuning ALPHA'S Electrified Vehicle Models Using Vehicle Validations
Using the architectures and ALPHA component input data described above, the HEV models
(PO, P2, and PowerSplit), and BEV models were developed, calibrated, tuned, and validated
using detailed test data measured in a laboratory from specific vehicles listed in Table 2-7 while
driven over the EPA city, highway, and US06 regulatory drive cycles.
Table 2-7: Table of test data vehicles used to validate ALPHA.
Model
Validation Vehicle
Notes
PO
Mild Hybrid
2013 Chevrolet
Malibu Eco
-Validation of ALPHA'S P0 mild hybrid model was previously
completed during the Midterm Evaluation. [9]
- Slight updates have been made since then based on data from chassis
testing done on 2018 Jeep Wrangler eTorque and 2020 Dodge Ram
eTorque vehicles.
PowerSplit
Strong Hybrid
2017 Toyota Prius
Prime PHEV
While this vehicle is a PHEV, the ALPHA validation of ALPHA'S
PowerSplit model primarily focused on "charge sustaining" operation.
P2
Strong Hybrid
2016 Hyundai
Sonata PHEV
While this vehicle is a PHEV, the ALPHA validation of ALPHA'S P2
hybrid model primarily focused on "charge sustaining" operation.
SP-P4 Strong
Hybrid &
PI I FA"
AWDF-150
While this vehicle is a PHEV, the ALPHA validation of ALPHA'S SP
hybrid model primarily focused on "charge sustaining" operation.
P2-P4 PHEV
A WD F-150 &
RAM 2500 trucks
While this vehicle is a PHEV, the ALPHA validation of ALPHA'S P2
hybrid model primarily focused on "charge sustaining" operation.
Battery Electric
Vehicle (BEV)
2018 Tesla Model 3
Each electrified vehicle model was tuned to achieve similar operational behavior for the
engine, transmission, electric motors, and battery, as observed in actual vehicle test data. For
example, Figure 2-14 compares data from the PowerSplit model against the corresponding
2-26
-------
measured test data on a 2016 Toyota Prius Prime. This validation process is similar to what was
done previously for conventional vehicles. (Newman, Kargul and Barba 2015a)
model v test diff = -1.8 %
c
rc
00
680
,210
2
; 200
¦ -100
700
720
740 760
Time [s]
model v test diff = -0.3 %
780
2
800
680 700 720 740 760 780 800
Time [s]
fi—n-
680 700 720 740 760 780 800
1 -50
780
680 700 720 740 760 780 800
Figure 2-14: Sample validation comparison of modeled versus measured data from a 2016
Toyota Prius Prime operating on the drive schedule between 680 to 820 seconds.
Table 2-8 summarizes the final results of the strong hybrid and BEV models. For the
PowerSplit strong hybrid model, the ALPHA simulated combined city-highway C02 grams per
mile was 1.7 percent higher than that of the 2017 Toyota Prius Prime driven on the
dynamometer. For the P2 strong hybrid model, the combined city-highway simulation results
were 3.6 percent higher than the 2016 Hyundai Sonata PHEV tested on the chassis dyno. Finally,
the combined results from the BEV model were 1.9 percent higher than the test data from the
2018 Tesla Model 3.
2-27
-------
Table 2-8: Percent difference of ALPHA vehicle validation simulation versus
benchmarking test data.
Model: Validation Vehicle Hot HW US06
UDDS
Combined Units
(hot-UDDS&HW)
Power Split Strong Hybrid:
2017 Toyota Priiis Prime PHEV 0.0% 3.8% -2.7%
1.8%
% Diff
C02
g/mile
% Diff
C02
g/mile
% Diff
kWh/mi
P2 Strong Hybrid:
2016 Hyundai Sonata PHEV 0.9% 6.8%
-1.4%
3.3%
Battery Electric Vehicle:
2018 Tcsla Model 3
3.8%
-0.5%
2.8%
1.8%
ALPHA Tuning of the SP-P4 Strong HFV/PHFV and P2-P4 PHEV models: Since the SP-P4
and P2-P4 PHEV models represent future vehicles, it was not possible to compare the computer
model results against either test vehicle or certification data. However, it was possible to
compare the ALPHA results for these models to the corresponding GT-Drive model results in the
REET study (Bhattachaijya, et al. 2023) as shown in Table 2-9. ALPHA was able to match the
GT-Drive results within +1-4 percent.
2-28
-------
Table 2-9: Comparison of ALPHA results from SP-P4 Strong HEV/PHEV and P2-P4
PHEV models to the results from the GT-Drive versions of these model in EPA's REET
study (Bhattacharjya, et al. 2023).
Series-Parallel (SP-P4) gCQ2/mile
UDDS
HwFET
US06
ALPHA
251.1
274.9
374.0
GT-DRIVE
244.1
267.2
363.7
F-150
difference
2.9%
2.9%
2.8%
Gasoline DHE
Parallel (P2-P4) gC02/mile
UDDS HwFET US06
ALPHA
249.0
292.5
398.3
GT-DRIVE
259.2
291.8
391.5
F-150
difference
-3.9%
0.2%
1.7%
Gasoline DHE
ALPHA
330.6
359.8
483.7
GT-DRIVE
326.1
358.9
466.1
RAM 2500
difference
1.4%
0.3%
3.8%
Gasoline DHE
ALPHA
310.4
336.7
453.8
GT-DRIVE
321.8
349.1
459.5
RAM 2500
difference
-3.5%
-3.6%
-1.2%
Diesel
Overall observations of ALPHA'S validation results: The differences of the combined
city-highway results between the test vehicle's data and ALPHA'S simulation fall within EPA's
goal for a validation. In its validation efforts, EPA typically targets results to be within +/- 4
percent difference between the test vehicle and the simulation results for the combined city and
highway cycles.
In some cases, for individual cycles the difference can be close to or above the targeted value
for acceptable reasons. There could be several reasons for differences between simulation and
actual validation test data, including:
2-29
-------
• The engine maps used for the validation are similar, but not identical, to the engine in
the tested vehicle. If the actual vehicle engine map were available as an ALPHA input,
the results might have been closer to the vehicle test data.
• It is possible that coast-down coefficient adjustments for electrified vehicles do not
adequately account for the losses that occur when the electric motor is always
connected to the input of the transmission. (Moskalik 2020)
• The as-driven speed trace in some cases contain additional signal noise that made it
unsuitable for simulation. In these cases, the target speed traces for the simulated
cycles were used instead, but these simulations would be missing the small variations
caused by the "true" as-driven trace.
• The precise amount of braking energy (regen) that can be captured varies from cycle-
to-cycle as well as with vehicle-to-vehicle.
Considering the factors listed above, and the good match between the operational behavior of
the model and the tested vehicle, the initial validations of these models are robust enough to be
suitable for their ultimate use to estimate fleet-wide emissions. The difference percentages of the
individual cycle results noted during the validation are small enough to meet the combined cycle
goal of the validation and not require any adjustments. The next chapter shows how well each of
the validated electrified vehicle models can simulate other variant vehicles (vehicles that are
technology-wise similar to the vehicles used to validate the model).
2.4.7.1 Verifying the Validated Strong Hybrid and BEV Models against Variant
Vehicles
Since ALPHA architecture models are intended to simulate a range of vehicles, it is helpful to
compare ALPHA results to data from multiple tests on multiple vehicles. Therefore, the next step
in the validation process was to verify the ALPHA model against a number of technically
similar, but different, vehicles (think of these other vehicles as "sibling" or "cousin" vehicles).
These variant vehicles were selected because they have very similar powertrain designs and
control strategies to the initial validation vehicle, yet they may be of different size and make.
Additionally, the certification data originates from different vehicles, drivers, equipment, and
laboratories, all of which increases the variability of the comparisons, and can yield a measure of
how well the validated model can simulate other vehicles.
Once each vehicle model was developed and tuned to provide similar behavior as its test
vehicle, CO2 (for hybrid operation) and energy consumption (for BEVs and PHEVs running in
charge depleting mode) results were compared for other "variant" vehicles from the same
manufacturer with very similar powertrain designs as the original validation vehicle. Since there
were no dynamometer test data for these variant vehicles, the ALPHA simulation results were
checked to see how closely they agreed with available vehicle Certification data. These results of
the ALPHA model validations and their variant verifications for the strong hybrids and BEVs are
summarized in Table 2-10.
2-30
-------
Table 2-10: Percent difference of variant vehicle ALPHA simulations versus
certification data.
Verification of
Variant Vehicles
City
HW
US06
Combined
(City &
HW)
units
#
vehs.
Power Split Strong Hybrid
Variants:
2017 Toyota Prius Prime
'PHEV
-4.8%*
-2.1%
-5.1%
-3.56%
Avg % diff C02 g/mi
5
2.9%*
2.5%
3.8%
2.54%
Std-dev of % diff C02 g/mi
P2 Strong Hybrid variants:
2016 Hyundai Sonata
PHEV
-0.7%*
5.6%
7.9%
1.92%
Avg % diff C02 g/mi
2
5.9",,*
3.2%
3.0%
3.61%
Std-dev of % diff C02 g/mi
Battery Electric Vehicle:
2018 Tesla Model 3
5.9%**
1.6%
n/a
3.91%
Avg % diff kWli/mi
12
2.3%**
2.7%
n/a
2.33%
Std-dev of % diff kWli/mi
* cold-start FTP ** warm-UDDS
The top row of Table 2-10 summarizes the average difference between ALPHA estimated
CO2 g/mile and Certification CO2 g/mile for five Toyota variants of the Prius Prime PowerSplit
design operating in charge sustaining mode. The comparison shows the average CO2 percent
difference over the three drive cycles (FTP, HW, and US06) to be -4.8 percent, -2.1 percent, and
-5.1 percent, respectively. The average percent difference of the combined (FTP-HW) CO2
values is shown to be -3.6 percent. The standard deviation of these combined averages is shown
to be 3.8 percent.
The center row of Table 2-10 summarizes the average difference between a P2 strong hybrid
vehicle's ALPHA estimated CO2 g/mile and its Certification CO2 g/mile for two Hyundai/Kia
variants of the Sonata P2 Hybrid design operating in charge sustaining mode. This comparison
shows the average CO2 percent difference over the three drive cycles (FTP, HW, and US06) to
be -0.7 percent, 5.6 percent, and -7.9 percent, respectively. The average percent difference of the
combined (cold FTP-HW) CO2 values is shown to be 1.9 percent. The standard deviation of
these combined averages is shown to be 3.6 percent.
The bottom row of Table 2-10 summarizes the average difference between a Tesla BEV's
ALPHA estimated energy consumption (kWh/mi) and its Certification energy consumption
(kWh/mi) for 12 variants of the Tesla Model 3 design. This comparison shows the average
kWh/mi percent difference between two available drive cycles (FTP and HW) to be 5.9 percent
and 1.6 percent, respectively; noUS06 Certification data were available for this comparison. The
average percent difference of the combined (FTP-HW) CO2 values is shown to be 3.9 percent.
The standard deviation of these combined averages is shown to be 2.3 percent.
Comparing the combined city-highway averages of the variant vehicle simulations in Table
2-10 to the vehicle validation combined averages in Table 2-8 shows a slight increase in
variability, which was expected given the validated model was tuned using a specific validation
vehicle, yet asked to estimate results for slightly different vehicles. There could be several
reasons for differences between simulation and actual certification test data, including:
• Certification values inherently contain some variability since they are driven and
measured in different laboratories. The typical test-to-test variation of chassis
2-31
-------
dynamometer certification testing can be +/-3 percent due to a variety of factors such
as different drivers, measurement equipment, fuel, and facilities.
• In addition, for hybrid and PHEV vehicles, certification testing follows the
recommended practice in the 2010 revision of SAE Standard J1711 (SAE 2010). This
standard allows a net energy change in the battery over a test of up to +/-1 percent of
the total fuel energy consumed, while not requiring the measured CO2 value to be
corrected (unlike in the ALPHA simulations). Depending on the average efficiency of
the engine and electrical system, this variation could alter fuel usage (and CO2
emissions) by +1-2 to 3 percent.
Considering the factors listed above, the variant vehicle simulation results in Table 2-10 are
quite good. With a large number of test vehicles, the test-to-test variations of certification data
would tend to cancel out as will be shown in Chapter 2.4.8 where ALPHA'S ability to model
large fleets is discussed.
Since the SP-P4 strong HEV/PHEV and P2-P4 PHEV models represent future vehicles that
are not yet in production, it was not possible to compare the ALPHA model results against
certification data of any variant vehicles.
2.4.7.2 P0 Mild Hybrid Validation Efforts
The ALPHA validation for P0 mild hybrid vehicles was done during the Midterm Evaluation
(Lee, et al. 2018), consequently there is no recent P0 mild hybrid vehicle validation data shown
in Table 2-11. Instead, a different approach was taken to validate the accuracy of the P0 model.
The first two rows of data in Table 2-11 summarize the differences of ALPHA CO2 simulations
of 24 conventional vehicles with both P0 mild hybrid and engine start-stop technology applied
compared to the ALPHA CO2 simulations of the same vehicles with neither technology applied.
The ALPHA simulation data shows an average combined (FTP-HW) CO2 reduction of 9.3
percent when applying both P0 mild hybrid and engine start-stop technologies to conventional
vehicles.
The next two rows of data in Table 2-11 document the differences between five comparisons
of EPA certification results of P0 mild hybrids with engine start-stop applied against the EPA
certification results of similar conventional vehicles without applying both the P0 mild hybrid
and engine start-stop technologies. The EPA certification data shows an average combined (FTP-
HW) CO2 reduction of 10.9 percent when applying both P0 mild hybrid and engine start-stop
technologies to a conventional vehicle. These results verify that ALPHA simulates P0 mild
hybrid combined with engine start-stop technologies within -1.6 percent.
2-32
-------
Table 2-11: Estimated CO2 reductions with both PO mild hybrid & engine start-stop
technologies applied to the comparable conventional vehicle.
MY 2019 PO Mild
Cold-
HW
US06
Combined
units
; #
Hybrids
Start
FTP
(cold-FTP
& HW)
vchs.
ALPHA of PO vs ALPHA
sim of conv vehicles
13.3%
2.1%
n/;i
9.3%
avg % diff CO2 for all pairs of sims.
24
2.4%
0.5%
n/;i
1.8%
std-dev of $diff C02 for all pairs of sims.
Cert of PO vs Cert of conv
vehicles
13.8%
5.1%
n/;i
10.9%
avg % diff CO2 for all pairs of Cert data
5
2.7%
2.0%
n/;i
1.5%
std-dev of $diff C02 for all pairs of Cert data
Difference of C02 averages
-0.5%
-3.0%
n/;i
-1.6%
difference of avg % diff C02
-
2.4.8 Verifying ALPHA'S Ability to Simulate Entire Fleets
To demonstrate that ALPHA is capable of modeling large fleets with a wide variety of
technologies, ALPHA3 was used to simulate the entire MY 2022 light-duty base year fleet. To
model the performance of these vehicles, data collected by EPA for compliance purposes,
together with information from other sources including laboratory vehicle benchmarking, were
used to calculate various metrics for vehicle and technology characteristics that are related to fuel
economy and GHG emissions. The process used was similar to that used by EPA in 2018.
(Bolon, et al. 2018)
2.4.8.1 Data Sources to Determine MY 2022 Light-Duty Fleet Parameters
Vehicle specification data that is relevant to characterizing emissions-reducing technologies
are available from multiple sources. Because these data sources were generally not originally
developed for this particular use, any single source will often provide only partial coverage of
vehicle models over the years of interest, and production volume data necessary for generating
aggregate statistics is often lacking. This chapter describes a methodology for consolidating data
from multiple sources, while maintaining the integrity of the original data.
The most basic obstacle to consolidating data sets is variation in how vehicles are classified in
different data sources. This might include variation in the level of detail as well as variation in
the particular dimensions along which vehicles are characterized. Even when various data sets
share a common categorization method, merging multiple sources may still be complicated when
one or more of the data sets does not include the entire range of vehicles.
The primary data source used by EPA to characterize the GHG performance of the existing
fleet is the certification data submitted by manufacturers to EPA's VERIFY database. The data
pertain mainly to vehicle emissions performance collected in dynamometer testing, and include a
general classification of engines, transmissions, and drive systems. Also included are vehicle
characteristics related to road loads: dynamometer target and set coefficients, road load
horsepower, and test weights. Additional data are obtained from the "2022 Test Car List"
available on EPA's "Data on Cars Used for Testing Fuel Economy" (U.S. EPA 2022e) and
FuelEconomy.gov (US DOE & EPA 2024) websites, both of which are publicly available.
2-33
-------
In addition to the information in datasets maintained by EPA, additional vehicle specifications
and technology details can be obtained through other public and commercially available sources
of vehicle data such as Edmunds.com, Wards Automotive (Penton) and AllData Repair (AllData
LLC).
For the MY 2022 base year light-duty fleet, there were a total of 1305 distinct vehicle model
types.
2.4.8.2 Vehicle Parameters
Using these aggregated data sources, a technology assignment of the powertrain for each
vehicle was made based on the nominal technology description in the data source. The categories
used to differentiate powertrain components are shown in Table 2-12.
Table 2-12: Powertrain components and categories.
Component Category
Level of electrification
Start-stop
Type of hybridization
Engine type
Transmission type
Number of gears
Cylinder deactivation
Engine power
Engine displacement
Engine number of
cylinders
Electric motor power
Batterv size
Applicable to
All vehicles
Conventional vehicles
Mild hybrids
Strong hvbrids/PHEVs
Non-BEVs
Conventional and mild hybrids
Strong hvbrids/PHEVs
BEVs
Step transmissions
Non-BEVs
Non-BEVs
Non-BEVs
Non-BEVs
BEVs and hybrids
BEVs and hvbrids
Values
Conventional, mild hybrid, strong hybrid, strong
PHEV. or battery electric vehicle
YorN
PO or PI
PowerSplit. P2. series-parallel, or series
dicscl. PFI naturally aspirated. GDI naturally
aspirated, turbocharged. supercharged, none
AT. CVT. DCT. manual"
Specialty
Direct drive
Number
Discrete, continuous, or none
Power (HP)
Displacement (liters)
3/4/6/8/10/12
Power (kW)
Energy (kWh)
In addition to the powertrain description, other vehicle parameters are necessary to simulate
individual vehicles within ALPHA. These parameters, shown in Table 2-13, were also pulled
from the data sources.
Table 2-13: Vehicle Parameters.
Parameter Values / Units
Equivalent test weight (ETW) lbm
Drive type FWD. RWD. or AWD
Vehicle coast-down target values (A, B, C) a (lbf). B (lbf/mph). C (lbf/mph2)
n/v ratio rpm/mph
Footprint Square feet
Production volume Number of units
Frame style Unibody v. body on frame
Towing capacity lbm
2-34
-------
2.4.8.3 Electrified Powertrain Model Assignments
Based on the level of electrification and the type of hybridization, vehicle model types in the
MY 2022 light-duty fleet were separated into individual groups to which the appropriate ALPHA
model was applied. These groups are shown in Table 2-14.
Generally, vehicles without electrified components were grouped as conventional vehicles,
while vehicles without internal combustion engines were grouped as BEVs. These include fuel
cell electric vehicles (FCEVs), which make up around 0.02 percent of the fleet. For hybrid
vehicles, those with PI or P0 architectures were considered mild hybrids and modeled with
ALPHA'S P0 model. For strong hybrids and PHEVs, vehicles with P2 architectures were
modeled with ALPHA'S P2 model, and the remainder of the hybrid fleet (the vast majority) was
modeled with the PowerSplit model.
Table 2-14: Vehicle model type assignments in MY 2022 light-duty fleet.
Vehicle architecture groups ALPHA model Number of vehicle model tvpes
Conventional vehicles, with or Conventional vehicle model 1018
without stop-start
Mild hybrids (PO and PI) PO model 91
Strong hybrids and PHEVs with P2 P2 model 60
architecture
All other strong hvbrids and PowerSplit model 40
PHEVs
BEVs and FCEVs BEV model %
2.4.8.4 Modeling Conventional Vehicles in the Fleet
To model conventional vehicles, available ALPHA maps for powertrain components were
assigned to each vehicle, depending on which map had attributes closest to the engine in the
specific vehicle being modeled. Engines in conventional vehicles were mapped to the ALPHA
input engine maps given in Table 2-3. The engine assignment was generally based on aspiration.
Previous investigation had determined that modern PFI and GDI engines have similar
performance, and thus these engines were grouped together as naturally aspirated (NA) engines.
Other groups were diesel engines, boosted engines, and Atkinson engines. For boosted engines,
some engine families with better performance were assigned to use an advanced turbocharged
engine.
For some engine categories, different engine maps were specified depending on whether the
modeled vehicle was categorized as a towing vehicle or not. For this purpose, all vehicles that
had an OEM-defined towing capacity and/or were large body-on-frame vehicles were classified
as towing/hauling, high load application (HLA) vehicles. This category encompassed large and
mid-sized pickup trucks, most large SUVs, and large vans. HLA vehicle model types with NA or
standard turbo engines were assigned ALPHA engines which reflected the towing requirement.
The assignment of ALPHA engines to conventional base year fleet vehicles is given in Table
2-15. For each vehicle model type, the engine model was scaled to match either the given power
of the engine (power scaling), or to match the engine displacement (displacement scaling) as
described in Chapter 2.4.6 (Dekraker, et al. 2017).
2-35
-------
Table 2-15: Assignments of engines used to simulate MY 2022 base year fleet conventional
vehicle model types, based on engines in Table 2-3.
Engine Categories
Modeled As
Sealing
ALPHA engine input
Diesel engines
Diesel engine
power
2020 GM 3.0L Duramax
PFI and GDI NA engines (non-
GDI engine
power
2013 Chevrolet 2.5L Ecotcc LCV
HLA) ^
PFI and GDI NA engines (HLA)
GDI engine
displacement
GTPowcr 2020 Ford 7.3L
Atkinson engines
Atkinson
power
2018 Toyota 2.5L A25A-FKS
Turbochargcd engines (non-HLA)
TDS engine
power
2013 Ford Ecoboost 1.6L
Turbochargcd engines (HLA)
TDS engine
power
2015 Ford EcoBoost 2.7L
Supercharged engines
TDS engine
displacement
2013 Ford Ecoboost 1.6L
Advanced turbochargcd engines
Adv. TDS
power
2016 Honda 1.5L L15B7
Transmissions in conventional vehicle model types were mapped to one of five automatic step
transmissions given in Table 2-5. Available production transmission varieties, including step
automatic transmissions (ATs), CVTs, DCTs, and manual transmissions were mapped to one of
these five modeled transmissions used in ALPHA. Although the behavior of the modeled TRX
ATs is not identical to other transmission varieties, the overall effect on powertrain efficiency
and thus GHG emissions is similar.
Losses in the transmission and differential were modified depending on whether the vehicle
was a front or rear wheel drive. This mapping is very similar to the process used by EPA in
earlier rulemakings (U.S. EPA 2016b). Losses in the transmission were scaled to the peak torque
of the engine.
Table 2-16: Transmissions used to simulate MY 2022 base year fleet conventional vehicles,
based on transmissions given in Table 2-5.
Transmission Categories Modeled As Souree/Notes
4- and 5-spd ATs. 5- and 6-spd manuals TRX 10 Fivc-spccd from 2007 Toyota Camrv
6-spd ATs TRX 11 Six-spccd GM 6T40
All DCTs. 7-spd manuals TRX 12 Six-spccd with advanced loss reduction
7-spd and above ATs. older CVTs TRX21 Eight-speed FCA 845RE
Newer CVTs TRX22 Eight-speed with advanced loss reduction
With the appropriate powertrain assigned, each vehicle was simulated in ALPHA over the
3-bag FTP and HWFET cycles, using the vehicle parameters in Table 2-13.
The grams per mile (g/mi) CO2 values from the ALPHA simulation were compared to
certification values; the production weighted averages of the differences, both in absolute g/mile
and in percent, are given in 2-17. A scatter plot of the ALPHA versus certification values is
shown in Figure 2-15. The sizes of the bubbles in the plot reflect production volumes for each
vehicle.
2-36
-------
Table 2-17: Conventional vehicle model type in the MY 2022 fleet - ALPHA CO2 g/mile
values versus certification CO2 g/mile.
FTP
HW
Combined
Production weighted average g/mile
-11.2 g/mile
+ 12.4 g/mile
-0.6 g/mile
Production weighted std. dev. g/mile
18.4 g/mile
13.3 g/mile
13.1 g/mile
Production weighted average percent
-3.0%
+5.2%
-0.1%
Production weighted std. dev. percent
5.2%
5.2%
4.5%
<
x
Cl
1
<
600
550
500
w
* •
JA
* m _
r ' •
41
• j
I*'
|P*
150 200 250 300 350 400 450 500 550
Combined Cycle g/mile C02: Certification
600
Figure 2-15: Conventional Vehicles in the MY 2022 fleet - ALPHA combined cycle CO2
g/mile values versus certification CO2 g/mile where bubble sizes reflect production
volumes.
The production weighted average of the ALPHA CO2 values is within 0.1 percent of the
certification values and is within 1 g/mile of the certification value average. The graph in Figure
2-15 clearly indicates there is some scatter in the results; however, over 60 percent of the
vehicles in the fleet were simulated within 10 g/mile of certification values, and nearly 90
percent of the fleet was simulated within 20 g/mile. Generally, the more substantial outliers (i.e.,
those model types where the ALPHA simulation was not as well matched to the certification
values) were low-volume vehicles.
2.4.8.5 Modeling Mild Hybrids in the Fleet
All mild hybrids were modeled using a P0 BISG model with engine start-stop, (Lee, et al.
2018) using the BISG motor from Table 2-4 and the 48V battery from Table 2-6. The mild
hybrids in the fleet are produced by multiple manufacturers with different operational strategies
2-37
-------
and configurations (PO v. PI configurations, and 48V v. 12V architecture). However, a single PO
model was used to represent all mild hybrid vehicles.
The engines and transmissions for mild hybrids were assigned and scaled in the same way as
for conventional vehicles. The motor power and battery pack capacity in the simulation were
scaled to match the values of these components in the vehi cle model type being simulated.
Each vehicle was simulated in ALPHA over the combined FTP and HWFET cycles using the
parameters in Table 2-13. The g/mile CO2 values from the ALPHA simulation were compared to
certification values; the production weighted average of the difference is given in Table 2-18. A
scatter plot of the ALPHA versus certification values is shown in Figure 2-16.
Table 2-18: PO mild hybrids in the MY 2022 fleet - ALPHA CO2 g/mile values versus
certification CO2 g/mile.
FTP
HW
Combined
Production weighted average g/mile
-5.1 g/mile
+25.2 g/mile
+8.5 g/mile
Production weighted std. dev. g/mile
21.6 g/mile
8.6 g/mile
14.0 g/mile
Production weighted average percent
-1.0%
+9.9%
+2.9%
Production weighted std. dev. percent
5.9%
3.4%
4.7%
450
<
x
Q_
400
O
O 350
OJ
E
300
QJ
U
& 250
-a
c
^ 200
o
u
150
• j
. •p
*
_ *
•
•
•
*
• P0 model types
'w
PI model types
150 200 250 300 350 400
Combined Cycle g/mile C02: Certification
450
Figure 2-16: P0 mild hybrids in the MY 2022 fleet - ALPHA combined cycle CO2 g/mile
values versus certification CO2 g/mile where both P0 and PI model types were simulated
using the ALPHA P0 model and bubble sizes reflect production volumes.
The same P0 architecture was used to simulate the performance of all mild hybrids; notably,
both P0 and PI architectures were represented by a P0 model. The P0 and PI model types are
illustrated separately in Figure 2-16. Even though the architecture in the actual vehicle model
type differed, the P0 model reasonably represented both POs and Pis. For all mild hybrids, the
production weighted average of the difference between ALPHA simulation and certification CO2
2-38
-------
values was 2.9 percent, or 8.5 g/mile. Individually, the PO and PI model types were simulated to
similar degrees of accuracy: for PO model types, the production weighted average difference was
2.9 percent, or 8.4 g/mile, and for PI model types, the same average difference was 3.1 percent,
or 9.4 g/mile.
2.4.8.6 Modeling Strong Hybrids in the Fleet
As shown in Table 2-14, the strong hybrids and PHEVs were divided into two categories,
covered separately in the following subchapters. The parallel P2 hybrids were modeled as P2s,
and the remainder of the strong hybrid fleet were modeled as PowerSplits. For many of these
strong hybrids, the engine was assumed to be a dedicated hybrid engine (DHE), utilizing either
an Atkinson cycle DHE or (in the case of turbocharged engines) a Miller cycle DHE, based on
the two dedicated hybrid engines given in Table 2-3. However, performance oriented strong P2
hybrids in the fleet do not contain dedicated hybrid engines; rather, their engines are more
similar to those in non-hybrid vehicles. For these vehicles, a non-DHE NA or turbo engine was
chosen with the process used for conventional vehicles. Likewise, the electric motors for strong
hybrids are based on the motors shown in Table 2-4, and the batteries are based on the batteries
from Table 2-6.
The range of strong hybrids in the fleet covers multiple manufacturers, vehicle applications,
hybrid configurations, and operational strategies. Additionally, not all strong hybrids have a
dedicated hybrid engine as modeled in ALPHA. However, it was judged that using these two
strong hybrid models would be reasonably representative of the fleet.
In a similar way to conventional vehicles, the engine model in each hybrid vehicle was
resized to match the given power of the vehicle engine as discussed above. The electric drive
motors were sized to provide power proportional to the engine power based on the configuration
of the original vehicle used to validate the P2 or PowerSplit model. Having a consistent ratio of
electric motor and engine sizes allowed the simulation to use the same control algorithms for
every vehicle model type, effectively representing the wide variation that exists in the original
fleet with just two models in ALPHA. Battery sizes were assigned according to the given value
for the vehicle from EPA's 2022 fleet parameter file (described in Chapter 2.4.8.1).
The same models were used to simulate the operation of non-plug-in HEVs and the charge
sustaining operation of PHEVs, with some minor differences in component sizing. PHEVs
generally have larger battery sizes, and thus the allowable SOC fluctuations during charge
sustaining mode were reduced. Additionally, for the P2 model, the nominal e-motor sizes for
PHEVs were increased to allow for all-electric operation over the drive cycles during charge
depleting mode.
2.4.8.6.1 PowerSplit modeling (HEVs and Charge-Sustaining-Mode PHEVs)
For the vehicle model types modeled as power splits, the engine and drive motor were
connected using a planetary gearset based on the Toyota Prius. Both e-motors in the system were
sized as a function of the rated engine power to keep the power values in the system
proportional.
Each vehicle was simulated in ALPHA over the 4-bag FTP and HWFET cycles, using the
vehicle parameters in Table 2-13. PHEVs were simulated in both charge sustaining and charge
depleting mode (discussed later in Chapter 2.4.8.6.3). The g/mile CO2 values from the ALPHA
2-39
-------
simulation were compared to certification values; the production weighted average of the
difference is given in Table 2-19.
Table 2-19: PowerSplit HEVs and PS PHEVs in Charge-Sustaining-Mode in the MY 2022
fleet - ALPHA CO2 g/mile values versus certification CO2 g/inile.
FTP
HW
Combined
Production weighted average g/mile
-8.1 g/mile
+4.9 g/mile
-2.2 g/mile
Production weighted std. dev. g/mile
5.6 g/mile
5.1 g/mile
4.6 g/mile
Production weighted average percent
-5.2%
+2.8%
-1.5%
Production weighted std. dev. percent
3.1%
3.2%
2.8%
300
<
CL
"Ai 250
O
o
_0J
£
"So 200
qj
o
>
u
~§ 150
15
E
o
u
100
• PowerSplit & Charge-Sustaining-
Mode PHEV vehicle model types
Series-parallel vehicle model types
100
150 200 250
Combined Cycle g/mile C02: Certification
300
Figure 2-17: PowerSplit HEVs and PS PHEVs in Charge-Sustaining-Mode in the MY2022
fleet - ALPHA combined cycle CO2 g/mile values versus certification CO2 g/mile values
where both PowerSplit and Series-parallel vehicle model types were simulated using the PS
model and bubble sizes reflect production volumes.
A scatter plot of the ALPHA versus certification values for vehicles modeled as PowerSplit
hybrids in the 2022 fleet is shown in Figure 2-17. The ALPHA PS model was used to simulate
both PowerSplit and series-parallel hybrids. The simulation results of the PowerSplit and series-
parallel hybrid types are illustrated separately in Figure 2-17. The PS model overall does quite
well in simulating the certification results, with an average difference under 2 percent and a low
standard deviation.
For PowerSplit vehicle model types, the production weighted average difference between
ALPHA simulation and certification CO2 values is very low, at -1.4 percent, or -2.2 g/mile. For
series-parallel vehicle model types, the same comparison difference is somewhat larger, at -3.7
percent, or -5.8 g/mile. This may be due to the limited number of vehicle model types available
2-40
-------
since those used produced simulated results close to the certification values. Although the
PowerSplit and series-parallel architectures are clearly different in practice, the ALPHA PS
model does produce fairly equivalent CO2 values when used to simulate either architecture.
2.4.8.6.2 P2 modeling (HEVs and Charge-Sustaining-Mode PHEVs)
For the vehicles modeled as a P2, the chosen engine was coupled to either a six- or eight-
speed transmission, depending on the number of gears in the modeled vehicles. These P2
transmissions were based on the TRX12 and TRX22, with the torque converter removed.
P2 vehicles were simulated in ALPHA over the 4-bag FTP and HWFET cycles, using the
parameters in Table 2-13. The g/mile CO2 values from the ALPHA simulation were compared to
certification values; the production weighted average of the difference is given in Table 2-20. A
scatter plot of the ALPHA versus certification values for P2 hybrid vehicles in the 2022 fleet is
shown in Figure 2-18.
Table 2-20: P2 HEVs and P2 PHEVs in Charge-Sustaining-Mode in the MY 2022 fleet -
ALPHA CO2 g/mile values versus certification CO2 g/mile.
FTP HW Combined
Production weighted average g/mile +2.6 g/mile +20.6 g/mile +10.7 g/mile
Production weighted std. dev. g/mile 22.0 g/mile 16.2 g/mile 16.0 g/mile
Production weighted average percent +2.6% +8.7% +5.0%
Production weighted std. dev. percent 8.2% 5.6% 6.2%
2-41
-------
400
<
300
_0J
E
"ao 250
ju
u
<3" 200
XI
0)
c
1 150
o
u
100
100 150 200 250 300 350 400
Combined Cycle g/mile C02: Certification
Figure 2-18: P2 HEVs and P2 PHEVs in Charge-Sustaining-Mode in the MY 2022 fleet -
ALPHA combined cycle CO2 g/mile values versus certification CO2 g/mile where bubble
sizes reflect production volumes.
Here, the ALPHA simulation on average has 5.0 percent, or 10.7 g/mile higher CO2 than the
certification values. It should be noted there are a low number of P2 model types, with only five
model types accounting for half of the total production volume. Figure 2-18 shows that two of
the highest volume vehicles have CO2 values over-predicted by ALPHA. If the unweighted
average were used in calculation, the difference is +2.1 percent, which indicates the sales mix,
particularly the highest volume model type, is influencing the sales-weighted average.
2.4.8.6.3 Charge depleting mode for PHEVs
The charge depleting operation of both PowerSplit and P2 PHEVs was also simulated.
ALPHA simulations over the certification drive cycles track direct current (D/C) energy
consumed from the battery, while certification values reflect the alternating current (A/C) energy
required to recharge the vehicle. Thus, the ALPHA D/C cycle simulation results were converted
into the equi valent A/C recharging energy required. The ratio of D/C electric energy consumed
to A/C energy used to charge the vehicle was assumed to be 0.87, based on an average of
available vehicle data from certification applications for battery electric vehicles. This factor was
applied to all PHEV simulation results to convert from D/C energy to A/C energy values.
These A/C energy values were compared to certification values. The operation of and results
from PowerSplit and P2 charge depleting operation were similar, and thus the results from both
architectures are presented together. The production weighted averages of the differences are
given in Table 2-21. A scatter plot of the ALPPIA versus certification values is shown in Figure
2-19 for both architectures.
2-42
-------
Table 2-21: PowerSplit and P2 PHEV both in Charge-Depleting-Mode in the MY 2022 fleet
- ALPHA kWh/100 miles values versus certification kWh/100 miles.
FTP
HW
Combined
Production weighted average kWli/100
miles
+0.37 kWh/100
miles
-1.43 kWh/100
miles
-0.44 kWh/100
miles
Production weighted std. dev. kWh/100
2.22 kWh/100
2.18 kWh/100
1.99 kWh/100
miles
miles
miles
miles
Production weighted average percent
+1.6%
-3.4%
-0.8%
Production weighted std. dev. percent
7.1%
5.8%
6.0%
^ 60
-------
from certification applications for battery electric vehicles. This factor was applied to all BEV
simulation results to convert to A/C energy values.
Fuel cell vehicle model types were also simulated using the BEV model. For FCEVs,
certification values were converted into an equivalent kWh/100 miles metric to have units
comparable to BEVs. For the simulation, a constant fuel cell system efficiency of 57 percent was
assumed, based on the fuel cell efficiency at low powers as reported by NREL (NREL 2019).
The ALPHA D/C cycle simulation results were converted into kWh/100 miles of hydrogen
energy required for each drive cycle using this factor.
Battery electric vehicles were simulated in ALPHA over the UDDS and HWFET cycles,
using the parameters in Table 2-13. The D/C kWh/100 miles values from the ALPHA simulation
were converted to effective A/C kWh/100 miles values by dividing by the appropriate factor of
0.87 (for BEVs) or 0.57 (for FCEVs). The resulting simulation values were compared to
certification values; the production weighted average of the difference is given in Table 2-22. A
scatter plot of the ALPHA versus certification values is shown in Figure 2-20.
Table 2-22: BEVs in the MY 2022 fleet - ALPHA kWh/100 miles values versus certification
Production weighted average kWh/100
miles
Production weighted std. dev. kWli/100
miles
kWh/100 miles
UDDS
+0.71 kWh/100
mile
1.68 kWli/100
miles
HW
+0.13 kWh/100
miles
1.48 kWh/100
miles
Combined
+0.45 kWh/100
miles
1.52 kWh/100
miles
+2.7%
Production weighted average percent
Production weighted std. dev. percent
+4.1%
6.6%
+ 1.2%
5.5%
5.9%
2-44
-------
o 15
u
15 20 25 30 35 40
Combined Cycle kWh/100 miles: Certification
45
Figure 2-20: BEVs in the MY 2022 fleet - ALPHA combined cycle kWh/100 miles values
versus certification kWh/100 miles where both BEV and FCEV vehicle model types were
simulated using a BEV model and bubble sizes reflect production volumes.
The production weighted average of the difference between ALPHA simulation and
certification consumption values for BEVs was 2.7 percent, or 0.45 kWh/100 miles. For FCEV
model types, the ALPHA simulation results were nearly identical, at 2.8 percent or 0.86
kWh/100 miles. Both FCEVs and BEVs are represented in Figure 2-20; however, the small
production volume of FCEVs make them difficult to distinguish in the figure.
2.4.8.8 Summary of ALPHA'S Ability to Simulate Entire Fleets
After validating the electrified vehicle models using test data from validation vehicles and
verifying that the models can simulate similar variant vehicles, the final validation step is to
verify that ALPHA can simulate entire vehicle fleets containing a wide range of different
powertrain technologies.
Table 2-23 summarizes production weighted average differences between the ALPHA
simulations and the certification values for each type of vehicle architecture across the MY 2022
fleet. The close match between ALPHA simulation and certification results verifies ALPHA
electrified vehicle models are fit for use in large-scale fleet-wide simulations. For example, the
production weighted average difference between ALPPIA simulations and certification values
across all the vehicle models in the entire MY 2022 fleet was 0.2 percent. An alternative method
of comparison, applying equal significance to each vehicle model type in the fleet, results in an
unweighted average difference of under 0.1 percent.
2-45
-------
Table 2-23: Summary of ALPHA Simulations vs Certification Values for MY 2022 Fleet
Yeliiele T> pe
Al.PII \
Model
hodueiion ueigliied
\\ erage Difference
ISclwcen ALPHA
Simulations and
Ccriificalion \ allies
I nwcigliied
\\ erage 1 )i fl'e re nee
Helwccu AI.IMIA
Simulations and
Ceilil'icalion \ allies
1:11111c \iy:o:: iieei
all
0 2"..
0 l"„
-0.1%
CV
-0.1%
Conventional Vehicles
Table 2-17
P0/P1 Mild Hybrids
P0
+2.9%
+3.5%
Table 2-18
PowerSplit & Series-Parallel
HEVs and PHEVS
(in charge-sustaining mode)
PS
-1.8%
-1.2%
Table 2-19
Parallel P2
HEVs and PHEVS
(in charge-sustaining mode)
P2
+5.0%
+2.1%
Table 2-20
PowerSplit and Parallel P2
PHEVs (both in charge-
depleting mode)
PS&P2
-0.8%
-2.0%
Table 2-21
Battery Electric Vehicles
BEV
+2.7%
-2.8%
Table 2-22
While on average, the ALPHA simulation values are close to the certification values in this
comparison of the entire MY 2022 fleet, there are some differences. There could be several
reasons for differences between simulation and actual certification test data due to variability in
the certification values, including:
• Certification values contain some variability because they are derived from measured
laboratory test data. The typical test-to-test variation of chassis dynamometer testing
can be +/-3 percent due to a variety of factors such as different drivers, measurement
equipment, fuel, and facilities.
• In addition, for hybrid and PHEV vehicles, certification testing for the MY 2022 fleet
followed the recommended practice in the 2010 revision of SAE Standard J1711 (SAE
2010). This standard allows a net energy change in the battery over a test of up to +/-1
percent of the total fuel energy consumed, while (unlike in the ALPHA simulations)
the measured CO2 value is not corrected. Depending on the average efficiency of the
engine and electrical system, this variation could alter fuel usage (and CO2 emissions)
by +1-2 percent or even +/-3 percent.
With large numbers of vehicle model types, these variations tend to cancel out (as seen in the
modeling of conventional vehicles, which contain by far the largest number of vehicle model
types). However, with smaller numbers of vehicle model types, as in the hybrid vehicle models,
the variability in certification values may influence the values seen in Table 2-23. This is
particularly apparent when just a few vehicles have relatively high production volumes, as seen
in the P2 simulations where only five vehicle model types account for half of the total production
2-46
-------
volume. In this case, if the unweighted average were used for comparison rather than the
production weighted average, the average difference is +2.1 percent rather than +5.0 percent.
The noticeable difference between production weighted and unweighted averages demonstrates
that most of this variation is caused by the few vehicle model types with the highest total
production volumes. A different mix of vehicle model types, or relative production volumes,
would likely result in a different average (due to the natural variations in certification test data as
discussed above) and thus the differences in Table 2-23 are within the expected range making the
ALPHA vehicle models suitable for their ultimate use for estimating current and future fleet-
wide emissions.
2.4.9 Peer-Reviewing ALPHA Electrified Models
After preparing ALPHA 3.0 to correctly simulate electrified vehicles, it was submitted to a
peer review process to examine its structure, operation, and simulation results to determine the
effectiveness of various vehicle technologies via simulation. The scope of the peer-review was
limited to the concepts and methodologies upon which the model relies and whether the model
can be expected to execute these algorithms correctly for the new electrified vehicle architectures
added to ALPHA. (ICF International 2022)
The peer review is centered on the five vehicle models detailed in Table 2-24.. The table
summarizes the configuration of each model provided for the peer review. The ETW and road
loads provided in the peer review were for a generalized mid-sized car and do not correspond to
any particular vehicle in the fleet. The Toyota Atkinson 2.5L engine was chosen based on the
base conventional vehicle and maintained for the electrified models to allow the CO2
performance of each model to be directly compared without the confounding factor of changing
engines. The transmission selected was a 6-speed automatic transmission (TRX12) and again
maintained for the P0 and P2 models. The PowerSplit, P2, and BEV models (including engine
and e-motor scaling) used for the peer review was the same as that described in the chapter
above.
Table 2-24: Details of ALPHA 3.0 models peer reviewed.
Model ETW
Conv.
P0
PS
PHEV
P2
PHEV
EV
3500
3500
3500
4250
Road
Load
(A, B, C
terms)
30. 0. 0.02
30. 0. 0.02
Engine Component
Name
engine_2018_T oyota_A25 AFKS
_2L5_Tier2.m (scaled to 150kw)
engine_2018_T ovota_A25 AFKS
_2L5_Tier2.m (scaled to 150kw)
3500 i 30.0.0.02 : engine_201 S Tovota A25AFKS
, _2L5_Tier2.m (scaled to 150kw)
30. 0. 0.02
30. 0. 0.02
engine_2018_T oyota_A25 AFKS
_2L5_Tier2.m (scaled to 150kw)
NA
Tran E-motor/EDU
s Component Name
TRX12 : NA
TRX12 emachine_2012 Hvundai
i _Sonata_8p5kW_270V_
BISG.m
Internal : MG1 and MG2:
to PS ! emachine_2010_Tovota_
model : Prius_60kW_650vjvlG2
EMOT.m
TRX12 : emachine_2011 Hvundai
i _Sonata_30kW_270V_E
MOT.m
9.5:1 ; emachine_IPM_150kW_
single 350V_EDU.m
speed
Engine and E-motor
scaling for
peer review vehicle
Engine: power scale
Engine: power scale
E-motor: power scaling
based on engine size (1 lkw
for peer review)
Engine: Displacement
scaling
E-motor: Power scaling
based on engine size (MG1
86kw MG2 106kw)
Engine: power scale
E-motor: power scaling
based on engine size (65kw)
E-motor: Power scaling
based on road load 150kw
2-47
-------
Each sub-model provided to the peer reviewers was validated against a combination of
internally and externally collected vehicle operational data while running the vehicle on a vehicle
dynamometer over the EPA city, highway, and US06 regulatory cycles (as described above in
Chapter 2.4.7). EPA's approach for validations was to use detailed 10Hz CAN and discrete
sensor vehicle benchmarking data to set up the model structure and tune it based on e-motor and
battery current and voltage, engine speed and load, battery SOC, etc., to generally achieve within
2-4 percent agreement with the CO2 measured over the city, highway, and US06 EPA regulatory
cycles, as shown in Table 2-8. Once the benchmarking test vehicle validation target was
achieved (generally after 3-6 months of work), the validated model was applied to variant
vehicles with the same powertrain design from the same manufacturer to achieve within 3-6
percent agreement on CO2 with EPA certification data over the combined FTP/Highway cycle.
Then the validated model was applied to the broader fleet of similar technology hybrid/BEV
vehicles to understand the variation in the hybrid/BEV performance (CO2 g/mile or kWh/mile)
across manufacturers.
2.4.9.1 Charge Questions for the ALPHA Peer Review
• Does EPA's overall approach to the stated purpose of the model (demonstrate
technology effectiveness for various fuel economy improvement technologies) and
attributes embody that purpose?
• What is the appropriateness and completeness of the overall model structure and its
components, such as:
• The breadth of component models/technologies compared to the current/future light-
duty fleet.
• The performance of each component model, including the reviewer's assessment of
the underlying equations and/or physical principles coded into that component.
• The input and output structures and how they interface with the model to obtain the
expected result, i.e., fuel/energy consumption and CO2 over the given driving cycles.
• The use of default or dynamically generated values to create reasonable models from
limited data sets.
• Does the ALPHA model use good engineering judgement to ensure robust and
expeditious program execution?
• Does the ALPHA model generate clear, complete, and accurate output/results (CO2
emissions or fuel efficiency output file)?
• Do you have any recommendations for specific improvements to the functioning or
the quality of the outputs of the model?
2-48
-------
2.4.9.2 Information Received from the Peer Review
General observations:
• The overall approach to the stated purpose of the model and attributes embodies the
goals as outlined.
• ALPHA model's structure and its components are sufficiently appropriate and
complete to achieve the stated purpose.
• The output results and output files are labeled appropriately and are relatively
complete. The results are generated across log files, console output, and figures, which
should provide users with a good amount of summary and detailed results.
In addition to the key themes and comments summarized here, reviewers provided numerous
other specific observations and recommendations for the ALPHA model in response to EPA's
individual charge questions, as documented in the peer review report (U.S. EPA 2023).
2.4.10 Estimating CO2 Emissions of Future Fleets
The variety of powertrain components (engines, transmissions, drive types, and electrification
architectures) modeled by ALPHA and described in Chapter 2.4.4 produces a range of
representative technology options available to manufacturers in the OMEGA model. In total, for
the final rule, the OMEGA model considers 542 unique technology combinations for each light-
duty vehicle: 360 conventional ICE packages, 120 mild hybrids, 26 hybrids, 33 PHEVs, and 3
BEV combinations.
To estimate CO2 emissions in future fleets, OMEGA uses a set of response surface equations
(RSEs) based on ALPHA simulation outputs (results). To define each RSE, technology packages
(consisting of specific combinations of components) were identified. Then, an ALPHA
simulation matrix run was created, sweeping vehicle parameters (road loads, vehicle power, and
weight) over a defined range so that the RSE could be applied to any vehicle. A unique ALPHA
simulation output is created for each vehicle parameter setting of the sweep. When applied over a
wide range of road loads and relative power levels, there are an infinite number of possible
combinations that can be characterized with response surface equations.
OMEGA also has several MDV technology options - 74 unique technology packages of
advanced ICE (gas and diesel) engines, range-extended P2P4 PHEVs, and BEVs - for
manufacturers to choose from in the medium-duty sector. Due to requirements for higher load
operation for extended periods of time for MDVs, many of the light-duty RSEs (for example:
power split hybrids, engines with cylinder deactivation, Atkinson engines, etc.) were not
considered candidate technologies for this use case.
2.4.10.1 Technology Packages used to create RSEs for OMEGA
ALPHA simulation outputs used to create the RSEs consisted of CO2 emissions or electric
energy consumption over each bag of the standard dynamometer cycles, FTP, HWFET, and
US06. Each RSE represents the results from simulating a single combination of technologies
known as a technology package across different combinations of vehicle parameters. The total
number of possible combinations of technologies are shown in Table 2-25, although not all the
2-49
-------
technology packages were used to make RSEs. EPA's website contains a copy of the ALPHA
output files that were generated to create the RSEs. Those output files contain data about each
technology used to create the vehicle technology packages for both light- and medium-duty
vehicles. The output files are located on EPA's ALPHA Tool webpage (U.S. EPA 2023a).
The components used to create each light-duty vehicle technology package are shown below
in Table 2-25.
• For conventional and mild hybrid (P0) vehicles, powertrain technology packages were
created for most combinations of engines and transmissions shown in Table 2-25.
SLA RSEs have front-wheel-drive, rear-wheel-drive, and all-wheel-drive variants.
HLA RSEs have rear-wheel-drive and all-wheel-drive variants. Finally, for each
combination, different packages were created (a) without stop-start technology, (b)
with stop-start technology, and (c) with a mild hybrid P0 technology.
For strong hybrid vehicles, technology packages were created (PowerSplit HEV, P2 HEV,
and SP-P2 PHEV models), using the dedicated hybrid engines in Table 2-25.
Additional technology packages were created for battery electric vehicles (BEVs) using the
electrification drive unit shown in Table 2-25.
2-50
-------
Table 2-25: List of Technology packages for LDV/LDT RSEs19.
l''.ll!iillc Niilllc
( nlllllMII'illilill
KM . < mil-
l)ri\r
KM '. < mil-
1 i;uiMiiiwiiiii
KM-'. < mil-
I'lci'liilKiiliiiii
KM-'. < imIi-
IM> I'm' ( niiM'iiiiiiiiiil \ chil lis:
1.?
2013 Chevrolet 2.5L Ecotec LCV Engine Reg E10 Fuel
GDI
(gas direct injection -SLA
only)
FWD
TRX10
(5-spd)
sso
(no stop-start)
2013 Chevrolet 2.5L Ecotec LCV Engine Reg E10 Fuel +
discrete CDA modifier
GDI DEAC D
(GDI+discrete CDA -SLA
only)
RWD
TRX11
(6-spd)
SSI
(stop-start)
2013 Chevrolet 2.5L Ecotec LCV Engine Reg E10 Fuel +
continuous CDA modifier
GDI DEAC C
(GDI+continuous CDA -SLA
only)
AWD
TRX12
(6-spd adv)
P0
(48V mild HEV)
2018 Toyota 2.5L A25A-FKS Engine Tier 3 Fuel
ATK
(Atkinson -SLA only)
TRX21
(8-spd)
2018 Toyota 2.5L A25A-FKS Engine Tier 3 Fuel +
continuous CDA modifier
ATKDEACC
(Atkinson+continuous CDA -
SLA only)
TRX22
(8-spd adv)
2013 Ford EcoBoost 1.6L Engine Tier 3 Fuel
TDS11
(TDSll:SLAonly)
2016 Honda 1.5L L15B7 Engine Tier 3 Fuel
TDS
(TDS12: SLA only)
2015 Ford EcoBoost 2.7L Engine Tier 3 Fuel
TDS
(HLA only)
2014 Chevrolet 4.3L EcoTec3 LV3 Engine LEVIII Fuel
GDI
(gas direct injection -HLA
only)
2014 Chevrolet 4.3L EcoTec3 LV3 Engine LEVIII Fuel +
discrete CDA
GDI DEAC D
(GDI+discrete CDA -HLA
only)
2014 Chevrolet 4.3L EcoTec3 LV3 Engine LEVIII Fuel +
continuous CDA modifier
GDI DEAC C
(GDI+continuous CDA -HLA
only)
Volvo 2.0L VEP LP Gen3 Miller Engine from 2020
Aachen Paper Octane Modified for Tier 3 Fuel
MILLER
(Miller)
2020 Ford 7.3L
PFI
(gas PFI large bore -HLA only)
RSI'.s Strong Hybrid Vehicles:
6
3
3
5
Toyota 2.5L TNGA Prototype Hybrid Engine from 2017
Vienna Paper Tier 3 Fuel
DHE
(Atkinson DHE for PS or P2
only-SLA)
FWD
(for PS & P2
only)
TRX12
(for P2 only)
PS
(PowerSplit HEV)
Geely 1,5L GHE Miller from 2020 Aachen Paper Tier 3
Fuel (PS, P2 only)
MILLER
(Miller DHE for PS or P2 only)
RWD
(for P2 only)
TRX22
(for P2 only)
P2
(P2 HEV)
Future 3.6L HLA Hybrid Concept Engine Tier 3 Fuel
MILLER
(Miller DHE for
SPP4 only -HLA only)
AWD
(for SPP4 &
P2 only)
TRXECVTF
(for PS only)
PS PHEV
(PS plug-in HEV)
2015 Ford EcoBoost 2.7L Engine Tier 3 Fuel
TDS
(for P2 & SPP4 -HLA only)
P2 PHEV
(P2 plug-in HEV)
2013 Chevrolet 2.5L Ecotec LCV Engine Reg E10 Fuel
GDI
(for P2 -SLA only)
SPP4HEV
(Series-parallel P4
HEV)
2020 Ford 7.3L
PFI
(gas PFI large bore for
P2 & SPP4 -HLA only)
SPP4PHEV
(Series-parallel P4
PHEV)
KM-'.* lor Ii;il1i-i\ l liTli ii \ i-hiili-^:
1
FWD
LDV/LDT BEV
EDU
RWD
AWD
19 Table note: combinations are not a factorial of all options (i.e., in some cases some combinations do not apply)
2-51
-------
A breakdown of components applied to candidate medium-duty vehicle technology
packages is shown below in Table 2-26. Medium-duty vehicles, especially pickup trucks, are
designed for extended high-load operation which requires a different subset of engines. EPA also
modeled the P2P4 PHEV strong hybrid option for medium-duty vehicles.
Table 2-26: List of Technology packages for medium-duty vans and pickups for RSEs.
Engine Name
Configuration
RSE Code
Drive
RSE
Code
Transmission
RSE Code
Electrification
RSE Code
RSEs for Conventional Vehicles:
3
2
5
2
2020 Ford 7.3L
PFI
(gas PFI large bore)
RWD
TRX10
(5-spd)
sso
(no stop-start)
Chevy Duramax 3.0L
DIESEL
(turbo diesel)
AWD
TRX11
(6-spd)
SSI
(stop-start)
2015 Ford EcoBoost 2.7L Engine Tier 3
Fuel
TDS
(TDS12:truck)
TRX12
(6-spd adv)
TRX21
(8-spd)
TRX22
(8-spd adv)
RSEs for Plug-In Hybrid Vehicles:
3
1
2
1
2020 Ford 7.3L
PFI (gas PFI large bore)
AWD
TRX12
(6-spd adv)
MDVP2P4REETPHEV
2015 Ford EcoBoost 2.7L Engine Tier 3
Fuel
TDS (TDS12:track)
TRX22
(8-spd adv)
Future 6.0L HLA Hybrid Concept Engine
Tier 3 Fuel
MILLER (Miller dedicated hybrid
engine)
RSE for Battery Electric Vehicles:
3
1
FWD
LDV/LDT BEV EDU
RWD
AWD
2.4.10.2 Vehicle Parameter Sweeps for each Technology Package
For each technology package, ALPHA 3.0 was used to provide data with which to construct
an RSE. To construct each RSE, a series of ALPHA simulations were performed using the same
technology package, but with different combinations of vehicle parameters, so that a single RSE
could be used to accurately characterize the performance of a range of different vehicles.
2.4.10.2.1 Swept Vehicle Parameters and Their Values
The vehicle parameters chosen for RSE development directly relate to vehicle parameters
used in certification dynamometer testing. These parameters were:
2-52
-------
• Equivalent test weight (ETW).
• Road load horsepower at 20 mph, calculated from the target coefficients (this value is
substantially dominated by rolling resistance losses).
• Road load horsepower at 60 mph, calculated from the target coefficients (this value is
substantially dominated by aerodynamic losses).
• Rated power of primary power source (engine or electric motor).
The calculated road loads at 20 mph and 60 mph were chosen to characterize vehicle losses
rather than the coefficient of rolling resistance and drag coefficient. The choice of road loads to
characterize losses ensures that all road load losses in base year vehicles are correctly accounted
for, while the choice of two widely separated speeds gives parameters combinations where road
loads dominated by rolling resistance (at low speed) and aero resistance (at high speed) can be
separately altered.
In choosing which combinations of parameters to simulate, combinations of parameters that
would not appear in the real fleet were avoided. For example, vehicles with very high ETW
would not also have low road loads, as both weight and road load are correlated to vehicle size.
With that in mind, rather than independently setting the value of each parameter, values of road
loads and engine/motor power were chosen to be proportional to ETW. In other words, the
parameters set were:
• Equivalent test weight (ETW)
• Road load horsepower at 20 mph / ETW (RLHP@20/ETW)
• Road load horsepower at 60 mph / ETW (RLHP@60/ETW)
• ETW / rated power (ETW/HP)
To determine the ranges of these parameters for light-duty RSEs, the values of ETW, target
coefficients, and rated power using the "2021 Test Car List" from EPA's publicly available
"Data on Cars used for Testing Fuel Economy" website (U.S. EPA 2022e). As shown in Figure
2-21, ETW values range from 2500 pounds to 7000 pounds, and the values of RLHP@20/ETW,
RLHP@60/ETW, and ETW/HP are roughly consistent across the span of ETW.
Additionally, the road loads at 20 mph and 60 mph are related, and thus so are the values of
RLHP@20/ETW and RLHP@60/ETW. When choosing parameters to simulate, only
combinations of RLHP@20/ETW and RLHP@60/ETW that were near the envelope of points
shown in Figure 2-21 were chosen.
2-53
-------
RLHP@20/ETW
RLHP@60/ETW
0.0014
0.0012
0.001
0.0008
0.0006
0.0004
0.0002
0
-tiliiii! i i:
4000 5000
ETW
ETW/Rated Power
35
30
25
20
15
10
5
•• •* t
¦'\i
0
9 t •
I ill
| 1
1 • • •
2000
3000
4000 5000
ETW
6000
7000
0.009
0.008
0.007
^ 0.006
o 0.005
UD
0.004
Q_
5 0.003
CC
0.002
0.001
0
0.008
0.007
0.006
S 0.005
* 0.004
0.003
2000
3000
4000 5000
ETW
v:-v
6000
7000
jpr
0.002
0.0003 0.0005 0.0007 0.0009 0.0011 0.0013
RLHP20/ETW
Figure 2-21: Relationships between vehicle parameters for the MY 2021 fleet.
For each RSE, di screte values of ETW, corresponding to test wei ght bins, were chosen
spanning from 3000 pounds to 10,000 pounds. For each ETW, a matrix of RLHP@20/ETW and
RLHP@60/ETW values were chosen which spanned the point cloud shown in Figure 2-21.
Combinations of ETW, RLHP@20/ETW, and RLHP@60/ETW that fell outside the range of
values in the fleet were eliminated from consideration.
Finally, for conventional vehicles, engine sizes were assigned so that the ETW/HP spanned
the values shown in Figure 2-21. The engine sizes for the LD applications were chosen from the
displacements listed in Table 2-27; for ML) applications only the V6 and V8 configurations were
used. For hybrid vehicles, the same set of engine sizes were used. For BEVs, electric motor sizes
were chosen in 50kW increments from 50 kW to 500 kW.
Table 2-27: Engine displacements used in RSE construction.
Engine configuration
Displacements
13
1.0L, 1.4L, 1.8L
14
1.6L, 2.2L, 2.8L
V6
2.5L 3.4L, 4.3L
V8
4.0L. 5.5L, 7.0L
2-54
-------
2.4.10.2.2 Values of Parameters Used for ALPHA Simulations
The ALPHA simulation uses various input parameters to set up the vehicle simulation. A
batch run was created for each RSE, using the powertrain configuration for that RSE. For each
combination of swept vehicle parameters, the powertrain was sized according to the engine
displacement or BEV EDU power specified.
The quadratic target coefficients (designated "A," "B," and "C" for the constant, linear, and
quadratic term, respectively) were determined using the values of RLHP@20 and RLHP@60. To
do so, the linear (B) coefficient was assumed to be 0.22 pounds/mph for LD applications and
0.96 for MD applications, representing the average value for that coefficient in the MY 2021
fleet. The constant (A) and quadratic (C) coefficients were then calculated to give the correct
values for RLHP@20 and RLHP@60. Although this method does constrain the range of
coefficients generated (as the linear term B is always the same), the resulting quadratic target
force curve accurately reflects a wide range of target curves. For example, recalculating the
target coefficients of the MY 2021 fleet with this methodology results in a difference between
the original and recalculated curves at 40 mph (midway between 20 and 60) of less than 4
percent for over 97 percent of the vehicles.
For hybrid vehicles, the electric machines and batteries used in the RSE ALPHA runs were
also scaled. For each hybrid configuration, the electric machine power was maintained as a
constant percentage of the engine power.
2.4.10.2.3 ALPHA Simulation Outputs for RSEs
The ALPHA runs for each RSE consisted of a series of simulations using a single powertrain,
but with different combinations of swept parameters. Each run simulated vehicle performance on
the FTP, HWFET, and US06 cycles.
For conventional and hybrid vehicles, the ALPHA 3.0 outputs consisted of CO2 emissions for
each phase of each simulated cycle. For electric vehicles, the ALPHA3 outputs consisted of
energy usage for each phase of each simulated cycle.
2.4.10.3 Transforming ALPHA Simulation Outputs into RSEs for OMEGA
The OMEGA model requires a complete set of full-vehicle efficiency simulations for the
entire vehicle fleet represented in this rule. To create this full set of simulations using a tool such
as the ALPHA model alone would require an unrealistic number of resources as millions of
simulation runs would be required to generate the resolution required to satisfy the requirements
of the OMEGA model.
To provide the necessary resolution for the OMEGA model while maintaining a realistic
number of ALPHA simulations, EPA implemented a peer reviewed (RTI International 2018)
Response Surface Methodology (RSM) (Kleijnen 2015). As described above, the inputs to the
RSM are a controlled set of ALPHA simulation outputs. The output from the RSM is a set of
Response Surface Equations (RSEs) suitable for the OMEGA model.
2-55
-------
2.4.10.3.1 Steps to Create an RSE from the RSM
For this example, 157 ALPHA model results were generated from FTP CO2 Bag 1
representing vehicles with GDI engines, Continuous DEAC, TRX21 Transmissions, FWD, and
Start-Stop.
Step 1: Compile the ALPHA model results. Table 2-28 contains a sample of the 157 results
from Bag 1 CO2 showing the 4 inputs and the CO2 output:
Table 2-28: Sample results.
RLHP20
RLHP60
HP_ETW
ETW
CO2
0.0005
0.003
0.032258
3250
204.8645
0.0005
0.0065
0.032258
3250
248.2313
0.00075
0.004
0.032258
4250
293.0054
0.00075
0.006
0.032258
3250
254.2941
0.00075
0.006
0.032258
4250
326.5969
0.001
0.005
0.032258
3250
252.8225
Step 2: Generate the RSE from a commercial or open-source product. EPA utilized the
popular open-source Python language (Foundation, Python 2023) including the sklearn library
(Foundation, sklearn 2023) to generate the RSE:
C02-RSE = (11.6954329620654+ RLHP20 * -19931.541254933 + RLHP60 * -
3972.87047794276 + HP_ETW * 491.637862683 + ETW * 3.59189981164081E-02 + RLHP20
* RLHP60 * -605147.450118866 + RLHP20 * HPETW * -114528.849261411 + RLHP20 *
ETW * 16.4326074843966 + RLHP60 * HP ETW * -14364.8208136198 + RLHP60 * ETW *
3.88487044872695 + HP_ETW * ETW * 9.75218337820661E-02 + RLHP20 * RLHP20 *
19630639.3025487 + RLHP60 * RLHP60 * 495588.873797923 + HP ETW * HP ETW *
700.465061662175 + ETW * ETW * -7.38489713757848E-09)
Step 3: Verify the output. Table 2-29 adds an additional column containing the results from
the RSE and Figure 2-22 shows all 157 ALPHA results vs 157 RSE results.
Table 2-29: Tabular results.
RLHP20
RLHP60
HPETW
ETW
CO2
CO2-RSE
0.0005
0.003
0.032258
3250
204.8645
203.0853
0.0005
0.0065
0.032258
3250
248.2313
247.1682
0.00075
0.004
0.032258
4250
293.0054
294.2893
0.00075
0.006
0.032258
3250
254.2941
252.7991
0.00075
0.006
0.032258
4250
326.5969
327.4422
0.001
0.005
0.032258
3250
252.8225
254.8888
2-56
-------
C02 - Bag 1
The validated RSE can now be used by the OMEGA model to generate a CO2 value (or
energy consumption rate, for BEVs) for any vehicle within the range of the controlled set of
ALPHA model simulation results. Utilizing the RSM results in a reduction of simulation and
storage resources by approximately a factor of 100.
2.4.11 Illustration of Vehicle-specific CQ2 Performance Compared to Footprint CO2
Targets
Because the standards are performance based, improvements in all vehicle and powertrain
technologies will contribute to a vehicle manufacturer's compliance.
Below, we show an example of projected vehicle-specific OMEGA compliance CO2 levels (in
g/mi), for the range of technology RSEs generated from ALPHA applied to the base year fleet in
MY 2032. Figure 2-23 and Figure 2-24 show the projected g/mi for MY 2032 cars and light
trucks, respectively, in comparison to the final standards footprint-based targets. These plots are
intended to illustrate the variety of technologies and vehicle compliance levels which exist in the
central case modeling results. Together, the combination of these vehicles represents one
possible compliance path for the industry, although the individual firms (and the industry as a
whole) may choose alternative combinations of technologies to comply with the standards.
2-57
-------
36 38 40 42 44 46 48 50 52 54 56 58 60
Footprint, sq ft
• BEV • HEV • PHEV • ICE 2032 tgt
Figure 2-23: MY 2032 Projected Vehicle Compliance Levels vs. Targets under Final
Standards - Cars
2-58
-------
400
36 38 40 42 44 46 48 SO 52 54 56 58 60 62 64 66 68 70 72 74 76
Footprint, sq ft
• BEV • HEV • PHEV • ICE 2032 tgt
Figure 2-24: MY 2032 Projected Vehicle Compliance Levels vs. Targets under Final
Standards - Trucks
As can be seen, PHEVs, as modeled20, are generally clustered around the individual vehicle
footprint-based targets, with some models above the targets and some below. PHEVs with
varying levels of all-electric range could have compliance CO2 levels higher or lower than these
projections. The figures also do not include the effect of sales weighting, nor do they preclude
any specific vehicle model from being part of a compliant fleet, as compliance is determined on
a fleetwide basis. For example, the distribution of car sales, by footprint, provided in Figure 2-25
shows that 97% of all car sales in MY 2032 are projected to have footprints of 52 square feet or
less. Figure 2-23 then supports EPA's judgment that, for at least the vast majority of the light
duty fleet, there exists a PHEV technology option that could meet each vehicle's individual
footprint-based target, and thus the industry could meet the final standards with only PHEVs.
® As we describe in Chapter 2.6.1.7 of this RIA, light-duty vehicle PHEV batteries were sized and modeled to
provide 40 miles of all-electric range capability over the US06 test.
2-59
-------
20%
10%
5%
I
£ bcr>«o
w jut .t. g jfc-'. fti 5
-------
2.5.2 Direct manufacturing costs
2.5.2.1 Battery cost modeling methodology
For this final rule, EPA updated the battery cost inputs to OMEGA. In the proposal analysis,
EPA had used ANL BatPaC 5.0 to develop base year (MY 2022) battery costs, expressed as a
cost per kWh as a function of battery gross energy capacity (kWh). Costs for future years were
then estimated by applying a cost reduction due to learning, based on cumulative Gigawatt-hours
(GWh) of battery production necessary to supply the number of BEVs that OMEGA placed in
the analysis fleet up to that analysis year.21 Finally, we applied additional cost reductions based
on our assessment of the future impact of the 45X cell and module production tax credit
provisions of the Inflation Reduction Act.
For this final rule analysis, we considered public comments that addressed the battery costs
and costing methodology used in the proposal. We also proceeded with additional study and
consideration of sources and planned new work that we described in the proposal. As a result, we
have made a number of updates to our battery cost estimates. These changes and the reasons for
them are discussed in more detail in Section IV.C.2 of the Preamble to the rulemaking. The
remainder of this chapter describes how battery costs were developed for the final rule analysis.
2.5.2.1.1 Battery sizing
Battery cost is an important input to the compliance analysis, which is performed using the
OMEGA model. When placing a PEV into the analysis fleet, the OMEGA model determines the
gross battery capacity (kWh) that the vehicle will require, based on its energy consumption
(kWh/mi), range target (mi), and other requirements. The direct manufacturing cost for a pack of
that capacity in a given year of the analysis is then estimated by use of a battery cost equation
(for 2023 through 2035) and a 1.5 percent reduction per year past 2035. The resulting battery
cost estimate becomes a term in the total direct manufacturing cost of the vehicle.
Determining the correct gross battery capacity is important because this determines the energy
consumption of the vehicle (which is in turn a result of battery weight) and also battery cost.
Gross battery capacity is generally a function of the desired electric driving range (a discussion
of OMEGA assumptions for BEV and PHEV ranges are provided in sections 2.6.1.6 and
2.6.1.7), the maximum fraction of gross battery capacity that is usable during a range test (usable
battery energy (UBE) fraction, in percent), and the on-road direct current (DC) energy
consumption of the vehicle. The driving range for BEVs is assumed to be 75 percent of the 55/45
2-cycle range which is consistent with higher volume BEVs in the market today and which we
expect to be more representative of future PEVs.22 Usable energy fraction is set at 0.95 for BEVs
and 0.75 for PHEVs.23 DC energy consumption is the average amount of on-road DC energy
required from the battery per mile driven on the relevant cycles. Note that this is not the same as
the energy consumption reported in the Fuel Economy guide, which includes charging losses
incurred between the grid power outlet and the battery. Charging losses are important for
calculating upstream emissions but must be excluded for battery sizing because battery capacity
21 The use of cumulative GWh of battery production as an input to the learning cost reduction was new to the
proposal for this rule.
22 While it varies by model, the current FE label range for BEVs is between 70-75 percent of the 2-cycle range.
23 HEV usable energy fraction is set at 0.50.
2-61
-------
for a given range is a function of DC energy consumption. DC energy consumption is derived
from a precomputed response surface generated by ALPHA results, using curb weight and other
vehicle attributes as inputs to the response surface. Curb weight is intimately tied to battery
capacity, via the effect of battery weight on total vehicle weight. This requires an iterative
process that OMEGA must perform in order to determine curb weight simultaneously with
arriving at the necessary battery size.
In the first step, OMEGA calculates the on-road DC energy consumption (Wh/mi) as a
function of vehicle parameters including curb weight (which includes a battery weight
determined by later steps). Next it estimates gross battery capacity, using DC energy
consumption, driving range, and usable energy fraction, according to the formula:
/Wh\ driving range in miles
(Wfl)Gr()ss Capacity ( ~) X — ~ —
\mi / oc energy consumption usable energy fraction
Next it estimates battery weight, using estimated gross capacity required (Watt-hours or Wh)
and a specific energy (Wh/kg) assigned to the given year of the analysis based on the 2023 ANL
battery costing study:
riw^ _ (WK)Gr0ss Capacity
ykg) Battery Weight ~ Wh/kg
Finally, it returns to the first step with the new battery weight, until the battery weight stops
changing (converges).
2.5.2.1.2 Battery costs for 2023 to 2035
To begin estimating cost for a pack in a given year of the analysis, OMEGA first requires
battery cost to be input as a pack-level cost per kWh, as a function of its gross capacity in kWh
and the model year.
As described in Preamble IV.C.2, input cost functions for Nickel-Manganese (Ni/Mn,
including NMC) and Iron phosphate (LFP) cathode chemistries were developed by Argonne
National Laboratory for reference by EPA and NHTSA in their respective rulemakings (ANL
2024). This work by ANL used the latest version of the BatPaC model and included a set of
custom inputs representing expected mineral prices over time (as forecast by Benchmark Mineral
Intelligence), and advancements in chemistry and manufacturing according to a technology
roadmap for 2023 to 2035 developed by ANL battery scientists. The resulting equations are
correlations of a larger data set across many battery specifications and capacities. They express
pack cost as a function of kWh and model year. The battery pack cost correlation equations, as
shown below, employ a different set of coefficients for each of PEV Ni/Mn, PEV LFP, and HEV
batteries; the variables x and y correspond to pack energy and model year, respectively.
g
%/kWh = A + — — D(y — 2023)eE^~2023)
X
2-62
-------
Table 2-30: ANL cost equation coefficients for HEV and PEV batteries ($50/hr labor).
Battery pack chemistry and application
A
B
C
D
E
Ni/Mn for HEV (5kWh or less)
122.9
509.6
0.7649
4.443
0.01018
Ni/Mn for PEV
128.9
1480
1.164
5.278
-0.0129
LFP for PEV
120.6
1535
1.148
10.04
-0.08346
Figure 2-26 and Figure 2-27 show examples of how these equations characterize the direct
manufacturing cost for PEV and HEV batteries from 2023 to 2035, using an example of a 100
kWh and 1.5 kWh pack, respectively.
160
140
120
100
JZ
5
80
60
40
20
0
2023 2025 2027 2029 2031 2033 2035
Year
Figure 2-26: NMC and LFP PEV battery pack direct manufacturing cost (100 kWh
example).
600
500
400
-C
| 300
200
100
0
2023 2025 2027 2029 2031 2033 2035
Year
Figure 2-27: Ni/Mn HEV battery pack direct manufacturing cost (1.5 kWh example).
Specific energy (Wh/kg) of a battery pack has an indirect effect on battery cost in its
determination of the weight of the battery and hence the amount of battery energy necessary to
2-63
-------
meet a given range target. The ANL work to project future costs of battery packs also extended
to the specific energy of the pack, which is reported by BatPaC in addition to the cost and other
results. EPA is using the equation and coefficients in the table below.
Wh/kg = 1000
B
A+-r- D(y - 2023)eE(y~2023)
Table 2-31: ANL cost equation coefficients for HEV and PEV batteries.
Battery pack chemistry and
application
A
B
C
D
E
Ni/Mn for HEV (5kWh or less)
5.22
13.398
0.941
0.359
-0.081
Ni/Mn for PEV
5.266
20.6
1.129
0.3537
-0.08158
LFP for PEV
6.602
26.62
1.016
0.3597
-0.09757
More information about the inputs and assumptions for the ANL study are available in the
ANL report (ANL 2024).
For PEV batteries, the input costs to OMEGA are a weighted average of the ANL cost
equations for Ni/Mn and LFP cathode chemistries. Specific energy is weighted in the same way.
To determine an appropriate share of LFP vs. NMC, we worked in conjunction with NHTSA.
Each agency consulted subscription forecasts of future LFP market share, EPA referring to
available data from Benchmark Mineral Intelligence (BMI) and NHTSA referring to available
data from Rho Motion. EPA's BMI access included a forecast of U.S. and North American
cathode powder production, and NHTSA's Rho Motion access included a forecast of battery
chemistry usage in future U.S. BEVs and PHEVs. EPA and NHTSA consider cathode powder
production and forecast battery usage in vehicles to represent two relevant perspectives on the
matter of future representation of LFP in the U.S. vehicle battery market. Both forecasts describe
a similar scenario in which LFP share rises for several years and levels off over time. We worked
with NHTSA to arrive at a generalized year by year estimate of LFP share potential by averaging
the forecasts of the two sources and smoothing the resultant curve.
Table 2-32 shows the result, indicating a gradual rise in LFP share from 2023 to 2029,
leveling off afterward at just under 20 percent. Representing gradual growth in share of LFP in
this way is consistent with public comments contending that rapid growth in battery demand may
place pressure on supplies and prices of critical minerals, as LFP is significantly less exposed to
these risks due to the absence of cobalt, nickel, and manganese.
2-64
-------
Table 2-32: U.S. PEV battery cathode chemistry market projections, 2023 to 2035.
Year
LFP share
Ni/Mn share
2023
8%
92%
2024
10%
90%
2025
16%
84%
2026
17%
83%
2027
18%
82%
2028
19%
81%
2029
19%
81%
2030
19%
81%
203 1
19%
81%
2032
19%
81%
2033
19%
81%
2034
19%
81%
2035
19%
81%
As described in Preamble IV.C.2, to represent the possible effect of temporarily elevated
mineral prices on battery cost in the short term, for the years 2023 through 2025 we kept battery
costs at the same level as for 2023. We note that since the 2021 rule, it has become increasingly
clear that mineral costs have risen sufficiently to interrupt historical trends of continuous battery
cost reduction. For example, the BNEF 2021 battery price survey indicated that the pace of
reduction had slowed considerably and predicted that costs may not reach $100/kWh (at pack
level) until 2024. The next year, the BNEF 2022 battery price survey reported that costs had
increased by 7 percent over the previous year, for the first time since the survey has been
conducted. At that time, elevated prices appeared likely to persist for some amount of time due to
increased mineral prices.
In late November 2023, the 2023 BNEF battery price survey indicated that battery prices had
resumed their downward trend, dropping 14 percent from 2022. EPA considered whether or not
this development warranted removal of the 2023 to 2025 cost plateau for the FRM analysis. In
consultation with DOE, ANL, and NHTSA, we decided to keep the plateau due to the prospect
for continued uncertainty in the short term regarding mineral prices. Since that decision was
made, we note that leading analyst firms continue to predict a sustained reduction in most
mineral costs through at least 2027, further suggesting that the plateau represents a conservative
assumption.
Also as described in Preamble IV.C.2, because the ANL cost equations define costs for future
years from 2023 through 2035, we discontinued our practice from the proposal of deriving these
costs by means of a learning curve equation.
More discussion of the OMEGA model and the OMEGA results can be found in Preamble
IV.C and elsewhere in this RIA. For additional discussion of the battery costing method and
2-65
-------
sources considered, and a comparison between the battery costs derived in this analysis and those
of the proposal, see Preamble Section IV.C.2.
2.5.2.1.3 Battery costs for 2036 to 2055
The ANL equations are limited to projecting battery pack costs from 2023 to 2035. Beyond
2035, such long-term estimates are by their nature more uncertain than shorter-term estimates
and are difficult to model by assuming a specific technology roadmap. Often, analysts model
costs over the longer term by applying an annual percentage cost reduction rate. We adopted this
approach for the years after 2035, applying a 1.5 percent per year reduction, which results in a
decline to about $60/kWh by 2055. Note that we apply this 1.5 percent per year reduction
regardless of whether it is the No Action or Action policy scenario being run in OMEGA. For
more discussion of the rationale for this learning rate, see Preamble IV.C.2.
2.5.2.1.4 Battery cost reductions due to Inflation Reduction Act
To reflect the anticipated effect of the Inflation Reduction Act on battery costs to
manufacturers, we applied a further cost reduction based on the Section 45X Advanced
Manufacturing Production Tax Credit. This provision of the IRA provides a $35 per kWh tax
credit for manufacturers of battery cells, and an additional $10 per kWh for manufacturers of
battery modules, as well as a credit equal to 10 percent of the manufacturing cost of electrode
active materials and another 10 percent for the manufacturing cost of critical minerals (all
applicable to manufacture in the United States). The credits, with the exception of the critical
minerals credit, phase out from 2030 to 2032.
Since the proposal, work was conducted by DOE and ANL to assess the rate of growth and
likely operating capacities of North American battery manufacturing plants announced or under
construction and relate this data to the likely access to 45X credits across the PEV fleet. This
work is cited and discussed in Section IV.C.2 of the preamble. For the purpose of modeling, we
conservatively estimated an average credit amount per kWh that can be realized across the
industry as a whole in each year from now until 2032. Section 2.6.8 of the RIA shows the
estimated yearly average credit amounts for 45X that were input to OMEGA. For a full
discussion of how EPA determined which 45X credits would be modeled and the values used in
OMEGA, see the preamble at section IV.C.2.
Figure 2-28 shows an example of the resulting effect on average pack direct manufacturing
costs (DMC) generated by OMEGA in the central case of the proposal, after application of the
45X credit. The 45X cell and module credits per kWh were applied not to the direct
manufacturing cost per kWh, but to the marked-up cost per kWh (that is, after multiplying the
direct manufacturing cost by the 1.5 retail price equivalent (RPE)). Because RPE is meant to be a
multiplier against the direct manufacturing cost, and the 45X credit does not reduce the actual
direct manufacturing cost at the factory but only compensates the cost after the fact, it is most
appropriate to apply the 45X credit to the marked-up cost. The 45X cell and module credits per
kWh were applied by first marking up the direct manufacturing cost by the 1.5 RPE factor to
determine the indirect cost (i.e., 50 percent of the manufacturing cost), then deducting the credit
amount from the marked-up cost to create a post-credit marked-up cost. The post-credit direct
manufacturing cost would then become the post-credit marked-up cost minus the indirect cost.
2-66
-------
w
tf>
o
u
o
to
a.
200
180
160
140
120
100
80
60
40
20
0
2022
2024
2026
2028 2030
Year
2032
2034
2036
¦ FRM w/RPE
• • FRM DMC
NPRM w/RPE
NPRM DMC
Figure 2-28: Volume weighted average pack direct manufacturing cost and marked-up cost
per kWh after application of 45X credit.
EPA did not apply a further cost reduction to represent the 10 percent electrode active
material or critical mineral production credits that are also available under 45X. Rationale for
this decision may be found in Section IV.C.2 of the preamble. The implementation of battery
costs as OMEGA inputs is described in Chapter 2.6.1.4 of this RIA for HEYs and Chapter
2.6.1.5 of this RIA for BEVs and PHEVs.
2.5.2.2 Non-Battery Cost Approach
For non-battery powertrain costs that were used in the proposed rule, EPA referred to a
variety of industry and academic sources, focusing primarily on teardowns of components and
vehicles conducted by leading engineering firms. For the final rule, we have largely updated
these costs with information from the FEV teardown and related work performed by FEY to
review the costs we had used in the proposal and develop scaling factors suitable for estimating
costs for different vehicle types across the analysis. The latter of this work is discussed in the
next section. In this section we primarily review the types of sources that EPA typically relies
upon for determining direct manufacturing costs in this and previous analyses.
The equations used in OMEGA for the non-battery electrified vehicle cost estimates used in
this final rule analysis may be found in Chapter 2.6 of this RIA.
While EPA relies on a variety of sources to establish direct manufacturing costs for vehicle
components, we have long considered teardown studies to be the preferred means for doing so.
EPA has previously estimated non-battery costs by commissioning teardown studies of early-
2-67
-------
stage EV technologies. For past rulemakings, we contracted with FEV North America to conduct
teardowns of electrified vehicle components that were available at the time in order to establish
direct manufacturing costs under high-volume production in a rigorous and transparent way.
Since then, the ongoing evolution of electrified vehicles has led to the emergence of improved
or entirely new components that benefit from improved manufacturing efficiencies, component
integration, and platform optimization. These developments call for ongoing updates to the way
we characterize and quantify vehicle costs. For example, in some cases, specific components that
we costed in the past may have become integrated with other components, or their design has
changed so they use less costly materials or can be manufactured in a more efficient way. The
vehicle platforms that incorporate these components may also have changed to optimize their
integration with the rest of the vehicle.
Third-party teardowns of vehicles and components have become more widely available from
a number of engineering firms, and at times we have consulted these sources to augment our
commissioned work. For example, to develop the costs that were used in the proposal to this
rule, EPA acquired several commercial teardown reports to inform these changes and to
represent the manufacturing cost of today's electrified vehicle components as accurately as
possible (Munro and Associates 2020a) (Munro and Associates 2021) (Munro and Associates
2020b) (Munro and Associates 2016) (Munro and Associates 2020c) (Munro and Associates
2018). EPA then worked jointly with CARB to analyze the data in these reports.
As previously described in the proposal to this rule, we also commissioned a new full-vehicle
teardown of two vehicles with FEV North America and have now completed this work (FEV
Consulting Inc. 2022). FEV tore down a 2021 Volkswagen ID.4 BEV and a 2021 Volkswagen
Tiguan, an ICE vehicle relatively equivalent to the ID.4 in size and function. This work has also
undergone peer review.
This project was initiated in part due to the realization that platform optimization is likely to
affect a variety of cost comparisons between ICE and BEV vehicles. For example, platform
optimization, particularly for BEVs, could potentially lead to differences in indirect costs that are
experienced by the manufacturer (such as assembly cost, certification cost, and calibration cost).
We also considered that the differences between a platform-optimized BEV and a platform-
optimized ICE vehicle might call for an absolute costing approach, instead of assuming that a
BEV can be costed as an ICE vehicle with ICE components removed and BEV components
added.
The study was therefore designed not only to provide an additional source for vehicle
component cost data, but also to inform several issues that are commonly cited with regard to the
difference in cost between conventional and battery electric vehicles. Because ICE vehicles and
BEVs are likely to be built on different dedicated platforms, the study was designed as a ground-
up study for which a complete costed bill of materials would be developed for every component
of each vehicle, including structural and other non-powertrain components. This would support
our intention to move to a costing regime based on absolute vehicle costs instead of relative or
incremental costs (as previously described in 2.5.1), and to allow comparisons to be made on a
vehicle-system basis to identify significant potential cost efficiencies attributable to a dedicated
BEV platform. FEV was also asked to comment on potential differences in indirect costs for
BEV design, certification, and calibration that might become apparent on a close inspection of
the components, their system integration, and their assembly characteristics. We also specified
2-68
-------
that a detailed labor assessment be performed for each component, in order to shed light on
differences in amount and type of assembly labor required for production. An additional task
under this work assignment was to evaluate the non-battery HEV and PEV costs EPA has
described under RIA Section 2.6.1, with respect to the cost values used and the method of
scaling these costs across different vehicle performance characteristics and vehicle classes.
Delivery of the primary teardown study results by FEV occurred in February 2023 and a peer
review was completed in mid-2023.
An additional FEV review of EPA's non-battery costs and scaling recommendations was
made available in a memo to the docket entitled "EV Non-Battery Cost Review by FEV."
Additional reports and presentations on this work have since been placed in the docket (Cherry
and Sherwood 2023). Through this work, we have updated our powertrain cost inputs for ICE
vehicles as well as non-battery costs for electrified vehicles.
2.5.2.3 Powertrain Cost Scaling Exercise for ICE, HEV, PHEV, and all Electrified
Vehicle Non-Battery Costs
As noted in a docket memorandum, EPA contracted FEV to conduct a scaling exercise to
develop up-to-date powertrain cost curves that could be used as inputs to OMEGA. (Cherry and
Sherwood 2023) As a result of that effort, we have updated nearly all of our powertrain costs,
including the non-battery technologies used in BEV, PHEV, and HEV powertrains. Chapter
2.6.1 of this RIA presents all of those new powertrain cost curves. In general, the updated cost
curves result in lower powertrain costs for nearly all powertrain technologies, with ICE
powertrain costs being reduced somewhat more than those for electrified powertrains. As a
result, the incremental costs when moving from ICE-only to any electrified powertrain have
increased somewhat since the NPRM. Importantly, the scaling effort provided ICE, HEV,
PHEV, and BEV powertrain costs that were generated using the same methodology.
2.5.3 Approach to cost reduction through manufacturer learning
Within OMEGA, learning factors are applied to technology costs as shown in Table 2-33.
These learning factors were generated with the expectation that learning on ICE body structure
technologies would slow, relative to their traditional rates, in favor of a focus on HEV, PHEV,
and BEV technologies.
Importantly, the learning factors shown are multiplicative scaling factors indexed to 2022.
The costs presented below in Chapter 2.6 represent first year costs and the learning factors
shown in Table 2-33 are applied to those first-year costs to arrive at costs for subsequent years.
Learning applied to BEV, PHEV, and HEV battery costs in MY 2036 and later was done as
described in Chapter 2.5.2.1.3.
2-69
-------
Table 2-33: Learning Factors Applied in OMEGA, Indexed to 2022a.
Model Year Tniditioiuil IC'IC l'owei'tr.iin & All Glider C osts III.Y, 1'III.Y & I5ICY i\on-l5sitteiy Costs
mid iNon-Tniditioiuil IC'IC l'owei'tr.iin Costs
2022
1.00
1.00
2023
1.00
0.86
2024
0.99
0.79
2025
0.99
0.74
2026
0.99
0.70
2027
0.98
0.67
2028 i
0.98
0.65
2029
0.98
0.63
2030
0.97
0.61
2031
0.97
1 0.59
2032
0.97
0.58
2033
0.96
0.57
2034
0.96
0.56
2035
0.96
] 0.55
2036
0.95
0.54
2037
0.95
0.53
2038
0.95
0.52
2039
' 0.95
0.51
2040
0.94
0.51
2041
0.94
0.50
2042
0.94
0.50
2043
0.94
0.49
2044
0.93
0.49
2045
0.93""
0.48
2046
0.93
0.48
2047
0.92
0.47
2048 !
0.92
0.47
2049
0.92
0.46
2050
0.92
: 0.46
2051
0.92
0.45
2052
0.91
0.45
2053
0.91
0.45
2054
0.91
; 0.44
2055
0.91
0.44
aLearning factors are indexed to 2022. Note that traditional ICE powertrain technologies include things like the low voltage battery and its
harness even if equipped on a BEV, PHEV, or HEV.
2.5.4 Indirect costs
To produce a unit of output, vehicle manufacturers incur direct and indirect costs. Direct costs
include cost of materials and labor costs. Indirect costs are all the costs associated with
producing the unit of output that are not direct costs - for example, they may be related to
production (such as research and development, R&D), corporate operations (such as salaries,
pensions, and health care costs for corporate staff), or selling (such as transportation, dealer
support, and marketing). Indirect costs are generally recovered by allocating a share of the costs
to each unit of goods sold. Although it is possible to account for direct costs allocated to each
unit of goods sold, it is more challenging to account for indirect costs allocated to a unit of goods
sold. To make a cost analysis process more feasible, markup factors, which relate total indirect
costs to total direct costs, have been developed. These factors are often referred to as a retail
price equivalent (RPE) markup.
EPA has frequently used cost multipliers to predict the resultant impact on costs associated
with manufacturers' responses to regulatory requirements. The best approach, if it were possible,
to determining the impact of changes in direct manufacturing costs on a manufacturer's indirect
2-70
-------
costs would be to estimate the cost impact on each indirect cost element. However, doing this
within the constraints of an agency's time or budget is not always feasible, or the technical,
financial, and accounting information to carry out such an analysis may simply be unavailable.
The RPE multiplier, or RPE markup factor, is based on an examination of historical financial
data contained in 10-K reports filed by manufacturers with the Securities and Exchange
Commission. It represents the ratio between the retail price of motor vehicles and the direct costs
of all activities that manufacturers engage in. The RPE markup provides, at an aggregate level,
the relative shares of revenues (Revenue = Direct Costs + Indirect Costs + Net Income) to direct
manufacturing costs as shown in Table 2-34. Using the RPE markup implicitly assumes that
incremental changes in direct manufacturing costs produce common incremental changes in all
indirect cost contributors as well as net income. However, a concern in using the RPE markup in
cost analysis for new technologies added in response to regulatory requirements is that the
indirect costs of vehicle modifications are not likely to be the same for different technologies.
For example, less complex technologies could require fewer R&D efforts or less warranty
coverage than more complex technologies. In addition, some simple technological adjustments
may, for example, have no effect on the number of corporate personnel and the indirect costs
attributable to those personnel. The use of a single RPE markup, with its assumption that all
technologies have the same proportion of indirect costs, is likely to overestimate the costs of less
complex technologies and underestimate the costs of more complex technologies.
Table 2-34: Retail Price Equivalent Factors in the Heavy-Duty and Light-Duty Industries
(Rogozhin 2009).
Cost Contributor
Contribution to Cost
Direct manufacturing cost
1.0
Warranty
0.03
R&D
0.05
Other (administrative, retirement, health, etc.)
0.36
Profit (cost of capital)
0.06
Retail price equivalent
1.50
To address this concern, modified multipliers were developed by EPA, working with a
contractor, for use in past EPA rulemakings. (Rogozhin 2009) Those modified multipliers were
referred to Indirect Cost Multipliers, or ICMs, and EPA applied low magnitude ICMs (i.e., the 1.5 RPE) to high
complexity technologies. This way, we could analyze the possible pathways toward compliance
with GHG regulations via, for example, application of many low complexity technologies versus
few high complexity technologies. In other words, we could weigh one technology against
another in a more finely tuned way.
The ICM approach served us well when dealing with incremental technology applications and
incremental costs for those technologies as was done in the 2010 and 2012 final rules (75 FR
25324 2010, 77 FR 62624 2012). However, as noted above, we no longer use that approach to
estimate compliance pathways. In contrast, since we now consider the whole vehicle and its total
cost and performance toward compliance, we no longer need the fine tuning of one technology
versus another that the ICM approach provided. As a result, for this analysis, we are using the
full RPE markup as the indirect cost markup as we did in our 2021 final rule (86 FR 74434
2021).
2-71
-------
2.6 Inputs and Assumptions for Compliance Modeling
The following subsections provide details on the technology cost methodology, including
equations and values applied in the OMEGA input files, for the final rulemaking.
2.6.1 Powertrain Costs
Table 2-35 shows the engine, exhaust system and, fuel system costs used for any vehicle
equipped with an internal combustion engine. Table 2-38 shows the driveline costs used for all
vehicles. Table 2-40 shows additional electrified driveline costs used for HEVs, PHEVs, and
BEVs. Note that, with the exception of the high voltage battery costs summarized in Figure 2-28
and the gasoline particulate filter (GPF) cost shown in Table 2-35, any cost denoted as having a
2022 cost basis is from the work discussed in Chapter 2.5.2.3. The high voltage battery costs,
which are denoted as having a 2022 dollar basis, are from the DOE/ANL work discussed in
Chapter 2.5.2.1. The GPF cost curve was developed by EPA and is discussed in Chapter 3.2.2.
Costs denoted with a dollar basis other than 2022 are from prior EPA work.
2-72
-------
2.6.1.1 Engine, exhaust and fuel system costs
Table 2-35: Engine, exhaust, and fuel system costs used for ICE, HEV and PHEV.
Item
; Engine
; Config
: Body
; Style
; Value
cylinder deac FC
i I
j -
; 169 * Markup
cylinder deac FC
r v
-
; 221 * Markup
cylinder deac PD
-
-
[ (-1.0603 * CYL ** 2 + 28.92 * CYL - 8.6935) * Markup
direct injection
i
-
; 212 * Markup
direct injection
: v"
-
j 319* Markup
Cooled EGR
-
-
! 100 * Markup
Engine block
I'i
-
! (324.71*LITERS + 680) * Markup
Engine block
rv
-
(246,87*LITERS + 1215) * Markup
Diesel engine block
-
-
f 1.5
scaler
Turbo Charger
f i
-
: 429 * Markup
Turbo Charger
! V
-
' 756 * Markup
VVT '
I
-
: 100 * Markup
YYT
; V
-
i 200 * Markup
Atkinson
-
-
; (4.907 * CYL ** 2 - 29.957 * CYL + 130.18) * Markup
Non-EAS
I
-
I 250 * Markup
Non-EAS
; v'""'
-
i 350 * Markup
Fuel storage
-
; pickup
! 571.45 * Markup
Fuel storage
-
; sedan
; 467.55 * Markup
Fuel storage
-
j CUV SUV
| 519.5* Markup
TWC substrate
-
-
(6.108 * LITERS * TWC SWEPT VOLUME + 1.95456) * Markup
TWC washcoat
-
-
i (5.09* LITERS* TWC SWEPT VOLUME) * Markup
TWC canning
-
-
: (2.4432 * LITERS * TWCJSWEPTVOLUME) * Markup
TWC swept volume
-
-
i 1.2 multiplier applied to engine displacement
TWC Pt grams/liter
-
-
1 0
TWC Pd grams/liter
-
-
| 2
TWC Rh grams/liter
-
-
i 0.11
TWC PGM
: (PT GRAMS PER LITER TWC * LITERS * TWC SWEPT VOLUME*
: PT USD PER OZ * OZ PER GRAM + PD GRAMS PER LITER TWC *
. LITERS * TWC SWEPT VOLUME * PD USD PER OZ * OZ PER GRAM
i RH GRAMS PER LITER TWC * LITERS * TWC SWEPT VOLUME *
i RH USD PER OZ * OZ PER GRAM) * Markup
Troy oz/gram
-
-
: 0.0322
PT USD" PER OZ
-
-
! 1030
PD USD PER OZ
-
-
: 2331
RH USD PER OZ
-
-
i 17981
Gasoline particulate
-
-
: (42.269 * LITERS + 22.213) * Markup
filter
Diesel EAS
-
-
700 * LITERS * Markup
Dollar
Basis
2022
2022
2006
2022
2022
2022
2022
2022
2022
2022
2022
2022
2010
2022
2022
2022
2022
'2022
2012
2012
2012
2022
2020
Markup
1.5
FC=full continuous cylinder deactivation; PD=partial discrete cylinder deactivation; LITERS=engine displacement in liters; GEARS=the
number of forward gears; CYL=number of cylinders; I=Inline configuration; V=V configuration; FWD=front wheel drive; RWD=rear wheel
drive; AWD=all wheel drive; EGR=Exhaust gas recirculation; VVT=Variable valve timing; EAS=exhaust aftertreatment system; Non-
EAS=exhaust system excluding the exhaust aftertreatment system; HVAC=heating, ventilation, air conditioning; TWC=three-way catalyst;
Pt=Platinum; Pd=Palladium; Rh=Rhodium.
2-73
-------
2.6.1.1.1 Cylinder Deactivation
The costs of partial discrete cylinder deactivation are based on past EPA analyses as shown in
Table 2-36.
Table 2-36: Cylinder Deactivation Costs used to generate a partial discrete cost curve for
OMEGA.
Item
DMC
Dollar Basis
Partial discrete. 3-cvlinder engine
76
2006
Partial discrete. 4-cylinder engine
76
2006
Partial discrete. 6-cylinder engine
136
2006
Partial discrete. 8-cylinder engine
152
2006
Using these values, the following cost curve was generated for use in OMEGA.
DeacPD = (-1.0603 x CYL2 + 28.92 x CYL - 8.6935) x Markup
Where,
CYL = the number of cylinders on the engine
Markup = the markup to cover indirect costs
The costs for full continuous cylinder deactivation are based on the FEV cost scaling work
and are calculated as shown in Table 2-35.
2.6.1.1.2 Atkinson Cycle Engine
The costs for Atkinson cycle engine (ATK) are based on costs used in past EPA analyses.
Those costs are shown in Table 2-37.
Table 2-37: Atkinson Cycle Engine Costs used to generate a cost curve for OMEGA.
Item DMC Dollar Basis
ATK. 3-cylinderengine 86 2010
ATK. 4-cvlindcr engine 86 2010
ATK. 6-cvlindcr engine 129 2010
ATK. 8-cvlindcr engine 204 2010
Using these values, the following cost curves were generated for use in OMEGA.
AtkinsonCycleEngine = (4.907 x CYL2 — 29.957 x CYL + 130.18) x Markup
Where,
CYL = the number of cylinders on the engine
Markup = the markup to cover indirect costs
2-74
-------
2.6.1.1.3 Gasoline Particulate Filter
The gasoline particulate filter (GPF) cost is a new cost for this analysis and has been updated
since the NPRM. This is described in detail in Chapter 3.2.2. The cost curve used in OMEGA is
shown below. Note that the GPF costs are applied only in the action case if GPFs are expected
for compliance with new gasoline PM standards.
GPF = (42.269 X LITERS + 22.213) X Markup
Where,
LITERS = the engine displacement in liters
Markup = the markup to cover indirect costs
2.6.1.1.4 Three-way Catalyst
OMEGA's three-way catalyst (TWC) costs are based largely on the approach used in the
light-duty highway Tier 3 criteria pollutant rule. In the Tier 3 rule, EPA presented cost curves to
estimate costs for the individual components of a TWC: the substrate; the washcoat; the canning;
and the platinum group metals (PGM, consisting of platinum (Pt), palladium (Pd) and rhodium
(Rh)). The four cost curves are shown below.
TWCsubstrate = (6.108 x 1.2 x LITERS + 1.955) x Markup
TWCwashcoat = (5.09 x 1.2 x LITERS) x Markup
TWCcanning = (2.4432 x 1.2 x LITERS) x Markup
, , TrovOi
TWCpGM = (Ptgpl X Pt^/Tr0yQZ + Pdgpl X Pd^^gyQz 4" Rhgpl X A ^ $/7Y (jy £)Z J X 1.2 X LITERS X gram
Where,
LITERS = the engine displacement in liters
1.2 = factor to account for the swept volume of the TWC (i.e., total TWC volume is 1.2x
engine displacement
Ptgpi = Platinum grams/liter, set to 0 in this analysis
Pdgpi = Palladium grams/liter, set to 2 in this analysis
Rhgpi = Rhodium grams/liter, set to 0.11 in this analysis
Pt%nmyOz = Platinum cost per Troy ounce, set to $1,030 in this analysis
2-75
-------
iV$/TroyOz = Palladium cost per Troy ounce, set to $2,331 in this analysis
i?//$/TroyOz = Rhodium cost per Troy ounce, set to $17,981 in this analysis
TroyOz = Troy ounces
TroyOz/gram = 0.0322, or 31.1 grams per Troy Oz
Markup = the markup to cover indirect costs
2.6.1.1.5 Diesel Exhaust Aftertreatment System
OMEGA's diesel exhaust aftertreatment system (diesel EAS) costs are structured for
consistency with the recent heavy-duty final rule (88 FR 4296 2023). The cost curve is as shown
below.
Diesel EAS = (700 x LITERS) x Markup
Where,
Diesel EAS = diesel exhaust aftertreatment system cost
LITERS = the engine displacement in liters
Markup = the markup to cover indirect costs
2-76
-------
2.6.1.2 Driveline System Costs for all Vehicles
Table 2-38: Driveline system costs for all vehicles.
Item
; Drive
; Engine
: Body Style
; Value
: Dollar
; System
; Config
; Basis
ICE transmission
] FWD
-
-
7(140.11 * GEARS+ 52.674) * Markup
] 2022
ICE transmission
I RWD
-
-
7 (140.11 * GEARS + 52.674) * Markup
I 2022
ICE transmission
I AW 1)
-
i pickup
: (140.11 * GEARS + 52.674 + 544.28) *
[ 2022
: Markup
ICE transmission
V AW 1)
-
| sedan
T (140.11 * GEARS + 52.674 + 281) * Markup
7 2022
ICE transmission
j AW 1)
-
; cuv suv
; (140.11 * GEARS+ 52.674+ 339.37)*
| 2022
! Markup
PO HEV transmission
; FWD
-
-
7 (140.11 * GEARS + 1.73 * KW_P0 + 178) *
'! 2022
; Markup
PO HEV transmission
j RWD
-
-
(140.11 * GEARS + 1.73 * KW_P0 + 178) *
72022
; Markup
PO HEV transmission
T AWD
-
-
7 (140.11 * GEARS + 1.73 * KW_P0 + 178)*
T 2022
; Markup
P2 HEV/PHEV transmission
7 FWD
-
-
i (140.11 * GEARS+1.73* KW_P2+178)*
T 2022
j Markup
P2 HEV/PHEV transmission
RWD
-
-
(140.11 * GEARS + 1.73 * KW_P2 + 178) *
I 2022
1 Markup
P2 HEV/PHEV transmission
7 AWD
-
-
7 (140.11 * GEARS + 1.73 *KW P2 + 178 +
7 2022
: 1.1097 *KW P4 + 323) * Markup
PS HEV/PHEV transmission
1 FWD
-
-
; 1587.77* Markup
1 2022
PS HEV/PHEV transmission
7 RWD
-
-
71587.77* Markup
7 2022
PS HEV/PHEV transmission
7 AWD
-
-
; (1587.77 + 1.1097 * KW P4 + 323)*
72022
; Markup
Start stop
-
-
-
| 200 * Markup
i 2022
High efficiency alternator
-
-
-
: (150)* Markup
2015
Low voltage battery
-
-
-
' 60 * Markup
] 2022
ICE low voltage harness
-
-
: pickup
; 60.31* Markup
; 2022
ICE low voltage harness
-
-
; sedan
50.26 * Markup
1 2022
ICE low voltage harness
-
-
; CUV SUV
50.26 * Markup
: 2022
BEV low voltage harness
: FWD;
-
: pickup
; 110.42 * Markup
T 2022
; RWD
BEV low voltage harness
I FWD;
-
; sedan
i 101.62 * Markup
; 2022
; RWD
BEV low voltage harness
7 FWD;
-
; cuv suv
; 100.44 * Markup
I 2022
| RWD
BEV low voltage harness
7 AWD
-
: pickup
7 157.75 * Markup
7 2022
BEV low voltage harness
7 AWD
-
; sedan
7143.48 * Markup
: 2022
BEV low voltage harness
AWD
-
: CUV SUV
; 145.17 * Markup
; 2022
HEV low voltage harness
-
-
i pickup
i 60.31* Markup
7 2022
HEV low voltage harness
-
-
sedan
j 50.26 * Markup
7 2022
HEV low voltage harness
-
-
I cuv suv
: 50.26 * Markup
7 2022
PHEV low voltage harness
-
-
: pickup
] 172.31* Markup ^ 7
I 2022
PHEV low voltage harness
-
-
; sedan
; 143.59 * Markup
7 2022
PHEV low voltage harness
-
-
i cuv suv
! 143.59 * Markup
7 2022
ICE Powertrain cooling
-
71
; pickup
s 334.04 * Markup
: 2022
ICE Powertrain cooling
-
7v '
; pickup
374.53 * Markup
7 2022
ICE Powertrain cooling
-
; i
i sedan
] 308.68 * Markup
; 2022
ICE Powertrain cooling
-
7 v
; sedan
! 346.09 * Markup
; 2022
ICE Powertrain cooling
-
i
; CUV SUV
j 308.68 * Markup
; 2022
ICE Powertrain cooling
-
! "V
: CUV SUV
346.09 * Markup
7 2022
HEV Powertrain cooling
-
; I
I pickup
7 519.13 * Markup
: 2022
HEV Powertrain cooling
-
] I
; sedan
] 493.78 * Markup
72022
HEV Powertrain cooling
-
71
cuv suv
J 493.78 * Markup
: 2022
HEV Powertrain cooling
-
j V
! pickup
: 559.63 * Markup
7 2022
HEV Powertrain cooling
-
7 v
: sedan
7 531.19 * Markup
2022
HEV Powertrain cooling
-
! V
I cuv suv
] 531.19 * Markup
! 2022
PHEV Powertrain cooling
-
I I
: pickup
538.5 * Markup
7 2022
PHEV Powertrain cooling
-
T I
i sedan
] 513.21 * Markup
7 2022
PHEV Powertrain cooling
-
; 'i
; CUV SUV
; 513.21 * Markup
2022
PHEV Powertrain cooling
-
7 v
; pickup
579.07 * Markup
; 2022
PHEV Powertrain cooling
-
f V
sedan
; 550.63 * Markup
1 2022
2-77
-------
Item
; Drive
; System
; Engine
; Config
: Body Style
; Value
: Dollar
; Basis
PHEV Powertrain cooling
; -
j V
; cuv suv
: 550.63 * Markup
; 2022
BEV Powertrain cooling
1 WD:
: RWD
-
i pickup
V 632.33 * Markup
f 2022
BEV Powertrain cooling
j FWD;
| RWD
-
: sedan
f 383.54 * Markup
; 2022
BEV Powertrain cooling
: FWD;
; RWD
-
| cuv suv
] 413.13 * Markup
: 2022
BEV Powertrain cooling
] AWD
-
: pickup
; 709.81 * Markup
j 2022
BEV Powertrain cooling
AW 1)
-
; sedan
I 456.71 * Markup
T 2022
BEV Powertrain cooling
[ AWD
-
; cuv suv
; 602.00 * Markup
! 2022
HVAC
-
-
-
r2.~6 * Markup
; 2022
A/C leakage
-
-
-
i (63) * Markup
! 2010
A/C efficiency
-
-
-
; (40) * Markup
: 2010
KW RDU share
-
-
-
| 0.65 * total e-machine power
KWFDU share
-
-
-
1 - KW RDU share
KW P0 share; KW P2 share
1 FWD;
-
-
: 1.0 * total e-machine power
(used for PS)
I RWD
KW P2 share
; AWD
-
-
0.5 * total e-machine power
KW P4 share
AWD
-
-
! 1-KWP2 share
Markup
-
-
-
l'L5
| PO, P2 and PS refer to HEV and PHEV architectures; FWD=front wheel drive; RWD=rear wheel drive; AWD=all wheel drive; KW_DU=total
i e-machine power; KW_RDU=rear drive unit power; KW_FDU=front drive unit power: KW_P2=primary motor power; KW_P4=secondary
; motor power; HVAC= heating, ventilation, air conditioning.
2.6.1.2.1 High Efficiency Alternator
OMEGA's high efficiency alternator cost is based on past EPA analyses and is calculated
according to the equation shown below.
HighEfficiencyAlternator = 150 x Markup
Where,
Markup = the markup to cover indirect costs
2.6.1.2.2 Air Conditioning
Air conditioning (A/C) system costs are based on past EPA analyses and are shown in Table
2-39.
Table 2-39: Air Conditioning System Costs in OMEGA.
Item DMC Do 1 lu r Basis
A/C efficiency improvements 40 * Markup 2010
A/C leakage control 63 * Markup 2010
Markup 1.5
OMEGA uses these costs as-is, other than applying the markup to account for indirect costs
and adjusting to the appropriate dollar basis.
2-78
-------
2.6.1.3 Electrified Driveline System Costs for HEV, PHEV, and BEV
Table 2-40: Electrified driveline system costs for HEV, PHEV, and BEV.
Item
Drive
System
Body Style
Value
Dollar
Basis
P0 & P2 HEV/PHEV
Gearbox
AWD
sedan
281 * Markup
2022
P0 & P2 HEV/PHEV
Gearbox
AWD
CUV SUV
339.37 * Markup
2022
P0 & P2 HEV/PHEV
Gearbox
AWD
pickup
544.28 * Markup
2022
BEV Gearbox
FWD;
RWD
pickup
544.28 * Markup
2022
BEV Gearbox
FWD;
RWD
sedan
281.00 * Markup
2022
BEV Gearbox
FWD;
RWD
CUV SUV
339.37 * Markup
2022
BEV Gearbox
AWD
pickup
2 * 544.28 * Markup
2022
BEV Gearbox
AWD
sedan
2 * 281.00 * Markup
2022
BEV Gearbox
AWD
CUV SUV
2 * 339.37 * Markup
2022
BEV e-motor
FWD
-
(1.1097*KW DU + 323.22) * Markup
2022
BEV e-motor
RWD
-
(1.1097*KW DU + 323.22) * Markup
2022
BEV e-motor
AWD
-
((0.77*KW_FDU+225.33) + (1.1097*KW_RDU + 323.22)) *
Markup
2022
HEV/PHEV Inverter
FWD
-
(1.26 * KW P2 + 559) * Markup
2022
HEV/PHEV Inverter
RWD
-
(1.26 * KW P2 + 559) * Markup
2022
HEV/PHEV Inverter
AWD
-
((1.26 * KW P2 + 559) + (1.26 * KW P4 + 559)) * Markup
2022
BEV Inverter
FWD
-
(1.26 * KW DU + 559) * Markup
2022
BEV Inverter
RWD
-
(1.26 * KW DU + 559) * Markup
2022
BEV Inverter
AWD
-
((1.26 * KW FDU + 559) + (1.26 * KW RDU + 559)) * Markup
2022
BEV high voltage
harness
FWD;
RWD
pickup
274.32 * Markup
2022
BEV high voltage
harness
FWD;
RWD
sedan
248.87 * Markup
2022
BEV high voltage
harness
FWD;
RWD
CUV SUV
244.87 * Markup
2022
BEV high voltage
harness
AWD
pickup
363.16 * Markup
2022
BEV high voltage
harness
AWD
sedan
337.21 * Markup
2022
BEV high voltage
harness
AWD
CUV SUV
333.71 * Markup
2022
HEV high voltage
harness
-
pickup
81.09 * Markup
2022
HEV high voltage
harness
-
sedan
67.58 * Markup
2022
HEV high voltage
harness
-
CUV SUV
67.58 * Markup
2022
PHEV high voltage
harness
-
pickup
246.09 * Markup
2022
PHEV high voltage
harness
-
sedan
205.07 * Markup
2022
PHEV high voltage
harness
-
CUV SUV
205.07 * Markup
2022
BEV/PHEV Charge cord
-
-
111 * Markup
2022
HEV DC-DC converter
-
-
250 * Markup
2022
BEV DC-DC converter +
onboard charger
-
-
57 * KW OBC * Markup
2022
PHEV DC-DC converter
+ onboard charger
-
-
57 * KW OBC * Markup
2022
KWOBC
-
-
For battery kWhdOO, KW OBC=l 1
For battery kWh>=100, KW OBC=19
BEV/PHEV High voltage
battery NMC
-
-
(128.9 + 1480 / (KWH ** 1.164) - 5.278 * (MODEL YEAR - 2023)
* e ** (-0.0129 * (MODEL YEAR - 2023))) * KWH * Markup
2022
BEV/PHEV High voltage
battery LFP
-
-
(120.6 + 1535 / (KWH ** 1.148) - 10.04 * (MODEL YEAR - 2023)
* e ** (-0.08346 * (MODEL YEAR - 2023))) * KWH * Markup
2022
2-79
-------
Item
Drive
System
Body Style
Value
Dollar
Basis
HEV High voltage
battery NMC
(122.9 + 509.6 / (KWH ** 0.7649) - 4.443 * (MODEL YEAR -
2023) * e ** (0.01018 * (MODEL YEAR - 2023))) * KWH *
Markup
2022
High voltage battery
learning
-
-
(1 - 0.015) ** (max(CALENDAR_YEAR, 2035) - 2035)
KW RDU share
-
-
-
KW FDU share
-
-
-
KW PO share; KW_P2
share (used for PS)
FWD;
RWD
-
-
KW P2 share
AWD
-
-
KW P4 share
AWD
-
-
Markup
-
-
1.5
FWD=front wheel drive; RWD=rear wheel drive; AWD=all wheel drive; KW DU=total e-machine power; KW RDU=rear drive unit power;
KW FDU=front drive unit power: KW P2=primary motor power; KW P4=secondary motor power; KWH=gross battery energy capacity;
KW OBC=onboard charger power; DC=direct current; NMC=nickel metal carbide; LFP=lithium iron phosphate. Note that "**" denotes an
exponent; e=2.718.
2.6.1.4 HEV and Mild HEV Battery Costs
As previously described in Chapter 2.5.2.1.2, OMEGA uses the HEV battery cost curve
developed by ANL and shown below. OMEGA uses this equation for both mild and strong
HEVs.
HEV (Ni/Mn) %/kWh =
122 9 + JZL ~ 4-443(MV ~ 2023)e-° 01018(MV-2023)
rC vV tl
x Markup
Where,
kWh = the gross energy capacity of the battery in kilowatt hours
MY = model year
Markup = the markup to account for indirect costs
2.6.1.5 BEV and PHEV Battery Costs
As described previously in Section 2.5.2.1.2, EPA is using a set of cost equations for PEV
batteries developed by ANL, which EPA uses for BEV and PHEV batteries. EPA is modeling
longer-range PHEVs, and the equations were developed by ANL to be inclusive of longer-range
PHEVs that would utilize battery capacities at the lower end of the range of kWh covered by the
equation.
BEV (NiMn) %/kWh =
1480
128 9 + 1 TT77I - 5- 278(MY - 2023)e-°0129(MV-2023)
kWh1164
BEV CLFP) %/kWh =
1535
120 6 + uwhi 148 ~ m 04(MY - 2023)e-° 08346(MY-2023)
KW tl
x Markup
x Markup
2-80
-------
PHEV (NiMn) %/kWh =
128.9 +
1480
kWhll 64
- 5.278(MK - 2023)e"0 0129(MF"2023)
x Markup
PHEV (LFP) $/fcWh =
120.6 +
1535
kWh1148
10.04(MK - 2023)e~0 08346(MF_2023)
x Markup
Where,
= the gross energy capacity of the battery in kilowatt hours
MY = model year
Markup = the markup to account for indirect costs shown in Table 2-34
Table 2-32 gives the assumed NMC/LFP share used in the OMEGA modeling for BEVs and
PHEVs. For HEV batteries we assume 100 percent NMC chemistry.
2.6.1.6 BEV Range Assumptions
For the purpose of compliance modeling, EPA assumed that all light-duty BEVs would be
designed with 300 miles of label range, which is calculated as 300 miles divided by a 0.75
adjustment factor used for fuel economy labeling purposes. The average range of MY 2022
BEVs exceeded 300 miles for the first time, as discussed in Chapter 3.1.1.3 of the RIA. EPA
assumed 150-mile range for its medium-duty BEV vans. The basis for this assumption is
discussed in Chapter 3.1.2 of the RIA.
2.6.1.7 PHEV Range Assumptions
Unlike ICE vehicles and BEVs, which operate under one dedicated fuel, PHEVs consume a
combination of electricity and gasoline. As is discussed throughout this rulemaking, the charge-
depleting range of a PHEV is directly related to the utility factor used to calculate each vehicle's
compliance CO2 value. Manufacturers have the flexibility to design PHEVs with a variety of
charge-depleting range, from a minimal all-electric range, up to strong PHEVs which might
exceed 100 miles of all-electric range. For the purposes of our compliance runs, EPA assumed
that manufacturers would design PHEVs that would provide enough range to qualify as ZEVs
under the ACC II and ACT programs being administered by California and participating Section
177 states. In OMEGA, EPA thus assumed that light-duty vehicle PHEV batteries would be
sized for 40 miles of all-electric range over the US06 cycle, while medium-duty PHEVs would
be sized to drive 75 miles over the UDDS while tested at ALVW.
2.6.1.8 Additional discussion of PHEV Architectures
EPA recognizes that PHEVs can provide significant reductions in GHG emissions and that
some vehicle manufacturers may choose to utilize this technology as part of their technology
portfolio.
2-81
-------
In general, EPA anticipates that individual component costs for PHEVs, such as power
electronics costs, P4 gearbox costs and AWD costs would be similar to that of BEVs. We also
expect that PHEVs, like BEVs, will have a variety of specific component sizings and
architectures oriented to specific uses, such as in passenger cars, pickup trucks, and medium-duty
vehicles.
For example, a series/parallel hybrid transmission for PHEVs may consist of:
• Motor-generator
• Starter-generator
• Clutch-pack to lock the ICE and starter generator to the motor generator for parallel
operation
An example of such a series/parallel hybrid drive system for transverse front-drive
applications is shown in Figure 2-29. Yamagishi and Ishikura provided a detailed description of
application of a similar series-parallel drive system to the Honda Clarity PHEV (Yamagishi and
Ishikura 2018). An application of this type of series/parallel drive to a front-engine/rear-drive
application would require use of a drive shaft and separate, rear-mounted differential. AWD
vehicles may include the cost of a series-parallel hybrid transmission with the addition of a P4
electric machine to either the front or rear depending on the application. One specific example is
an LDT4 with OMEGA size-class 7 vehicle with:
• A combined electric drive system power of 240 kW
• A P4 induction machine for front axle electric-only drive
• A 3.0L Miller Cycle engine coupled to a series/parallel (S/P) drive single-speed
transmission with a drive shaft and rear differential for rear axle drive
• The figure is adapted from a presentation by Prof. J.D. Kelly, Weber State University
(Kelly 2020).
2-82
-------
Figure 2-29: Example of a series-parallel hybrid drive system for a transverse/front-drive
application,
2.6.2 Glider Costs
Glider cost curves in OMEGA represent three different body-styles: sedan, CUV/SUV, and
pickup; two different structure styles: unibody and ladder frame; two different primary materials:
steel and aluminum; as well as non-structural elements. The relevant curves used in OMEGA are
shown in Table 2-41. Note that "structure_mass_lbs" term shown in the table is determined
according to the structure mass curves shown in Table 2-42.
Note that, unlike past EPA GHG analyses, OMEGA no longer models mass as a compliance
strategy in discrete percentages of mass reduction. Instead, OMEGA calculates mass based on
the factors shown in Table 2-42, with the compliance strategy decision based primarily on steel
versus aluminum structure. Footprint may also change which would impact the mass of the
vehicle, but the mass associated with potential footprint changes is a secondary effect of the
footprint decision. In other words, footprint does not change as a result of mass reduction
strategies and, instead, mass may change as a result of footprint strategies. The cost of the
resultant mass is estimated using the equations shown in Table 2-41.
2-83
-------
Dollar
Basis
2020
2020
2020
2020
2020
2020
2020
2020
2020
2020
2020
2020
2020
Table 2-41: Glider Costs in OMEGA.
Body-slvlc Structure DMC
Material
Sedan Steel (1.5 * slruclurcjnassjbs + 1500) * markup
Sedan Aluminum (3.4 * slruclurcjnassjbs + 1500) * markup
CUV/SUV Slccl (1.5 * slruclurcjnassjbs + 1700) * markup
CUV/SUV Aluminum (3.4 * slruclurcjnassjbs + 1700) * markup
CUV/SUV Slccl ((1.5 * slruclurcjnassjbs + 550) + (1.5 * (0.66 *
slruclurcjnassjbs) + 2000)) * markup
CUV/SUV Aluminum ((1.5 * slruclurcjnassjbs + 550) + (3.4 * (0.66 *
slruclurcjnassjbs) + 2000)) * markup
Pickup Slccl ((1.5 * slruclurcjnassjbs + 550) + (1.5 * (0.66 *
slruclurcjnassjbs) + 2000)) * markup
Pickup Aluminum ((1.5 * slruclurcjnassjbs + 550) + (3.4 * (0.66 *
slruclurcjnassjbs) + 2000)) * markup
Pickup Slccl (1.5 * structure mass lbs + 1700) * markup
Pickup Aluminum (3.4 * slruclurcjnassjbs + 1700) * markup
Sedan Various (24.3 * delta footprint + 2.4* delta footprint *
(vehicle.heightin - vehicle.ground clearance in)):
markup
CUV/SUV Various (24.9 * dcllafoolprinl + 2.6 * dcllafoolprinl *
(vehicle.height in - vchiclc.ground clcarancc in)):
markup
Pickup Various (18.2 * dclla foolprinl + 2.1 * dclla foolprinl *
(vehicle.height in - vchiclc.ground clcarancc in)):
markup
1.5 RPE markup to account for indirect costs
2-84
-------
Table 2-42: Mass Calculations in OMEGA.
Item
Body-style
Structure
Material
Value
Null structure mass
Sedan
Ladder
2.2 * (5.5045 * footprint +
105.4)
Null structure mass
Sedan
U nibodv
2.2 * (5.5045 * footprint +
105.4)
Null structure mass
CUV/SUV
Ladder
2.2 * (7.7955 * footprint +
127.48)
Null structure mass
CUV/SUV
U nibodv
2.2 * (10.077 * footprint -
76.528)
Null structure mass
Pickup
Ladder
2.2 * (7.7955 * footprint +
127.48)
Null structure mass
Pickup
U nibodv
2.2 * (10.077 * footprint -
76.528)
Structure mass lbs
Steel
liullslruclurcniass
Structure mass lbs
Ladder
Aluminum
(0.63 * 0.66 +0.34 )*
nullslruclurcniass
Structure mass lbs
U nibodv
Aluminum
0.65 * null slruclurc niass
Delta glider non-slruclurc mass
Sedan
(15.1 * dclla foolprinl + 2.3 *
delta footprint *
(vehicle.height -
\ chiclc.groimdclcaraiicc)/12)
Delia glider non-structure mass CUV/SUV (17.3 * dcllafoolprinl + 2.5 *
dclla_Ibolprinl *
(vehicle.height -
\ chiclc.groundclcarancc)/12)
Delta glider non-slruclurc mass Pickup (18.1 * dclla foolprinl +1.9*
dclla foolprinl *
(vehicle, height -
vehicle ground_clearance)/12)
Note: footprint is in square feet; height and ground clearance are in inches; mass \ ilucs uc m pounds; 2.2
converts kilograms to pounds
2.6.3 Consumer demand assumptions and PEV acceptance
OMEGA estimates the share of PEVs demanded within each of three body styles as a function
of the relative consumer generalized costs for PEV and ICE vehicles, and a PEV acceptance
parameter, which we refer to as a shareweight parameter. The shareweight parameter changes
over time to account for factors that are not included in the generalized costs, such as greater
access to charging infrastructure or greater availability and awareness of PEVs. The
determination of consumer generalized costs and share weights for ICE vehicles and PEVs is
described in more detail in Chapter 4.1.
2.6.4 Producer decision modeling and constraints for technology adoption
2.6.4.1 Redesign schedules
Consistent with past rulemakings, EPA has included redesign cycles as a constraint to restrict
introduction of new technology for a given vehicle model to the cadence of a typical product
cycle time. As implemented for this rule, OMEGA may only redesign a vehicle every 5 years for
unibody vehicles and every 7 years for body-on-frame vehicles. An example of behavior for one
manufacturer's vehicles can be seen in Figure 2-30.
2-85
-------
2040
II ,
i. ^
2015
2010
2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 2036 2037 2038 2039 2040
Model Year
Figure 2-30: Redesign Years for Select Vehicles.
EPA has populated its base year vehicles file with the year of last redesign for each model in
the light duty fleet. Applying OMEGA's redesign constraints above with the distribution of
redesign years across the industry yielded the following distribution of redesigns on a sales basis.
Table 2-43 provides a count of the discrete vehicle models and sales in MY 2032, the year in
which they were last redesigned, and the corresponding sales volume that was redesigned in
prior years. As can be seen, there is a fairly even distribution of vehicle model redesign years.
Note that many vehicles which were redesigned in MY 2026 were eligible for another redesign
in MY 2031.
Table 2-43: MY 2032 Vehicles: Year of Last Redesign.
Year
# of Models
Total Sales
% of Sales
Redesigned
2026
10
83,818
1%
2027
38
823,932
5%
2028
201
2,601,172
17%
2029
284
2,849.875
19%
2030
284
3,250,190
21%
2031
358
4,136,099
27%
2032
284
1,641,242
11%
Totals
1459
15,386,328
100%
2.6.4.2 Materials and mineral availability
The development of a constraint on PEV production, which is based on an upper bound on
GWh of battery production in each year of the analysis, can be found in Section 3 .1.5 of this
RIA.
Table 2-44 shows the limits, in terms of maximum industry GWh (available for production of
U.S. vehicles) that resulted from this assessment and that are applied in OMEGA.
2-86
-------
Table 2-44: Industry Maximum Battery Production Limits (GWh), by Model Year.
MY
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
GWh limit
79
150
261
372
483
580
720
860
1000
1100
1200
1300
1400
1500
2.6.4.3 Credit Banking
A number of individual manufacturers have accumulated banked credits from MY 2022 and
earlier vehicles. These existing credits, shown in Table 2-45, are available for use as a
compliance flexibility to carry-forward and offset debit generation in a future year. Given that
manufactures have not previously allowed credits to expire, EPA considers it reasonable to
assume that manufacturers will utilize these historical credits. OMEGA allows the use of these
historical credits for strategic debit generation in the early years of the analysis.
Credits generated during the analysis years of MY 2023 and later (i.e. not historical credits)
are banked and carried forward to be available for use in years where the manufacturer does not
comply due to, for example, strategic decision making related to considerations such as redesign
intervals, available battery production, consumer demand or other modeled constraints. Debits
generated during a particular analysis year are first offset by the use of any available credits from
prior years. If any debit remains, the manufacturer will attempt to achieve a lower Mg CO2 value
in the subsequent year than is required by the standards. Those additional credits will be used as
carry back credits to offset the prior debit. The maximum credit carry back timeframe is three
years, as defined by the standards, so the amount of the offset is either one third, one half, or the
entire debit amount depending on whether the number of years remaining before the debit would
become past due is 3, 2, or 1 year, respectively.
2-87
-------
Table 2-45: Light-duty Vehicle Historical Mg CO2 Credit Bank, by Model Year Vintage.
Credit Vintage (Model Year)
2017
2018
2019
2020
2021
2022
79
150
261
372
483
580
Aston Martin
681
9.470
5.489
5,302
BMW
2.259.136
138.811
281.656
117.904
582,705
297.241
Ferrari
1.047
4.192
Ford
1.854.402
2.118.166
1.118.271
General Motors
4.021.303
10,000,000
Honda
2.938.779
8.192.781
5.365.503
2.868.950
4.005.904
4.599.960
Hyundai
1.924.619
JLR
Kia
1.374.461
60,900
-416.089
Lucid
97.261
Ma/da
95.569
Mercedes Ben/.
4.123.493
Mitsubishi
144.541
105.497
56.866
476.697
559.669
Nissan
1.170.458
1.100.000
1.400.000
Rivian
1.385.539
Stellantis
6.406.741
1.107.0481
8443.887
12.000.000
Subaru
2.156.402
2.561.015
3.261.822
3.041.737
2.859.900
1.097.819
Tcsla
1.766
53.704
208.003
13.928
60.395
Toyota
1.944.036
2.126.707
1.596.028
16.66.470
2.592.586
2.380.010
Volvo
78.996
778.606
316.651
215.898
831.916
1.119.171
2.6.4.4 Credit Trading and Credit Market Efficiency
OMEGA uses a two-pass approach to model compliance. On the first pass, the individual
manufacturers are consolidated to represent the industry as a whole. The resulting compliance
decisions imply "perfect trading" of credits in which the application of technology is applied to
individual vehicles in the most cost-effective manner across the entire industry. On the second
pass, each manufacturer is modeled in isolation, with their Mg CO2 goal (the CO2 target with
some offset) in a given year determined from the results from the first pass. To account for a
possible future scenario where some portion of industry-wide credits are allowed to expire even
as additional technology adoption is required for compliance, OMEGA has a "Credit Market
Efficiency" parameter. This parameter value determines the degree to which debit-generating
manufacturers from the first modeling pass seek lower Mg CO2 values in the second pass.
If the Credit Market Efficiency parameter is set to zero, each manufacturer attempts to meet
their own Mg CO2 certification target value without benefiting from the option to purchase
credits from other manufacturers. While EPA considers this scenario to be very unlikely given
the history of a robust credit market over the past decade, we present results for a "No Trading"
sensitivity in RIA Sections 12.1.4 and 12.2.4 for light and medium duty vehicles. If the Credit
Market Efficiency is set to one, then the manufacturer Mg CO2 goal for the second pass is set to
the achieved value from the "perfect trading" first pass. If the Credit Market Efficiency is set
somewhere in-between zero and one, then "imperfect" trading occurs. As implemented in
2-88
-------
OMEGA, manufacturers that were under their Mg CO2 target (i.e. generating credits) in the first
pass will continue to do so on the second pass, and their Mg CO2 goal will be based on their
achieved value from the first pass. Those manufacturers that were over their Mg CO2 target (i.e.
generating debits) in the first pass will be required to apply additional technology to close the
gap between their compliance target and achieved Mg CO2. In this case, vehicle production costs
for those manufacturers will be somewhat higher in the second pass due to the lost credits.
For this rulemaking, EPA has used a Credit Market Efficiency of 0.8, so that over target (debit
generating) manufacturers will attempt to close the gap with the compliance target by 20 percent
on the second pass. This provides some additional credits for debit generating manufacturers to
carry forward and act as a reserve in the event that credits are not purchased in some future year,
for any reason. EPA derives this 0.8 value from the 5-year credit life, so that the accumulation of
20 percent reserve credits generated in each year would enable the complete offset of a deficit if
no credits are purchased in the fifth year. In our modeling results, the net effect of the credit
generators continuing to generate the same credits, and debit generators applying additional
technology on the second pass is that there will be some excess of credits that will expire,
thereby increasing overall compliance costs in a representation of an imperfect credit market.
2.6.4.5 Producer Generalized Costs and compliance cost minimization
The producer module in OMEGA determines vehicle technology packages and mix of
powertrain types in order to minimize costs while satisfying consumer demand, regulatory
requirements, and constraints on vehicle production. More specifically, OMEGA's producer
decisions minimize "generalized cost", which includes the cost of manufacturing a vehicle and
the producer's assumptions of consumer valuations for key vehicle characteristics that are
relevant to modeling compliance. For this rulemaking, EPA has included the producer-assumed
consumer valuation of fuel cost and vehicle size (footprint) in the producer generalized cost
equation.
We estimate the producer-assumed consumer valuation of fuel cost based on 2.5 years of
driving 15,000 miles per year, which is the same value we use for the actual consumer valuation
of fuel costs in the new vehicle sales response. As discussed in Chapter 4.4, while there is some
evidence that actual fuel cost valuation by consumers may be higher, EPA views the 2.5 year
estimate as within the range of the literature, and appropriately conservative since lower-end
estimates like this one will tend to indicate less emissions-reducing and fuel-saving technology
adoption in the absence of stringent standards.
OMEGA allows manufactures to adjust vehicle footprints, which will influence the modeled
vehicle weight, manufacturing cost, and emissions. Depending on the definition of the footprint-
based standards, the CO2 emissions target may also be affected. In terms of a compliance
approach which minimizes generalized costs, manufacturers may benefit in some cases by
increasing vehicle size, and in other cases by decreasing vehicle size. EPA's main intention for
this modeling feature is to assist in the development of footprint curve slopes, with the aim of
determining a size-neutral standard which does not incentivize manufacturers to shift the average
size of vehicles as a compliance strategy. As discussed in preamble Section I.l.iii, we have used
a consumer valuation of $200 per square foot for increases in vehicle size. This valuation is not
intended to capture the effect of differences in vehicle classes, and we did not aim to model
substantial changes to vehicle footprints. Thus, we have constrained the allowable footprint
2-89
-------
changes to a narrow range of plus or minus 5 percent relative to the base year value for each
vehicle.
OMEGA's consumer module considers vehicle price as one of the important elements for
modeling consumer purchase decision and powertrain choice (i.e. ICE, PHEV, or BEV), as
discussed in Chapter 4.1 and Chapter 4.3. The producer has the ability to influence the price of a
vehicle offered, even with no change in production cost, through the use of incentives and cross-
subsidies. OMEGA implements an approach where revenue is held constant (relative to no cross-
subsidies) while applying cross-subsidies to the powertrain types within each of the three market
categories defined by body styles as sedan/wagons, crossover/SUVs, and pickups. For example,
if a cost-minimizing compliance strategy involves relatively more BEVs, similar PHEVs, and
fewer ICE vehicles, the producer module will generate a set of vehicle offerings where BEV
prices are marked down and ICE prices are marked up while maintaining the sales-weighted
average price of that vehicle segment. In practice this means that the lower-market-share vehicles
will have higher dollar-per-vehicle price changes than the higher-market-share vehicles. If the
market shares were equal, then the price increase of one segment would equal the price decrease
of the other segment.
OMEGA users can specify a limit to the range of possible cross-subsidies. For this analysis,
EPA constrained cross-subsidies to be within 5 percent of the marked up vehicle manufacturing
cost. This value lies within the recent historical range of purchase incentives of roughly 3 to 8
percent of vehicle transaction price (Kelley Blue Book 2023a) (Kelley Blue Book 2023b). EPA
considers that this is likely a conservative (i.e. low) limit of pricing flexibility, given additional
evidence for within-model year price changes for individual nameplates, where manufacturers
have made recent adjustments of 10 percent or more for both price decreases (Autoweek 2023)
and price increases (GenSao 2022).
OMEGA considers IRA incentives that are provided directly to the consumer, specifically the
IRA's 30D purchase incentives, as being applied on top of the producer's cross-subsidized price.
These may work to either enhance or act against the producer's cross-subsidy. For example, if a
producer's cost-minimizing strategy involves selling fewer BEVs in a given year than is
demanded by consumers in the absence of cross-subsidies, they could mark up BEV prices while
simultaneously the IRA purchase incentive may bring down the consumer's net purchase cost.
2.6.5 Manufacturing capacity
In addition to availability of critical minerals, the ability to perform final assembly of vehicles
that use them could also be understood as a potential constraint on increased production of
BEVs. However, EPA notes that major manufacturers are already building a large amount of
assembly plant manufacturing capacity both in the U.S. and abroad to meet future demand for
these vehicles, and these efforts are poised to continue (see the discussions of announced
manufacturer plans and investments in section I. A.2 and IV.C.2 of the preamble). Unlike critical
minerals which have fundamental constraints on their production due to limited presence of these
resources as well as a relatively long lead time for increasing their extraction, vehicle assembly
capacity can respond relatively quickly to the necessary investment commitments. Given the
relatively long lead time before MY 2027 when the standards would begin, EPA did not
specifically impose a limit on vehicle assembly capacity. However, as described in Chapter 3.1.5
of this RIA, EPA did represent a reasonable rate of battery manufacturing ramp-up by using
information about battery manufacturing facilities announced or in operation, and estimates of
2-90
-------
lithium availability, to develop a constraint on annual GWh battery demand for use by OMEGA.
For more discussion of manufacturing capacity and critical minerals, please see Section IV.C.7
of the preamble and Chapter 3.1 of this RIA.
2.6.6 Fuel and electricity prices used in OMEGA
OMEGA uses liquid fuel and electricity prices to estimate generalized costs as part of the
compliance modeling algorithm. See Table 2-46 and Table 2-47Table 2-47, respectively.
OMEGA also uses these fuel prices in estimating fuel expenditures and fuel savings that are
included in the benefit-cost analysis results presented in Chapter 9 of this RIA. The OMEGA
compliance model makes use of only the retail price while the effects calculations make use of
both the retail and the pretax prices.
Note that OMEGA uses different electricity prices in different parts of the model. For
compliance costs, electricity and liquid fuel prices play a role in producer/consumer decision-
making in that they impact the fuel cost portion of the generalized cost. For compliance,
OMEGA uses AEO 2023 electricity prices since those are the prices we believe consumers
would be most familiar with and those are the prices AEO uses to generate future fleet mixes
which are used to define the future body style and regulatory class mix projections in OMEGA's
new vehicle market input file. For effects calculations, we use the electricity prices generated in
the Integrated Planning Model and Retail Price Model work presented in Chapter 5 of this RIA.
Those prices include impacts of the Inflation Reduction Act along with other market forces
driving renewables into the grid mix while driving down electricity prices. Note also that our No
Action effects and our Action effects use unique electricity prices as modeled by IPM.
2-91
-------
Table 2-46: AEO2023 Liquid Fuel Prices Used in OMEGA Compliance and Effects
Modeling (2022 dollars).
Gasoline Diesel
Calendar
Pre-tax
Retail
Pre-tax
Retail
Year
($/gallon)
($/gallon)
($/gallon)
($/gallon)
2027
2.53
3.05
3.20
3.74
2028
2.53
3.05
3.10
3.63
2029
2.54
3.06
3.11
3.65
2030
2.56
3.07
3.12
3.65
2031
2.56
3.06
3.15
3.67
2032
2.58
3.08
3.17
3.69
2033
2.59
3.09
3.19
3.71
2034
2.60
3.10
3.20
3.71
2035
2.61
3.10
3.23
3.74
2036
2.65
3.13
3.24
3.74
2037
2.66
3.14
3.26
3.76
2038
2.67
3.15
3.28
3.78
2039
2.68
3.16
3.29
3.78
2040
2.69
3.17
3.30
3.79
2041
2.69
3.16
3.33
; 3.81
2042
2.71
3.17
3.33
3.81
2043
2.70
3.16
3.35
3.83
2044
2.72
3.18
3.34
3.82
2045
2.72
3.18
3.35
3.82
2046
2.78
3.27
3.39
3.90
2047
2.78
3.27
3.40
3.91
2048
2.81
3.30
3.41
; 3.91
2049
2.82
3.30
3.42
3.92
2050
2.85
3.33
f 3.43
3.92
2051
2.88
3.36
3.43
3.93
2052
2.91
3.39
3.43
3.93
2053
2.95
3.42
3.44
3.93
2054
2.98
3.45
3.44
3.93
2055
3.01
3.48
3.45
; 3.93
2-92
-------
Table 2-47: Electricity Prices used in OMEGA Compliance and Effects Modeling (2022
dollars).
: Compliance Model ; Effects No Action ; Effects Action
Calendar =
Retail
: Pre-tax
Retail
Pre-tax
Retail
Year
($/kWh)
; ($/kWh)
: ($/kWh)
($/kWh)
i ($/kWh)
2027 ;
0.130
j 0.112
f 0.119
0.112
r 0.118
2028
0.129
0.113
j 0.119
0.112
: 0.118
2029
0.128
! o.iio
: 0.116
0.109
1 0.116
2030 :
0.129
; o.io7
1 0.113
0.107
f 0.113
" 2031 j
0.129
; 0.106
i' 0.113
0.107
j 0.113
2032
0.130
0.106
0.112
0.106
0.113
2033
0.132
f 0.105
0.111
0.106
: 0.112
2034
0.133
: 0.104
; o.iio
0.106
; 0.112
2035
0.134
1 0.103
0.109
0.106
: 0.112
2036
0.134
0.103
r o.io9
0.106
I 0.112
2037
0.135
r o.io3
! 0.110
0.106
i 0.112
2038
0.137
; 0.104
! 0.110
0.106
i 0.112
2039 ;
0.138
1 0.104
1 0.110
0.105
j 0.112
2040
0.139
: 0.104
1 0.110
0.105
f 0.111
2041
0.140
T 0.103
0.109
0.105
! 0.111
2042
0.140
1 0.102
; 0.108
0.104
! 0.110
2043 ]
0.140
7" 0.102
: 0.107
0.104
; 0.110
2044
0.141
] 0.101
0.107
0.103
; 0.109
2045
0.141
0.100
| 0.106
0.102
; 0.108
2046 1
0.141
r 0.099
j 0.105
0.102
0.108
2047
0.140
T 0.098
r 0.104
0.101
I" 0.107
2048
0.140
: 0.098
: 0.103
0.100
: 0.106
2049 I
0.140
1 0.097
r 0.102
0.099
! 0.105
2050
0.139
: 0.096
i 0.101
0.098
j 0.104
2051
0.137
T 0.096
0.101
0.098
j 0.104
2052
0.136
"i 0.096
' 0.101
0.098
: 0.104
2053 V
0.135
] 0.096
; 0.101
0.098
i 0.104
2054 r
0.134
0.096
0.101
0.098
I 0.104
2055
0.133
: 0.096
; 0.101
0.098
: 0.104
2-93
-------
2.6.7 Gross Domestic Product Price Deflators
To adjust all monetary inputs used in OMEGA to a consistent dollar basis, OMEGA uses the
gross domestic product (GDP) implicit price deflators shown in Table 2-48. These deflators were
generated by the Bureau of Economic Analysis, Table 1.1.9, revised on April 27, 2023.
2.6.8 Inflation Reduction Act
OMEGA explicitly accounts for two elements of the Inflation Reduction Act in compliance
modeling: the IRS Section 45X production tax credit of up to a combined $45 per kWh for
battery cells and modules, and the combined effect of the 30D Clean Vehicle Credit and the 45W
Commercial Clean Vehicle credit. Note that the No Action and Action scenarios in OMEGA use
identical treatment of IRA incentives.
The 45X production tax credit is treated within the modeling as a reduction in direct
manufacturing costs, which in turn is assumed to result in a reduction in purchase price for the
consumer after the application of the Retail Price Equivalent (RPE) factor. The credit phases out
by statute from 2030 through 2032. For discussion of how EPA estimated these values, see
section IV.C.2 of the preamble.
Table 2-48: Gross domestic product implicit price deflators,
Calendar GDP Implicit Price Deflator
Year
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
78.025
79.783
81.026
82.625
84.843
87.504
90.204
92.642
94.419
95.024
96.166
98.164
100
101.751
103.654
104.691
105.74
107.749
110.339
112.318
113.784
118.895
127.224
2-94
-------
The resulting value of the credit applied in OMEGA, in terms of dollars per kWh of gross
battery capacity, is shown in Table 2-49. These represent an average credit amount across the
industry as a whole.
Table 2-49: IRA Battery Production Tax Credits in OMEGA.
Year Tax credit value ($/kWh) % of maximum available credit
2023 ;
$22.50
50.0%
2024
$24.11
53.6%
2025
$25.71
57.1%
2026
$27.32
60.7%
2027 1
$28.93
64.3%
2028
$30.54
67.9%
2029
$32.14
71.4%
2030
$25.31
75%
203 1
$19.69
87.5%
2032
$11.25
100%
2033 :
$0
-
The IRS 30D and 45W Clean Vehicle Credits are reflected in the modeling through their
effect on vehicle purchase costs, and therefore have an influence on the shares of PEVs
demanded by consumers. The purchase incentive is assumed to be realized entirely by the
consumer and does not impact the producer generalized cost value or the manufacturing cost.
While the restrictions imposed by the IRA on the 30D credit (income, MSRP, critical mineral
content, and manufacturing content) limit the vehicles which are eligible for the full $7,500
incentive under 30D, we expect that manufacturers will work to increase the number of vehicles
that qualify over time due to the high marketing value of the credit. Further, we expect that the
IRS 45W Clean Commercial Vehicle Credit, which is not subject to many of the restrictions on
the 30D credit, will likely impact a significant portion of PEV sales, through fleet purchases and
also through reduced cost of vehicle leasing to consumers. For these reasons, we have
conceptualized the purchase incentive as a combination of the average value 30D and 45 W
credits realized per PEV across the new vehicle fleet as a whole. The resulting values of the
credit applied in OMEGA are shown in Table 2-50. For discussion of how EPA estimated these
values, see section IV.C.2 of the preamble. We have also assessed sensitivities on the IRA
assumptions as described further in Chapter 12.1.4 of this RIA.
Table 2-50: IRS 30D and 45W Clean Vehicle Credit in OMEGA.
Model
Year Combined 30D and 45W Value % of maximum available credit
2023 $2925 39%
2024 $3225 43%
2025 $3300 44%
2026 $3300 44%
2027 $3600 48%
2028 $3750 50%
2029 $3900 52%
2030 $4125 55%
2031 $5075 68%
2032 $6000 80%
2033 ] $0
2-95
-------
Chapter 2 References
75 FR 25324. 2010. (May 7). https://www.govinfo.gOv/content/pkg/FR-2010-05-07/pdf/2010-
8159.pdf.
77 FR 62624. 2012. (October 15). https://www.govinfo.gov/content/pkg/FR-2012-10-
15/pdf/2012-21972.pdf.
86 FR 74434. 2021. (December 30). https://www.govinfo.gov/content/pkg/FR-2021-12-
30/pdf/202 l-27854.pdf.
88 FR 4296. 2023. (January 24). https://www.federalregister.gov/documents/2023/01/24/2022-
27957/control-of-air-pollution-from-new-motor-vehicles-heavy-duty-engine-and-vehicle-
standards.
ANL. 2024. Cost Analysis and Projections for U.S.-Manufactured Automotive Lithium-Ion
Batteries. Argonne National Laboratory.
Autoweek. 2023. Ford Drops F-150 Lightning Prices, Citing Incoming Production Bump. 7 17.
Accessed 3 6, 2024. https://www.autoweek.com/news/industry-news/a44564656/ford-150-
lightning-prices-2023/.
Bhattachaijya, Shuvodeep, Paul Chambon, Graham Conway, Robert Mitchell, Piyush Bhagdikar,
Kevin Whitney, Tom Reinhart, Kevin Hoag, Joseph McDonald, and Shabbir Ahmed. 2023.
Heavy-light-duty and Medium-duty Range-extended Electric Truck Study. Contract Report, Ann
Arbor, Michigan: US EPA.
Bolon, Kevin, Andrew Moskalik, Kevin Newman, Aaron Hula, Anthony Neam, and and
Brandon Mikkelsen. 2018. "Characterization of GHG Reduction Technologies in the Existing
Fleet." SAE Technical Paper 2018-01-1268. doi: 10.4271/2018-01-1268.
Cherry, Jeff, and Todd Sherwood. 2023. "Availability of Cost Estimation Curves and Data
Generated by FEV." U.S. EPA, August.
Dekraker, Paul, John Kargul, Andrew Moskalik, Kevin Newman, Mark Doorlag, and and Daniel
Barba. 2017. "Fleet-Level Modeling of Real World Factors Influencing Greenhouse Gas
Emission Simulation in ALPHA." SAE Int. J. Fuels Lubr. 10 (1): 217-235. doi: 10.4271/2017-01-
0899.
FEV Consulting Inc. 2022. "Cost and Technology Evaluation, Conventional Powertrain Vehicle
Compared to an Electrified Powertrain Vehicle, Same Vehicle Class and OEM." Contract Report
for the U.S. EPA, Contract No. 68HERC19D00008.
Foundation, Python Software. 2023. "Python." https://www.python.org/.
—. 2023. "sklearn." https://pypi.org/project/sklearn/.
2-96
-------
GenSao. 2022. Tesla Motors Club. 4 7. Accessed 3 6, 2024.
https://teslamotorsclub.com/tmc/threads/another-price-increase-tonight-04-06-
2022.263312/#post-6594645.
ICF International. 2022. EPA-420-R-23-004: Peer Review of Electrified Vehicle Simulations
within EPA's ALPHA Model. US EPA.
Kelley Blue Book. 2023a. Average New Car Price Falling; Incentives Making a Comeback. 3 8.
Accessed 03 6, 2024. https://www.kbb.com/car-news/average-new-car-price-falling-incentives-
making-a-comeback/.
—. 2023b. U.S. New-Vehicle Prices Down Year over Year for Third Straight Month; Market
Shifts toward Buyers as Incentives Climb, Profit Margins Squeezed. 12 11. Accessed 3 6, 2024.
https://www.prnewswire.com/news-releases/us-new-vehicle-prices-down-year-over-year-for-
third-straight-month-market-shifts-toward-buyers-as-incentives-climb-profit-margins-squeezed-
302011706.html.
Kleijnen, Jack P.C. 2015. "Response Surface Methodology." In Handbook of Simulation
Optimization, by Michael C. Fu, 81-104. New York, NY: Springer, doi: 10.1007/978-1-4939-
1384-8.
Lee, SoDuk, Jeff Cherry, Michael Safoutin, Anthony Neam, Joseph McDonald, and Kevin
Newman. 2018. "Modeling and Controls Development of 48 V Mild Hybrid Electric Vehicles."
SAE Technical Paper 2018-01 -0413. doi: 10.4271/2018-01 -0413.
Moskalik, Andrew. 2020. "Using Transmission Data to Isolate Individual Losses in Coastdown
Road Load Coefficients." SAE International Journal of Advances and Current Practices in
Mobility 4 (1): 2156-2171. doi: 10.4271/2020-01-1064.
Munro and Associates. 2018. "2017 Tesla Model 3 Cost Analysis."
Munro and Associates. 2020c. "2020 Tesla Model Y Cost Analysis."
Munro and Associates. 2020b. "3 Inverter Side-by-Side Analysis."
Munro and Associates. 2021. "6 Inverter Side-by-Side Analysis."
Munro and Associates. 2016. "BMW i3 Cost Analysis."
Munro and Associates. 2020a. "Twelve Motor Side-by-Side Analysis."
Newman, Kevin, John Kargul, and Dan Barba. 2015a. "Benchmarking and Modeling of a
Conventional Mid-Sized Car Using ALPHA." SAE Technical Paper 2015-01-1140.
doi: 10.4271/2015-01-1140.
Newman, Kevin, John Kargul, and Dan Barba. 2015b. "Development and Testing of an
Automatic Transmission Shift Schedule Algorithm for Vehicle Simulation." SAE International
Journal of Engines 8 (3). doi: 10.4271/2015-01-1142.
NREL. 2019. "Fuel Cell Electric Vehicle Durability and Fuel Cell Performance NREL/TP-5400-
73011." https://www.nrel.gov/docs/fyl9osti/73011.pdf.
2-97
-------
Rogozhin, A.,et al. 2009. "Using indirect cost multipliers to estimate the total cost of adding new
technology in the automobile industry." International Journal of Production Economics.
RTI International. 2018. Peer Review of EPA's Response Surface Equation Report, Report EPA-
420-R-18-006. EPA. https://nepis.epa.gov/Exe/ZyPURL.cgi?Dockey=P 100VK3Z.txt.
SAE. 2010. "Recommended Practice for Measuring the Exhaust Emissions and Fuel Economy of
Hybrid-Electric Vehicles, Including Plug-in Hybrid Vehicles." SAE Standard J1711.
U.S. EPA. 2023a. Advanced Light-Duty Powertrain and Hybrid Analysis (ALPHA) Tool.
Accessed February 2, 2023. https://www.epa.gov/regulations-emissions-vehicles-and-
engines/advanced-light-duty-powertrain-and-hybrid-analysis-alpha.
—. 2022c. "ALPHA Documentation 0.2.0 documentation." Accessed 11 17, 2022. https://epa-
alpha-model.readthedocs.io/_/downloads/en/latest/pdf/.
—. 2023b. Combining Data into Complete eMotor ALPHA Maps, https://www.epa.gov/vehicle-
and-fuel-emissions-testing/combining-data-complete-emotor-alpha-maps.
—. 2023c. Combining Data into Complete Engine ALPHA Maps. Accessed February 2, 2023.
https://www.epa.gov/vehicle-and-fuel-emissions-testing/combining-data-complete-engine-alpha-
maps.
U.S. EPA. 2016. "Cost Reduction through Learning in Manufacturing Industries and in the
Manufacture of Mobile Sources, EPA-420-R-16-018."
https://nepis.epa.gov/Exe/ZyPDF.cgi?Dockey=Pl 00PUSX.PDF.
—. 2022e. "Data from Cars used for Testing Fuel Economy." https://www.epa.gov/compliance-
and-fuel-economy-data/data-cars-used-testing-fuel-economy.
—. 2016a. "EPA-420-R-16-020: Proposed Determination on the Appropriateness of the Model
Year 2022-2025 Light-Duty Vehicle Greenhouse Gas Emissions Standards under the Midterm
Evaluation." November. https://nepis.epa.gov/Exe/ZyPDF.cgi?Dockey=PlOOQ3DO.pdf.
—. 2016b. "EPA-420-R-16-021: Proposed Determination on the Appropriateness of the Model
Year 2022-2025 Light-Duty Vehicle Greenhouse Gas Emissions Standards under the Midterm
Evaluation: Technical Support Document." November. Accessed November 3, 2022.
https://nepis.epa.gov/Exe/ZyPDF.cgi?Dockey=Pl 00Q3L4.pdf.
—. 2017a. "EPA-420-R-17-001: Final Determination on the Appropriateness of the Model Year
2022-2025 Light-Duty Vehicle Greenhouse Gas Emissions Standards under the Midterm
Evaluation." https://nepis.epa.gov/Exe/ZyPDF.cgi?Dockey=P 100QQ91.pdf.
—. 2017b. "EPA-420-R-17-002: Final Determination on the Appropriateness of the Model Year
2022-2025 Light-Duty Vehicle Greenhouse Gas Emissions Standards under the Midterm
Evaluation: Response to Comments." January.
https://nepis.epa.gov/Exe/ZyPDF.cgi?Dockey=Pl 00QQ9Y.pdf.
U.S. EPA. 2023. External Peer Review of EPA's OMEGA Model. Peer Review Report, Ann
Arbor: EPA.
—. 2024. OMEGA Documentation. https://omega2.readthedocs.io/en/2.5.0/.
2-98
-------
U.S. EPA. 2023. Peer Review of Electrified Vehicle Simulations within EPA's ALPHA Model.
Peer Review Report, Ann Arbor: EPA.
—. 2022b. Technical Publications and Presentations Concerning Benchmarking. Accessed 11 17,
2022. https://www.epa.gov/vehicle-and-fuel-emissions-testing/benchmarking-advanced-low-
emi s si on-li ght-duty-vehicl e-technol ogy.
U.S. EPA; U.S. DOT-NHTSA; CARB. 2016. "EPA-420-D-16-900: Draft Technical Assessment
Report: Midterm Evaluation of Light-Duty Vehicle Greenhouse Gas Emission Standards and
Corporate Average Fuel Economy Standards for Model Years 2022 - 2025." July.
https://nepis.epa.gov/Exe/ZyPDF. cgi/P100OXEO.PDF?Dockey=P 100OXEO.PDF.
US DOE & EPA. 2024. Download Fuel Economy Data. Accessed 02 07, 2024.
https://www.fueleconomy.gov/feg/download.shtml.
Yamagishi, Tomoya, and Takashi Ishikura. 2018. "Development of Electric Powertrain for
CLARITY PLUG-IN HYBRID, SAENo. 2018-01-0415." Int. J. Alt. Power. (SAE) 7 (3): 323-
333. doi: 10.4271/2018-01-0415.
Zhuanga, W., S. Li (Eben), X. Zhang, D. Kum, Z. Song, G. Yin, and F. Ju. 2020. "A survey of
powertrain configuration studies on hybrid electric vehicles." Journal of Applied Energy 262
(114553). doi: 10.1016/j.apenergy.2020.114553.
2-99
-------
Chapter 3: Analysis of Technology Feasibility
This chapter summarizes our assessment of the feasibility of the final greenhouse gas (GHG)
and criteria pollutant emission standards. It includes a description of the wide range of emissions
control technologies considered for criteria pollutant exhaust and evaporative emissions, GHG
emissions control, and on-board diagnostics.
EPA has assessed the feasibility of the standards in light of current and anticipated progress
by automakers in developing and deploying new emissions-reducing technologies. The primary
body of our assessment of these topics resides in the preamble, where we develop our
conclusions regarding technology feasibility including availability of advanced technologies,
PEV feasibility, critical minerals, manufacturing capacity, and mineral security, as well as other
aspects of feasibility of compliance with the standards. This Section 3.1 serves to provide a
review of key topics relating to these aspects of feasibility and additional information we
considered relating to GHG technology application to light- and medium-duty vehicles. For
references and full discussion please refer to the specific preamble sections cited in each section
and where applicable, other relevant sections of the preamble and RIA.
Section 3.1.1 reviews recent trends and feasibility of the wide range of light-duty vehicle
technologies that manufacturers have available to meet the standards. Similarly, Section 3.1.2
discusses recent trends in medium-duty vehicle technology, focusing primarily on electrification.
Section 3.1.3 reviews major highlights of our consideration of PEV technology feasibility,
largely referring to the main arguments presented in Section IV.C.7 of the preamble to this rule.
Section 3.1.4 provides a review of major highlights of our consideration of critical minerals,
battery cell manufacturing, and mineral security, and also refers to Section IV.C.7 of the
preamble. Section 3.1.5 provides technical detail on our development of a constraint on PEV
market penetration used in OMEGA, based on our assessment of lithium availability and battery
manufacturing capacity.
3.1 Vehicle Technologies and Trends
3.1.1 Light-Duty Vehicle Technologies and Trends
3.1.1.1 Advanced ICE technologies
Compliance with the EPA GHG standards over the past decade has been achieved
predominantly through the application of advanced technologies to internal combustion engine
(ICE) vehicles. For example, in the analyses performed for the 2012 rule (77 FR 62624 2012),
the Draft Technical Assessment Report (TAR) for the Midterm Evaluation (MTE) of the 2022-
2025 standards (U.S. EPA, CARB, U.S. DOT NHTSA 2016), the 2016 Proposed Determination
(U.S. EPA 2016), and the 2021 rule (86 FR 74434 2021), a significant portion of EPA's analysis
included an assessment of technologies available to manufacturers for achieving compliance
with the standards. Advanced ICE technologies were identified as playing a major role in
manufacturer compliance with the emission reductions required by those rules.
Innovation in the automobile industry has led to a wide array of technology available to
manufacturers to achieve goals for performance, fuel economy and CO2 emissions (U.S. EPA
2023). Figure 3-1 illustrates manufacturer-specific technology usage for model year 2022, with
larger circles representing higher usage rates (U.S. EPA 2023). These technologies are all being
used by manufacturers to, in part, reduce CO2 emissions and increase fuel economy. Each of the
3-1
-------
fourteen largest manufacturers have adopted several of these technologies into their vehicles,
with many manufacturers achieving very high penetrations of several technologies. It is also
clear that manufacturers' strategies to develop and adopt new technologies are unique and vary
significantly, as each manufacturer is choosing technologies that best meet the design
requirements of their vehicles and the needs of their customer base.
Engine technologies such as turbocharged engines (Turbo) and gasoline direct injection (GDI)
allow for more efficient engine design and operation. Cylinder deactivation (CD) allows for use
of only a portion of the engine when less power is needed, while stop/start systems can turn off
the engine entirely at idle to save fuel. Hybrid vehicles use a larger battery to recapture braking
energy and provide power when necessary, allowing for a smaller, more efficiently operated
engine. The hybrid category includes strong hybrid systems that can temporarily power the
vehicle without engaging the engine and smaller "mild" hybrid systems that cannot propel the
vehicle by electric power alone. Transmissions that have more gear ratios, or speeds, allow the
engine to more frequently operate near peak efficiency. Two categories of advanced
transmissions are shown in Figure 3-1. In model year 2022, hybrid vehicles reached a new high
of 10 percent of all production. This increase was mostly due to the growth of hybrids in the
truck SUV and pickup vehicle types.
Tesla-
100%
Hyundai -
25%
72%
23%
64%
37%
9%
4%
Honda -
39%
69%
30%
64%
36%
77%
10%
Kia-
27%
68%
28%
53%
49%
3%
6%
Subaru-
31%
100%
94%
73%
0%
Toyota -
6%
78%
38%
42%
25%
22%
2%
Nissan -
16%
85%
66%
32%
12%
2%
Mazda -
24%
100%|
44%
0%
VW-
81%
93%
3%
90%
72%
17%
8%
E3MW-
97%
97%
96%
58%
29%
8%
Mercedes -
90%
97%
3%
97%
64%
30%
3%
Ford-
78%
88%
21%
4%
91%
60%
6%
4%
GM-
49%
94%
49%
11%
72%
76%
2%
Stellantis -
13%
8%
27%
1%
96%
47%
17%
5%
Manufacturers -
37%
73%
16%
26%
59%
50%
10%
7%
—i 1 1 1 1 1 1 1
Turbo GDI or Cylinder CVT 7+Gears Non-Hybrid Hybrid PHEV/
GDPI Deactivation StopStart EV/FCV
Figure 3-1: Manufacturer Use of Key Technologies in Model Year 2022.
3-2
-------
3.1.1.2 Hybrid Electric Technologies
Hybrid electric vehicles (HEVs) were first introduced in the U.S. marketplace in model year
2000 with the Honda Insight and the Toyota Prius. As more models and options were introduced,
hybrid production increased to 3.8 percent of all vehicles in model year 2010, before declining
somewhat over the next several years. However, in model year 2022 hybrid production reached a
new high at 10.2 percent and is projected to reach 13.6 percent in model year 2023, as shown in
Figure 3-2 (U.S. EPA 2023).
The growth in hybrid vehicles is largely attributable to growth outside of the sedan/wagon
vehicle type. In model year 2020 the production of hybrids in the truck SUV category (largely
mild HEVs) surpassed the production of sedan/wagon hybrids for the first time and did so by
more than 50 percent. Hybrids also began to penetrate the pickup and minivan/van vehicle types.
However, there remain very few hybrid car SUVs. Sedan/wagon hybrids accounted for only 25
percent of all hybrid production in model year 2022.
The growth of hybrids in the pickup vehicle type is largely due to the introduction of "mild"
hybrid systems that are capable of regenerative braking and many of the same functions as other
hybrids but utilize a smaller battery and an electrical motor that cannot directly drive the vehicle.
These mild hybrids accounted for about 40 percent of hybrid production in model year 2022.
3-3
-------
12 5%
10 0%
Vehicle Type
Sedan/Wagon
Car SUV
Tmc< SUV
| Minivan'Vao
¦ Pickup
I 7554
c
o
o
u
O 5.0%
2 5%
0%
2000
2005
2010 2015
Model Year
¦ I
2020
Figure 3-2: Gasoline Hybrid Engine Production Share by Vehicle Type.
3.1.1.3 Plug-in Electric Vehicle Technologies
The previously described trend in application of BEV and PHEV technologies to light-duty
vehicles is evidence of a continuing shift toward electrification across the vehicle industry. As
described in detail in the Executive Summary of the preamble (Section I.A.2), recent trends in
market penetration of PEVs show that demand for these vehicles in the U.S. is increasing, as the
production of new PEVs (including both BEVs and PHEVs) is roughly doubling each year. As
also described at length in that section, manufacturers have increasingly allocated large amounts
of new investment to electrification technologies i ncluding HEVs, PHEVs and BEVs. For more
discussion of these rapidly increasing trends, see Section l.A .2 of the preamble.
The production of BEVs and PHEVs has increased rapidly in recent years. Prior to model year
2011, BEVs were available, but generally only in small numbers for lease in California. In model
year 2011 the first commercially available PHEV, the Chevrolet Volt, was introduced along with
the Nissan Leaf BEV. Many additional models have been introduced since, and in model year
2022 combined BEV/PHEV production reached almost 7 percent of all new vehicles. Combined
BEV and PHEV production is projected to reach a new high of 12 percent of all production in
model year 2023. The trend in BEVs, PHEVs, and FCEVs are shown in Figure 3-3 (U.S. EPA
3-4
-------
2023).
12%
Plug-In Hybrid EV
Electric Vehicle
Fuel Cell Vehicle
9% -
-------
10%
75%
5%
2 5%
0%-
Figure 3-4: Electric Vehicle Production Share by Vehicle Type.
Vehicle Type
Sedan/Wagon
0 Car SUV
Truck SUV
i Minivan/Van
I Pickup
.1.1
2010
2015
2020
Model Year
-------
10%
7.5%
2
CD
£
w
c
o
o
~
T>
O
5%-
2.5%
0%-
Vehide Type
Sedan/V&gon
¦ Car SUV
Truck SUV
E Mini van .'Van
2010 2D1S 2020
Model Year
Figure 3-5: Plug-In Hybrid Vehicle Production Share by Vehicle Type.
Figure 3-6 (U.S. EPA 2023) shows the range and fuel economy trends for EVs and PHEVs.
The average range of new BEVs has climbed substantially. In model year 2022 the average new
BEV range is 305 miles, or more than four times the range of an average BEV in 2011. The
range values shown for PHEVs are the charge-depleting range, where the vehicle is operating on
energy in the battery from an external source. This is generally the electric range of the PHEV,
although some vehicles also use the gasoline engine in small amounts during charge depleting
operation. The average charge depleting range for PHEVs has remained largely unchanged since
model year 2011.
Along with improving range, the fuel economy of electric vehicles has also improved as
measured in miles per gallon of gasoline equivalent (mpge). The fuel economy of electric
vehicles increased by about 10 percent between model years 2011 and 2022 (a decrease in 2022
due to the introduction of larger vehicles). The combined fuel economy of PHEVs has been more
variable but is about 30 percent lower in model year 2022 than in model year 2011. This
decrease may be attributable to the growth of truck SUV PHEVs.
3-7
-------
Range (mi) Fuel Economy (mpge)
Model Year
Figure 3-6: Charge Depleting Range and Fuel Economy for BEVs and PHEVs.
Figure 3-7 (U.S. EPA 2023) shows the model year 2022 production volume of BEVs, PHEVs
and FCEVs. More than 1.3 million BEVs, PFLEVs, and FCVs were produced in the 2022 model
year. Of those vehicles, about 78 percent were BEVs, 22 percent were PFEVs, and less than 1
percent were FCEVs.
3-8
-------
450-
400-
350-
¦ Electric Vehicle
¦ Plug-In Hybrid Electric Vehicle
Fuel Cell Vehicle
300-
250 J
o
o
o.
c
o
o
3 200-
o
E.
0.
150J
100-
50-
0-
li i l_ ...Mill
1—i—r
n r~
"i—i 1 r
/
V
Figure 3-7: Model Year 2022 Production of BEVs, PHEVs, and FCEVs.
3.1.2 Medium-duty Vehicle Technologies and Trends
The medium-duty sector is also experiencing a shift toward electrification in a similar manner
to the light-duty sector and within several important market segments. As cited in Section I.A of
the preamble, numerous commitments to produce all-electric medium-duty delivery vans have
been announced by large fleet owners including FedEx, Amazon, and Wal-Mart, in partnerships
with various OEMs. This shift to full electrification from a fleet that is currently predominantly
gasoline- and diesel-powered suggests that the operators of these fleets consider full
electrification as the best available and most cost-effective technology for meeting their mission
objectives, while also reducing the emissions from their business operations. Owing to the large
size of these vehicle fleets, this segment alone is likely to represent a significant portion of the
future electrification of the medium-duty vehicle fleet.
3-9
-------
As described in Chapter 1.2.2.1 of the Regulatory Impact Analysis and within section III. A of
the preamble, the Agency is using the term "Medium-duty vehicle" (MDV) for the first time
within its regulations. MDVs are comprised of the following weight categories:
• Class 2b - 8,501 pounds to 10,000 pounds rated gross vehicle weight (GVWR)
• Class 3 - 10,001 to 14,000 pounds GVWR
MDV also excludes what EPA defines as medium-duty passenger vehicles (MDPVs), which
are regulated along with light-duty vehicles and trucks. For more information, please refer to
section III.A.l of the preamble. MDVs can either be "incomplete" chassis cabs onto which
customized body work or beds are added after their original manufacture or are "complete"
pickup trucks or vans. Examples of incomplete vehicles customized for specific applications
include motorhomes, ambulances, wreckers, panel vans, flatbeds, etc. See Figure 3-8. In model
year 2020, less than 5 percent of MDV sales were incomplete vehicles, with the remainder being
complete.
Figure 3-8: Examples of incomplete MDV chassis finished with customized bodies for
specific applications.
MDV pickup trucks are generally built with heavier frames and designed with sufficient
powertrain, brake and suspension systems to support significantly higher towing capability than
found in light-duty pickup trucks. MDV pickup truck applications have considerable tow
capability, which can be in excess of 20,000 pounds gross combined weight rating (GCWR)
pickups with gasoline engines and can be over 40,000 pounds GCWR for pickups with diesel
engines. MDV vans have comparable payload carrying ability to MDV pickups; however, they
3-10
-------
typically have significantly lower tow capability with GCWR comparable to, or less than, many
light-duty pickups.
There are both diesel engine and spark-ignition gasoline engine applications in MDV. Their
shares of MDV sales are shown for both pickups, vans, and incomplete vehicles in Table 3-1.
Both gasoline and diesel engines used in van applications and many gasoline engines used in
pickup truck applications are derived from light-duty applications. Examples include the:
• Mercedes Benz OM654 diesel engine in the MY2023 Sprinter Van (engine family
shared with the C-Class and E-Class passenger cars and GLC CUV sold outside the
U.S.)
• Mercedes Benz M274 turbocharged GDI engine in the MY2023 Sprinter Van (engine
family shared with the C-Class and E-Class passenger cars and GLC light-duty CUV)
• Ford 3.5L EcoBoost in the MY2015-2023 Transit Van (engine family shared with the
2011-2016 Ford F150 light-duty pickup)
• GM LWN diesel engine in the Chevrolet and GMC vans (engine family shared with
Chevrolet Colorado light-duty pickup)
• RAM 6.4L Hemi in the RAM 2500 and 3500 pickups (engine family shared with
RAM light-duty pickups, and Dodge, Jeep and Chrysler passenger cars and light-duty
CUVs)
• GM L8T naturally aspirated GDI engine used in Chevrolet Silverado 2500HD and
3500HD pickups and G3500 vans and sharing the GM "Generation V" V8 engine
family with many GM light-duty trucks, CUVs, SUVs and some passenger cars.
Table 3-1: Percentage of MY2020 sales and sales volumes of pickup, van, and incomplete
MDVs by fuel type.
Pickups Vans
Fuel Type* Gasoline Diesel
MY2020 sales share 24.2% 37.1%
MY2020 sales 213.7% 327.488
Gasoline Diesel
30.4% 3.7%
269.038 32.351
Incomplete Vehicles Grand
Total
Gasoline Diesel
4.5% 0.1% 100%
40.043 978 883.694
*Other sources of powertrain energy, including electrification, accounted for <4% of MDV sales in MY2020.
While many gasoline engine families used for pickup truck applications share engine families
and/or key design elements with light-duty applications, in some cases engine block materials
may shift from cast aluminum alloy in light-duty applications to cast iron for MDV applications
(e.g., GM L8T engine). In other cases, engine families are solely used in MDV and are also
shared with heavier weight-class trucks above MDV, for example Ford's 7.3 L Super-Duty
naturally aspirated, port-fuel-injected, gasoline engine used in the F250 and F350 MDV pickups,
which is also used in the heavier Ford F450/550/600 and F650/750. Diesel equipped MDV
pickup trucks are equipped with 6L and larger engines, some of which have peak torque ratings
3-11
-------
in excess of 1000 ft-lbs. Diesel engines used in MDV pickup trucks have no light-duty
counterparts and most also share engine families with significantly heavier classes of vehicles
(e.g., weight classes 4 through 7) (40 CFR 86.1803-01 2023).
The use of commercial vans for last mile delivery in the U.S. has grown significantly since
the start of the global COVID-19 pandemic, primarily through the growth of e-commerce.24 In
the U.S., 2021 e-commerce sales totaled $870 billion, which represents an increase of over 14
percent from 2020, and over 50.5 percent compared to 2019. U.S. e-commerce represented just
over 13.2 percent of all retail sales in 2021 (U.S. Census Bureau 2022). Globally, the automotive
market supporting e-commerce was valued at over $66 billion in 2021 and is expected to grow to
over $75 billion by the end of 2022 and to over $213 billion by 2029. (Fortune Business Insights
2022). Based on the results of a recent pilot study of the electrification of commercial delivery
vans and step vans, the North American Council for Freight Efficiency identified this segment as
"100% electrifiable" (North American Council for Freight Efficiency 2021).
Vans using dedicated battery-electric vehicle (BEV) architectures are beginning to enter the
U.S. market. The first mass-produced models became available for MY2023 and additional
production volume and models have been announced for MY2024. Initial dedicated BEV van
chassis have been predominantly targeted towards parcel delivery and include the GM
BrightDrop Zevo 400 and Zevo 600; and the Rivian EDV 500 and EDV 700 (Rivian 2023)
Figure 3-9 (BrightDrop 2023). Both GM and Rivian share key electric powertrain and battery
storage components between their light-duty and/or MDPV BEV products and their dedicated
BEV commercial van products, which provides improved economies of scale for their
commercial BEV MDV vans. EPA does not require manufacturers to report the electric range of
MDVs, however manufacturers and key customers (e.g., Amazon and FedEx) appear to be
targeting approximately 150 miles of range based upon public data for battery pack capacity of
approximately 135 kWh for the EDV700, approximately 115-kWh for the Zevo 400, and
standard capacity of approximately 115-kWh for the Zevo 600 with an optional 165-kWh
capacity (Seabaugh 2022) (BrightDrop 2022) (Battery Design 2022).25
24 Commercial transactions, including retail sales, conducted electronically on the internet.
25 BrightDrop useable pack capacity calculated from: public data on GM ultium prismatic NCMA cells at 103 Ah
cell capacity, 3.7 VDC nominal cell voltage; public data on GM Ultium modules at 24 cells per module; and
BrightDrop public data on the availability of 14 module and 20 module Ultium battery packs (Battery Design 2022)
(BrightDrop 2022).
3-12
-------
Figure 3-9: Rivian EDV 700 (left) and GM BrightDrop ZEVO 600 MDV (right) vans
operated by Amazon and FedEx, respectively.
Although no PHEV pickup truck or MDV applications are yet in production, EPA believes
the PHEV architecture may lend itself well to future applications, particularly MDV pickup truck
applications at or below 10,000 pounds GVWR and MDV vans used outside of last-mile delivery
applications and thus have included PHEV MDV within our analysis for the final rule. One
major manufacturer, Stellantis, has announced that it will introduce a PHEV pickup for MY2024,
although it is still unclear if the pickup will be an MDPV or an MDV (Stellantis 2023). A MDV
PHEV pickup architecture could potentially provide several benefits: some amount of zero
emission electric range (depending on battery size); increased total vehicle range during heavy
towing and hauling operations using both charge depleting and charge sustaining modes
(depending on ICE-powertrain sizing); job-site utility with auxiliary power capabilities similar to
portable worksite generators, and the efficiency improvements normally associated with strong
hybrids that provide regenerative braking, extended engine idle-off, and launch assist for high
torque demand applications. Depending on the vehicle architecture, PHEVs used in MDV pickup
applications may also offer additional capabilities, similar to BEV pickups, with respect to
torque control and/or torque vectoring to reduce wheel slip during launch in trailer towing
applications. In addition, PHEVs may help provide a bridge for commercial consumers that may
not be ready to adopt a fully electric MDV pickup.
EPA has completed contract work to investigate likely technology architectures of both
PHEV and internal combustion engine range-extended electric light-duty and MDV pickup
trucks that provided data for the final rule. This study is summarized in Chapter 3.5.2 of the RIA.
Costs for potential PHEV designs for this application are outlined in Chapter 2.6.1.4 of the RIA.
While the agency anticipates that electrification of vans will be a cost-effective compliance
strategy for meeting the GHG and criteria pollutant standards, vehicle manufacturers may also
choose to improve their conventional, ICE-based vehicles. MDV GHG emissions can be reduced
via improving powertrain efficiency or by making improvements to road loads through improved
aerodynamics, reduced tire rolling resistance and reduced vehicle weight. For a summary of
conventional MDV GHG emissions control technology, please refer to Chapter 2.5 of the Heavy-
duty Phase 2 GHG Regulatory Impact Analysis. MDV emissions that contribute to criteria air
pollutants can be reduced by improvements to engine management systems, fuel systems,
evaporati ve emissions control systems, catalyst systems, and via the addition of modern exhaust
filtration systems such as the gasoline particulate filter (GPF). Many of the anticipated controls
3-13
-------
for future MDVs share significant design elements with criteria pollutant emissions controls used
for light-duty applications and are discussed in more detail in Chapter 3.2 of this RIA.
3.1.3 Review of Light- and Medium-Duty PEV Feasibility
In this section we briefly review the major highlights of our assessment of feasibility of PEV
technology.
The primary body of our PEV feasibility assessment resides in the preamble Section IV,
where we analyze and cite relevant evidence and form our conclusions regarding PEV feasibility
as well as other aspects of feasibility of compliance with the standards. For references and full
discussion please refer to the specific preamble sections cited below and where applicable, other
relevant sections of the preamble and RIA.
The technology trends outlined in the previous sections show that among other technologies,
BEV and PHEV technologies are being increasingly employed across the fleet in both light-duty
and medium-duty applications. This trend also serves as evidence that BEVs and PHEVs are
seen not only as an effective and feasible means to comply with emissions regulations but also as
an effective and attractive solution that can serve the functional needs of a large portion of light-
and medium-duty vehicle buyers. This represents an opportunity to accelerate needed reductions
in criteria pollutant and GHG emissions by encouraging continued uptake of these technologies
in the U.S. light- and medium-duty vehicle fleet.
As noted previously, zero- and near-zero emissions technologies are more feasible and cost-
effective now than at the time of prior rulemakings. The developments in vehicle electrification
that have brought this about are driven in part by the industry's need to compete in a diverse
market, as zero-emission transportation policies continue to be implemented across the world.
Sections I.A.2 and IV.C.l of the preamble provide a comprehensive analysis of recent events in
the growth of electrification of the automotive sector, and established a number of important
points, which are reviewed briefly below. Citations for the content in this section can be found in
the parallel discussions in Sections I.A.2 and IV.C.l of the preamble, unless specifically cited
here.
One observation of that discussion was that growth in vehicle electrification is likely being
driven in part by automakers' need to compete in a diverse global marketplace in which many
countries are continuing to implement emissions-reducing or zero-emission transportation
policies. Specifically, at least 20 countries across the world, as well as numerous local
jurisdictions in the U.S. and abroad, have announced plans to shift all new passenger car sales to
zero-emission vehicles in the coming years — Norway by 2025; Austria, the Netherlands,
Denmark, Iceland, India, Ireland, Israel, Scotland, Singapore, Sweden, and Slovenia by 2030,
Canada, Chile, Germany, Thailand, and the United Kingdom by 2035, and France, Spain, and Sri
Lanka by 2040. Many of these announcements extend to light commercial vehicles as well, and
several also target a shift to 100 percent all-electric medium- and heavy-duty vehicle sales
(Norway targeting 2030, Austria 2035, and Canada and the United Kingdom 2040). In addition,
in March 2023 the European Union approved a measure to phase out sales of ICE passenger
vehicles in its 27 member countries by 2035. Together, about half of annual global light-duty
sales are in countries with various levels of zero-emission vehicle targets by 2035, up from about
25 percent in 2022. As of late 2023, 17 automotive brands globally had announced corporate
targets for phasing out ICE technology, representing 32 percent of the global automotive market.
3-14
-------
In 2023, 22 percent of new car registrations in the European Union were either BEVs or PHEVs,
led by Norway which reached about 80 percent BEV and 89 percent combined BEV and PHEV
sales. California finalized the Advanced Clean Cars II (ACC II) rule that specifies, by 2035, all
new light-duty vehicles sold in the state are to be zero-emission vehicles. Twelve additional
states have adopted all or most of the zero-emission vehicle phase-in requirements under ACC II,
including Colorado, Delaware, Maryland, Massachusetts, New Jersey, New Mexico, New York,
Oregon, Rhode Island, Vermont, Virginia, and Washington.
In addition to spurring industry development of BEV and PHEV technology, developments
such as these suggest a growing global consensus that BEV and PHEV technologies are broadly
feasible as emissions-reducing technologies. For additional details and citations regarding the
domestic and global developments that we have briefly reviewed here, please refer to Sections
I.A.2 and IV.C.l of the preamble.
Preamble Sections I.A.2 and IV.C. 1 also establish that demand for these vehicles in the U.S.
is rapidly increasing, even under current standards. The production of new PEVs (including both
BEVs and PHEVs) is growing steadily, expected to be 11.8 percent of U.S. light-duty vehicle
production in MY 2023, up from 6.7 percent in MY 2022, 4.4 percent in MY 2021 and 2.2
percent in MY 2020. In California, new light-duty PEV sales have reached 25.1 percent through
the third quarter of 2023, after reaching 18.8 percent in 2022, up from 12.4 percent in 2021. The
number of PEV models available for sale in the U.S. has grown from about 24 in MY 2015 to
about 60 in MY 2021 and over 180 in MY 2023, with offerings in a growing range of vehicle
segments. MY 2023 BEVs and PHEVs are now available as sedans, sport utility vehicles, and
pickup trucks, with the greatest offering in the crossover/SUV segment.
U.S. new PEV sales in 2023 surpassed 1.4 million, an increase of more than 50 percent over
the 807,000 sales that occurred in 2022. This represents 9.1 percent of new light-duty passenger
vehicle sales in 2023, up from 6.8 percent in 2022 and 3.2 percent the year before. Despite talk
of a reduced rate of growth in PEV sales in 2023 (see Preamble I.A.2), the growth trend seen in
previous years continues to be seen in 2023 data (see Figure 1 of Preamble I.A.2).
Before the Inflation Reduction Act (IRA) became law, analysts were already projecting that
significantly increased penetration of plug-in electric vehicles would occur in the United States
and in global markets. As fully discussed in Preamble I.A.2, projections have suggested rapid
growth. In 2021, IHS Markit predicted a nearly 40 percent U.S. PEV share by 2030. Projections
made in 2022 by Bloomberg New Energy Finance (BNEF) suggested that under then-current
policy and market conditions, and prior to the IRA and this final rule, the U.S. was on pace to
reach 43 percent PEVs by 2030, and when adjusted for the effects of the IRA, this estimate
increased to 52 percent. Another study by the International Council on Clean Transportation
(ICCT) and Energy Innovation that includes the effect of the IRA projects that the share of BEVs
will increase to 56 to 67 percent by 2032. Although the assumptions and other inputs to these
forecasts vary, they point to greatly increased penetration of electrification across the U.S. light-
duty fleet in the coming years, without specifically considering the effect of increased emission
standards under this rule.
A similar trend was seen in forecasts reviewed for the global market, showing that growth in
PEV sales in the U.S. is part of a global trend. Global light-duty passenger PEV sales surpassed
10 million in 2022, up from 6.6 million in 2021, bringing the total number of PEVs on the road
to more than 26 million globally. For fully-electric BEVs, global sales rose to 7.8 million in
3-15
-------
2022, an increase of about 68 percent from the previous year and representing about 10 percent
of the new global light-duty passenger vehicle market. Leading sales forecasts predict that PEV
sales will continue to accelerate globally in the years to come. For example, in June 2023,
Bloomberg New Energy Finance reported that global PEV sales were 10.5 million in 2022 and
forecasted that annual sales will rise to 27 million in 2026 (implying an annual growth rate of
about 27 percent from 2022), with total global PEV stock rising from 27 million in 2022 to more
than 100 million by 2026.
In the preamble discussions we also cited extensive evidence that, while ICE vehicles and
HEVs together retain the largest share of the market, the year-over-year growth in U.S. PEV
sales suggests that an increasing share of new vehicle buyers are concluding that a PEV is the
best vehicle to meet their needs. PEV owners often describe specific advantages of PEVs as key
factors motivating their purchase, such as responsive acceleration, improved performance and
handling, quiet operation, lower cost of ownership, and the ability to charge at home. A 2022
survey by Consumer Reports shows that, even at a time when many consumers are not yet as
familiar with BEVs as with ICE vehicles, more than one-third of Americans would either
seriously consider or definitely buy or lease a BEV today if they were in the market for a vehicle.
Because familiarity with BEVs promotes acceptance (see for example section IV.C.6 of the
preamble and RIA Chapter 4), this share is expected to rapidly grow as familiarity increases in
response to increasing numbers of BEVs on the road and growing visibility of charging
infrastructure. The U.S. Bureau of Labor Statistics has indicated that growing consumer demand
and growing automaker commitments to electrification are important factors in the growth of
PEV sales and that growth will be further supported by policy measures including the BIL and
the IRA. Most PEV owners who purchase a subsequent vehicle choose another PEV, and often
express resistance to returning to an ICE vehicle after experiencing PEV ownership. Many
analysts believe that as PEVs continue to increase in market share, PEV ownership will continue
to broaden its appeal as consumers gain more exposure and experience with the technology and
with the benefits of PEV ownership, with some analysts suggesting that rapidly accelerating PEV
adoption will then result.
We also cited evidence that, while PEVs are typically offered at a higher price than
comparable ICE vehicles at this time, the price difference for BEVs, which have only an electric
powertrain, is widely expected to narrow or disappear as the cost of batteries and other
components fall in the coming years. Among other evidence, we noted that an emerging
consensus suggests that purchase price parity is likely to begin occurring by the mid- to late-
2020s for some vehicle segments and models, and for a broader segment of the market on a total
cost of ownership (TCO) basis. By some accounts, a compact car with a relatively small battery
(for example, a 40 kilowatt-hour (kWh) battery and approximately 150 miles of range) may
already be possible to produce and sell for the same price as a compact ICE vehicle. For larger
vehicles and/or those with a longer range (either of which necessitate a larger battery), many
analysts expect examples of price parity to increasingly appear over the mid- to late-2020s.
Prospects for price parity improve greatly when considering state and federal purchase
incentives. For example, the 30D Clean Vehicle Credit under the IRA provides a purchase
incentive of up to $7,500, effectively making some BEVs more affordable to buy today than
comparable ICE vehicles. Kelley Blue Book already estimates that the lowest TCO for the full-
size pickup and luxury car classes of vehicle are BEVs. Based on average annual mileage,
BloombergNEF states that in the U.S., electric SUVs have already achieved lower TCO than
3-16
-------
similar ICE vehicles, and for higher mileages, BEVs have lower TCO than similar small,
medium, and large ICE vehicles. Because businesses tend to pay close attention to TCO of
business property, TCO parity of BEVs is likely to be of particular interest to commercial owners
and fleet operators.
As further evidence of the feasibility of BEVs and PHEVs as an emissions-reducing
technology, we also cited a large number of announcements made by all of the major automakers
in the past several years, signaling a rapidly growing shift in product development focus toward
electrification. Section I.A.2 and IV.C.l of the preamble introduces and provides citations for
many of these announcements.
General Motors announced plans to become carbon neutral by 2040, including an effort to
shift its light-duty vehicles entirely to zero-emissions by 2035. In March 2021, Volvo announced
plans to make only electric cars by 2030, and Volkswagen announced that it expects half of its
U.S. sales will be all-electric by 2030. In April 2021, Honda announced a full electrification plan
to take effect by 2040, with 40 percent of North American sales expected to be fully electric or
fuel cell vehicles by 2030, 80 percent by 2035 and 100 percent by 2040. In May 2021, Ford
announced that they expect 40 percent of their global sales will be all-electric by 2030. In June
2021, Fiat announced a move to all electric vehicles by 2030, and in July 2021 its parent
corporation Stellantis announced an intensified focus on electrification, including both BEVs and
PHEVs, across all of its brands. Also in July 2021, Mercedes-Benz announced that all of its new
architectures would be electric-only from 2025, with plans to become ready to go all-electric by
2030 where possible. In August 2021, many major automakers including Ford, GM, Stellantis,
BMW, Honda, Volkswagen, and Volvo, as well as the Alliance for Automotive Innovation,
expressed continued commitment to their announcements of a shift to electrification, and
expressed their support for the goal of achieving 40 to 50 percent sales of zero-emission vehicles
by 2030. In December 2021, Toyota announced plans to introduce 30 BEV models by 2030. In
August 2023, Subaru announced that its previous plan to target 40 percent combined HEVs and
BEVs was being revised to 50 percent BEVs globally by 2030.
In addition to BEV technology, some automakers have also indicated a strong role for PHEVs
in their product planning. For example, Toyota continues to anticipate PHEVs forming an
increasing part of their offerings, and Stellantis will be introducing a plug-in version of its Ram
pickup for MY 2024. As discussed in more detail in Section IV.C. 1 of the preamble, the number
of PHEV and BEV models has steadily grown and manufacturer announcements signal the
potential for significant growth in the years to come. We also showed a tabulation of these and
other OEM announcements that indicates that the sales collectively implied by such
announcements to date would amount to about 49 percent new light-duty zero-emission vehicle
sales in the U.S. by 2030.
In the second half of 2023, some automakers announced changes to previously announced
investment plans and made statements suggesting increased attention to PHEVs or HEVs in their
future product plans. For example, in mid-2023, Ford paused construction (later restarted) of
their recently announced battery plant in Marshall, Michigan, and in November 2023 announced
a reduction in the size of the plant from 50 GWh to 20 GWh. Later in 2024 Ford also signaled a
growing interest in producing HEVs and a shift from large BEV SUVs toward smaller BEVs,
while General Motors indicated increased attention toward producing PHEVs in addition to
BEVs. We review several other examples of announced adjustments to previously announced
3-17
-------
investment or product plans in Section I. A.2 of the preamble. There we also noted that some
industry analysts had connected such adjustments to a purported drop in PEV demand or a
weakening of manufacturer interest in investing in PEV technology. In that discussion, we
considered these developments carefully, particularly in light of the larger context of information
about manufacturer plans, including comments submitted by the manufacturers on this
rulemaking and our ongoing engagement with the manufacturers. Overall, EPA finds that the
recent announcements do not reflect a significant change in manufacturer intentions regarding
PEVs generally or specifically through the 2027-2032 timeframe of this rule. For more
discussion on this topic, refer to section I.A.2 of the preamble.
In the Preamble we also cited evidence documenting numerous commitments to produce all-
electric medium-duty delivery vans, which have been announced by large fleet owners including
FedEx, Amazon, and Wal-Mart, in partnerships with various OEMs. For example, Amazon has
deployed thousands of electric delivery vans in over 100 cities, with the goal of 100,000 vans by
2030. Many other fleet electrification commitments that include large numbers of medium-duty
and heavier vehicles have been announced by large corporations in many sectors of the
economy, including not only retailers like Amazon and Walmart but also consumer product
manufacturers with large delivery fleets (e.g. IKEA, Unilever), large delivery firms (e.g. DHL,
FedEx, USPS), and numerous firms in many other sectors including power and utilities, biotech,
public transportation, and municipal fleets across the country. As another example, Daimler
Trucks North America announced in 2021 that it expected 60 percent of its sales in 2030 and 100
percent of its sales by 2039 would be zero-emission.
We also provided numerous citations showing evidence that these announcements and others
like them continue a pattern over the past several years in which most major manufacturers have
taken steps to aggressively invest in zero-emission technologies and reduce their reliance on the
internal-combustion engine in various markets around the globe.
One cited analysis indicated that 37 of the world's automakers are planning to invest a total of
almost $1.2 trillion globally by 2030 toward electrification, a large portion of which would be
used for construction of manufacturing facilities for vehicles, battery cells and packs, and
materials. This would support up to 5.8 terawatt-hours of battery production and 54 million
BEVs per year globally. Another cited analysis showed that a significant shift in North American
investment is occurring toward electrification technologies, with more than 90 percent ($36
billion of about $38 billion) of total automaker manufacturing facility investments announced in
2021 being slated for electrification-related manufacturing in North America, with a similar
proportion and amount on track for 2022.
In September 2021, Toyota announced large new investments in battery production and
development to support an increasing focus on electrification, and in December 2021, announced
plans to increase this investment. In December 2021, Hyundai closed its engine development
division at its research and development center in Namyang, South Korea in order to refocus on
BEV development. In summer 2022, Hyundai announced an investment of $5.5 billion to fund
new battery and electric vehicle manufacturing facilities in the state of Georgia, and recently
announced a $1.9 billion joint venture with SK to fund additional battery manufacturing in the
U.S. In 2023, Ford announced plans for a new battery plant in Michigan, part of $17.6 billion in
investments in electrification announced by Ford and its partners since 2019. By mid-2023 the
International Energy Agency indicated that as of the previous March, major manufacturers had
3-18
-------
announced post-IRA investments in North American supply chains totaling at least $52 billion,
mostly in battery manufacturing, battery components and vehicle assembly. By January 2024, a
White House accounting of BIL and IRA investments indicated that the total had increased to at
least $155 billion. The U.S. Department of Energy indicates this represents over $120 billion in
over 200 new or expanded minerals, materials processing, and manufacturing facilities and over
$35 billion in over 140 new or expanded sites for EY assembly, EV component, or charger
manufacturing.
The following chart (Figure 3-10) more clearly illustrates how these and many other instances
of North American investments in battery and electric vehicle and component manufacturing
have added up to a picture of steady and robust growth in investment commitments in recent
years (ANL 2024a).
300
ST 250
yy
c
a>
E
8 200
u
>-
Q.
a.
150
100
£•
a)
ro
m 50
~ EVSE
0 EV Components
~ EV Assembly
0 Pack
¦ Cell
¦ Other Components
~ Separator
E9 Electrolyte
~ AAM
~ CAM
~ Minerals
1/1/2021 7/1/2021 1/1/2022 7/1/2022 1/1/2023 7/1/2023 1/1/2024
Figure 3-10 North American battery and electric vehicle investments classified by
manufacturing product
Manufacture and sale of PEVs is anticipated to grow significantly in the United States over
the next decade as provisions of the Inflation Reduction Act of 2022 (IRA) continue to take
effect (Public Law 117-169 2022). The IRA has key provisions that will reduce the cost of PEVs
to consumers, reduce the cost of battery manufacturing in the U.S. for automakers, and foster
significant emissions reductions from the U.S. electric power sector. Manufacturing-related
provisions include the Domestic Manufacturing Conversion Grant Program, Advanced
Technology Vehicle Manufacturing Program, expanded authorities for the DOE Loan Programs
Office, and the 45X Advanced Manufacturing Production Credit. Incentives for the purchase of
clean vehicles include the 30D Clean Vehicle Tax Credit, 45W Clean Commercial Vehicle
3-19
-------
Credit, and the 25E credit for purchase of used clean vehicles. The 30C credit also addresses
aspects of charging equipment cost. There are also power sector provisions such as the Clean
Electricity Production and Investment Tax Credits as well as incentives for renewable electricity
generation and grid battery storage. An Existing Nuclear Production Tax Credit is oriented to
extending nuclear EGU service life. There is also a 45Q Carbon Capture and Storage Tax Credit.
For further discussion of the impacts of the IRA on the electric power sector, please refer to
Chapter 5.2.3 of the RIA. Our modeling assumptions regarding IRA manufacturing and
consumer purchase credits are discussed in more detail in Preamble IV.C.2 and elsewhere.
PEVs, which include BEVs and PHEVs, can fill the same role as ICE vehicles for most if not
all light- and medium-duty vehicle owners. PHEVs have both an electric powertrain and a
gasoline engine and as previously discussed in section IV.C.l of the preamble, are particularly
good candidates for some demanding heavy-duty applications, although BEVs are capable of
these uses as well. BEVs rely solely on the battery for energy and so differ in some ways from
ICE vehicles and PHEVs in the way they are refueled. Obviously, charging a BEV is not exactly
the same as refueling an ICE vehicle with gasoline. While BEVs generally take longer to charge
than an ICE vehicle takes to refuel, charging does not need to be attended and for many users can
be done at home. BEVs can also be used in cold and hot climates. Just as with ICE vehicles, cold
or hot weather can increase energy consumption due to use of cabin heating and cooling and
defrosting. Because BEVs are more efficient, less waste heat is available to heat the cabin, so the
energy must come from the battery. However, climate control requires only a fraction of the
energy needed to drive a vehicle, and most of today's BEVs have a substantial driving range,
often 300 miles or more, that can easily accommodate climate control needs. Heat pump
technology can reduce energy consumption further and is increasingly being used in BEV
climate control systems. Cold weather can also affect charging speed, but most BEVs are
programmed to warm the battery while charging in cold temperatures and have thermal
management systems to manage battery temperatures.
In addition to feasibility of PEV technology itself, EPA has also performed extensive analysis
of other factors in PEV feasibility beyond the vehicle itself. These include the availability and
projected growth of charging infrastructure both at home and in public places, the ability of the
electric grid to support the additional electric demand for charging, and consumer acceptance of
PEVs, among many other factors.
We expect that through 2055 the majority of light and medium duty PEV charging will occur
at home, but we recognize the need for additional public charging infrastructure to support
anticipated levels of PEV adoption. As discussed in preamble Section IV.C.5 of the preamble
and RIA Chapter 5.3, charging infrastructure has grown rapidly over the last decade, and
investments in charging infrastructure continue to grow. Based on our evaluation of the record,
EPA has found that the market for charging is already responding to increased demand through
investments from a wide range of public and private entities, and that it is reasonable to expect
the market will continue to keep up with demand. We further anticipate these final standards will
encourage additional investments in charging infrastructure.
EPA does not find that the increase in electricity consumption associated with modeled
increases in PEV sales will adversely affect reliability of the electric grid, and, as explained in
Section IV of this preamble and Chapter 5 of the RIA, more widespread adoption of PEVs could
have significant benefits for the electric power system.
3-20
-------
Our modeling also incorporates constraints related to consumer acceptance. Under our central
case analysis assumptions, the model anticipates that consumers will in the near term tend to
favor ICE vehicles over PEVs when two vehicles are comparable in cost and capability. Taking
into account individual consumer preferences, we anticipate that PEV acceptance and adoption
will continue to accelerate as consumer familiarity with PEVs grows, as demonstrated in the
scientific literature on PEV acceptance and consistent with typical diffusion of innovation.
Adoption of PEVs is expected to be further supported by expansion of key enablers of PEV
acceptance, namely increasing market presence of PEVs, more model choices, expanding
infrastructure, and decreasing costs to consumers. See also Section IV.C.5 of the preamble and
RIA Chapter 4.
More detail about our technical assessment, and the assumptions for the production feasibility
and consumer acceptance of PEVs is provided in Section IV of this preamble, and Chapters 2, 3,
4, and 6 of the RIA.
In considering feasibility of the standards, EPA also considered the impact of available
compliance flexibilities on automakers' compliance options, as well as constraints posed by the
typical cadence of manufacturer redesign cycles. In Section V.B of the preamble, we described
how EPA's technical assessment accounts for redesign limits.26 Once a redesign opportunity is
encountered, we have assumed limits to the rate at which a manufacturer can ramp in the
transition from an ICE to a BEV vehicle. We have also applied limits to the ramp up of battery
production, considering the time needed to increase the availability of raw materials and expand
battery production facilities. These limits as they are applied in OMEGA are discussed in
Chapter 2 of the RIA.
We also consider feasibility from the perspective of critical mineral availability,
manufacturing capacity for battery cells and related products and components, and mineral
security, which are introduced in Section IV.C.7 of the preamble and further examined in the
next section (3.1.4) of this RIA.
In Section V.B of the preamble, we cite many of our findings on feasibility to assess the
overall technological feasibility and sufficiency of lead time necessary for manufacturers to meet
the standards using the technologies that are available to them. Our assessment shows that there
is sufficient lead time for the industry to deploy existing technologies, including increasing
proportions of PEV technology, more broadly and thereby successfully comply with the final
standards. There we also describe the levels of ICE vehicle, HEV, PHEV, and BEV penetration
indicated by our compliance analysis. The central analysis combined with the various
sensitivities we perform (for example, high and low battery costs, among others) show that
manufacturers can comply with the standards with varying percentages of each technology.
3.1.4 Additional Background on Critical Minerals and Manufacturing
This RIA section provides (a) a brief review of some of the major points and evidence that
EPA examined in reaching the conclusions presented in Section IV.C.7 of the preamble, and (b)
additional background information that we considered regarding critical minerals. EPA has also
26 In our compliance modeling, we have limited vehicle redesign opportunities in our compliance modeling to every
7 years for pickup trucks, and 5 years for all other vehicles.
3-21
-------
carefully considered the findings of the March 2024 ANL report "Securing Critical Materials for
the U.S. Electric Vehicle Industry" (ANL 2024b) as well as the February 2024 ANL Report
"Quantification of Commercially Planned Battery Component Supply in North America through
2035" (ANL 2024a)
This Section 3.1.4 is meant to serve primarily as a review. The primary body of our
assessment of battery cell and cell component manufacturing, critical minerals, and mineral
security resides in the Preamble IV.C.7, where we analyze and cite all relevant evidence and
form our conclusions regarding these topics as they relate to compliance with the standards. For
references and full discussion please refer to the specific preamble sections cited below and
where applicable, other relevant sections of the preamble and RIA.
3.1.4.1 Review of Key Developments Considered
In IV.C.7 of the preamble, we considered issues related to manufacturing capacity, critical
minerals, and mineral security from the perspective of industry's ability to comply with the
standards. In that discussion, we reviewed the key themes of public comments that we received
on these topics and described the additional research we had conducted to address these
comments and represent the latest and best additional information since the issuance of the
proposal. There we presented the primary evidence and developed our conclusion that issues
related to manufacturing capacity and critical mineral availability will not prevent manufacturers
from meeting the standards, and that the standards can be met without adverse impact on national
security.
The preamble discussion establishes a number of key observations about the status of critical
minerals and manufacturing capacity, and the outlook for development of the supply chain in
response to industry investment and government policy. This section provides a review of some
of our key observations. For a full discussion of the evidence we considered and how we
developed our findings, see Preamble IV.C.7.
We noted that about 57 percent of cells and 84 percent of assembled packs sold in the U.S.
from 2010 to 2021 were manufactured in the U.S., and due to continued production largely by
the same OEMs represented in the data, as well as the large amount of announced U.S. capacity
under construction or planned, this is likely to continue to be the case going forward. This
suggests that PEV production in the U.S. need not be heavily reliant on foreign manufacture of
battery cells or packs as PEV penetration increases.
Many automakers are building battery and cell manufacturing facilities in the U.S. and are
also taking steps to a secure a supply of minerals and commodities either domestically or through
FTA or allied trade partners to supply production for these plants. Our analysis of constructed
and planned plant capacity for assembly of battery cells indicates that battery manufacturing
capacity does not appear to pose a critical constraint to increased production of PEVs to meet
anticipated globally or domestic demand. Domestically, our analysis of current capacity and
construction announcements made by the major automakers and suppliers indicates that the U.S.
will have more than enough cell manufacturing capacity to supply U.S. demand under the final
standards.
We also drew observations regarding which minerals are of greatest concern as a potential
constraint on PEV production during the time frame of the rule. Mineral demand for ICE catalyst
3-22
-------
production is relatively stable and would not be expected to increase as a result of electrification.
Rare earths used in permanent magnet motors have potential alternatives in the use of induction
machines or other electric machine technologies that do not require rare-earth magnets, or in the
use of advanced ferrite or other advanced magnets. On a quantity basis and probably also on a
value basis, battery minerals are likely to be the most important mineral-related factor of
consideration for PEV production during the time frame of the rule. Of these, the most attention
is commonly given to lithium, nickel, cobalt, and graphite. Manganese is also used, but due to
available world supplies and relative level of importance across the clean energy sector, it is not
designated as critical by DOE.
Currently, most mining and refining of these minerals occurs outside of the U.S. The U.S.
does not lack significant natural deposits of some of these minerals, for example graphite and
lithium, but relatively little mining and refining capacity is currently in operation or remains
undeveloped. The development of mining and refining capacity in the U.S. is a primary focus of
industry toward building a robust domestic supply chain for electrified vehicle production. For
example, LG Chem has announced plans for a cathode material production facility in Tennessee,
said to be sufficient to supply 1.2 million high-performance electric vehicles per year by 2027.
We review other announcements like this in the preamble and in this section.
We also noted that further development of a domestic mineral supply chain will be
accelerated by the provisions of the Inflation Reduction Act (IRA) and the Bipartisan
Infrastructure Law (BIL), as well as ongoing efforts by the Executive Branch. The IRA offers
sizeable tax provisions that incentivize domestic production of batteries and critical minerals,
including a $7,500 Clean Vehicle Credit for vehicles manufactured in North America that use
domestically produced components and mineral products, and production tax credits that apply
to domestically produced cells, modules, electrode active materials, and critical minerals, that
can reduce battery manufacturing cost by a third or more. Among much other funding, the BIL
provides $7.9 billion to support development of the domestic supply chain for battery
manufacturing, recycling, and critical minerals. Provisions extend across critical minerals mining
and recycling research, USGS energy and minerals research, rare earth elements extraction and
separation research and demonstration, and expansion of DOE loan programs in critical minerals
and recycling. Through these provisions DOE is actively working to prioritize points in the
domestic supply chain to target with accelerated development, and rapidly funding those areas
through numerous programs and funding opportunities. With BIL funding and matching private
investment, more than half of the capital investment that the Department of Energy's Li-Bridge
alliance considers necessary for supply chain investment to 2030 has already been committed.
Specifically, a large amount of funding for battery production is being offered by the federal
government through IRA tax credits, loans through the DOE Loans Program Office, and DOE
Office of Manufacturing and Energy Supply Chains (MESC), as seen in the following table
(ANL 2024a):
3-23
-------
Table 3-2 Summary of Funding Programs for U.S. Battery Production
Program
Battery Materials
Processing Grants &
Battery Manufacturing
and Recycling Grants
(MESC)
Domestic Manufacturing
Conversion Grants
(MESC)
ATVM (LPO)
Title 17 (LPO)
48C Qualifying
Advanced Energy Tax
Credit (IRS, MESC)
45X Advanced
Manufacturing
Production Tax Credit
(IRS)
Funding Allocated*
~$1.9B
$0
Total Available**
~$4.1B
$2B
~$15.9B
S398.6M
$0
~$49.8B
$60B
$10B
No limitation
Period of Availability
2022-2026; Until
Expended***
To remain available
through 9/30/2031
No restriction
No restriction
Until expended
For critical minerals:
permanent For other
items: full credit
available between 2023-
29 with phase down
from 2030-32
Project Examples
CAM and A AM
production, separator
production, precursor
materials production,
battery cell production.
Eligible projects include
facilities to produce
components for electric
vehicles.
Battery cell production,
lithium carbonate
production, AAM
production, foil
production, CAM
production.
Zinc bromine battery
energy storage systems.
Eligible projects include
production and recycling
of clean energy
technologies, critical
minerals processing and
recycling.
Eligible projects include
battery components,
critical minerals,
inverters, components
for solar and wind
energy technology.
*Funding announced since 2021, as of February 2024, for projects related to the scope of the cited ANL study (cells, packs, CAM, AAM,
electrolyte, foil, separator, precursor materials). Includes conditional commitments (LPO only)
"For grants, the total available is the total allocated subtracted from the allocation, and indicates how much grant funding is left. For LPO, this
number represents approximate loan authority available as of January 2024, reported by LPO.
***For the purposes of this table, the Battery Materials Processing Grants & Battery Manufacturing and Recycling Grants are combined. These
two programs are authorized separately in the IIJA. Their periods of availability are listed respectively.
A substantial portion of this supporting industrial policy is still unfolding. This includes final
rulemaking and Treasury guidance for various details of the IRA tax credits; the submission,
selection, and award of second round of funding from the Battery Materials Processing and
Manufacturing Grants program by January 2025 (IIJA section 40207) and the 48C tax credit
(Qualifying Advanced Energy Project Credit), and, respectively, final interpretive guidance and
rulemaking from the Department of Energy and the Department of the Treasury on Foreign
Entities of Concern (FEOC) and Excluded Entities for the 30D tax credit (Clean Vehicle Credit).
We also noted the following observations about forecast global supplies of refined critical
minerals. An analysis by the Department of Energy's Li-Bridge based on data from a leading
mineral analysis firm indicated that no shortage of cathode active material is expected globally
through 2035. Despite recent short-term fluctuations in price, leading analyst firms currently are
projecting mineral prices to stabilize through the second half of the 2020s. Prices for battery
minerals have fallen considerably in the last year and forecasts to 2029 indicate that prices are
expected to remain stable at or slightly lower than current levels. This forecast stability also
suggests that industry sentiment does not expect a critical long-term shortage to develop given
current resource identification and the current level of investment activity to develop resources
as well as increase manufacturing capacity for all important inputs to cell production. For more
discussion of these trends with specific references and examples, see Preamble IV.C.7.
3-24
-------
EPA also noted that the global minerals industry is already anticipating and preparing for
accelerated growth in demand for critical minerals resulting from already-existing expectations
of greatly increased global PEV production and sales in the future, as well as expectations of
growing demand for these materials in other areas of clean energy and decarbonization. Thus, in
the context of evaluating the impact of the standards on demand for critical minerals and
development of the domestic supply chain, EPA recognizes that much of the anticipated growth
in global mineral demand stems not from the incremental effect of the standards but from these
ongoing forces that are already driving the global industry to increase mineral production.
Relatedly, EPA noted also that the IRA, the BIL, and ongoing activity on the part of
Executive Branch agencies are actively addressing the need for further development of the
domestic supply chain to supply growing demand for critical minerals. The provisions of the
IRA and BEL were in fact developed with the intent of growing the domestic supply chain for
critical minerals and related products and to achieve mineral security as the industry pursues
clean energy technology. Accordingly, EPA expects that the BIL and IRA will be very helpful
toward meeting incremental needs of the supply chain under the standards.
Supply & Demand-side Incentives
Extract
Refine/Process Recycle
Manufacture
Assemble
Drive/Charge
11JA 40207 Battery
Grants
Advanced Technology
Vehicle Manufacturing
Loans
Domestic Manufacturing
Conversion Grants
Qualifying Advanced
Energy Project Credit
(48C) Program
45X Advanced
Manufacturing
Production Tax Credit
30D Clean Vehicle Credit
45W Commercial Clean
Vehicle Credit
25E Previously-Owned
Clean Vehicle Credit
30C Alternative Fuel
Vehicle Refueling
Infrastructure Credit
Critical Minerals, Constituent Materials, Battery Components
Critical Minerals, Constituent Materials, Battery Cells, EV Wiring
LMHDV Electric Vehlcies and Components
Critical Materials
-
¦
Electric Grid Mod. Equipm. & Components,
LMHDEV Tech, Components. Materials
Critical Minerals (10%),
Electrode Active Materials (10%)
Battery Cells ($35/kWh)
Modules ($10/kWh)
Critical Minerals from U.S. & FTA Countries from N. America Battery Components from N. America
$3,750x2 = $7,500
$7,500 <14,000 lbs.
$40,000 2 14,000 lbs.
$4,000
$l,000/item home
$100,000/item
commercial
Figure 3-11: Breadth of BIL and IRA supply and demand side incentives underway to
build the supply chain.
This section has provided only a brief review of some key highlights related to manufacturing
capacity and critical minerals. The primary body of our assessment of battery cell and cell
component manufacturing, critical minerals, and mineral security resides in the Preamble IV.C.7,
where we analyze and cite all relevant evidence and form our conclusions regarding these topics
as they relate to compli ance with the standards. For references and full discussion please refer to
the specific preamble sections cited below and where applicable, other relevant sections of the
preamble and RIA.
3-25
-------
3.1.4.2 Background on Global Distribution and Production of Critical Minerals
Here we provide background on the worldwide sources of critical minerals as they are
currently understood. This discussion only presents additional background information that EPA
considered in its analysis of critical minerals, in addition to that examined and discussed in the
primary discussion. For our primary discussion on critical minerals, see section IV.C.7 of the
preamble.
Section IV.C.7 of the Preamble discussed several critical minerals important to battery cell
production including lithium, nickel, cobalt, graphite, and manganese. In that section we
described these minerals and their processing into battery materials as being widely but not
evenly distributed across the world.
As shown in Figure 3-12 the IEA estimates that in 2019 about 50 percent of global nickel
production occurred in Indonesia, Philippines, and Russia, with the rest distributed around the
world. Nearly 70 percent of cobalt originated from the Democratic Republic of Congo, with
some significant production in Russia and Australia, and about 20 percent in the rest of the
world. More than 60 percent of graphite production occurred in China, with significant
contribution from Mozambique and Brazil for another 20 percent. About half of lithium was
mined in Australia, with Chile accounting for another 20 percent, and China about 10 percent
(IEA 2022).
Share of top three producing countries in total production for selected minerals and fossil fuels, 2019
Oil
o Natural gas
United States
Copper
Nickel
Cobalt
2
Graphite
i
Rare earths
Lithium
Platinum
Saudi Arabia Russia
United States
Russia Iran
Chile Peru
China
Indonesia
Philippines
Russia
Russia Australia
Australia
Mozambique Brazil
Myanmar
Chile
South Africa
Russia
0%
20%
40%
60%
100%
Figure 3-12: Share of top three producing countries for critical minerals and fossil fuels in
2019 (IEA).
According to the Administration's 100-day review under E.O. 14017, of the major actors in
mineral refining, 60 percent of lithium refining occurred in China, with 30 percent in Chile, and
10 percent in Argentina. 72 percent of cobalt refining occurred in China, with another 17 percent
distributed among Finland, Canada, and Norway. 21 percent of Class 1 nickel refining occurred
in Russia, with 16 percent in China, 15 percent in Japan, and 13 percent in Canada (The White
House 2021). Similar conclusions were reached in an analysis by the International Energy
Agency, shown in Figure 3-13.
3-26
-------
Share of processing volume by country for selected minerals, 2019
0% 20% 40% 60% 80% 100%
Figure 3-13: IEA accounting of share of refining volume of critical minerals by country
(IEA 2022).
Since the proposal, the Department of Energy (DOE) worked with ANL to provide an
independent analysis of the outlook for critical minerals including nickel, cobalt, graphite,
lithium, and manganese (ANL 2024b). ANL consulted the latest available announcements and
forecasts to develop an up-to-date assessment of activity in advancing the availability of these
minerals on a global and domestic basis. Findings from this work and how we considered them
in our assessment are discussed in Section IV.C.7 of the preamble to this rule. A summary of
some high level takeaways from this work relating to the distribution of minerals across the
world are provided here.
Although the U.S. has nickel reserves, and opportunity also exists to recover significant nickel
from mine waste remediation and similar activities, at present it is more convenient to import it
from established producers in other countries, with 68 percent coming from Canada, Norway,
Australia, and Finland, countries with which the U.S. has good trade relations; according to the
USGS, ample reserves of nickel exist in the U.S. and globally, potentially constrained only by
processing capacity (The White House 2021). ANL notes that currently, there is no Class I
(battery grade) nickel production or refining in the U.S, and that there has been an influx of
investment by China in Indonesia, a major global producer of nickel (ANL 2024b).
The U.S. has numerous cobalt deposits, but few are developed, although some have produced
cobalt in the past; about 72 percent of U.S. cobalt consumption is currently imported (USGS
2020). ANL notes that China controls about 50 percent of cobalt production in the Democratic
Republic of Congo (DRC), a major global producer of cobalt.
Similar observations may be made about graphite. The U.S. has significant deposits of natural
graphite, but graphite has not been produced in the U.S. since the 1950s and significant known
resources remain largely undeveloped (USGS 2022). ANL notes that China dominates natural
graphite production and has been a major source of U.S imports. ANL also indicates that
meeting U.S. demand with natural supply from free trade agreement (FTA) and Mineral Security
Partnership (MSP) (State Department 2023) countries is unlikely in the near term, but medium
term synthetic graphite scaling has potential to mitigate graphite risk (Reuters 2023). Another
3-27
-------
concern is that in 2023, China imposed an export permit requirement on graphite, which will
temporarily reduce graphite exports due to a 45-day application period for permits, and suggests
that graphite exports from China may be controlled in the future. However, at this time it is not
clear that this requirement will meaningfully impact exports over the long term, as similar permit
requirements have existed on other exports; Wood Mackenzie reports that a change to material
flows is unlikely, and that a graphite supply chain outside of China is rapidly developing (Wood
Mackenzie 2023).
With regard to nickel, cobalt and graphite, ANL also identifies potential enabling approaches
to mitigate the risks that they identify. For nickel, continued economic partnership and trade with
non-FTA countries with significant capacity, such as Indonesia, Philippines, Botswana, South
Africa, Papua New Guinea, Madagascar, Tanzania, and Zambia provide an avenue to securing
supply. Efforts to strengthen battery recycling in the U.S and ally nations is also identified, as
well as collaborative efforts with FTA and MSP partners to ensure mining project success (for
example, financing promising projects in FTA and non-FTA countries. In the longer term, ANL
also identifies use of battery chemistries that use less or no nickel. With regard to cobalt, the
same approaches are identified, including economic partnership and trade with non-FTA
countries including Indonesia, Philippines, Zambia, Papua New Guinea and Madagascar. For
graphite, ANL identifies similar economic partnership and trade objectives, as well as
strengthening synthetic graphite production capacity in the U.S. and ally nations.
ANL assesses that domestic lithium production is currently limited, but the next decade could
see a surge from promising projects that are already underway, potentially satisfying domestic
demand and allowing the U.S. to become a global leading producer of lithium depending in part
on the progress of permitting and other contingencies common to any new mining operations. As
described in Preamble IV.C.7, the U.S. government is actively working through various
programs to streamline U.S. mining as well as promote and pursue partnerships and resource
development opportunities in FTA countries, MSP countries, and allies. ANL also notes that in
both the near and medium term, a significant portion of domestic lithium demand can be met by
lithium in the U.S and in FTA countries, with several MSP partners likely to add capacity. ANL
identifies several potential mitigation approaches for any remaining risk, including collaborative
efforts with FTA and MSP partners to ensure mining project success in the U.S, FTA and non-
FTA countries, pursuing offtake agreements for stockpiling lithium from U.S producers to
alleviate downward price pressure that could discourage development of new sources, and
strengthening recycling in the U.S. and ally nations.
ANL also extended its analysis to manganese, which EPA did not individually assess in the
proposal as being of critical concern. ANL assesses that currently, there is no production of
battery grade manganese in the U.S., but there are six facilities that process manganese ore in the
U.S., and agrees with our initial assessment that compared to expected U.S. demand, manganese
supply is not critical. However, most of the significant production is in non-FTA countries,
primarily Australia, South Africa, and Gabon, with the latter currently accounting for two-thirds
of U.S. manganese ore imports.
For nickel, ANL noted that there are significant efforts to scale nickel supply in FTA
countries, with a majority of early-stage exploration projects in Australia and Canada. While the
probability of these early-stage projects to add to capacity by 2035 is low, they indicate global
efforts are taking place to scale nickel availability to meet global demand. Exploration efforts for
3-28
-------
nickel in the U.S. specifically are currently limited, reflecting current geological sources and
technology. About 50% of global mining is in Indonesia. Most Class I (battery grade) nickel is
cost effective to produce from sulfide deposits, with the majority of supply likely to be originally
sourced from Indonesia, Australia, Canada, Russia, and Philippine mines. Currently there is no
Class I nickel production in the U.S., but ANL anticipates that some nickel mining supply could
be refined into battery grade. ANL identifies several U.S. policy levers that could support build
out of nickel battery grade refining and recycling capabilities, for example, the DOE Loan
Programs Office (LPO), and funding under the Bipartisan Infrastructure Law (BIL) and Inflation
Reduction Act(IRA). ANL concludes that international trade is likely to be important to
strengthening U.S. supply of nickel.
Like nickel, there are significant efforts to scale cobalt supply in the FTA countries. Cobalt
and nickel tend to be co-located and co-produced, so the same projects that produce nickel often
also produce cobalt. The majority of early stage and exploration projects are in Australia and
Canada. While probability of these early-stage projects to add to capacity by 2035 is small, they
indicate global efforts to scale cobalt to meet global demand. Exploration efforts for cobalt in the
U.S. is limited. While the Democratic Republic of Congo (DRC) is and will continue to be a key
global source of cobalt mining supply (currently about 70 percent of global cobalt mining
supply), other promising sources outside DRC include Indonesia and Australia. Importantly, the
majority of global mined cobalt is currently refined in China. Cobalt production in the U.S. is
very limited and there is no cobalt refinery, but several efforts exist to support build out of
domestic cobalt refining. DOE concludes that trade with non-FTA countries including allies in
addition to FTA and MSP partners will be key to securing cobalt supply for those lithium-ion
chemistries that use it. ANL additionally identifies policy levers to promote both domestic
production and supply partnerships with these nations. ANL also notes that other chemistries
such as LFP do not need cobalt and are suitable for many applications.
Graphite has similarities to nickel and cobalt in that there are significant efforts to scale
graphite supply in the FTA countries. The majority of early stage and exploration projects are in
Canada and Australia. There is no current U.S. production of graphite from mine sources, and
exploration efforts are currently limited. China is likely to continue to be the leading producer of
natural graphite in the world, while in the near term other countries such as Tanzania,
Mozambique, Canada, and Australia are likely to increase supply, diversifying global supply
away from China. This suggests that in the near term and medium term, a significant portion of
natural graphite will be in non-FTA countries. FTA countries (Canada and Australia) are likely
to add natural graphite capacity over the medium term. The earliest U.S. production of natural
graphite is anticipated in 2025 from Coosa County, Alabama with capacity of 7,500 metric tons
per year, with the biggest project anticipated to come online in 2028 from Graphite Creek with
capacity of 51,000 tons per year. In addition to natural graphite from mine sources, synthetic
graphite shows promising opportunities with the earliest project anticipated to come online in
2024. International trade will likely continue to play a crucial role in securing graphite supply.
Currently, the major U.S. source of imports other than China include Canada, Mexico,
Madagascar, Brazil, and Mozambique, and these sources are likely to continue to be major
sources of import for U.S. manufacturers.
Regarding lithium, DOE finds that there are significant efforts to scale lithium supply both
domestically and also in the FTA countries. The majority of early stage and exploration projects
are in Australia, Canada, and the U.S. DOE assesses that the U.S. is well positioned in securing
3-29
-------
lithium materials domestically, particularly if all projects underway (particularly later stage
projects) are successful. Global lithium mining supply is anticipated to more than double in the
next five years. In fact, if lithium demand does not match this supply, it could lead to oversupply
and create downward price pressure. Several U.S. projects are in the construction stage,
including at Fort Cady, Thacker Pass, Rhyolite Ridge, and King Mountains, with others
undergoing prefeasibility or feasibility studies, e.g., Great Salt Lake. Through such projects the
U.S. lithium supply is expected to more than double by 2025, and the U.S. is poised to become a
global key player in lithium industry if all ongoing projects come to fruition and can overtake
current key players such as Australia, Argentina and Chile. The majority of U.S. lithium
production is likely to come from brines, which are relatively cheaper to produce compared to
lithium from spodumene deposits. Both in the near term and the medium term a significant
portion of lithium will be available domestically and in FTA countries, likely enough to meet
domestic demand. Several FTA and MSP partners, such as Canada and Germany, are likely to
add capacity over the medium term, further strengthening U.S. lithium availability. DOE
assesses that the U.S. largely has sufficient lithium supply to meet domestic demand of battery
manufacturers under a number of reasonable demand scenarios. Only in the near term will the
U.S. likely depend on imported lithium, and sufficient additional capacity exists in FTA
countries to meet this import demand. Specifically, international trade will continue to be
important in the next three years as the U.S. scales domestic production; from 2025, if all U.S.
projects currently underway commence production and scale as expected, the U.S. may have
sufficient lithium to meet domestic manufacturer demand with an opportunity to be a net
exporter of lithium.
With regard to manganese, DOE assesses that both in the near term and medium term, a
significant portion of manganese will be available domestically from non-FTA countries. While
capacity in FTA and MSP partners is concentrated in a few countries such as Australia, Canada,
and India, it is likely to be sufficient to meet U.S. demand in both the near and medium term.
Conversely, because there is limited outlook for manganese production in the U.S. due to the
poor quality of ore prospects, the U.S. is likely to depend on FTA-imported mining supply to
meet domestic demand for the foreseeable future.
It is important to note that where U.S. mineral sources may be limited, importation from FTA
countries is consistent with mineral sourcing requirements for the 30D clean vehicle credit. Even
where minerals from non-FTA countries are used, this content does not prevent these vehicles
from being sold in the U.S. market, and the mineral content they represent will become domestic
content when recycled. DOE has performed additional analysis relating its domestic critical
mineral findings to its projections of 30D/45W uptake which EPA uses in its analysis, and this
analysis is described in the ANL critical minerals report (ANL 2024b).
In its assessment, DOE notes that a number of uncertainties affect every forward-looking
assessment of mineral and manufacturing trends, and EPA has considered this fundamental layer
of uncertainty which could cause these projections to prove either optimistic or pessimistic. It is
well known in the forecasting and cyclical commodities industries that price volatility can be
driven by demand or supply, or both. While oversupply is positive for battery cost in the short
term, it can lead investors to consider some development projects uneconomical. Slow demand
growth can be a factor in lower-than-expected prices. In a near term perspective, something as
simple as a temporary drop in consumer demand due to changes in economic fundamentals can
contribute to such a situation. Over the medium to long term, the same impact can result from
3-30
-------
changes in policy, or technology disruption (e.g., substitution of one mineral for another, or
alternative chemistries that eliminate the mineral as mentioned previously). Another uncertainty,
particularly in the U.S., is permitting for new mining projects, which can take several years.
Financing is also subject to uncertainty, as mining is considered to be a relatively high-risk
investment that pays off over a long time frame, subject to the uncertain factors above. Political
and social risks, such as war, changes in trade policy, and labor disputes are additional factors. In
some cases, the location of a mine may be remote, leading to potential difficulties in attracting
qualified labor.
While all of these uncertainties can have an impact on future projection of progress in mineral
production, it is also true that all mineral production currently in operation has transcended these
risks, often in periods of far less rapid growth in demand of the minerals involved. With the
importance of battery minerals in many different sectors that are relevant to reducing pollution
and GHG, demand for these minerals is rapidly growing. As these uncertainties are well
understood to accompany most if not all mining investments, EPA does not consider these
factors to be uniquely restrictive of the ability of the global industry to develop mineral
production capacity in response to what is widely understood to be an era of robust demand.
This section has reviewed some supplemental information related to critical mineral sources.
The primary body of our assessment of battery cell and cell component manufacturing, critical
minerals, and mineral security resides in the preamble Section IV.C.7, where we analyze and cite
all relevant evidence and form our conclusions regarding these topics as they relate to
compliance with the standards. For references and full discussion please refer to the specific
preamble sections cited below and where applicable, other relevant sections of the preamble and
RIA.
3.1.4.3 Enabling Approaches on Strengthening Supply Chains
The 2024 ANL study titled "Securing Critical Minerals for the U.S. Electric Vehicle Industry"
(ANL 2024b) was developed and published concurrently with the development of the final rule.
EPA has carefully considered the extensive findings of this work and considers it thorough and
up to date, representing some of the best available information on the status and outlook for
critical mineral availability now and during the time frame of the rule. EPA has cited this study
frequently in the preamble and in this RIA as a key reference in developing our outlook for
critical minerals in the context of compliance with the standards.
In examining mineral supply and demand, the ANL study also pays close attention to the
primary uncertainties relevant to increasing mineral production and availability, as well as a
number of enabling approaches that U.S. government and industrial actors are pursuing as part of
a broad strategy to further increase domestic critical mineral supply. These efforts are generally
important to understanding how minerals and related products can be accessed reliably through a
combination of domestic sources and through global partners including our FTA partners, MSP
partners, and allies. We briefly mention some of these activities in section IV.C.7 of the
preamble. Here we provide a further summary of some of the approaches specifically identified
in the ANL study, to provide a sense of some of the enabling activities that are currently being
pursued by the U.S. government. Citations, additional detail and further examples of current
activities are available in the ANL study.
3-31
-------
Collaboration with trading partners is a major focus of attention. This involves
diversifying supply chains beyond existing free trade agreements by strengthening
trade with potential countries that have or could have significant capacity, as well as
joint efforts with MSP partners to ensure the success of mineral projects in member
countries through coordinated financial assistance, mobilizing both government and
private capital, providing technical expertise, and streamlining ESG standards to
include traceability standards. Collaboration could extend to financing promising
projects within non-FTA countries by approaches such as leveraging existing and new
interagency efforts across various agencies and departments such as State, Commerce,
DOE, USAID, US DFC, USTDA and EXIM, in collaboration with the private
financing sector.
Improving the permitting process for critical minerals projects is another thrust of
activity. The Biden-Harris Permitting Action Plan (May 2022) and subsequent
implementation guidance (March 2023) identifies key steps including: acceleration of
permitting through early cross-agency coordination, establishing clear timeline goals
and tracking, engaging in early and meaningful outreach and communication with
Tribal Nations, States, territories, and local communities, improving agency
responsiveness, technical assistance, and support, and adequately resourcing agencies
and using the environmental review process to improve environmental and community
outcomes.
Stockpiling and supply chain readiness is another focus. Strategic stockpiles can serve
as a buffer against potential disruptions. This approach could also protect domestic
projects to develop mining and recycling from intentional oversupply (product
dumping) by actors aimed at reducing global competition. Efforts around stockpiling
are already in progress. For example, DOD, DOE, and the State Department are laying
the foundation for a new interagency process for stockpiling minerals. Other efforts to
stabilize supply chain volatility and uncertainty include better data tracking and
sharing, alert systems, and international partnerships to respond to supply chain
disruptions.
Increasing domestic recycling capacity can be a strong factor in reducing future need
for new critical minerals. The Federal Consortium for Advanced Batteries (FCAB)
developed a National Blueprint for Lithium Batteries, outlining near-term objectives to
achieve the goal of scaling end-of-life reuse and recycling for minerals. The DOE has
also announced $37 million in available funding to improve the economics and
industrial ecosystem for battery recycling.
Advanced recycling techniques such as direct recycling can offer lower costs when,
commercialized and scaled. The BIL funds research and development for advanced
recycling; DOE has already announced more than $45 million for advanced recycling
projects, including direct recycling.
3-32
-------
Identifying non-traditional sources of critical materials that are available domestically,
such as industrial by-products and mining waste streams, can also help meet minerals
demand over time. The U.S. government is supporting efforts to fund research into
non-traditional sources of minerals: for example, in February 2024, DOE announced it
would invest $17 million into projects to recover minerals from coal-based resources,
and in November 2023 USGS announced $2 million to 14 states to study critical
minerals in mine waste. Research suggests that resource recovery from coal and
mining waste may also help remediate abandoned mines.
Workforce development can be promoted by coordination and collaboration with
academic institutions and training centers to develop the next-generation workforce to
serve the potentially growing domestic mining sector. For example: DOE, in
collaboration with DOL, AFL-CIO, and other partners, launched the Battery
Workforce Initiative through the National Energy Technology Laboratory (NETL) to
develop training up and down the battery supply chain. Talon Metals and the United
Steelworkers have also announced a joint workforce development partnership for the
Tamarack Nickel Project. The FY24 NDAA directs that the Defense Department study
the feasibility for and plan for the creation of a University Affiliated Research Center
for Critical Minerals which would assess institutional capabilities and investments
needed for workforce development to support needs related to critical materials. The
Department of Commerce, through the CHIPS Act is funding workforce development
across the battery supply chain in Missouri, New York, and Nevada.
Strengthening environmental, social and governance (ESG) implementation can be
key to reducing risk for mining projects to improve chance of production and reduce
impact. Strategies include pursuing robust consultation with communities near where
mining resources are located, and adherence to strong labor, human rights and
environmental practices. Internationally, some USG efforts already exist to advance
ESG compliance and to improve environmental and social outcomes of minerals
development. DOE's Advanced Research Projects Agency-Energy (ARPA-E) is
funding 16 projects across 12 states that aim to increase mineral yield while
decreasing energy and emissions from mineral extraction. The U.S. Department of
Labor (DOL) offers resources to companies looking to mitigate risks related to labor
violations and programs to raise awareness and address international ESG. Through
the IPEF Supply Chain Agreement, the U.S. is also engaged in a Labor Rights
Advisory Board to promote worker rights across supply chains. The "Presidential
Memorandum on Advancing Worker Empowerment, Rights, and High Labor
Standards Globally" directs the Secretaries of State, Labor, Energy, Treasury,
Homeland Security, and Commerce along with the Administrator of USAID and the
U.S. trade representative to address labor rights across global supply chains.
Community and tribal engagement is also important to addressing potential conflict
between communities and mining interests, which increase risk and uncertainty to all
stakeholders when mineral resources are identified for development. Many grants and
3-33
-------
loans provided by the Department of Energy under BIL and IRA require applicants to
submit a Community Benefit Plan, which is evaluated at 20 percent of the overall
application; community agreements such as Community Benefit Agreements (CB As)
and other community and workforce agreements are strongly encouraged by these
programs, which may provide funding to mining and materials processing initiatives.
The DOE also sponsors programs that incentivize the transition of defunct mines into
clean energy sites, including the Biden Administration's $500 Million Program to
Transform Mines Into New Clean Energy Hubs and the Qualifying Advanced Energy
Project Credit (48C) Program.
ANL also notes, as EPA noted in the preamble discussion, that mineral substitution holds
strong potential to reduce dependence on some minerals. For example, cobalt and nickel in
cathode materials can be eliminated by iron phosphate which is appropriate for some
applications and is already taking place. Sodium-based chemistries currently under development
and beginning to enter production could reduce the need for lithium. Electrode compositions that
improve energy density, safety, and cost, as well as new battery technologies like solid-state and
Li-metal, can also help over the long term.
Through these and other examples the ANL study provides evidence of currently identified
approaches, current activity, and relevant goals that are accompanying efforts to promote and
maintain a secure supply of minerals and related products for U.S. industry.
3.1.5 Modeling Constraint on Rate of PEV Technology Penetration
This section details how EPA developed a modeling constraint to represent a potential
production-based limit on the rate of penetration of PEV technology into the fleet during the
timeframe of the standards.
In modeling potential PEV penetration into the fleet as a result of the standards, EPA
considered how best to represent any limitations that are likely to be imposed by the supply
chain. Potential constraints on availability of minerals that are used in the manufacture of PEVs
are particularly relevant to projecting practical limits on the rate of penetration of PEVs into the
fleet in the future. EPA considered data from industry analysts, including Wood Mackenzie and
Benchmark Minerals Intelligence, to pursue a quantitative and qualitative understanding of the
future availability of these critical battery minerals during the time frame of the rule.
From a modeling perspective, the question of how to constrain the modeled rate of BEV
penetration to remain within limits imposed by the developing supply chain is an important one.
As part of the rulemaking analysis, EPA uses its OMEGA model to identify compliance
pathways (in other words, applications of available technology to the fleet) by which
manufacturers can meet the standards. The OMEGA model selects among available advanced
vehicle technologies and applies them to the fleet in the most cost-effective way, given the cost
of each technology and its effectiveness at achieving manufacturer compliance within the fleet
averaging structure of the program. Although BEV technology has a higher absolute cost than
many other technologies, it is particularly attractive to manufacturers because BEVs achieve zero
tailpipe emissions and are credited with such under the compliance accounting. On the other
hand, there is likely to be a limit to the rate at which BEV technology can phase into the fleet due
to various constraining forces such as growth in consumer acceptance, timing of refresh/redesign
3-34
-------
cycles, activation of battery cell and pack manufacturing capacity, and critical mineral
availability, particularly in the early years of the analysis. If these constraints were not
represented, the OMEGA model might select BEV technology at a rate that results in a faster
penetration of BEVs into the fleet than these real-world constraints might practically allow. EPA
implemented several constraints in the OMEGA model to account for these factors.
Consumer acceptance is discussed in more detail in Chapter 4 of this RIA, and its
representation by means of S-curves in the OMEGA model is discussed in Chapter 2.6.3.
Refresh/redesign cycles are also represented in the OMEGA model and are discussed in more
detail in Chapter 2.6.4 of this RIA as well as IV.C of the Preamble.
To implement a constraint representing the potential for battery manufacturing capacity or
critical mineral availability to constrain PEV production, EPA implemented a constraint in terms
of Gigawatt-hours (GWh) of lithium-ion battery production per year that could be available for
BEVs supplying the U.S. new vehicle fleet. As described below, we developed this constraint by
considering estimates of existing and announced battery production capacity in North America
and comparing these forecasts to estimates of projected lithium supply and demand. We
considered available data on forecast battery manufacturing capacity, global lithium demand, and
global lithium chemical production. As all such estimates concern prediction of future events and
are by nature uncertain (particularly in the out years), we adopted a simplified approach that
provided what we consider to be a reasonable and conservative view of future PEV production
capacity as constrained by manufacturing and mineral supply.
We selected lithium supply as the primary mineral-based limiting factor for several reasons.
In Preamble IV.C.7 we noted that cobalt, nickel, and manganese are important to today's leading
battery chemistry formulations, but we also note that there is some flexibility in choice of
cathode minerals, and in many cases, opportunity will exist to reduce cobalt and manganese
content or to employ iron-phosphate cathode chemistries that do not utilize nickel, cobalt, or
manganese. Graphite is used as the anode of most current and near-term PEV battery
chemistries, and all require lithium in the form of lithium carbonate or lithium hydroxide in the
electrolyte and the cathode. The role of natural graphite in many cases can be served by artificial
graphite or highly refined hard or amorphous carbon. However, lithium has no substitute in
commercially produced automotive applications at this time (however, see the discussion of
alternatives to lithium under development, later in this section). Although the common
chemistries vary in their need for either lithium hydroxide or lithium carbonate, either can
potentially be produced from available lithium sources.
Further, and as described in greater detail in the preamble to the original proposal, we
considered projections of cobalt, nickel, and lithium supply and demand published in 2022 by the
International Energy Agency (IEA), which concluded that supply of cobalt and nickel should be
sufficient to meet demand between 2020 and 2030 for the two most likely demand scenarios
modeled, while lithium demand may begin to approach available supply after 2025. By contrast,
as also described in the original proposal, projections made by DOE in November 2022 indicate
that global supplies of cathode active material (and incidentally, lithium chemical products) are
within the range of expected global demand through 2035.
In the preamble to the original proposal we described the development of our assessment that
battery mineral supply is likely to be sufficient to meet the standards. The data examined also
suggested that, among the primary critical minerals needed for battery manufacturing, growth in
3-35
-------
demand for lithium would likely be the first to approach available supply, if a battery mineral
shortage were to be encountered. Accordingly, we focused on lithium availability as a potential
limiting factor on the rate of growth of PEV production, and thus the most appropriate basis for
establishing a modeling constraint on PEV penetration into the fleet over the time frame of the
rule.
With regard to battery manufacturing capacity in the U.S., we considered estimates of
announced manufacturer plans and currently installed capacity as reported in mid-2022 by S&P
Global and in late 2022 by Argonne National Laboratory (ANL 2022). S&P Global indicated
that U.S. battery manufacturing capacity will reach 382 GWh by 2025 (S&P Global 2022a), and
580 GWh by 2027. (S&P Global 2022b).
Although these forecasts suggest that planned manufacturing facility capacity could be a
potential basis for a modeling constraint on battery production, this would not reflect the
possibility that operating capacity could be constrained by mineral availability. We thus sought
to condition these production capacities by comparing them to estimates of global lithium supply
and demand.
Here it is relevant to note that, although the Inflation Reduction Act incentivizes use of
domestically sourced and processed mineral products, it only ties these products to availability of
the related tax incentives (primarily the Clean Vehicle Credit under 30D) and does not prohibit
use of imported mineral products by manufacturers that cannot secure domestic sources. Thus, it
is the global supply for lithium, not only domestically sourced supply, that potentially constrains
battery production.
We then referred to proprietary projections of lithium chemical capacity obtained from Wood
Mackenzie through a service subscription (Wood Mackenzie 2022). Forecast lithium production
in tons per year was reported as lithium carbonate equivalent (LCE) which EPA converted to
GWh of gross battery capacity using a widely accepted conversion factor. As a first-order,
conservative approximation of lithium availability to supply U.S. demand, we first subtracted the
Wood Mackenzie projections of U.S. lithium demand from their projections of global demand, to
estimate a "rest of world" (ROW) lithium demand trajectory. We then calculated the difference
between the ROW demand trajectory and the Wood Mackenzie high and low estimates of global
lithium chemical production. This difference was taken to represent a hypothetical lithium
production capacity that would be available to the U.S. market, assuming that ROW demand was
satisfied first, and growing demand did not generate a demand response among lithium suppliers
beyond what is already represented in the forecast. This is a conservative assumption, as market
forces would ultimately play some role in determining distribution, and increased demand would
likely result in higher prices and greater market certainty for investment in additional supply
capacity. We also noted that the resulting availability curve would be most applicable to earlier
years, as the data on which it was based does not represent likely industry response to increased
lithium demand as the market continues to grow. For this reason, we do not depict the supply
curves beyond 2027 due to the lack of modeled demand response in the underlying data.
Figure 3-14 shows the S&P and Argonne battery plant production capacity estimates plotted
against the calculated lithium carbonate equivalent (LCE) production capacity potentially
available to the U.S. market after global projected demand is satisfied (in estimated battery GWh
equivalent).
3-36
-------
1750
1500
1250
1000
5
0
750
500
250
A'"wk
A
l i
~ C
~ E
~
£
;
f*
• • Excess LCE
N.A. GWh (ANL 2022)
—B— N.A. LDV GWh (ANL 2024)
O US cell mfg(S&P 2022)
OMEGA limit
Central case
— — — Pathway ES
.....Pathway C
2020 2022 2024 2026
2028
Year
2030 2032 2034 2036
Figure 3-14: Limit on battery GWh demand implemented in OMEGA, compared to
projected battery manufacturing capacity and excess LCE supply.
We then examined this data to establish a conservative but reasonable limit on GWh battery
supply for use by the OMEGA model. First, we noted that the S&P estimate of U.S. battery
manufacturing plant production capacity, which due to its earlier date of origin is conservative
with respect to today and extends only to 2027, is well beneath the low estimate of hypothetical
"excess" lithium supply. This suggests that lithium supply is more than sufficient to sustain the
S&P estimate of U.S. plant operation at full capacity.
The ANL 2022 accounting of U.S. plant capacity is larger than the S&P accounting, reflecting
the pace of newer announcements. It begins to exceed the low estimate of excess lithium supply
in 2026 but is still well within the upper limit afterward.
As a conservative bound on battery production capacity for the OMEGA model, we thus
followed the S&P trajectory to 2027 at 580 GWh. This trajectory stays within lower expected
lithium excess capacity for the first several years, when limited time is available for new capacity
to come on-line.
Past 2027, estimates of "excess" lithium as a difference between ROW demand and a current
accounting of global supply become less informative, because a demand response is not built
into the supply data. Therefore, uncertainty about the supply-demand balance against ROW
demand increases as the time horizon increases. In general, analysts believe that as demand for a
mineral commodity remains strong over time, investment in mining operations and exploration
will consistently increase, which leads to unknown or previously unprofitable geological sources
to become available (Sun, Ouyang and Hao 2022). In resonance with this fact, we noted that it
3-37
-------
seems very unlikely from an investment point of view that manufacturers and battery suppliers
would plan to construct plant capacity to come online in 2030 if it exceeds their expectation of
availability of mineral products to supply the plant's production. Given the amount of lead time
in the time frame past 2027 and the current level of activity in development of supply chain
capacity across the world, we considered it reasonable to expect that the ANL estimate for 2030
at 1320 GWh should be feasible to supply. However, to retain a conservative estimate to reflect
uncertainty, for the purpose of creating a constraint line we retained the previous estimate of
1000 GWh ANL had provided at the time of the proposal and which we had used in the proposal.
We then continued a similar rate of increase to 2035 at 1500 GWh. Passing through these three
defined points results in an almost linear growth rate that we then adopted as the annual battery
GWh production limit for OMEGA modeling purposes, which is unchanged from the proposal.
We flattened the limit at 1500 GWh after 2035 due to lack of data for that time period.
Since the proposal, an updated and more detailed estimate by Argonne National Laboratory
was published in March 2024 (ANL 2024a). This included a 36 month delay time from
anticipated plant opening date to operation at full capacity using a linear ramp up. This study is
described in more detail in section IV.C.2 and IV.C.7 of the preamble. The figure above shows
the ANL 2024 estimate for North American cell manufacturing capacity that is designated for
light-duty vehicles (excluding uses for heavy-duty vehicles and stationary applications). These
later and more specific estimates of North American manufacturing capacity continue to be in
excess of the demand projected by the final standards and Pathways B and C. Because during the
time frame of the rule the projected GWh demand under the central case and the alternative
pathways B and C do not exceed the updated ANL 2024 manufacturing capacity projection nor
the GWh constraint that was developed for the proposal, the GWh constraint is not controlling on
the results. EPA therefore did not consider it necessary or useful to modify the GWh constraint
for the FRM analysis as there was no evidence of reduced lithium availability that would suggest
the constraint should be tightened and it was not controlling on the result.
Here it is also important to note, again, that our estimate of lithium available to the U.S., as
the difference between currently anticipated global lithium supply and ROW demand, would be
expected to be a conservative estimate because it quantifies only currently known sources of
lithium that would not be subject to demand elsewhere, and does not reflect the development of
additional sources over time.
The numeric values for the annual GWh limit input to OMEGA are provided in RIA Chapter
2.6.4.2. More details on how OMEGA calculates BEV battery capacities that are summed to a
fleet GWh production capacity is provided in RIA Chapter 2.5.2.1.1.
3.2 Criteria and Toxic Pollutant Emissions Standards
EPA is finalizing changes to criteria pollutant emissions standards for both light-duty vehicles
and medium duty vehicles (MDV). Light-duty vehicles include LDV, LDT, and MDPV.
NMOG+NOx changes for light-duty vehicles include a declining fleet average standard, the
elimination of higher certification bins, a requirement for the same fleet average standard to be
met across four test cycles (25°C FTP, HFET, US06, SC03), a change from a fleet average
NMHC standard to a fleet average NMOG+NOx standard in the -7°C FTP test, and three
NMOG+NOx provisions aligned with the CARB Advanced Clean Cars II program.
3-38
-------
NMOG+NOx. changes for MDV include a fleet average that declines from 2027-2033 in the
early compliance program or steps down in 2031 in the default program, the elimination of
higher certification bins, a requirement for the same fleet average emissions standard to be met
across four test cycles (25°C FTP, HFET, US06, SC03), and a new fleet average NMOG+NOx
standard in the -7°C FTP. EPA is also finalizing in-use standards for spark ignition and
compression ignition MDV with GCWR above 22,000 pounds from proposed Alternative 2 that
are consistent with MY 2031 and later California chassis-certified MDV in-use emissions
standards.
EPA is finalizing a continuation of light-duty vehicle and MDV fleet average FTP
NMOG+NOx standards that include both ICE-based and zero emission vehicles in a
manufacturer's compliance calculation. Performance-based standards that include both ICE and
zero emission vehicles are consistent with the existing NMOG+NOx program as well as the
GHG program. EPA considers the availability of battery electric vehicles as a compliance
strategy in determining the appropriate fleet average standards. Given the cost-effectiveness of
BEVs for compliance with both criteria pollutant and GHG standards, EPA anticipates that most
(if not all) automakers will include BEVs in their compliance strategies. However, the standards
continue to be a performance-based fleet average standard with multiple paths to compliance,
depending on choices manufacturers make about deployment of a variety of emissions control
technologies for ICE as well as electrification and credit trading.
EPA is finalizing a PM standard of 0.5 mg/mi for light-duty vehicles and MDV that must be
met across three test cycles (-7°C FTP, 25°C FTP, US06), a requirement for PM certification
tests at the test group level, and a requirement that every in-use vehicle program (IUVP) test
vehicle is tested for PM. The 0.5 mg/mi standard is a per-vehicle cap, not a fleet average.
EPA is finalizing CO and formaldehyde (HCHO) emissions requirement changes for light-
duty vehicles and MDVs including transitioning to emissions caps (as opposed to bin-specific
standards) for all emissions standards, a requirement that CO emissions caps be met across four
test cycles (25°C FTP, HFET, US06, SC03), and a CO emissions cap for the -7°C FTP that is the
same for all light-duty vehicles and MDVs.
EPA is finalizing a refueling standards change to require incomplete MDVs to have the same
on-board refueling vapor recovery standards as complete MDVs.
EPA is finalizing that light-duty vehicle 25°C FTP NMOG+NOx credits and -7°C FTP
NMHC credits (converting to NMOG+NOx credits) may be carried into the new program. MDV
25°C FTP NMOG+NOx credits may only be carried into the new program if a manufacturer
selects the early compliance pathway. New credits may be generated, banked, and traded within
the new program to provide manufacturers with flexibilities in developing compliance strategies.
The finalized criteria pollutant phase-in schedules apply to NMOG+NOx bin structure, -7°C
NMOG+NOx, PM, CO, HCHO, -7°C CO, and three light-duty vehicles provisions aligned with
CARB ACC II, and standards for MDV with GCWR above 22,000 pounds.
Chapters 3.2.1 through 3.2.4 summarize the standards being finalized for light-duty vehicles
and MDV for NMOG+NOx, PM, CO, and HCHO, and in-use standards for high GCWR MDV.
Chapter 3.2.5 and Section III.D.2 of the preamble demonstrate the feasibility of the
NMOG+NOx standards including the NMOG+NOx light-duty vehicle provisions aligned with
3-39
-------
the CARB ACC II program. Chapter 3.2.6, together with Section III.D.3 of the preamble,
demonstrate emission control technology and measurement procedure readiness, emissions
benefits, and the importance of the three certification test cycles for the PM standards, as well as
quantifying GPF cost and impact on CO2 emissions. Section III.D.4 of the preamble
demonstrates the feasibility of the CO standards. Chapter 3.2.7 and Section III.D.6 of the
preamble demonstrate the feasibility of the refueling emissions standards.
3.2.1 NMOG+NOx Standards
The final NMOG+NOx standards, including declining fleet averages, the elimination of
higher certification bins, a requirement for the same fleet average emissions standard to be met
across four test cycles (25°C FTP, HFET, US06, SC03), a change from a fleet average NMHC
standard to a fleet average NMOG+NOx standard in the -7°C FTP test, three NMOG+NOx
provisions aligned with CARB ACC II for light-duty vehicles, and MDV in-use emissions
standards, are discussed in Section III.D.2 of the preamble. For reference, the final NMOG+NOx
standards for light-duty vehicles and MDV are shown in the tables below.
3.2.1.1 NMOG+NOx Bin Structure for Light-Duty and MDV
The final bin structure for light-duty vehicles and MDV is shown in Table 3-3. The upper six
bins (Bin 75 to Bin 170) are only available to MDV. For light-duty vehicles, the finalized bin
structure removes the two highest Tier 3 bins (Bin 160 and Bin 125) and adds new bins such that
the bins increase in 5 mg/mi increments from Bin 0 to Bin 70. For MDV, the finalized bin
structure also moves away from separate bins for Class 2b and Class 3 vehicles, adopting light-
duty vehicle bins along with higher bins only available to MDV. In part due to comments
received from MDV manufacturers, the final MDV-only bins have been harmonized with bins
used for compliance with California chassis-certified MDV standards with the exception of
elimination of any bins higher than Bin 170. The highest bin was also changed from Bin 160 to
Bin 170 to better align with the California ACC II program and to serve as a cap on MDV
emissions.
3-40
-------
Table 3-3: Light-duty vehicle and MDV NMOG+NOx bin structure.
Bin
NMOG+NOx (mg/mi)
Bin 170*
170
Bin 150*
150
Bin 125*
125
Bin 100*
100
Bin 85*
85
Bin 75*
75
Bin 70
70
Bin 65
65
Bin 60
60
Bin 55
55
Bin 50
50
Bin 45
45
Bin 40
40
Bin 35
35
Bin 30
30
Bin 25
25
Bin 20
20
Bin 15
15
Bin 10
10
Bin 5
5
BinO
0
*MDV only
3.2.1.2 Light-Duty NMOG+NOx Standards and Test Cycles
The finalized NMOG+NOx fleet average standards for MY 2027 and later light-duty vehicles
are shown in Table 3-4. The same bin-specific numerical standard must be met across four test
cycles: 25°C FTP, HFET, US06, and SC03. Vehicles that are not part of the phase-in percentages
described in Section III.D.l of the preamble are considered interim vehicles, which must
continue to demonstrate compliance with all Tier 3 regulations with the exception that all
vehicles (interim and those that are part of the phase-in percentages) contribute to the Tier 4
light-duty vehicle NMOG+NOx declining fleet average described shown in Table 3-4.
Table 3-4: LDV, LDT and MDPV NMOG+NOx standards for 25°C FTP, US06, HFET and
SC03.
Model Year
LDV, LDT 1-2 (GVWR < 6000 lb)
LDT3-4 (GVWR 6001-8500 lb) and MDPV
NMOG+NOx (mg/mi)
NMOG+NOx (mg/mi)
default
early
2026*
30*
30*
30*
2027
25
30*
25
2028
23
30*
23
2029
21
30*
21
2030
19
15
19
2031
17
15
17
2032 and later
15
15
15
* Tier 3 standards provided for reference
3-41
-------
3.2.1.3 NMOG+NOx Standards for MDV
The final MDV NMOG+NOx standards are shown in Table 3-5 for optional early compliance
and in Table 3-6 for default compliance. The CAA requires 4 years of lead time and 3 years of
standards stability for heavy-duty vehicles, and MDV fall under the CAA definition for heavy-
duty vehicles. Under default compliance, MDVs will continue to meet Tier 3 standards through
the end of MY 2031 and then MDVs will proceed to meeting a 75 mg/mi NMOG+NOx standard
in a single step in MY 2031 (Table 3-6) in order to comply with CAA provisions for regarding
standards stability. Under default compliance, MDV may not carry forward Tier 3 NMOG+NOx
credits into the Tier 4 program. The optional early compliance path has declining NMOG+NOx
standards that gradually phase-in from MY 2027 through MY 2033. MDV manufacturers opting
for early compliance may carry forward Tier 3 NMOG+NOx credits into the Tier 4 program
when Tier 3 is closed out, up to the end of the Tier 3 five-year credit lifeTable 3-5.
Note that the phase-in percentages from Section III.D. 1 .i of the preamble also apply. MDV
that are not part of the phase-in percentages summarized in Section III.D. l.ii of the preamble and
are considered interim vehicles, which must continue to demonstrate compliance with all Tier 3
standards and regulations with the exception that all vehicles (interim and those that are part of
the phase-in percentages) contribute to the Tier 4 MDV NMOG+NOx declining fleet average.
Certification data show that for MY 2022-2023, 75 percent of sales-weighted Class 2b and 3
gasoline vehicle certifications were below 120 mg/mi in FTP and US06 tests. Diesel-powered
MDVs designed for high towing capability (i.e., GCWR above 22,000 pounds) had higher
emissions, however 75 percent were still below 180 mg/mi NMOG+NOx. The year-over-year
fleet average FTP standards for MDV and the rationale for the manufacturer's choice of early
compliance and default compliance pathways is described in Section III.D. 1 of the preamble. For
further discussion of MDV NMOG+NOx feasibility, please refer to Chapter 3.2 of the RIA.
The final MDV NMOG+NOx standards are based on applying existing light-duty vehicle
technologies, including electrification, to MDV. As with the light-duty vehicle categories, EPA
anticipates that there will be multiple compliance pathways, such as increased electrification of
vans together with achieving 120 mg/mile NMOG+NOx for ICE-powered MDV. Present-day
MDV engine and aftertreatment technology allows fast catalyst light-off after cold-start followed
by closed-loop A/F control and excellent exhaust catalyst emission control on MDV, even at the
adjusted loaded vehicle weight, ALVW [(curb + GVWR)/2] test weight, which is higher than
loaded vehicle weight, LVW (curb + 300 pounds) used for testing light-duty vehicles. Diesel
MDV are adopting more advanced SCR systems for NOx emissions control that incorporate
dual-injection systems for urea-based reductant similar to SCR systems that have been developed
to meet more stringent NOx standards for MY 2024 and later heavy-duty engine standards in
California and federal MY 2027 and later heavy-duty engine standards. The final MDV
standards begin to take effect after 2031. While the originally proposed date of 2030 for default
compliance was fully consistent with the CAA section 202(a)(3)(C) lead time requirement for
these vehicles, EPA delayed implementation in the final rule to provide additional lead time
based in part on comments received from auto manufacturers expressing concerns that additional
lead time was important for compliance. Similarly, the early compliance pathway was delayed
by one year relative to our proposal.
3-42
-------
Table 3-5: MDV fleet average NMOG+NOx standards under the early compliance
pathway!.
Model Year
NMOG+NOx (mg/mi)
Class 2b
Class 3
2026
178*
247*
2027
175
2028
160
2029
140
2030
120
2031**
100
2032**
80
2033 and Later**
75
¦f Please refer to section III.D. 1 for further discussion of the early compliance and default
compliance pathways
* Tier 3 FTP fleet average standards provided for reference
** MDV with a GCWR greater than 22,000 pounds must also comply with additional moving
average window (MAW) in-use testing requirements
Table 3-6: MDV fleet average chassis dynamometer FTP NMOG+NOx standards under
the default compliance pathway*.
Model Year
MDV NMOG+NOx (mg/mi)
Class 2b
Class 3
2026
178*
247*
2027
178*
247*
2028
178*
247*
2029
178*
247*
2030
178*
247*
2031**
75f
2032**
75f
2033 and later
75f
¦f Please refer to section III.D. 1 for further discussion of the early compliance and default compliance
pathways
* Tier 3 FTP fleet average standards provided for reference
** MDV with a GCWR greater than 22,000 pounds must also comply with additional moving average
window (MAW) in-use testing requirements
If a manufacturer has a fleet mix with relatively high sales of MDV BEV, that would ease
compliance with MDV NMOG+NOx fleet average standards for MDV ICE-powered vehicles.
An option also remains for manufacturers of high GCWR MDV to choose engine-certification as
a light-heavy-duty engine as an additional compliance flexibility. This would allow some
manufacturers to choose the option of moving vehicles with the highest towing capability out of
the fleet-average chassis-certified standards and into the heavy-duty engine program. If a
manufacturer has a fleet mix with relatively low BEV sales, then improvements in NMOG+NOx
emissions control for ICE-powered vehicles would be required to meet the fleet average
standards and/or more capable high GCWR MDV could be moved into the heavy-duty engine
program. Improvements to NMOG+NOx emissions from ICE-powered vehicles are feasible with
available engine, aftertreatment, and sensor technology, and has been shown within an analysis
of MY 2022-2023 MDV certification data (see RIA Chapter 3.2). Fleet average NMOG+NOx
3-43
-------
will continue to decline to well below the final Tier 3 NMOG+NOx standards of 178 mg/mi and
247 mg/mi for Class 2b and 3 vehicles, respectively.
The final standards require the same MDV numerical standards be met across all four test
cycles, the 25°C FTP, HFET, US06, and SC03, consistent with the approach for light-duty
vehicles described in Section III.D.2.iii of the preamble. This would mean that a manufacturer
certifying a vehicle to bin 75 would be required to meet the bin 75 emissions standards for all
four cycles.
Meeting the same NMOG+NOx standard across four cycles is an increase in stringency from
Tier 3, which had one standard over the FTP and less stringent bin standards for the HD-SFTP
(weighted average of 0.35xFTP + 0.28xHDSIM + 0.37xSC03, where HDSIM is the driving
schedule specified in 40 CFR 86.1816-18(b)(l)(ii)). Previous MDV control technologies allow
closed-loop A/F control and high exhaust catalyst emissions conversion throughout the US06
and SC03 cycles, so compliance with higher numerical emissions standards over these cycles is
no longer needed. Manufacturer submitted certification data and EPA testing show that Tier 3
MDV typically have similar NMOG+NOx emissions in US06 and 25°C FTP cycles, and
NMOG+NOx from the HFET and SC03 are typically much lower. Testing of a 2022 F250 7.3L
at EPA showed average NMOG+NOx emissions of 56 mg/mi in the 25°C FTP and 48 mg/mi in
the US06. Manufacturer-submitted certifications show that MY 2021+2022 gasoline Class 2b
trucks achieved, on average, 69 mg/mi in the FTP, 75 mg/mi in the US06, and 18 mg/mi in the
SC03. MY 2021+2022 gasoline Class 3 trucks achieved, on average, 87 mg/mi in the FTP and
25 mg/mi in the SC03.
Several Tier 3 provisions will end with the elimination of the HD-SFTP and the combining of
bins for Class 2b and Class 3 vehicles. First, Class 2b vehicles with power-to-weight ratios at or
below 0.024 hp/pound may no longer replace the full US06 component of the SFTP with the
second of three sampling bags from the US06. Second, Class 3 vehicles may no longer use the
LA-92 cycle in the HD-SFTP calculation but will instead have to meet the NMOG+NOx
standard in each of four test cycles (25°C FTP, HFET, US06, and SC03). Third, the SC03 may
no longer be replaced with the FTP in the SFTP calculation.
The final standards do not include relief provisions for MDV 25°C FTP NMOG+NOx
certification at high altitude conditions (1520-1720 m), as is being finalized for light-duty
vehicles. Modern engine systems can use idle speed, engine spark timing, valve timing, and other
controls to offset the effect of lower air density on catalyst light-off at high altitudes.
EPA is also setting a new -7°C FTP NMOG+NOx fleet average standard of 300 mg/mi for
gasoline and diesel MDV. EPA testing has demonstrated the feasibility of a single fleet average -
7°C FTP NMOG+NOx standard of 300 mg/mi across light-duty vehicles and MDV. EPA did not
include EVs in the assessment of the fleet average standard and therefore EVs and other zero
emission vehicles are not included and not averaged into the fleet average -7°C FTP
NMOG+NOx standards.
Since -7°C FTP and 25°C FTP are both cold soak tests that include TWC operation during
light-off and at operating temperature, it is appropriate to apply the same Tier 3 useful life to
both standards. Additional discussion on the feasibility of the standards can be found in Chapter
3.2.5.
3-44
-------
3.2.2 PM Standard for Light-Duty and Medium-Duty Vehicles
The final PM standard is presented in Section III.D.3 of the preamble. GPF technology, GPF
benefits and feasibility of the standard, importance of test cycles, a GPF cost, and GPF impact on
CO2 emissions are discussed in RIA Chapter 3.2.6.
For reference, the final light-duty PM standard is shown in Table 3-7 and the final MDV PM
standard is shown in Table 3-8.
Table 3-7: Light-duty PM standard.
Test Cycle
Tier 3 Standard (mg/mi)
Final PM Standard (mg/mi)
25°C FTP
3
0.5
US06
6
0.5
-7°C FTP
Not applicable
0.5
Table 3-8: MDV PM standard.
Test Cycle
Tier 3 Standard (mg/mi)
Final PM Standard (mg/mi)
25°C FTP
8 (Class 2b)
10 (Class 3)
0.5
US06
10 over SFTP (Class 2b)
7 over SFTP (Class 3)
0.5
-7°C FTP
Not applicable
0.5
3.2.3 CO and Formaldehyde (HCHO) Standards
The final CO and formaldehyde (HCHO) standards are described in detail within Section
III.D.4 of the preamble. For reference, the light-duty vehicle standards are shown in Table 3-9
and the MDV standards are shown in Table 3-10.
Table 3-9: Light-duty CO and HCHO standards.
CO cap for 25°C FTP. HFET. SC03 (g/mi) 1.7
CP cap for US06 (g/mi) 9.6
CO cap for -7°C FTP (g/mi) 10.0
HCHO cap for 25°C FTP (g/mi) 4
Table 3-10: MDV CO and HCHO standards.
CO cap for 25°C FTP. HFET. SC03 (g/mi) 3.2
CP cap for US06 (g/mi) 25
CO cap for -7°C FTP (g/mi) 10.0
HCHO cap for 25°C FTP (g/mi) 6
CO standards for the 25 °C FTP, SC03, and US06 are harmonized with those of the California
LEV IV program for light-duty vehicles, and with California Class 2b standards for all MDV.
There is currently significant overcompliance within the Tier 3 program for CO over the FTP.
CO emissions over the 25 °C FTP are very low due to the need to maintain low cold-start and
running NMOG emissions in order to meet FTP NMOG+NOx standards. The SC03 is a hot
running cycle with moderate load, and thus has comparable or lower CO emissions compared to
3-45
-------
the 25°C FTP. The US06 has higher CO emissions, but only has a 27% weighting under the Tier
3 composite SFTP standards. An analysis of 2024 certification data shows significant
overcompliance with Tier 3 FTP CO standards even when averaging across all bin levels (see
Table 3-11).
Table 3-11: Average" FTP, SC03, US06 and composite SFTP CO emissions for MY 2024
test groups certified to Tier 3 that overlap with Tier 4 Standard Bins
Class
Cert. Bin
FTP
CO
Cert.
Level
(g/mi)
: a(n-l)
: Tier 3
; FTP CO
i Standard ;
(g/mi
SC03
CO
Cert.
Level
(g/mi)
| a(n-l)
US06
CO
Cert.
Level
i (g/mi)
i a(n-l)
SFTP
CO
Cert.
Level
(g/mi)
: a(n-l)
Tier 3
SFTP
CO
, Standard
; (g/mi)
LD
30
; 0.27
: 0.21
1.0
0.41
; 0.31
; 0.96
1.3
0.52
f 0.56
4.2
LD
50
u.51
i 0.30
1.7
0.55
0.41
1.1
2.0
0.69
; o.8o
' 4.2
LD j
70
i 0.47
0.29
1.7
0.43
1 0.30
1.0
r i.i
0.61
1 0.51
4.2
Class 2b ;
170
: 0.70
I 0.08
4.2
0.78
0.59
1.7
2.1
0.99
; 0.82
i ' 12
* Average and standard deviation [a(n-l)] at each bin level for all MY2024 spark-ignition test groups for which Tier 3 Bins overlap with
Tier 4 bins. Note that no MY 2024 data was found within current certification data for LD Tier 3 Bin 20 and insufficient data was found
for SFTP calculation for Class 2b Bin 150.
Assuming certification levels of 0.5 g/mi over the FTP and SC03 would allow compliance
with the 4.2 g/mi CO composite LD SFTP standard of 4.2 g/mi with US06 CO emissions results
as high as 13.7 g/mi. Similarly, assuming FTP and SC03 CO certification levels of 0.8 for MDV
would allow compliance with the 12 g/mi CO composite Class 2b SFTP standard of 12 g/mi with
US06 CO emissions results as high as 40 g/mi.
The LEV IV program caps US06 emissions at approximately what would be allowable under
the Tier 3 composite SFTP standard if FTP and SC03 are near the FTP standard for CO (Table
3-12). When also considering the NMOG levels needed to comply with NMOG+NOx standards,
this serves to cap back sliding on the US06 that might occur under an SFTP composite standard.
Table 3-12: Comparison of LEV IV CO standards calcluated as a composite SFTP to the
Tier 3 SFTP composite CO Standards
LEV IV Standards
Vehicle
Category
Bin
FTP CO
(g/mi)
SC03 CO'"
(g/mi)
US06 CO
(g/mi)
SFTP
(Composite
calculated from
LEV IV
standards)
3.408
3.912
9.304
10.024
T3 SFTP
Standards
4.2
4.2
12
12
LD 0-30 1 1 9.6
LD 35-70 1.7 1.7 9.6
2B 75-150 3.2 3.2 25
2B 170 4.2 4.2 25
SFTP 0.35 0.37 0.28
Composite
Weighting:
* LEV IV does not set SC03 standards for MDV. For purposes of comparison to Tier 3. SC03 was set to the FTP CO standard, which is
consistent with the LEV IV CO standards for LD.
3-46
-------
3.2.4 In-use Standards for High GCWR Medium-Duty Vehicles.
The Agency proposed requiring high GCWR MDVs, defined as MDV with a gross combined
weight rating (GCWR) above 22,000 pounds, to be subject to heavy-duty engine certification
instead of chassis-certification for criteria air pollutant standards. Within the proposed rule, the
Agency asked for comment on three alternatives to engine certification of high GCWR MDV:
1. MDV above 22,000 pounds GCWR would comply with the MDV chassis dynamometer
standards with the introduction of additional engine-dynamometer-based standards over the
Supplemental Emissions Test as finalized within the Heavy-duty 2027 and later standards;
2. MDV above 22,000 pounds GCWR would comply with the MDV chassis dynamometer
standards with additional in-use testing and standards comparable to those used within the
California ACC II;
3. Introduction of other test procedures for demonstration of effective criteria pollutant
emissions control under the sustained high-load conditions encountered during operation above
22,000 pounds GCWR.
The Agency is adopting Alternative 2 for the final rule. Alternative 2 includes PEMS-based
moving-average-window in-use standards that are harmonized with California in-use standards
for chassis-certified MDV. The Agency is not finalizing mandatory engine certification for
compliance with criteria pollutant emissions standards for high GCWR MDV; however, there is
still an option that allows manufacturers to choose compliance with light-heavy-duty engine
standards for high GCWR MDV in lieu of compliance with MDV test procedures and standards.
For further information, please see Section III.D. 5 of the preamble.
3.2.4.1 Background on California ACC II/LEVIV Medium-Duty Vehicle In-use
Standards
As part of ACC II and LEV IV programs, California established in-use testing requirements
for chassis-certified LEV IV MDV with a GCWR greater than 14,000 pounds using PEMS-based
moving average window (MAW) in-use standards. California's in-use test procedures and
standards for chassis-certified MDV are based upon California's MAW in-use test procedures
and standards for heavy-duty engines. Under California's program, chassis-certified diesel MDV
with a GCWR greater than 14,000 pounds meet NOx, NMHC, CO, and PM in-use emissions
standards over a three-bin MAW (3B-MAW) with bins representing idle operation (less than or
equal to 6 percent engine load), low-load operation (above 6 percent engine load and less than or
equal to 20 percent engine load) and medium-high operation (above 20 percent engine load) at
up to GCWR. Chassis-certified gasoline MDV with a GCWR greater than 14,000 pounds attest
to meeting NOx, NMHC, CO, and PM in-use emissions standards over a single MAW (1B-
MAW) at up to GCWR (California Environmental Protection Agency, Air Resources Board
2022). Note that under these provisions, chassis certified MDV with a GCWR greater than
14,000 pounds are required to meet g/bhp-hr MAW standards instead of g/mi MAW standards
and use a FTP CO2 family certification level (FCL) calculated either from chassis dynamometer
test results or engine dynamometer test results. The chassis dynamometer FCL definition uses
OBD torque data collection together with CO2 emissions measurement during chassis-
dynamometer testing. The California MDV in-use standards also include a conformity factor
3-47
-------
(CF) for in-use compliance that is multiplied by each emissions standard. The CF is set to 2.0 for
MYs 2027 through 2029. The CF is set to 1.5 for MY 2030 and subsequent model year vehicles.
3.2.4.2 Background on Federal MAW Standards and Procedures for Light-Heavy-
duty Engines and California Harmonization with Federal Standards
In January 2023, the Agency finalized MAW in-use test procedures and NOx, PM, HC and
CO in-use standards for heavy-duty diesel engines based upon a two-bin moving average
window (2B-MAW) instead of California's 3B-MAW. The Federal 2B-MAW standards also
applied a separate temperature correction to light-heavy-duty diesel engine (LHDDE) NOx
standards than the temperature correction used for medium- and heavy-heavy-duty diesel
engines. The Agency established IB-MAW test procedures for gasoline heavy-duty engines
comparable to the California procedures, however the Agency did not establish IB-MAW
standards for heavy-duty gasoline engines.
The Federal 2B-MAW procedures for diesel engines are based upon two 300-second moving
average window (MAW) operational bins. Bin 1 represents extended idle operation and other
very low (< 6 percent) load operation where exhaust temperatures may drop below the optimal
temperature for aftertreatment function. Bin 2 represents higher load operation (>6 percent). The
California 3B-MAW procedures differ chiefly by dividing Bin 2 into Bin 2 and Bin 3, with Bin 2
representing operation from 6 percent to 20 percent load and Bin 3 having operation at greater
than 20 percent load.
Within the Federal in-use procedures, CO2 emissions rates normalized to the maximum CO2
rate of the engine are used as a surrogate for engine power within the bin definitions. The
maximum CO2 rate is defined as the engine's rated maximum power multiplied by the engine's
CO2 family certification level (FCL) for the FTP certification cycle.
In June 2023, a final agreement was signed by representatives of the California Air Resources
Board (CARB), the Truck and Engine Manufacturers Association, Cummins, Daimler Truck,
General Motors, Hino, Isuzu, Navistar, PACCAR, Stellantis, and Volvo. As part of this
agreement, CARB proposed adopting the Federal 2B-MAW test procedures and standards from
40 CFR part 1036 for diesel heavy-duty engines with no changes to California's IB-MAW
standards and procedures for gasoline heavy-duty engines. It is not yet clear when California
harmonization with the Federal 2B-MAW will be finalized by CARB.
California has previously maintained consistent MAW standards and procedures between
their in-use medium-duty chassis-certified Tier IV program and their medium-duty engine-
certified program, however it is also unclear when chassis-certified MDV will be moved to 2B-
MAW procedures and standards within California's program.
3.2.4.3 In-Use Testing Requirements for Chassis-Certified High GCWR Medium-
Duty Vehicles Using the Moving Average Window (MAW).
The agency is finalizing criteria pollutant standards based on chassis-certification along with
additional in-use standards for high GCWR MDV. We are also finalizing optional engine-
certification for high GCWR MDV, however we are not finalizing a requirement for engine-
certification for these vehicles. See Section III.D.5.iv of the preamble for further description of
the option to certify engines under 40 CFR part 1036. The rest of this section describes the in-use
3-48
-------
standards and procedures for high-GCWR MDVs certified to criteria pollutant emission
standards under 40 CFR part 86, subpart S.
The agency is finalizing in-use standards for MY 2031 and later high GCWR MDVs
consistent with the California provisions for certification and in-use standards for chassis-
certified MDV based on moving average windows (i.e., Alternative 2 in the proposal).
Consistent with the proposal, note that this differs from the California program with respect to
applicability. The Federal in-use standards only apply for MDV with a GCWR greater than
22,000 pounds whereas the California program applies for vehicles above 14,000 pounds
GCWR.
Manufacturers of high GCWR diesel MDV may choose between compliance with CARB's
adopted 3B-MAW standards or EPA's adopted 2B-MAW standards.
The final in-use test procedures and standards for high GCWR MDV are based upon Federal
heavy-duty in-use test procedures and standards for light-heavy-duty engines with some
modifications that include:
1. Optionally allow FCL to be derived entirely from chassis dynamometer testing, emissions
measurement, and OBD data collection.
2. Addition of an option for 3B-MAW standards and procedures for high GCWR diesel
MDV. Note that the 3B-MAW standards specified in this final rule incorporate California's full-
phase-in conformity factor of 1.5.
3. Addition of IB-MAW standards for high GCWR spark-ignition MDV.
The high GCWR gasoline MDV standards are summarized in Table 3-13. High GCWR diesel
3B-MAW standards and off-cycle bin definitions are summarized in Table 3-14 and Table 3-15.
High GCWR diesel 2B-MAW standards and off-cycle bin definitions are summarized in Table
3-16 and Table 3-17. Note that the 2B-MAW standards also include a PEMS accuracy margins,
which are summarized in Table 3-18. The 2B-MAW and 3B-MAW NOx standards, including
any applicable accuracy margins and temperature corrections, are compared in Figure 3-15 and
Figure 3-16. For further details regarding the finalized high GCWR MDV in-use standards,
please refer to 40 CFR 86.1811-27(e). For further details regarding the finalized high GCWR
MDV in-use test procedures, please refer to 40 CFR 86.1845-04(h).
Table 3-13: Spark-Ignition Standards for Off-Cycle Testing of High GCWR MDV.
NOx HC PM CO
mg/hphr mg/hphr mg/hphr g/hphr
30 210 7.5 21.6
: a Standards already include a conformity factor of 1.5 and Accuracy Margins
i do not apply. _
: b There is no applicable temperature condition, Tamb, for spark-ignition
i vehicles certifying to moving average window standards.
: c In-use standards for spark-ignition vehicles are not divided into separate
; operation bins.
3-49
-------
Table 3-14: Compression-Ignition Standards for Off-Cycle Testing of High GCWRMDV
Over the 3B-MAW Procedures.
Off-cvclc NOx HC mg/hphr PM mg/hp hr CO g/hp hr
Bina.b.c
Bin 1 7.5 g/lir
Bin 2 75 mg/hp hr 21 7.5 23.25
Bin 3 30 mg/hp hr 21 7.5 23.25
a Vehicles optionally certifying to 3-bin moving average window standards.
: b Standards already include a conformity factor of 1.5 and Accuracy Margins do not apply.
c There is no applicable temperature condition. Tamb. for vehicles certifying to 3-bin moving
: average window standards.
Table 3-15: Criteria for 3B-MAW Off-Cycle Bins.
Bi" Normalized CO2 emission mass over the 300 second lest interval
1 mC02,norm.testinterval < 6.00%.
Bin 2 6.00% < mC02,norm,testinterval 5; 20.00%
^ mC02,norm.testinterval > 20.00%.
Table 3-16: Compression-Ignition Standards for Off-Cycle Testing Over the 2B-MAW.
Off-cycle NOx'1 Temperature HC PM CO
Bin" adjustment0 mg/hphr mg/hphr g/hphr
Bin 1 10.0 g/hr (25.0-Tamb) • 0.25
Bin 2 58 (25.0-Tamb) • 2.2 120 7.5 9
mg/hphr
; a Vehicles and engines certifying to 2-bin moving average window standards,
b Use Accuracy Margins from 40 CFR 1036.420 (a).
c Tamb is the mean ambient temperature over a shift-day, or equivalent. Adjust the off-cycle NOx standard for Tamb below j
25.0 °C by adding the calculated temperature adjustment to the specified NOx standard. Round the temperature adjustment
to the same precision as the NOx standard for the appropriate bin. If you declare a NOx FEL for the engine family, do not
apply the FEL scaling calculation from 40 CFR 1036.104(c)(3) to the calcul t d t perature adjustment.
Table 3-17: Criteria for 2B-MAW Off-Cycle Bins.
Bin Normalized CO2 emission mass over the 300 second test interval
1 mC02,norm.testinterval < 6.00%.
^ mC02,norm.testinterval > 6.00%.
Table 3-18: Accuracy Margins for In-Use Testing Over the 2B-MAW.
NOx HC PM co
Bin 1 0.4 g/hr
Bin 2 5 mg/hp hr 10 mg/hp hr 6 mg/hp hr 0.025 g/hp hr.
3-50
-------
4
2
0
0 5 10 IS 20 25 30 35 40
Wn ("C)
Figure 3-15: 2B-MAW Bin 1 In-use NOx standard with Ambient Temperature Correction
and PEMS Accuracy Margin compared to 3B-MAW Bin 1 In-use NOx standard.
3-51
-------
120
100
80
£ 50
40
107mg/hp hr
20
G Q
2B-MAW Bin 2
~- - • +
3B-MAW 6m 2
- •
3B MAW Bin 3
75 ng/hp hr (Bin 2}
63 ng/hp hr
30 ng/hp hr (Bin 3|
0
0 5 10 15 20 25 30 35 40
i..., rc)
Figure 3-16: Figure 8: 2B-MAW Bin 2 In-use NOx standard with Ambient Temperature
Correction and PEMS Accuracy Margin Compared to 3B-MAW Bin 2 and Bin 3 In-use
NOx standard.
3.2.4.4 Optional High GCWR Medium-Duty Vehicles Engine Certification
The final rule includes the option for engine-based certification to emission standards for both
spark ignition and compression ignition (diesel) engines, and complete and incomplete vehicles
that have GCWR above 22,000 pounds. Engine certification would require compliance with all
the same engine certification criteria pollutant requirements and standards as for MY 2027 and
later engines installed in Class 4 and higher heavy-duty vehicles, including the recently adopted
NOx, HC\ PM, and CO standards, useful life, warranty and in-use requirements. 88 FR 4296
(Jan. 24, 2023). Complete MDVs would still require chassis dynamometer testing for
demonstrating compliance with GHG standards as described in Section III.C.3 of the preamble
and are included within the fleet average MDV GHG emissions standards along with the other
MDVs. Manufacturers would have the option to certify incomplete MDVs to GHG standards
under 40 CFR 86.1819, or under 40 CFR parts 1036 and 1037. Note that existing regulations, see
40 CFR 1037.150(1), allow a comparable dual-testing methodology, which utilizes engine
dynamometer certification for demonstration of compliance with criteria pollutant emissions
standards while maintaining chassis dynamometer certification for demonstration of compliance
with GHG emissions standards under 40 CFR 86.1819.
3.2.5 Feasibility of ICE-Based Vehicle NMOG+NOx Standards
The current light-duty vehicle Tier 3 standards will be fully phased-in by MY 2025 and will
result in a fleet average standard for passenger cars and light trucks of 30 mg/mi FTP
NMOG+NOx. This means on average, the light-duty fleet will be certified to Tier 3 Bin 30 in
MY 2025. As discussed in Section III.D.2.iii of the preamble, the declining FTP NMOG+NOx
3-52
-------
fleet average in this proposal is fully feasible with the introduction of zero emission vehicles
such as BEVs. There are many pathways to NMOG+NOx compliance with lower levels of BEV
penetration that would require some level of ICE aftertreatment improvement (examples
provided in Chapter 3.2.5.1 below). Even with BEV penetrations as low as 35 percent (e.g., as
projected in our No Additional BEVs sensitivity) and considering many existing ICE vehicles
already emit below 30 mg/mile, manufacturers would comply with the NMOG+NOx standard
with minimal aftertreatment improvements for their remaining ICE vehicles.
Continued reductions in ICE-based vehicle emissions can provide an alternative pathway to
compliance (or, at a minimum, offset the number of BEVs a vehicle manufacturer might choose
to produce) to meet the standards. Many ICE vehicles today already meet the MY 2032
NMOG+NOx standards over the FTP. EPA reviewed the MY 2023 Annual Certification Data for
Vehicles (U.S. EPA 2023) and identified 39 vehicle models with ICE NMOG+NOx emissions
performance on the FTP below 15 mg/mi, four of which are certified to 10 mg/mi or less (Table
3-19). Beyond the vehicles listed in Table 3-19, an additional 60 models are certified to between
15 mg/mi and 20 mg/mi.
3-53
-------
Table 3-19: Examples of NMOG+NOx MY 2023 certification emissions that are less than
15 mg/mi.
Manufacturer
Model
C ,'crtificd
NMOCJ+NOX
(g/ini)
Vehicles Certified at 10 ing/mi or less
Mercedes-Hen/.
N 5X11 e IMA I II
0.006
Hyundai
Sonata Hybrid
0.010
Kin
Niro Hybrid
0.010
Kin
NiroPHEV
0.010
Vehicles Certified at 15 ing/mi or less
I5MW
X5 xDrive45e
0.011
KOMOC O
Escape PHEV
0.011
KOMOC O
Corsair AWD PHEV
0.011
Ilvuiuliii
Sonata Hybrid
0.011
Kill
Niro Hybrid
0.011
Kill
Soul
0.011
Kill
Forte 5
0.011
Toyota
Prius Prime
0.011
I5MW
X5 xDrive45e
0.012
KOMOC O
Escape
0.012
KOMOC O
Escape AWD HEV
0.012
KOMOC O
Corsair AWD PHEV
0.012
Ilvuiuliii
Kona
0.012
Kill
NiroPHEV
0.012
Kin
Soul
0.012
Mercedes-Hen/.
GLC 300 IMA I K
0.012
Mercedes-Hen/.
N 5X11 e IMA I II
0.012
Toyota
Prius AWD XLE/LTD
0.012
Volkswagen
Q3
0.012
I5MW
X3 xDrive30i
0.013
KOMOC O
Escape PHEV
0.013
Ilvuiuliii
Santa Fe PHEV
0.013
Kin
Sportage PHEV
0.013
Kin
Sorento PHEV
0.013
Kin
Soul
0.013
Kin
Forte 5
0.013
Toyota
Corolla Cross AWD
0.013
Toyota
NX 350 AWD
0.013
Toyota
Corolla XSE
0.013
I5MW
X5 xDrive45e
0.014
I5MW
John Cooper Works Conv.
0.014
c;m
Malibu
0.014
Ilvuiuliii
Elantra
0.014
Volkswagen
Tiguan AWD
0.014
Volkswagen
Taos
0.014
The Agency also analyzed emissions certification data of MY 2022 and MY 2023 emissions
families for medium-duty vehicles (MDVs). The emissions family certification data are
graphically represented in Figure 3-17 for gasoline and diesel MDV vans and pickups using a
"box-and-whisker" plot (Frigge, Hoaglin and Iglewicz 1989) (Tukey 1977) (Benjamini 1988).
The upper and lower boxes correspond to the first and third quartiles (the 25th and 75th
percentiles), respectively, of the NMOG+NOx emissions data for each MDV category. The
horizontal line between each set of upper and lower boxes represents median emissions and the
"x" represents mean emissions. The upper vertical line or "whisker" extends from the median to
the highest value that is within 1.5X inner quartile range (IQR) of the median, where IQR is the
distance between the first and third quartiles. The lower "whisker" extends from the median to
the lowest value within 1.5X IQR of the median. A certification emissions data point was
3-54
-------
considered an outlier if it exceeded a distance of 1.5 times the IQR below the 1st quartile or 1.5
times the IQR above the 3rd quartile and is represented as a "dot" in the "box-and-whisker" plot.
The analysis found significant compliance headroom for MDVs below the current Tier 3
NMOG+NOx emissions standards for Class 2b and Class 3 MDVs27, with median NMOG+NOx
emissions of approximately 100 mg/mi for gasoline pickups, approximately 80 mg/mi for
gasoline vans, and approximately 130 mg/mi for diesel vans. Median emissions for diesel
pickups were approximately 170 mg/mi. Application of NOx control technologies that will be
implemented on Class 4 through Class 8 trucks for compliance with 2027 and later emissions
standards, such as dual-SCR, active thermal management measures, and passive thermal
management design, can also be applied to MDV diesels in order meet new MDV NMOG+NOx
emissions standards. A thorough discussion of these technologies may be found within the
Regulatory Impact Analysis for the 2027 and Later Heavy-Duty Engine and Vehicle Standards
Final Rule (U.S. EPA 2022).
¦ Gasoline Vans ¦ Gasoline Pickups Q Diesel Vans ~ Diesel Pickups
Figure 3-17: MY2022-2023 MDV box and whisker plot showing the
interquartile range (boxes) and data within 1.5X of the interquartile range (whiskers) for
NMOG+NOx certification data.
EPA recognizes that compliance headroom is a concern for vehicle manufacturers. Vehicle
manufacturing variation, test to test variations, and test location variables all contribute to a
manufacturer's desire to have 30 to 40 percent compliance headroom when submitting data and
vehicles to EPA for certification. However, given the low emissions demonstrated by current
MY 2023 LD vehicles and MY2022 and MY2023 MDVs, EPA believes that manufacturers will
be able to utilize the lower bins finalized in this rulemaking and maintain their target compliance
headroom. Certification of ICE-based vehicles to the lower bins in combination with the
introduction of an increasing number of PEVs into the fleet average provides a feasible
compliance pathway to meet the declining FTP NMOG+NOx fleet averages for both LD
vehicles and MDVs.
27 Tier 3 FTP NMOG+NOx emissions standards are 178 mg/mi for Class 2B and 247 mg/mi for Class 3 MDVs.
3-55
-------
3.2.5.1 Technologies that can reduce NMOG+NOx emissions
Multiple technologies and control strategies are available to reduce tailpipe emissions from
ICE. Included below is a survey of possible changes that can be made to aftertreatment systems,
engine operation, catalyst heating, and hybrid/PHEV control strategy to reduce NMOG+NOx
emissions. These technology changes primarily focus on improving three-way catalyst (TWC)
performance and/or engine-out emissions in the first 60 to 100 seconds of the FTP. During this
time period, the catalyst is still relatively cold; it is estimated that up to 70-plus percent of total
tailpipe cycle emissions are created over the first sixty seconds of the FTP (Warkins, Tao and
Lyu 2020), while 90 percent are created over bag 1 of the FTP (Moses-DeBusk, Storey, et al
2023).
However, some of these technologies also serve to reduce NMOG+NOx over the remaining
bags of the FTP where the catalyst is already warm. Implementing these technologies also serve
to reduce NMOG+NOx over the other test cycles (such as the US06) that are run completely
under warm conditions. Vehicles with certification data over both the FTP and US06 cycles tend
to have lower US06 values, averaging 97 percent of the FTP value (see Figure 3-18). In
particular, vehicles with the lowest FTP NMOG+NOx emissions - below about 20 mg/mi - have
US06 certification values that are primarily at the same as or lower than FTP values (as indicated
by the shaded triangle in Figure 3-18).
0.1
1
^2 0.08
to
o
(/i
z>
- 0.06
_aj
CD
" 0.04
x
O
2
+
(J
0 0.02
1
0
NMOG+NOx cert level: FTP (g/mi)
Figure 3-18: Comparison between FTP and US06 NMOG+NOx certification values for MY
2023 vehicles with FTP certification values below 100 mg/mile.
MY 2023 NMOG + NOx certification values
•
•
•
•
•
• •
•
•
•
/ - •
/
••
• 4
• 1.
•.Vi
• y
* / •
r a
•
• Conventional
+ MHEVs
•
•
4
i •• •
>
• •
•
Strong hybrids
+ PHEVs
•
/V v •
%
•
0 0.02 0.04 0.06 0.08 0.1
3-56
-------
3.2.5.2 Changes in aftertreatment system hardware
Altering the hardware in the aftertreatment system, particularly the TWC itself, can improve
emissions performance.
3.2.5.2.1 Lower mass catalysts
Reducing the thermal mass of the three-way catalyst reduces the energy required to heat the
catalyst, reduces the time required for light-off, and thus reduces the tailpipe emissions
associated with the engine cold start. Decreasing total catalyst mass can be accomplished by
reducing substrate weight or washcoat amount.
To reduce substrate weight, Corning has introduced high-porosity, low mass catalyst
substrates commercially (Corning 2023). Testing on multiple vehicles demonstrated a faster
catalyst temperature increase and earlier light-off, culminating in an average 15 percent
reduction in both NMHC and NOx for vehicles with FTP emissions values ranging from 10 to 40
mg/mi (Warkins, Tao and Lyu 2020). Testing with a similar high porosity substrate on a pickup
with a 5.3L gasoline engine showed a 22 percent reduction in NMHC + NOx, from 25.3 mg/
mile to 19.7 mg/mi (Asako, et al. 2022). Similar results (25 percent NOx reduction and 21
percent NMHC over an RDE cycle) were obtained by researchers from NGK and Ford
(Nakasumi, et al. 2023).
Alternatively, the mass of the washcoat can be reduced. Researchers from Honda tested a
catalyst where the washcoat was reduced by 40 percent, leading to a 25 percent reduction in
NMOG+NOx over the FTP (Nakanishi, et al. 2019). To maintain the performance of the
reduced-washcoat catalyst after aging, improvements were made in the carrier materials and a
phosphorus trapping material was added.
3.2.5.2.2 Higher surface area catalysts
Increasing the cell density in a catalyst increases the geometric surface area, leading to higher
conversion efficiency and better emissions performance. Researchers from NGK and Ford have
shown that a high cell density in the close-coupled catalyst improves NMHC + NOx emissions
both during a cold start and during later warm phases, with a 6 percent NOx reduction and 14
percent NMHC over an RDE cycle (Nakasumi, et al. 2023). They also tested high-porosity
substrates, showing that a combination of higher porosity (and thus reduced heat capacity) with
higher cell density (and thus increased surface area) can reduce both cold emissions but also
overall cycle emissions.
3.2.5.2.3 Advanced washcoat and PGM technology
Higher platinum group metal (PGM) loading in the washcoat can increase catalyst efficiency,
particularly in the light-off phase between 150 °C and 350 °C. For example, Maurer reports that
with higher PGM loading compared to a state-of-the-art catalyst, a NOx conversion rate of 99
percent can be reached at a 50 °C cooler catalyst temperature (Maurer, Yadla, et al. 2020).
Researchers from Johnson Matthey report that increasing the amount of rhodium loading on a
TWC has a relatively large effect for a small increase in rhodium, with the temperature of 50
percent conversion (T50) decreasing by 20 °C for both HC and NOx with a small increase in Rh
(Cooper and Beecham 2013). Similarly, Alikin and Vedygin found that increasing rhodium
percentage in the TWC can potentially reduce the T50 by 60 °C for both HC and NOx (Alikin
3-57
-------
and Vedyagin 2016). Researchers at Ford and Johnson Matthey looked at multiple formulations
of a catalyst, finding that the use of rhodium supported on alumina, along with an overlayer of
either titanium or zirconium, substantially reduced the T90 temperature (C. Lambert 2018)
(Theis, Getsoian and Lambert 2018). In their experimentation, they found that a rhodium loading
of a titanium overlayer of 0.5 percent produced optimum results. Compared to a commercial
catalyst, T90 was reduced by 90 °C for HC and over 100 °C for NO.
3.2.5.2.4 HC traps, NOx adsorbers, and catalyzed filters
Another concept for aftertreatment hardware is the addition of an HC trap. HC traps
specifically target the emissions produced at engine start, before the TWC lights off. To
investigate this, researchers looked at placing an HC trap in the exhaust upstream of the TWC
(Maurer, Yadla, et al. 2020). This showed that the trap adsorption efficiency was 75 percent, and
nearly 85 percent immediately after the initial cold start. Another concept looked at adding a HC
trap downstream of the TWC, along with a GPF (Moses-DeBusk 2021). Experiments were done
on pickup trucks with 2.7L turbo V6 and 5.3L naturally aspirated V8 engines. An HC trap alone
was found to reduce total HCs by 65 percent over the first bag of FTP, while the addition of a
GPF to the exhaust reduced total HCs by 77 percent over the first bag of FTP.
Another study looked at adding an HC and NOx adsorber to the catalyst (Gao, et al. 2012).
Simulations with hybrid vehicles over a cold-start UDDS showed a two- to three-fold reduction
in tailpipe HC over the cycle and a 30 to almost 50 percent reduction in NOx.
A somewhat unconventional application of a HC trap was performed with a VW vehicle
(Moser, et al. 2021). The original vehicle had two TWCs, one close-coupled to the turbocharger,
and the other downstream under the floor. Replacing the underfloor TWC with a HC trap
reduced the NMHC over the FTP from 8 mg/mi to about 5.5 mg/mi. Replacing the close-coupled
TWC with a HC trap instead reduced the NMHC to 5 mg/mi. The same study optimized the
usage and placement of the catalyst materials in the remaining under-floor TWC. The
combination of the close-coupled HC trap and optimized under-floor TWC resulted in a
reduction in NMHC + NOx over the FTP from 29.4 mg/mi NMHC + NOx to 13.7 mg/mi. With
the same configuration, NMHC + NOx over the US06 cycle was maintained below 10 mg/mi.
The addition of a catalyzed gasoline particulate filter can also decrease tailpipe emissions.
Researchers installed a catalyzed, passive regenerating, wall-flow GPF in a Ford Focus
downstream of the TWC (Chan, et al. 2016). Over the FTP, this configuration reduced NOx by
34 percent (from 17.6 mg/mi to 11.5 mg/mi) and THC by 38 percent (from 11.8 mg/mi to 7.3
mg/mi). Because of the placement, these reductions were smaller over the cold start because the
catalysts on the GPF did not reach the light-off temperature early enough. However, over the
US06, the same configuration reduced NOx by 88 percent (from 47.2 mg/mi to 5.7 mg/mi) and
THC by 54 percent (from 25.9 mg/mi to 12.0 mg/mi). Similar results are seen when installing a
second stage catalyst in the same location, with NMOG reductions of 86 percent and NOx
reduction of 20% over the US06 (Roy, et al. 2018).
3.2.5.3 Changes in engine operation
Below is a survey of possible changes when made to engine operation have the potential to
reduce NMOG+NOx emissions.
3-58
-------
3.2.5.3.1 Changing valve timing for initial engine start
Changing valve timing during a cold engine start can reduce the amount of pollutants which
are emitted by the engine. Researchers from MIT studied the effects of negative valve overlap
(NVO), combining late intake valve opening (IVO) combined with early exhaust valve closing
(EVC) (Rodriguez and Cheng 2016). They found that delayed IVO created lower in-cylinder
pressures and thus improved spray flash boiling due to the lower in-cylinder pressures, while
improving mixing and potentially reducing wall wetting. The early EVC traps more high
temperature residuals, improving spray vaporization. The combination reduced engine-out HC
from the cold start by 30 percent (and PM by 28 percent). The same strategy reduces engine-out
NOx from the cold start by 59 percent due to reduced effective compression ratio and thus lower
peak temperature during the first 2 engine cycles, and the increased residual gas fraction after the
2nd cycle (Rodriguez and Cheng 2017). Other researchers found similar results, where
increasing NVO to 138 degrees reduces NOx by up to 95 percent and HC by 20 percent (Zhu, et
al. 2013).
Similarly, changing both opening and closing timing of the exhaust valve can also reduce
cold-start engine-out emissions. Early exhaust valve opening (EVO) increases exhaust gas
temperature, while the early EVC enhances fuel vaporization and air/fuel mixing. The
combination can reduce start-up hydrocarbon emissions by 27 percent (Bohac and Assanis
2004).
Other researchers looked at the effects of injection, spark, and valve timing, combined with
the higher cranking speeds made possible with hybrid powertrains. Over the first few cycles, a
1600 rpm cranking speed combined with late intake first injection, highly retarded spark timing,
and high valve overlap conditions reduced the engine-out HC emissions by 94 percent compared
to the baseline conditions (Khameneian, et al. 2022).
Another potential strategy for reducing tailpipe emissions is cylinder deactivation. Cylinder
deactivation is typically used to reduce fuel consumption and GHG, but it can also be an enabler
for early catalyst heating. For example, rolling cylinder deactivation using Tula's Dynamic Skip
Fire (DSF) improves the combustion stability due to higher cylinder load, thus allowing more
spark retard and higher exhaust temperatures. A study was done on the potential for DSF to
increase exhaust gas temperature during a cold start (Luo, et al. 2020). Combustion phasing was
retarded by 20 degrees with no degradation of combustion stability, resulting in an increase in
catalyst temperature of 100 °C in the first 17.5 seconds. Depending on the DSF pattern chosen,
cold start THC can be reduced by 9 percent to 19 percent, and cold start NOx by up to 50
percent.
3.2.5.3.2 Engine Speed
A typical 12V starter motor will accelerate the engine to only about 300 rpm before the first
injection and ignition event. The low speed causes some issues with combustion, such as bad
evaporation and homogenization and fuel enrichment (Menne, et al. 2022). Increasing the engine
speed at the first ignition event avoids these problematic factors and reduces emissions from the
first few engine cycles. This would require more powerful starter motors, operating at 48V or
higher, such as those seen in mild and strong hybrid vehicles. These 48V systems could be
implemented for engine start in conventional vehicles, with the advantage of the GHG reduction
opportunities afforded by a mild hybrid system.
3-59
-------
3.2.5.4 Addition of active catalyst heating
Active catalyst heating involves the installation of additional hardware in the aftertreatment
system that is used to preheat the catalyst before and immediately after engine start. The heater
can be either electrically powered or can be a fuel burner. Below is a summary of possible active
catalyst heating changes that can more quickly heat the catalyst to operating temperature and
reduce NMOG+NOx emissions.
3.2.5.4.1 Electrically heated catalysts
Electrically heated catalysts (EHC) increase the temperature of the catalyst before and/or just
after engine ignition, causing the catalyst to reach light-off temperature more quickly than it
would otherwise. These systems often include an air pump to transfer heat from the heating
element to the catalyst without requiring air flow from the engine (Maurer, Kossioris, et al.
2023). Although EHCs can be powered using the 12V electrical system, the low voltage limits
the power that can be applied to the heater and thus limits the rate of temperature rise. Therefore,
EHCs are more effective with 48V or higher electrical systems (Jean and Goncalves 2023)
(Laurell, et al. 2019). EHCs are already commercially available from suppliers such as Vitesco
(Bargman, et al. 2021) (Continental 2015) and Faurecia (Jean and Goncalves 2023) (Faurecia
2023).
Researchers from Corning tested a 48V electrical heater integrated as closely as possible to
the first TWC, then used the EHC to pre-heat it before running over the WLTC (Kunath, et al.
2022). Pre-heating initiated approximately 15 seconds before engine ignition resulted in an
earlier temperature increase in and light-off of the catalyst, reducing both THC and NOx in the
first bag by 30-40 percent. The system was also tested with a supplemental air supply system to
accelerate the heat-up of the TWC. With the supplemental air injection, the TWC was heated
more quickly, resulting in a 70-75 percent reduction in THC and an 80 percent reduction NOx.
Researchers from Volvo and Continental found similar results with a system containing a 48V
electrical heater without a supplemental air supply. Here, pre-heat began approximately 12
seconds before engine ignition, which reduced both HC and NOx by 50 percent during the initial
30 seconds of engine run time.
3.2.5.4.2 Electrically Heated catalysts in HEVs and PHEVs
Electrically heated catalysts can work particularly well with HEVs or PHEVs for two reasons:
first, the higher voltage batteries provide a ready source of power for the EHC, allowing higher
power and quicker temperature rise than achievable with 12V systems. High voltage HEVs may
require a DC/DC converter to step the voltage down to an appropriate value, but these converters
are already commercially available (Green Car Congress 2021). The second reason is that a
hybrid vehicle can be driven exclusively with the electric motor at the beginning of the cycle,
allowing additional time for the catalyst to heat up before the initial engine start. The additional
time can be very beneficial to emissions reduction; for example, Maurer showed that while a 15
second preheating time significantly reduces tailpipe NOx, increasing the preheat time to 30
seconds nearly eliminates the NOx (Maurer, Kossioris, et al. 2023).
In one experimental demonstration, installation of an electric heater on a Mitsubishi Outlander
PHEV showed that total NMHC + NOx over the FTP could be reduced from 22 mg/mi to 12
mg/mi (Jean and Goncalves 2023). The authors note that this level of performance requires either
3-60
-------
a higher voltage power source to the heater (for example a 48 V source rather than 12V) to
provide adequate heating power during a cold start, or the additional time afforded by the
delayed engine start in a HEV.
In another demonstration, Toyota showed that the application of an electrically heated catalyst
to a Prius Prime PHEV along with a motored engine start sequence (where the engine is brought
up to idle speed by the electric motor and motored for several seconds prior to ignition) and a
catalyst control to avoid over-heating resulted in FTP NMOG+NOx emissions of less than 3
mg/mi (Kawaguchi, et al. 2019).
3.2.5.4.3 Exhaust burners
Like electric heaters, exhaust burners are used to heat the catalyst before engine ignition.
Burners have the advantage of not requiring higher voltage sources to supply sufficient heat, and
thus may be more adaptable to non-hybrid applications. The burner itself also produces small
amounts of pollutants, on the order of 4.5 mg of HC and 7.1 mg NOx over 15 seconds of burner
operation (Maurer, Kossioris, et al. 2023). However, the addition of a small carbon canister for
HC desorption from the burner exhaust may ameliorate this (Maurer, Yadla, et al. 2020).
One experimental investigation into a burner with a pre-catalyst showed a significant increase
in temperature of the catalyst before engine start (Clenci, et al. 2022). This investigation
incorporated a substantial pre-heating time (approximately 100 seconds), but catalyst
temperatures were above light-off well before the end of the pre-heat time.
3.2.5.5 Solutions for Hybrids and PHEVs
The incorporation of hybrids and PHEVs into the fleet can present challenges to emissions
reduction due to extended engine off operation eventually requiring engine restarts. In particular,
high-power cold starts seen in PHEVs may yield higher NMOG+NOx exhaust gas emissions
(Pham and Jeftic 2018). However, as noted above in Chapter 3.2.5.1.2.2, the availability of the
electric drive motor can enable engine start and re-start strategies that reduce cold-start
emissions. In fact, the majority of the low NMOG+NOx emission vehicles listed in Table 3-19
are strong hybrid or plug-in hybrid vehicles. Higher voltage batteries used by hybrids and
PHEVs may also enable electrically heated catalysts, as discussed in Chapter 3.2.5.1.3.1.
A further examination of all MY 2023 certification values (U.S. EPA 2023), seen in Figure
3-19 as a function of engine displacement, shows that HEV and PHEV emissions for
NMOG+NOx tend to be lower than corresponding conventional and mild hybrid vehicles
(MHEV). This suggests that, rather than allowing the high-power cold starts to inflate vehicle
emissions, manufacturers are taking advantage of the emissions reduction possibilities available
when the electric drive motor is used in conjunction with the engine. Additionally, Figure 3-18
shows that HEVs and PHEVs tend to have even lower NMOG+NOx emissions over the US06
cycle than over the FTP.
3-61
-------
MY 2023 NMOG + NOx FTP certification values
• Conventional + MHEVs
~ Strong hybrids + PHEVs •
- I : • .. !
a o.i • • #
-------
obtained by combining more than one strategy. Moreover, multiple strategies can react
synergistically to produce greater benefits.
For example, researchers from Hyundai suggest that "almost zero" emissions over the FTP
are possible by applying multiple aftertreatment solutions to selectively reduce emissions (Kim,
et al. 2018). They identified three areas that contribute the vast majority of NMOG+NOx
emissions: cold-start NOx, cold-start HC, and fuel-cut NOx emissions that occur later in the
cycle when the engine cuts fuel and delivers fresh air into the aftertreatment system. To combat
the first, the researchers used a cold-NOx TWC to act as a NOx adsorber during the first 15
seconds of engine operation to substantially reduce cold-start NOx. They combined this with a
fuel-cut NOx TWC that was able to eliminate 74 percent of the NOx emissions over bag 2 of the
FTP, and then added an underfloor HC trap and EHC which was able to reduce cold start HC
emissions by 40 percent, this addressing all of the primary causes of high NMOG+NOx.
In a similar way, Toyota combined a heated catalyst in a Toyota Prius Prime with the electric
operation of the PHEV at startup and a motored engine start. This reduced emissions on the FTP
to less than 2 mg/mi NMOG and less than 1 mg/mi NOx and reduced emissions on the WLTC to
less than 10 mg/mi NMHC and around 1 mg/mi NOx (Kawaguchi, et al. 2019). Similarly, FEV
combined an engine timing change to reduce NOx with a heated catalyst and a NOx trap in an
HEV. This reduced emissions to 4 mg/mi HC and well under 1 mg/mi NOx on the WLTC, and 5
mg/mi HC and well under 1 mg/mi NOx on a "worst case" dynamic RDE cycle (Maurer, Yadla,
et al. 2020).
These results of these experimental investigations show that NMOG+NOx values well under
15 mg/mi are achievable.
3.2.5.6.1 Current ICE Emissions at -7°C FTP
As described in preamble Section III.D.2.iii, EPA is replacing the existing -7°C FTP NMHC
fleet average standard of 300 mg/mi for LDV and LDT1, and 500 mg/mi fleet average standard
for LDT2-4 and MDPV, with a single NMOG+NOx fleet average standard of 300 mg/mi for
LDV, LDT1-4 and MDPV for the -7°C FTP cycle. NMOG should be determined as explained in
40 CFR 1066.635.
-7°C FTP NMOG+NOx emissions were measured from several current vehicles and results
are shown in Table 3-20. The LDV and LDT4 vehicles were significantly under the 300 mg/mi
standard being finalized and the MDV exceeded it. Electrification and the technologies described
in RIA Chapter 3.5.1 are readily available for further reductions in fleet average -7°C FTP
NMOG+NOx emissions.
Table 3-20: -7°C FTP NMOG+NOx emissions measurements at EPA.
Vehicle
NMOG+NOx (mg/mi)
2021 Corolla 2.0L
138 ± 12
2019 F150 5.0L
220 ±4
2021 F150 3.5L Powerboost HEV
184 ± 16
2022 F250 7.3L
311 ±62
3-63
-------
3.2.5.6.2 Feasibility of a Single Numerical Standard for FTP, HFET, SC03 and
US06
Table 3-21 below provides a comparison of FTP, HWFE, SC03 and US06 test results for
several MY 2023 light-duty vehicles that represent a broad spectrum of vehicle types and
conventional powertrain technologies. For most of the vehicles identified the FTP results are
higher than the HWFE, SC03 and US06 test results showing that a single standard is feasible and
already being met by some manufacturers. There are several examples where SC03 or US06
results are higher than the FTP results. The data shows that the FTP and the US06 are the most
stringent standards because the of the FTP cold start and the US06 because of higher power
requirements and potential enrichment. The HWFE and SC03 cycles are less stringent due to the
lack of cold start and lower power demands.
Table 3-21: Comparison of FTP, HFET, SC03, US06 cert test results for MY 2023 LD
vehicles.
Manufacturer
Vehicle
Manufacturer Reported NMOG+NOx Values
FTP (g/mi)
HWFE (g/mi)
SC03 (g/mi)
US06 (g/mi)
BMW
X4 xDrive 30i
0.020
0.008
0.008
0.014
BMW
I3s REX
0.014
0.020
0.012
0.011
BMW
540i xDrive
0.036
0.020
0.031
0.029
Ford
Corsair
0.035
0.009
0.09
0.030
Ford
Ranger
0.052
0.033
0.05
0.090
Ford
Explorer
0.038
0.025
0.03
0.030
Ford
F150
0.026
0.014
0.017
0.041
GM
Terrain
0.013
0.001
0.011
0.005
GM/Cadillac
XT6
0.026
0.002
0.008
0.005
GMC
K10 Sierra 4WD
0.026
0.005
0.014
0.008
Hyundai
Genesis
0.038
0.014
0.013
0.056
Hyundai
Elantra
0.037
0.015
0.028
0.072
Kia
Sportage
0.036
0.017
0.036
0.024
Kia
Sorento
0.032
0.016
0.03
0.039
Nissan
Altima
0.015
0.006
0.019
0.017
Porsche
Cayenne Turbo
0.072
0.034
0.050
0.050
Volkswagen
Audi Q3
0.008
0.002
0.009
0.012
Volkswagen
Tiguan A WD
0.017
0.002
0.009
0.008
Volkswagen
Jetta GLI
0.017
0.003
0.016
0.009
Average
0.029
0.013
0.025
0.030
As the result of this change, EPA expects light-duty vehicles to have lower emissions over a
broader area of vehicle operation. Present-day engine, transmission, auxiliary and aftertreatment
control technologies allow closed-loop A/F control and good emissions conversion throughout
the HWFE, US06 and SC03 cycles; as a result, higher emissions standards over these cycles are
no longer justified. Overall, approximately 60 percent of the of test group / vehicle model
certifications from MY 2021 have higher NMOG+NOx emissions in the FTP as compared to the
US06, supporting the conclusion that a single standard is feasible and appropriate.
3-64
-------
3.2.6 Particulate Matter Emissions Control
The final PM standard and its phase-in are presented in Section III.D.3 of the preamble. An
overview of GPF technology is provided in RIA Chapter 3.2.6.1. GPF benefits and the feasibility
and readiness of GPF technology and measurement procedure readiness are demonstrated in RIA
Chapter 3.2.6.2. The importance of the PM test cycles is described in Chapter 3.2.6.3. GPF cost
is discussed in RIA Chapter 3.2.6.4. The impact of GPF on CO2 emissions (energy) is presented
in RIA Chapter 3.2.6.5.
3.2.6.1 Overview of GPF Technology
Gasoline particulate filter (GPF) technology is not new. It has been used in series production
on all new pure GDI type approvals (new vehicle models) in Europe since 2017 (WLTC and
RDE test cycles) and on all new pure GDI vehicles in Europe since 2019 (WLTC and RDE test
cycles) to meet a 6xlOu #/km solid particle number (PN) standard. All new gasoline vehicles in
China have had to meet the same 6xlOu #/km solid PN limit over the WLTC test since 2020,
and in the WLTC and RDE starting in 2023. In India, BS6 stage 2 requires all new pure GDI
vehicles to also meet the 6xlOu #/km solid PN limit in the MIDC (Indian version of NEDC) and
RDE since April 2023. U.S., European, and Asian manufacturers and suppliers have extensive
experience with applying GPF technology to series production vehicles and at least six
manufacturers are assembling at least ten vehicle models with GPFs in the U.S. for export to
other markets. There have been approximately 100 million gasoline particulate filters (GPFs)
installed in light-duty vehicles worldwide.
GPFs require that manufacturers design for safe interaction between hot GPF surfaces and
other vehicle components, systems, and the environment, like the integration of turbochargers,
catalysts, and exhaust components into a vehicle. There are no known safety incidents associated
with the millions of GPFs that have been in use in Europe, China, and India, and the same is
expected in the U.S. market.
GPFs being used in Europe and Asia, and those expected to be used in the U.S. to meet the
0.5 mg/mi PM standard, use a ceramic honeycomb structure with alternating channels plugged at
their inlet and outlet ends (Figure 3-20). GPFs use Cordierite for its low coefficient of thermal
expansion and thermal shock tolerance. GPF substrates typically have 45-65 percent porosity,
10-25 |im median pore size, 6-12 mil (1 mil = 1/1000 inch) wall thickness, and 200-300 cpsi
(cells per square inch) cell density. GPF substrates can be manufactured in various diameters,
lengths, and shapes (e.g., round or oval).
Wall flow filters allow exhaust gas to flow through porous filter walls while particulates are
captured in or on the wall (Figure 3-20). Gasoline engine-out particulates (typically from <10 to
300 nm) are smaller than GPF mean pore size (typically 10-25 jam) but particles are captured at
high filtration efficiencies across their size range by Brownian diffusion (effective for small
particles, <30 nm), interception (intermediate particles), and inertial impaction (large particles).
3-65
-------
Diagram courtesy of Coming
Figure 3-20: Wall-flow GPF design.
A clean GPF initially captures particulates within its pore structure (depth filtration mode);
then at high levels of soot loading, additional particulates form a soot layer (soot cake) on the top
of the wall (soot cake or surface filtration mode). Filtration efficiency improves rapidly with
initial soot and ash loading (Lambert, et al. 2017), then levels off at high filtration efficiency at
high soot loading. GPF backpressure increases with soot and ash loading. Operation at low levels
of soot loading is more challenging for high PM filtration because the GPF cannot rely on stored
soot to assist with filtration, but newer designs, e.g., MY2022 GPFs, use a different pore
structure at the surface of the wall than beneath it to achieve high filtration efficiency with low
flow resistance even without soot or ash loading.
Both bare and catalyzed GPFs are used extensively in series production. Catalyzed GPFs
typically use a washcoat containing Pd and Rh for TWC-type activity. Catalyzed GPFs reduce
the temperature needed to oxidize stored soot and convert gaseous criteria emissions like a TWC
does. A catalyzed GPF can replace one of the TWCs on a vehicle, potentially reducing system
cost. Optimizing filtration, backpressure, and gaseous emissions light-off, however, can be more
challenging with a catalyzed GPF as compared to a bare GPF.
Accumulated soot in a GPF is oxidized to CO2 and H2O in the presence of sufficient
temperature and oxidants (mostly O2 in gasoline engines). Significant rates of GPF regeneration
are observed above 600°C for a bare GPFs (Borger, et al. 2018) and above 500°C for a catalyzed
GPFs (Saito, et al. 2011). In most applications, normal vehicle operation results in sufficiently
high temperature, and deceleration fuel cut-off (DFCO) supplies the GPF with sufficient O2,
resulting in passive regeneration. If a vehicle is only operated at very low load conditions or is
not allowed to warm up, a differential pressure sensor on the GPF can sense imminent GPF
overloading and initiate an active regeneration in which engine settings are adjusted to increase
GPF temperature and supply it with sufficient O2. Active GPF regeneration strategies are
discussed in (van Nieuwstadt, et al. 2019).
GPFs are sometimes installed close to the engine in a "close-coupled" position, immediately
following the TWC, to promote passive regeneration and fast light-off of a catalyzed GPF. Other
times GPFs are installed farther from the engine, in an "underfloor" location, for packaging
3-66
-------
reasons. The lower exhaust gas temperature in underfloor GPFs also reduces backpressure for a
given GPF size and geometry because cooler exhaust has higher density.
GPF size, design, and installation relative to the engine must be considered for the GPF to
have sufficient PM filtration efficiency, sufficiently low backpressure, sufficient ash loading
capacity, fast light-off if the GPF washcoat is relied upon for gaseous criteria emissions
conversion, and good regeneration characteristics. Unlike soot that is oxidized after being
captured by the GPF, ash accumulates on the GPF, typically for the life of the vehicle. Thus, ash
capacity is one factor that determines GPF size for a given application.
GPFs are like diesel particulate filters (DPF) in certain respects. Both GPFs and DPFs are
wall-flow filters that use a ceramic honeycomb substrate with alternating channels plugged at
their inlet and outlet ends to filter particulates. But GPFs operate at higher exhaust gas
temperatures, lower soot loadings, lower exhaust gas O2 and NO2 concentrations, and only see
elevated exhaust gas O2 concentrations during DFCO events. High exhaust gas temperature tends
to keep GPFs at lower soot loading through frequent passive regeneration, making high filtration
efficiency harder to achieve in GPFs, especially in applications that frequently operate at high
load. Newer GPF designs that use different pore structure on the surface have largely addressed
low filtration efficiency at low soot and ash loading. Low soot loading of GPFs results in lower
backpressure than DPFs. GPFs require low heat capacity to make use of relatively short bursts of
elevated O2 during DFCO events, so Cordierite has become the dominant GPF substrate
material. DPFs benefit from higher heat capacity to accommodate larger and less frequent
regeneration events involving larger amounts of soot and high flow rates of exhaust O2 making
silicon-carbon a popular DPF substrate material.
GPFs have a strong record with respect to robust operation and durability since their
introduction into mass production in Europe and China. The first GPFs introduced into series
production have not experienced the failures that troubled early DPFs introduced into series
production, in part because the higher exhaust gas temperatures seen by GPFs promote frequent
passive regeneration, avoiding larger, less frequent regeneration events seen by DPFs that store
larger amounts of soot and have high exhaust O2 flow.
GPF technology has been studied extensively for more than a decade and there exists
extensive literature on GPF. GPF technology review articles include (Saito, et al. 2011), (Joshi
and Johnson 2018), (Boger and Cutler 2019).
3.2.6.2 GPF Benefits and Feasibility of the Standard
This section quantifies the emissions benefits of GPF technology and provides additional
detail, relative to Section III.D.3 .iii of the preamble, on the feasibility of using existing GPF
technology and existing PM test procedures in meeting the final PM standard. The section begins
by describing test vehicles, GPFs, laboratory procedures, and PM sampling. Reductions in PM
mass, elemental carbon (EC), and polycyclic aromatic hydrocarbon (PAH) over a composite
drive cycle are then presented. After that, cycle-specific reductions in PM mass emissions from
three GPF-equipped vehicles are shown. Finally, laboratory to laboratory PM measurement
reproducibility is discussed.
3-67
-------
3.2.6.2.1 Setup and Test Procedures
Testing was performed using five chassis dynamometer test cells at three organizations (EPA,
ECCC, FEV) and five test vehicles in stock and GPF configurations. Test vehicles included
light-duty vehicles and MDV powered by naturally aspirated and turbocharged PFDI (port and
direct fuel injection), DI (direct injection), and PFI (port fuel injection) gasoline engines. GPF-
equipped vehicles used series-production GPFs from MY 2019 and MY 2022 GPF. GPFs used
catalyzed and bare substrates, and they were installed in close-coupled and underfloor
configurations.
The five chassis dynamometer test cells used in the demonstration included three test cells at
EPA National Vehicle and Fuel Emissions Laboratory (NVFEL) in Ann Arbor, one test cell at
ECCC in Ottawa, and one test cell at FEV in Auburn Hills. -7°C FTP tests were performed at
EPA (one test cell), ECCC, and FEV. 25°C FTP and US06 tests were performed at EPA (three
test cells), ECCC, and FEV. Two test vehicles were tested at all organizations, while three
vehicles were only tested at EPA.
All five test cells used in the demonstration were designed to be compliant with 40 CFR parts
1065 and 1066. In each test cell, vehicle exhaust gas is diluted in a constant volume sampler
(CVS) full-flow dilution tunnel system. Heated particulate filter samplers draw dilute exhaust
through a coarse particle separator (-2.5 |im cut at sampling conditions) and 47 mm PTFE
membrane filters [e.g., Measurement Technology Laboratories (MTL) PT47DMCAN],
PM filters were conditioned at 22±1°C, 9.5±1°C dew point for a minimum of 1 hour before
being weighed, before and after being loaded with PM. Filters were weighed using a
microbalance (e.g., Mettler-Toledo XPU2) while being surrounded by strips of Po210 (e.g., 6
strips of 500 |iCi placed on top of and around the filter on the microbalance) for static charge
removal. EPA used an MTL A250 robotic autohandler for filter weighing; ECCC and FEV labs
used manual filter weighing.
To increase sample filter loading and increase signal to noise ratio for GPF-equipped tests,
test cell sampling settings were adjusted relative to test settings typically used to measure Tier 3
levels of PM emissions. For GPF-equipped tests, 1) Dilution factor (DF) was set to the lower or
middle part of the CFR-allowable range of 7-20. 2) 25°C FTP and -7°C FTP tests were mostly
run using a single filter, as allowed by 40 CFR 1066.815(b)(5). 3) In many tests, filter flow was
increased from a typical setting of -58 slpm to -65.25 slpm in phases 1 & 2 and -87 slpm in
phases 3 & 4 to increase filter loading and maintain proper phase weighting using flow
weighting, while staying below the maximum allowable filter face velocity (FFV) of 140 cm/s as
specified by 40 CFR 1066.110(b)(2)(iii)(C). 4) Many of the 25°C FTP and -7°C FTP tests were
run as 4-phase FTP tests as opposed to 3-phase FTP tests, although in hindsight, phase 4 didn't
add much PM mass to the sampling filter and may not be worth the extra test time. Additionally,
to further increase PM filter loading, some of the GPF-equipped tests used double sampled US06
tests, which is not specified in the CFR. In retrospect, a standard single sampled US06 would
have been sufficient for demonstrating compliance with the 0.5 mg/mi standard.
Tier 3 certification fuel was used for 25°C FTP and US06 testing, and Tier 3 winter
certification fuel was used for -7°C FTP testing at all three organizations. Engine oil was
conditioned in each vehicle for a minimum of 600 miles prior to emissions sampling to stabilize
the oil (Christianson, Bardasz and Nahumck 2010).
3-68
-------
The newest of the five test vehicles was an MDV Tier 3 bin 200 MY 2022 Ford F250 with a
naturally aspirated 7.3L V8 PFI engine. It was tested at an ETW of 8000 lb. The F250 was tested
in stock and GPF configurations. Vehicle mileage at the start of testing was 2700 miles. For GPF
testing, series-production MY 2022 GPFs, one for each bank, were installed downstream of the
stock TWCs where the resonator is normally mounted. The GPF used bare substrates of c|)6.443"
x 6" (3.21 L each), 200 cpsi, 8 mil wall thickness. GPFs were aged through 1500 miles of road
driving prior to emissions sampling. GPF pressure drop and temperatures were logged.
The second newest test vehicle was a LDT4 Tier 3 bin 70 MY 2021 Ford F150 HEV with a
turbocharged (PowerBoost) 3.5L V6 PFDI engine. It was tested at an ETW of 6000 lb. The F150
HEV was only tested in GPF configuration. Vehicle mileage at the start of testing was 5000
miles. A series-production MY 2022 GPF was installed after the Y-pipe in place of the resonator.
The GPF used a bare substrate of c|)6.443" x 6" (3.21 L), 200 cpsi, 8 mil wall thickness. The GPF
was aged through 1500 miles of road driving prior to emissions sampling. GPF pressure drop and
temperatures were logged.
The third newest test vehicle was an LDV Tier 3 bin 30 MY 2021 Toyota Corolla with a
naturally aspirated 2.0L 14 PFDI engine. It was tested at an ETW of 3375 lb. The Corolla was
only tested in stock (no GPF) configuration. Vehicle mileage at the start of testing was 5800
miles.
The fourth newest test vehicle was an LDT4 Tier 3 bin 125 MY 2019 Ford F150 with a
naturally aspirated 5.0L V8 PFDI engine. It was tested at an ETW of 5000 lb. The 2019 F150
was tested in stock and GPF configurations. Vehicle mileage at the start of testing was 6700
miles. For GPF testing, a series-production aftertreatment system from a MY 2019 European
Ford Mustang 5.0L replaced the stock aftertreatment system on the F150. The Mustang
aftertreatment system uses a ccl (close-coupled, position 1) TWC and a cc2 catalyzed GPF for
each bank of the engine. The stock aftertreatment system uses a ccl TWC and a cc2 TWC for
each bank. The Mustang GPFs are c()5.2" x 3.3" (1.15 L each), 300 cpsi, 12 mil wall thickness.
The Mustang aftertreatment system was aged through 1500 miles of road driving prior to
emissions sampling. GPF pressure drop and temperatures were logged.
The oldest test vehicle was an LDT4 Tier 2 bin 4 MY 2011 Ford F150 with a turbocharged
(Powerboost) 3.5L V6 DI engine. It was tested at an ETW of 5500 lb. The 2011 F150 was tested
in stock and GPF configurations. Vehicle mileage at the start of testing was 21,100 miles. For
GPF testing, a series-production MY2019 GPF was installed after the Y-pipe in place of the
resonator. The GPF used a catalyzed substrate of c|)5.66" x 4" (1.65 L), 300 cpsi, 12 mil wall
thickness. The GPF was aged through 600 miles of dynamometer driving prior to emissions
sampling. GPF pressure drop and temperatures were logged.
GPF operation was characterized over a range of soot loadings, but because GPFs are required
to comply with the PM standard in any state of soot loading, only results from low-soot-loading
tests (which are worst case with respect to tailpipe PM) are included in the following
demonstration of meeting the PM standard. GPFs were regenerated before each set of tests by
using a sawtooth regeneration cycle. The sawtooth cycle used a series of vehicle accelerations to
raise the GPF to high temperature, each followed by a DFCO to provide the GPF with oxygen
for regenerating stored PM.
3-69
-------
3.2.6.2.2 PM Mass, BC, and PAH Emissions Reductions over a Composite Drive
Cycle
Results shown in this section are from a MY 2011 F150 that was retrofit with an underfloor
catalyzed MY 2019 GPF. Additional details of the vehicle, GPF, and test setup are described in
Chapter 3.2.6.2.1 and in (Bohac and Ludlam, Characterization of a Lightly Loaded Underfloor
Catalyzed Gasoline Particulate Filter in a Turbocharged Light Duty Truck 2023). Emissions
were quantified over a composite test cycle, comprised of vehicle operation at 60 mph cruise
control, 25°C FTP, HFET, and US06. Results are shown in total emissions mass per total
distance of the test cycle. Tailpipe emissions were quantified a) without a GPF, b) with the GPF
in a lightly loaded state with the GPF predominantly in the depth filtration mode
(Konstandopoulos 2008) with 0.1-0.6 g/L, (grams soot per liter of GPF substrate volume), and
c) with the GPF predominantly in the soot-cake or surface filtration mode (Konstandopoulos
2008) with 1.7-2.0 g/L.
Composite cycle PM emissions are shown in Figure 3-21. PM was reduced by 94 percent with
the GPF in a lightly loaded state and 98 percent with the GPF in a heavily loaded state.
Additional results and discussion, including cycle-specific PM reductions, can be found in
(Bohac, Ludlam and Martin, et al. 2022).
bfl
E
8
7
6
5
4
3
2
1
0
I no GPF
I GPF, 0.1-0.6 g/L soot
I GPF, 1.7-2.0 g/L soot
98% reduction
Figure 3-21: Composite cycle PM reduction at low and high GPF soot loading.
Elemental carbon (EC) emissions without a GPF and with the GPF in a lightly loaded state
(0.1-0.6 g/L soot loading) are shown in Figure 3-22. EC was reduced by 100.0 percent in the 60
mph, 25°C FTP, and HFET cycles, and by 98.5 percent in the US06 cycle. EC measurements
were performed using 47 mm quartz fiber filters (Pall Tissuquartz 7202) and a Sunset Laboratory
model 5L OCEC Analyzer running National Institute for Occupational Safety and Health
(NIOSH) method 870.
3-70
-------
EC emissions quantified in this study and airborne black carbon (BC) studied by climate
scientists have different operationally defined definitions, but they are closely related and often
used as surrogates (Bond, Doherty and Fahey 2013).
8
60mph FTP HWFET US06
Figure 3-22: Cycle-specific EC reduction.
Another significant benefit of GPF technology is the reduction of PAH emissions. To quantify
PAH emissions reductions, filter-collected PAH were sampled onto 47 mm quartz fiber filters
(Pall Tissuquartz 7202) and gas-phase PAH were sampled using sorbent tubes (Carbotrap C+F).
PAHs on filter punches and sorbent tubes were thermally desorbed, cryofocused, and speciated
(Agilent 6890/5973 GCMS operated in selected ion mode). 26 PAHs ranging from naphthalene
to coronene were quantified. Additional sampling and analysis details can be found in (Bohac,
Ludlam and Martin, et al. 2022).
PAH emissions reductions are shown in Figure 3-23. Measurements were performed with the
GPF in lightly loaded state (0.1-0.6 g/L soot loading) and in a heavily loaded state (1.7-2.0 g/L
soot loading). Filter-collected PAH emissions (those collected by the PM sampling filter) were
reduced by over 99 percent and gas-phase PAH emissions (those passing through the PM
sampling filter) were reduced by about 55 percent. The percentage reduction in gas-phase PAH
emissions may be less for bare GPFs.
3-71
-------
ru
-------
QUO
3
I
-a
Q)
+->
u
_a>
"o
u
v
1.8
1.5
1.2
0.9
0.6
0.3
0.0
I no GPF
IGPF, 0.1-0.6 g/L soot
I GPF, 1.7-2.0 g/L soot
lL,
CD
c
aj
ro rc ru ro
aj a;
c c
oj _aj
>•
q. q. a.
CD CD CD
^ _E _E
Q_ Q-
03 03
C C
*>• ^
^ u
-£ <
c
_c
_c
LL
03
-t-J
-t-J
c
c
c
_c
OJ
<
03
+->
CD
-C
>•
Q_
_c
aj
c oc
03
L_
o
aj aj aj
c c c
aj d) oj
¦ >¦ ¦£:
*- c
-C -c TO
U U ill
>• O
_C =5
aj
c
aj
Q.
aj cd
c c
aj cd
aj
tu nj Q- u +j
aj
c
aj
CQ
¦g
o o
N N
c C
aj oj
CQ CD
O O
M M
C C
-------
_>
'+->
— T
(U
*- a.
Ve
CM
O *
> ^
X
o
1.2
1.0
0.8
0.6
0.4
0.2
0.0
¦ no GPF
¦ GPF, 0.1-0.6 g/L soot
¦ GPF, 1.7-2.0 g/L soot
99.8% 99.8%
Figure 3-25: Cancer potency weighted toxicity of 20 filter-collected PAHs with no GPF,
lightly loaded GPF, and heavily loaded GPF.
3.2.6.2.3 Cycle-Specific Reduction in PM Mass Emissions from GPF
Application to Three Vehicles
Reductions in cycle-specific PM mass emissions resulting from the adoption of GPF
technology is discussed in this subsection. Three vehicle examples are presented: a MY 2019
F150 5.0L, a MY 2021 F150 HEV 3.5L Powerboost, and MY 2022 F250 7.3L.
The first test vehicle is a MY 2019 F150 5.0L that was tested stock and with a MY 2019
European Ford Mustang 5.0L aftertreatment system. PM emissions are shown in Figure 3-26.
This GPF system reduced PM emissions by 91 percent, 90 percent, and 77 percent in the -7°C
FTP, 25°C FTP, and US06 cycles, respectively. The testing was conducted with the GPFs in a
lightly loaded state. The lightly loaded state was achieved by running a sawtooth GPF
regeneration cycle after several tests were completed. The sawtooth cycle used a series of vehicle
accelerations to raise the GPF to high temperature, each followed by a DFCO to provide the GPF
with oxygen for oxidizing stored PM. Older technology GPFs like the one used in on this test
have lower filtration efficiency at low soot loading than newer GPFs used on the other two test
vehicles subsequently described in this subsection. Figure 3-26 shows that filtration efficiency
was lowest in the US06, which was caused by the passive regeneration that occurs in this cycle.
Additional details of the vehicle and GPFs are provided in Chapter 3.2.6.2.1.
3-74
-------
8
7
¦ no GPF
¦ GPF
6
2
1
0
-7°C FTP
25°C FTP
US06
Figure 3-26: PM emissions from a MY 2019 F150, with and without a MY 2019 GPF.
The second test vehicle is a MY 2021 F150 HEV Powerboost that was tested with and
without a MY 2022 bare underfloor GPF. PM emissions are shown in Figure 3-27. The MY 2022
GPF reduced PM emissions by 99 percent, 96 percent, and 96 percent in the -7°C FTP, 25°C
FTP, and US06 cycles, respectively. The GPF was fully regenerated immediately before each
day of testing using a sawtooth GPF regeneration cycle. The GPF results shown here are worst
case with respect to PM filtration because testing was preceded by a full GPF regeneration, so
the GPF was evaluated with almost no soot. Filtration efficiency of the MY 2022 GPF was
significantly higher than what was achieved with the MY 2019 GPF shown in Figure 3-26,
especially in the US06. Additional details of the vehicle and GPFs are provided in Chapter
3.2.6.2.1.
3-75
-------
QJO
E
8
7
6
5
4
3
2
1
0
±o
¦ no GPF
¦ GPF
99%
reduction
96%
96%
-7°C FTP
25°C FTP
US06
Figure 3-27: PM emissions from a MY 2021 F150 HEY, with and without a MY 2022 GPF.
The third test vehicle is a MY 2022 F250 7.3L that was retrofit with two MY 2022 GPFs, one
for each engine bank. PM emissions are shown in Figure 3-28. The MY 2022 GPFs reduced PM
emissions by 98 percent, 78 percent, and 98 percent in the -7°C FTP, 25°C FTP, and US06
cycles, respectively. The GPF was fully regenerated immediately before each day of testing
using a sawtooth GPF regeneration cycle. The GPF results shown are worst case with respect to
PM filtration because testing was preceded by a full GPF regeneration, so the GPF was tested
with almost no soot.
Filtration efficiency of the MY 2022 GPFs on the MY 2022 F250 was nearly identical to the
filtration efficiency of the MY 2022 GPF on the MY 2021 F150 HEV in the -7°C FTP and US06
cycles. Filtration efficiency in the 25°C FTP test was higher on the MY 2021 F150 HEV than on
the MY 2022 F250, but the extremely low GPF-equipped levels of PM, around 0.04 to 0.06
mg/mi makes precise PM mass measurements more challenging.
Tailpipe PM was significantly lower with the MY 2022 GPFs as compared to the MY 2019
GPF, especially in the US06 cycle where passive GPF regeneration occurs. Additional details of
the vehicle and GPFs are provided in Chapter 3.2.6.2.1.
3-76
-------
ao
Q_
Figure 3-28: PM emissions from a MY 2022 F250, with and without MY 2022 GPFs.
3.2.6.2.4 Laboratory Round Robin Reproducibility
Two test vehicles were tested by all three organizations in a round robin test program. The
MY 2021 Ford F150 HEV was tested with a MY 2022 GPF and the MY 2021 Toyota Corolla
2.0L was tested stock without a GPF. Vehicle and GPF details are described in Chapter 3.2.6.2.1.
Laboratory-specific PM measurements are shown for the 2021 F150 HEV with GPF in Figure
3-29 and for the 2021 Corolla without GPF in Figure 3-30, for three test cycles (-7°C FTP, 25°C
FTP, US06).
The 2021 F150 HEV with GPF easily meets the 0.5 mg/mi standard across all three test cycles
at all three organizations with a significant compliance margin, even though the results are not
background corrected although that is allowed by the CFR, and GPF tests were performed with
little or no soot on the GPF by running a sawtooth prep cycle as described in Chapter 3.2.6.2.1.
Test-to-test and lab-to-lab variability is evident, but the PM averages and variations are
sufficiently low relative to the final 0.5 mg/mi standard to allow robust compliance of the 2021
F150 HEV with 2022 GPF in any of the labs and across all three test cycles.
The 2021 Corolla without GPF meets the 0.5 mg/mi standard in most but not all of the 25°C
FTP and US06 testing and fails the standard in the -7°C FTP in every test and every lab. The
PFDI engine achieves low PM in most of the two warm tests but the cold temperature of the -
7°C FTP results in dramatically elevated PM. The next section (Chapter 3.2.6.3) discusses
elevated PM from the -7°C FTP test.
8
7
6
5
4
3
2
1
0
I no GPF
I GPF
±a
98%
reduction
78%
-7°C FTP
25°C FTP
US06
3-77
-------
4.0
3.5
3.0
2.5
1
2.0
bfl
E,
1.5
1.0
Q.
0.5
0.0
-7°C FTP
25°C FTP
US06
PM standard
^ m-
— ___
^ ^
u
_£2
_ro
<
CL
u
u
u
J3
_ro
<
CL
u
u
u
LD
J3
_ro
<
CL
u
u
u
Figure 3-29: Test-to-test and lab-to-lab variability for 2021 F150 HEV with 2022 GPF.
4.0
3.5
3.0
2.5
1" 2.0
w> 1 c
& 1-5
S 1.0
CL
0.5
0.0
-7°C FTP
25°C FTP
US06
PM standard
r-
u u >
_Q U LU
m U U-
LT) X
-Q _Q
_fD _nj
< <
CL Q_
u >
U LU
U LJ-
LT) X
-Q _Q
_fD _nj
< <
CL Q_
u >
U LU
U LJ-
Figure 3-30: Test-to-test and lab-to-lab variability for 2021 Corolla.
3.2.6.3 Importance of Test Cycles
The -7°C FTP test is crucial to the PM standard because it addresses uncontrolled cold PM
emissions in Tier 3 and -7°C is an important real-world temperature common through much of
the United States during winter months. Also, absent the -7°C FTP test, vehicles would not
3-78
-------
achieve PM reductions commensurate with what GPF technology offers across a wide range of
operating conditions. Without the -7°C FTP test cycle, vehicles would not have low PM under all
operating conditions.
Based on EPA testing, PM emissions in the -7°C FTP are significantly higher than those
demonstrated during a 25°C FTP test (e.g., Figure 3-26, Figure 3-27, Figure 3-28, and the first
two figures in Section III.D.3 of the preamble). In addition to controlling cold weather PM
emissions that were uncontrolled in Tier 3, the -7°C FTP test differentiates Tier 3 levels of PM
from GPF-level PM.
PM is elevated in the -7°C FTP test because heavy species in gasoline have very low vapor
pressure at cold temperatures, making them difficult to vaporize on cold engine surfaces. For
example, as shown in Figure 3-31, the vapor pressure of toluene and n-decane, two
representative heavy species of gasoline, are reduced by 6.5X and 12X, respectively, as
temperature decreases from 25° to -7°C. Early examples of peer-reviewed literature showing
cold ambient temperature (including -7°C) increases PM mass and solid PN from a GPF-
equipped vehicle include (T. W. Chan 2013) and (T. W. Chan 2014).
40
03
_Q
35
£
CJ
30
to
25
to
OJ
CL
20
o
Q.
15
03
>
CJ
10
c
OJ
5
o
H
0
toluene
6.5X reduction
25
-7 C
n-decane
1 9Y rarli iM-inn
25 -> -
¦7°C
-10
4.0
3.5
3.0
2.5
2.0
1.5
1.0
0.0
0 5 10
Temperature (°C)
15
20
25
03
-Q
-------
In Tier 3, most Class 2b vehicles used the US06 cycle in the HD-SFTP, while low power to
weight Class 2b vehicles and Class 3 vehicles used the LA92 cycle in the HD-SFTP. The final
PM standard requires all light-duty vehicles and MDV to certify using the same cycles: -7°C
FTP, 25°C FTP, and US06. Requiring the US06 for all Class 2b and 3 vehicles ensures that some
GPF regeneration occurs during the test cycle and ensures high GPF filtration under all operating
conditions, even during and after a GPF regeneration. Without the US06 test, GPF regeneration
may not occur during certification test cycles, allowing high PM emissions during high load
operation such as trailer towing. If a Class 2b/3 vehicle is unable to follow the US06 trace, it
must be driven at maximum effort.
3.2.6.4 GPF Cost
An updated GPF cost model was developed for the FRM to estimate direct manufacturing
cost (DMC) of a bare GPF and associated hardware in the exhaust system of a gasoline-powered
light-duty vehicle or MDV where the GPF is installed downstream of the TWCs in its own
aftertreatment enclosure (can). The FRM GPF cost model has been incorporated in the OMEGA
model.
A bare GPF installed downstream of the TWCs may have higher DMC than a catalyzed GPF
that replaces a TWC because the bare downstream GPF requires an additional substrate,
substrate matting, and can. However, some or all of the additional DMC of a bare downstream
GPF may be offset by enabling a reduction in total precious metal content because the precious
metal content can all be used on the lower heat capacity walls of the TWCs that warm up faster
after an engine start. Overall, it is believed that the FRM GPF cost model estimates a DMC that
is either slightly higher or similar to the DMC of a catalyzed GPF that replaces a TWC.
Indirect costs (IC), including R&D and markup, are separately calculated by OMEGA.
OMEGA estimates the IC of a bare downstream GPF in the same way as it does for other
emissions control components, so these ICs are not included in the FRM GPF DMC model
discussed below.
The FRM GPF DMC model is based on an ICCT GPF cost analysis for a bare "stand-alone"
GPF (Minjares and Sanchez 2011) and includes several updates and changes. The DMC model
considers costs for the GPF substrate, housing, accessories, pressure sensor, labor and overhead,
machinery, and warranty. The GPF DMC model uses a GPF swept volume ratio (GPF volume to
engine volume) of 0.80 as compared to a 0.55 swept volume ratio used in the NPRM GPF DMC
model and in the 2011 ICCT GPF cost analysis. The larger swept volume ratio is based on an
updated EPA GPF/vehicle database, input from a GPF supplier, and an ICCT PM/GPF fact sheet
(Isenstadt 2023). The swept volume ratio of 0.80 is representative and falls between that of a
European 2019 Ford Mustang (0.46) and a European 2023 VW T-Roc (0.93).
Substrate and housing costs scale with GPF volume. The substrate cost in the 2011 ICCT
analysis is reduced by 30 percent (from 30 $/literGPF to 21 $/literGPF) based on information
from substrate suppliers and reflects manufacturing learning. Accessories, pressure sensor, labor
and 40 percent overhead, and machinery costs are a fixed dollar amount per vehicle ($39.58 in
2011 dollars). Warranty costs are 3 percent of all the above-mentioned costs. A production
volume discount of 20 percent is then applied, and finally, total cost is converted from 2011 to
2022 dollars (multiplier of 1.296).
3-80
-------
Results from the FRM GPF DMC model are shown in Figure 3-32 for engines ranging from
1.0 to 7.0 using GPFs with swept volume ratios of 0.80 (representative ratio used in FRM
OMEGA analysis), 0.46 (a low swept volume ratio used by the 2019 Mustang), and 0.93 (a high
swept volume used by the 2023 T-Roc).
225
200
175
150
125
100
75
50
25
0
w
(/>
o
u
\J 39^
.S0^se
6 \\ke 2019
Mustang
' '
1.0 2.0 3.0 4.0 5.0 6.0
engine displacement (L)
Figure 3-32: GPF direct manufacturing cost estimates.
7.0
3.2.6.5 GPF Impact on CO2 Emissions
Integrating GPF technology into vehicle aftertreatment systems has the potential to increase
CO2 emissions (energy use) in two ways: during active GPF regeneration, and from increased
backpressure. Active regeneration can increase CO2 emissions while the engine burns more fuel
to add heat to the exhaust gas. However, based on discussions with vehicle manufacturers and
GPF suppliers, and supported by testing conducted by EPA, most production vehicles will rarely
or never need to use active GPF regeneration because systems with close-coupled GPFs or
underfloor GPFs with insulated exhaust pipes (e.g., double wall) naturally cause sufficiently high
GPF temperature for passive GPF regeneration. Increased CO2 due to active regeneration is
therefore considered negligible in this analysis. The following discusses the effect of GPF
backpressure on CO2 emissions.
GPF pressure drop (i.e., backpressure) and CO2 increase were measured on four test vehicles
across three test cycles (-7°C FTP, 25 °C FTP, US06) presents a summary of key vehicle and
GPF specifications. Additional vehicle details are provided in Chapter 3.2.6.2.1.
3-81
-------
Table 3-22: Vehicle and GPF specifications.
MY2022
F250 7.3L
MY2021 F150 3.5L
Powerboost HEV
MY2019
F150 5.0L
MY2011
3.5L Ecoboost
GPF
model year
2022
2022
2019
2019
GPF type and
location
bare
underfloor
bare
underfloor
catalyzed
close-coupled
catalyzed
underfloor
GPF size
(L)
6.42 (total for two)
3.21
2.30 (total for two)
1.65
GPF volume / engine
displacement (-)
0.88
0.92
0.46
0.47
GPF volume / ave
US06 power (L/kW)
0.199
0.115
0.107
0.065
GPF c|) x L
(in)
6.443 x 6 (each)
6.443 x 6
5.2 x 3.3 (each)
5.66x4
GPF cell density
(cpsi)
200
200
300
300
GPF wall thickness
(mil)
8
8
12
12
Average GPF pressure drop for each test cycle and vehicle is shown in Figure 3-33. Average
GPF pressure drop is highest in the US06 because this cycle demands the highest average power
and has the highest average exhaust flow rate. Average GPF pressure drop is similar for the -7°C
FTP and 25°C FTP because these cycles use the same drive trace. The 2011 F150 showed
slightly higher GPF pressure drop in the -7°C FTP as compared to the 25°C FTP, presumably
because powertrain friction increases at cold temperatures before the powertrain warms up.
Figure 3-33 does not show GPF pressure drop for the 2019 F150 because the GPF differential
pressure sensor was installed on this vehicle after -7°C FTP testing was conducted.
ro
Q_
Q_
<
Ll_
Q_
0)
>
2022 F250 7.3L
2021 F150 HEV3.5L
2019 F150 5.0L
2011 F150 3.5L
0
l.i.
-7°C FTP 25°C FTP US06
Figure 3-33: Cycle-average GPF pressure drop as a function of test cycle.
3-82
-------
Figure 3-34 shows GPF pressure drop in the US06 decreases asymptotically as the ratio of
GPF volume to average US06 power increases. Larger GPF volume provides more GPF wall
area for exhaust flow, and lower average US06 power results in reduced exhaust flow volume
(due to reduced exhaust mass flow and lower exhaust temperature).
The results shown in Figure 3-33 and GPF volume to average US06 power tabulated in Table
3-22 demonstrate how for each test cycle (-7°C FTP, 25°C FTP, US06), average GPF pressure
drop increases with decreasing ratio of GPF volume to average power in the US06. Based on a
review of several European GPF-equipped production vehicles and discussions with GPF
suppliers, GPF volumes of the 2022 F250 and the 2021 F150 HEV are within typical production
ranges for these vehicles, while GPF volumes of the 2019 F150 and 2011 F150 are relatively
small for these vehicles, despite the GPF system on the 2019 F150 test vehicle coming from a
European 2019 series production Mustang that uses the same engine displacement as the 2019
F150 test vehicle.
ro
Q_
^4
to
o
I/) „
Q- 9
< Z
Ll_
Q_
<3 1
0)
>
0
0.00 0.04 0.08 0.12 0.16 0.20 0.24
ratio of GPF size to US06 power (L/kW)
Figure 3-34: Cycle-average GPF pressure drop as a function of the ratio of GPF size to
average power required to drive US06 cycle.
Higher GPF pressure drop increases the work that an engine must do to expel exhaust gas
through the exhaust system. To maintain commanded power to follow a drive trace, the throttle
is opened more, and this reduces intake pumping loss, partially offsetting the increased exhaust
pumping work. The net effect is expected to be a slight reduction in brake thermal efficiency and
a slight increase in CO2 emissions. A more detailed discussion of intake pumping loss partially
offsetting exhaust pumping work can be found in (Bohac and Ludlam, Characterization of a
Lightly Loaded Underfloor Catalyzed Gasoline Particulate Filter in a Turbocharged Light Duty
Truck 2023).
Table 3-23 shows the change in measured CO2 emissions for each test cycle when GPFs were
added, when results are averaged across the four test vehicles. Averaging across four test
3-83
-------
vehicles results in CO2 increases between 0.0 percent for the 25°C FTP and 0.9 percent for the
US06. Since two of the test vehicles were equipped with relatively undersized GPFs, these
average CO2 increases are likely higher than on productions vehicles with typical GPF volumes.
Figure 3-35 breaks down the increase in CO2 emissions that was averaged across the four test
vehicles, on a vehicle and test cycle basis. Doing so results in more data scatter, resulting in only
the two light blue bars indicated being statistically significant to 95% confidence.
Table 3-23: Change in measured CO2 emissions for each test cycle when GPFs are added,
averaged across four test vehicles (2022 F250, 2021 F150 HEV, 2019 F150, 2011 F150).
Test Cycle
CO2 Increase (%)
-7°C FTP
0.6
25°C FTP
0.0
US06
0.9
no
3
Li-
fe 2
E
o 1
p<0.05
o
(/>
ro
a
u
c
O
u
0
-1
-2
-3
I
¦
/ ^
¦ 1 ¦
¦
-7°C FTP 25°C FT
p
US06
1
1
¦ 2022 F250 7.3L
1 2021 F150 HEV
¦ 2019 F150 5.0L
¦ 2011 F150 3.5L
1
Figure 3-35: CO2 increase caused by added GPF. Only the two light blue bars indicated are
statistically significant to 95% confidence (p<0.05).
Considering the analyses summarized in Table 3-23 and Figure 3-35, it is estimated that
integrating GPFs into vehicle aftertreatment systems causes less than 1 percent increase in CO2
emissions in the -7°C FTP, 25°C FTP and US06 cycles.
3.2.7 Refueling Standards for Incomplete Spark-Ignition Vehicles
The agency is adopting a requirement for incomplete medium-duty vehicles to meet the same
on-board refueling vapor recovery (ORVR) standards that currently apply for complete vehicles.
Incomplete vehicles have not been required to comply with the ORVR requirements because of
the potential complexity of their fuel systems, primarily the filler neck and fuel tank. Unlike
complete vehicles, which have permanent fuel system designs that are fully integrated into the
3-84
-------
vehicle structure at time of original construction by manufacturers, it was believed that
incomplete vehicles may need to change or modify some fuel system components during final
assembly by secondary manufacturers. For this reason, it was determined that ORVR might
introduce a complexity for the upfitters that is unnecessarily burdensome.
In observations by the agency of current ORVR-equipped vehicles and their incomplete
versions, the agency believes that the fuel system designs are almost identical with only the
ORVR components removed for the incomplete version. The complete and incomplete vehicles
appear to share the same fuel tanks, lines, and filler tubes. Extensive differences between the
original manufacturer's designs and the upfitter modifications to the fuel system, while expected,
have not been observed. This supports the conclusion that almost all incomplete vehicles can
comply with the same ORVR standards as complete vehicles, with the addition of the same
ORVR components that are already installed on counterpart complete vehicles.
In current practice, manufacturers of the original incomplete vehicles identify certain
modifications of the fuel system that are not allowed by the upfitter. This is because the
incomplete vehicle manufacturer is responsible for all current evaporative requirements (2-day,
3-day, running loss, etc.) and almost any modification to the fuel system could compromise
compliance with those program standards. There is also an aspect of compliance with crash and
safety requirements that prevent upfitters from making changes to the fuel system components.
For these reasons, with rare exception, the fuel system design and installation are completed by
the original vehicle manufacturer. The exception that the agency observed is that some
incomplete vehicles do not have the filler tube permanently mounted to a body structure until the
upfitter adds the finishing body hardware (i.e.; flatbed, box). In these cases, the upfitter is limited
to only attaching the filler tube to the added structure, but they must maintain the original
manufacturer design specifications that are part of the certified configuration for meeting EPA
evaporative emission standards.
3.2.7.1 Technologies to Address Evaporative and Refueling Emissions
As exhaust emissions from gasoline engines continue to decrease, evaporative emissions
become an increasingly significant contribution to overall HC emissions from gasoline-fueled
vehicles. Opportunity exists to extend the usage of the refueling evaporative emission control
technologies already implemented in complete medium-duty gasoline vehicles to the counterpart
incomplete gasoline vehicles. The primary technology we are considering is the addition of
ORVR, which was first introduced to the chassis-certified light-duty and medium-duty
applications beginning in MY 2000. See 65 FR 6698 (Feb. 10, 2000). An ORVR system includes
a carbon canister, which is an effective technology designed to capture HC emissions during
refueling events when liquid gasoline displaces HC vapors present in the vehicle's fuel tank
during refueling. Instead of releasing the HC vapors into the ambient air, ORVR systems recover
these HC vapors and store them for later use as fuel to operate the engine.
The fuel systems on incomplete medium-duty gasoline vehicles are very similar to those on
complete medium-duty vehicles that are already required to incorporate ORVR. Fuel tanks on
these incomplete vehicles are almost always identical to the fuel tanks on counterpart complete
vehicles. There may be occasional optional larger fuel tanks requiring a greater ORVR system
storage capacity, and possibly some unique accommodations for dual tanks (e.g., separate fuel
filler locations), but we expect they will maintain a similar design. Figure 3-36 presents a
schematic of a standard ORVR system.
3-85
-------
Carbon
Canister
Filler Pipe
Multi-Function
Control Valves
Fill Limit
Vent Valve
Figure 3-36: Schematic of an ORVR system28.
3.2.7.2 Filler Pipe and Seal
In an ORVR system, the design of the filler pipe connecting the nozzle entry to the fuel tank
is integral to how fuel vapors displaced during a fuel fill will be handled. The filler pipe is
typically sized to handle the maximum allowable fill rate of liquid fuel while also integrating
either a mechanical or liquid seal to prevent fuel vapors from exiting through the filler pipe to the
atmosphere. A dual fuel tank chassis configuration may require a separate filler pipe and seal for
each fuel tank.
The mechanical seal is typically located at the top of the filler neck where the fuel nozzle is
inserted. The hardware piece forms a seal against the fuel nozzle by using some form of a
flexible material, usually plastic, that makes direct contact with the fuel nozzle to prevent fuel
vapors from exiting the filler pipe as liquid fuel is pumped into the fuel tank. In the case of
capless systems, this seal may be integrated into the spring-loaded seal door that opens when the
nozzle is inserted. There are concerns with a mechanical seal's durability due to wear over time,
and its ability to maintain a proper seal with unknown nozzle integrity and variations outside of
design tolerances.
Liquid seals depend on the shape of the filler pipe to be continuously full across the full
diameter of the filler pipe, either inside the fuel tank or close to the fuel tank entry. The liquid
28 Stant ORVR System http://stant.com/orvr/orvr-systems/
3-86
-------
seal prevents fuel vapors from escaping up and out through the filler pipe. By creating a column
of liquid fuel in the filler pipe, the liquid seal does not require a mechanical contact point with
the fill nozzle to prevent escape of vapors. The liquid seal has been the predominant sealing
method implemented in the regulated fleet in response to the ORVR requirements.
3.2.7.3 ORVR Flow Control Valve
As described above, the sealing of the filler pipe prevents the fuel vapors from escaping into
the ambient air; however, the fuel vapors that are displaced by the incoming liquid fuel need to
be routed to the canister. To properly manage the large volume of vapors during refueling, most
ORVR systems have implemented a flow control valve that senses when the fuel tank is getting
filled with fuel and triggers a low-restriction flow path to the canister. This flow path is
specifically used only during the refueling operation and is unique in that it provides the ability
to quickly move larger volumes of fuel vapors into the tank than is normally required when not
refueling. The flow control valve will allow this larger flow volume while refueling, but then
return to a more restrictive vapor flow path while driving, while parked for overnight diurnals,
and at other times.
The flow control valve is generally a fully mechanical valve system that utilizes connections
to the fuel tank and filler pipe to open and close vapor pathways with check valves and check
balls and pressure switches via diaphragms. The valve may be integrated into the fuel tank and
incorporate other aspects of the fuel handling system ("multi-function control valve" in Figure
3-36) including roll-over valve, fuel and vapor separators to prevent liquid fuel from reaching the
canister, and other fuel tank vapor control hardware. Depending on the design, the filler pipe
may also be integrated with the flow control valve to provide the necessary pressure signals. A
dual fuel tank chassis configuration may require a separate flow control valve for each fuel tank.
3.2.7.4 Canister
The proven technology to capture and store fuel vapors is activated carbon. This technology
has been used in vehicles for over 50 years to reduce evaporative emissions from sources such as
fuel tanks and carburetors. When ORVR was originally discussed, existing activated carbon
technology was determined to be appropriate to capture and store refueling related fuel vapors.
This continues to be the case today, as all known ORVR-equipped vehicles utilize some type of
activated carbon.
The activated carbon is contained in a canister made from a durable material that can
withstand the fuel vapor pressures, vibration, and other durability concerns. For vehicles without
ORVR systems, canisters are sized to handle evaporative emissions for the three-day diurnal test
with the canister volume based on the fuel tank capacity. A dual fuel tank chassis configuration
may require a separate canister for each fuel tank.
3.2.7.5 Purge Valve
The purge valve is the electro-mechanical device used to remove fuel vapors from the fuel
tank and canister by routing the vapors to the running engine where they are burned in the
combustion chamber. This process displaces some amount of the liquid fuel required from the
fuel tank to operate the engine and results in a small fuel savings. The purge valve is controlled
by the engine's emission control electronics with the goal of removing the necessary amount of
captured fuel vapors from the canister to prepare it for subsequent fuel vapor handling needs of
3-87
-------
either the next refueling event or vapors generated from a diurnal event. All on-road vehicles
equipped with a canister for evaporative emission control utilize a purge valve. Depending on the
design, a dual fuel tank chassis configuration may require a separate purge valve for each fuel
tank.
3.2.7.6 Design considerations for Unique Fuel Tanks
Commercial gasoline trucks may incorporate several fuel tank options that require unique
ORVR design considerations. While most commercial vehicle fuel tanks are similar to the
already ORVR-compliant complete medium-duty vehicles, some commercial vehicles include
larger tank sizes (up to 50 gallons) or may have a dual tank option. As described above, the
canister sizing will be a function of the required amount of fuel vapor handling during refueling.
Larger fuel tanks will require larger canisters with more activated carbon than historically found
in other gasoline vehicles. Some design challenges will likely exist in designing the canister
system to handle the large vapor volumes while balancing the restriction to flow through the
larger canisters.
Dual fuel tank systems, which have very limited availability, may also require some unique
design considerations. Typically, the canister is located close to the fuel tank to properly and
efficiently manage the refueling fuel vapors. Dual fuel tanks may require duplicate ORVR
systems to have the necessary flexibility to manage the refueling vapors, particularly since the
fuel tanks are filled independently through separate filler pipe assemblies.
3.2.7.7 Onboard Refueling Vapor Recovery Anticipated Costs
Incomplete medium-duty vehicles are not currently required to meet ORVR. There are four
main equipment components and strategies incomplete medium-duty vehicles need to update to
implement ORVR: Increased working capacity of the carbon canister to handle additional vapors
volumes during refueling, flow control valves to manage vapor flow pathway during refueling,
filler pipe and seal to prevent vapors from escaping, and the purge system and management of
the additional stored fuel vapors. The associated direct manufacturing costs for these updates are
summarized below. No labor cost was identified so the direct manufacturing cost is equal to the
piece cost plus tooling cost (per piece). ORVR requirements apply starting in model year 2030
for incomplete medium-duty gasoline vehicles. For our cost analysis, we assumed all gasoline
medium-duty vehicles identified as incomplete light heavy-duty trucks in MOVES will have an
average fuel tank capacity of 35 gallons.
Capturing the increased vapor volume from the vapor displaced during a refueling event will
require canisters to increase working vapor capacity approximately 15 to 40 percent, depending
on the fuel tank size and other specifications for individual vehicle systems. Manufacturers can
achieve greater working capacity by increasing the canister volume using conventional carbon. A
typical evaporative canister has approximately 2.6 liters of conventional carbon to capture
overnight diurnal evaporative emissions for a 35-gallon fuel tank. An increase in required
capacity to allow for capturing refueling vapors results in the need for an additional 1 liter of
conventional carbon. A change in canister volume to accommodate additional carbon includes
increased costs for retooling and additional canister plastic material, as well as design
considerations to fit the larger canister on the vehicle. However, because these medium-duty
vehicles almost always have a complete version already required to comply with refueling
3-88
-------
standards, the necessary larger canister sizes are already produced and available, which likely
avoids the need for new tooling.
An alternative to retooling for a larger single canister would be to add a second canister for
the extra canister volume. Several smaller-volume canisters are available on the market today.
Another approach, based on discussions with canister and carbon manufacturers, can be achieved
by using a higher adsorption carbon along with modifications to compartmentalization within the
existing canister plastic shell that will increase the canister working capacity without requiring a
larger canister size.
There are also two primary technologies used to prevent vapors from escaping into the
atmosphere through the filler neck and around the fuel nozzle area when the vehicle is refueling
that can affect the canister vapor capacity design requirements: A mechanical seal that makes
direct physical contact with the refueling nozzle to create a nozzle-to-filler neck seal; or a liquid
seal further down in the filler pipe that uses the liquid fuel mass flowing down the filler pipe and
entering the tank to hydraulically prevent vapors from migrating back up the fill pipe. There is
approximately a 20 percent reduction in activated carbon required for a mechanical seal. While
mechanical seals are not currently the preferred technology, manufacturers facing the design
options for accommodating larger fuel volumes and the need for a larger matching evaporative
canister may opt for a mechanical seal design. We share our assumptions and cost estimates for
both seal options in Table 3-24 and Table 3-25. A dual tank may require two seals if dual filler
necks are used instead of a single filler neck and transfer pump to move fuel between the two
tanks.
The second required equipment update would be to install flow control valves, which may be
integrated into existing roll-over/vapor lines. The flow control valves are needed to manage the
vapors during the refueling event by providing a low restriction pathway for vapors to enter the
canister for adsorption and storage on the carbon materials. We anticipate that vehicles would
require on average one valve per vehicle, which would be approximately $6.50 per valve. A dual
tank system may require a flow control valve system per tank depending on the design approach.
Thirdly, as mentioned above, a filler pipe and seal system would be needed for each filler
neck to keep the vapors contained during refueling. Manufacturers have the option of a
mechanical seal that costs approximately $10.00 per seal and a liquid seal. The liquid seal is a
design feature of the filler neck, which means that it has no direct cost, but it may require
hardware modifications to provide enough back pressure to stop the fuel flow when the tank is
full. In some cases, incomplete vehicles share the same filler tube design with the refueling
requirements compliant complete version, in which case there would again be no cost for
upgrading the system to manage automatic shutoff
Lastly, manufacturers may need to address the engine control of the canister purge rates. This
update would include calibration improvements and potentially additional hardware to ensure
adequate purge volumes to maintain an appropriate canister state to prepare for further vapor
loading from diurnal and subsequent refueling events. However, if the incomplete version shares
engines and fuel systems with the complete vehicle, the development of calibrations for the
required purge volumes has likely already been completed, eliminating any need for further
changes or development work. If required for a dual tank system, an extra purge valve may be
needed if the two-tank system maintains independent canisters instead of a single common
canister as observed in dual-tank, single canister light-duty applications. Table 3-24 shows our
3-89
-------
calculations estimating the amount of extra canister size for conventional carbon for a 35-gallon
tank, using Tier 3 core evaporative requirements (i.e., 2-day and 3-day SHED) as a baseline.
Currently under Tier 3 requirements the canister and purge strategy are sized for the 3-day
diurnal test and designed to meet the Bleed Emission Test Procedure (BETP) requirements.
During the diurnal test, the canister is loaded with hydrocarbons over two or three days, allowing
the hydrocarbons to load a conventional carbon canister (1500 GWC, gasoline working capacity)
at a 70 g/L effectiveness. During a refueling event, which takes place over a few minutes, the
vapor from the gas tank is quickly loaded onto the carbon in the canister with an ORVR system,
causing the efficiency of the canister loading to drop to 50 g/L effectiveness. Typically, a design
safety margin includes an additional 10 percent carbon to ensure adequate performance over the
life of the system. Therefore, even though there is typically less fuel vapor mass generated and
managed during a refueling event than is generated over a three-day diurnal, the amount of
carbon that is necessary to contain the vapor is higher for a refueling event.
Table 3-24: ORVR Specifications and Assumptions used in the Cost Analysis for
Incomplete Medium-Duty Vehicles.
Tier 3
Baseline
Diurnal
Diurnal Heal Build 72-%°F
RVP
Nominal Tank Volume
Fill Volume 40%
Air Ingestion Rale
Mass Vented per heal build, g/d 60
Mass Vented per refueling event
Hot Soak Vapor Load 2.5
Mass vented over 48-hour test 114
Mass vented over 72-hour test 162
1500 GWC. g/L" 70
Excess Capacity 10%
Estimated Canister Volume Requirement, liters'1
48-hour Evaporative only 1.8
72-hour Evaporative only 2.5
Total of 72-hour + ORVR0
a Efficiency of conventional carbon.
b Canister Volume = l.l(mass vented)/ 1500 GWC (Efficiency),
c ORVR adds .3 liters and 1 liter for Mechanical Seal and Liquid Seal respectively.
ORVR Filler Neck Options
Mechanical Seal Liquid Seal
ORVR
80°F
9 psi
35 gallons
10% to 100%
o%
128
50
10%
2.8
13.50%
158
50
10%
3.5
3-90
-------
Table 3-25: Estimated Direct Manufacturing Costs for ORVR Over Tier 3 as Baseline.
Liquid Seal Mechanical Seal
New Canister New Canister
Additional Canister $10 $4
Costs
Additional Tooling" $0.50 $0.50
Flow Control Valves $6.50 $6.50
Seal $0 $10
Total" $17 $21
a Assumes the retooling costs will be spread over a five-year period.
b Possible additional hardware for spitback requirements. Note that these manufacturing costs do not include a markup representative of
Retail Price Equivalent values.
3.3 On-board Diagnostics
EPA regulations state that onboard diagnostics (OBD) systems must generally detect
malfunctions in the emission control system, store trouble codes corresponding to detected
malfunctions, and alert operators appropriately. EPA adopted (as a requirement for an EPA
certificate) the 2013 California Air Resources Board (CARB) OBD regulation, with certain
additional provisions, clarifications, and exceptions, in the Tier 3 Motor Vehicle Emission and
Fuel Standards final rulemaking. 40 CFR 86.1806-17; 79 FR 23414 (Apr. 28, 2014). Since that
time, CARB has made several updates to their OBD regulations and continues to consider
changes periodically. In this rule, EPA is updating to the latest version of the CARB OBD
regulation; California's 2022 OBD-II requirements are part of (Title 13 § 1968.2 California Code
of Regulations 2022). This is accomplished by adding a new 40 CFR 86.1806-27 for vehicles
built after 2027 model year and only adding requirements to that section that are not in the new
CARB regulation.
3.4 PHEV Accounting
3.4.1 Final Approach for the Revised PHEV Utility Factor
EPA is finalizing its proposed change to the light-duty vehicle PHEV Fleet Utility Factor
(FUF) curve used in CO2 compliance calculations for PHEVs. To address concerns about
adequacy of lead time for the early years of the program, we are delaying the application of the
revised FUF until MY 2031, as further discussed below in this section.
A fleet utility factor provides a means of accounting for a PHEV's operation using electricity,
known as the Charge Depleting Mode with respect to the total mileage that a PHEV travels. The
Charge Depleting mode is dependent on two significant factors. The first factor is the size or
capacity of the battery. Typically, a PHEV with a larger battery will have more all-electric range,
all other vehicle attributes being equal. The second factor is an owner's propensity to charge the
battery. SAE J2841 states explicitly that the UF represented in SAE standard assumes that a
PHEV is fully charged at least once per day. Recent literature (Aaron Isenstadt, Zifei Yang,
Stephanie Searle, John German 2022), (Seshadri Srinivasa Raghavan & Gil Tal 2022) and data
(California Air Resource Board [OBD data records] 2023, California Air Resource Board [OBD
data records] 2022) have identified that the current utility factor curves overestimate the fraction
of driving that occurs in charge depleting operation. Other literature (Aaron Isenstadt, Zifei
3-91
-------
Yang, Stephanie Searle, John German 2022), (Seshadri Srinivasa Raghavan & Gil Tal 2022) also
concludes that vehicles with lower charge depleting ranges have even greater discrepancy in CO2
emissions. The current SAE J2841 FUF curve and the finalized FUF curve are shown in Figure
3-37. The finalized FUF curve represents a modest change of about 11 percent from SAE J2841
FUF curve while the averages of FUF values by the SAE J2841 FUF curve are approximately 55
percent higher than those by the "BAR Regression Fit" curve between 0 and 180-mile 2-cycle
combined GHG emission-certified CD ranges using November 2023 CARB dataset. In contrast,
the labeled ranges are already reduced by 30 percent from 2-cycle combined GHG emission-
certified CD ranges for CAFE (Corporate Average Fuel Economy) standards compliance and
labeling the all-electric ranges and CD ranges of PHEV vehicle stickers.
CD Ranges [miles], GHG Emission Certified
Figure 3-37: SAE J2841 FUF and finalized FUF for PHEV compliance.
3.4.2 Overview of BAR dataset
The November 2023 CARB BAR OBD PHEV dataset for the finalized FUF curve has
approximately 8,800 PHEV vehicles, from 43 PHEV models with 90 PHEV model variants, and
over 169.4 million vehicle miles traveled. About 79 percent of PHEV vehicles, representing
approximately 140.4 million miles traveled, are from in-state ownership transfers. The remaining
21 percent of PHEV vehicles, representing around 23.1 million miles traveled, are from out-of-
state vehicle registrations. The filtered dataset has 42 PHEV models and 89 model variants, and
approximately 8,600 individual vehicles that travelled approximately 163.2 million miles. The 42
PHEV models and about 8,600 filtered data points are approximately 98 percent of the 43 PHEV
models and about 8,800 vehicles in the unfiltered data.
3-92
-------
3.4.2.1 Descriptions of Data Source and Filtering Method
For the FRM data analysis, we used a data set from BAR (California Air Resource Board
[OBD data records] 2023) that has been updated from the data used for the NPRM. This
November 2023 dataset contains an additional year of PHEV activity on top of the NPRM's
October 2022 dataset. C ARB filtered the updated data to include only PFIEV vehicles, and to
exclude vehicles with less than 3,000 miles of total lifetime distance traveled. Also, vehicles that
CARB identified as having incorrectly logged data in the lifetime distance OBD data fields used
for this analysis were excluded. EPA filtered out a few additional vehicles where data necessary
for determining whether the vehicle was being imported from out of state were missing.
The odometer reading data field was not used for filtering the updated data set. EPA
concluded that a mismatch between the technician-recorded odometer reading and the OBD data
was not indicative of any issue with our use of the OBD lifetime distance traveled data fields for
calculating FUF. Similarly, the 20 percent window filtering between the total grid energy into the
battery pack and total grid energy consumption during CD operation was not used in filtering the
November 2023 CARB dataset. After further inspection of the data, EPA identified underlying
reasons for why a difference of more that 20 percent would not be indicative of OBD data issues.
For example, a number of vehicles in the BAR data set do not have any EV distances at engine
off charge depleting operation, even when total grid energy consumption (from wall-charging) is
reported, due to use cases such as portable electric energy storages and other outdoor activity
power supply.
Total Grid Energy into Battery (kWh) < 200 kWh, # Samples: 2408/8573
0 10000 20000 30000 40000 50000 60000 70000 80000 90000 150000
Total Lifetime Distance Traveled (mile)
Figure 3-38: Lifetime Total Grid Energy into Battery (less than 200-kWh)
Figure 3-38 shows PFtEV vehicles with over 3000 miles travelled that were charged less than
200-kWh total grid energy over their lifetime. The 200-kWh total grid energy charging to the
battery packs can be achieved less than a month when charging a small 8.8-kWh PFLEV battery
pack daily. Of these vehicles, about 28 percent were filtered out when the 20 percent window
3-93
-------
filtering between the total grid energy into the battery pack and total grid energy consumption
during CD operation was applied. Therefore, this filter was not used for calculating various real-
world usages and new applications of the PHEV vehicles and battery packs.
3.4.2.2 Minimum VMT and Sample-Size Sensitivities
To investigate data sampling sensitivities, we used various minimum VMT values for filtering
data using the October 2022 BAR OBD data (California Air Resource Board [OBD data records]
2022). As shown in Figure 3-39, the relative FUFs over the SAE J2841 FUFs are not
significantly different at various minimum VMT filtering
Relative FUF by Models
OOOOOOOOOOOOOOOOI
oooooooooooooooooo
OOOOOOOOOOOOOOOOOO
>000009000000000000
oooooooooooooooooo
o o oc oooooooooooooo
>000000000000000000
oooooooooooooooooo
000000000000000
OOOOOOOOOOOOOOOOOO
>000000000000000000
oooooooooooooooooo
oooooooooooooooOc
OOOOOOOOOOOOOOOOOO
OOOOOOOOOOOOOOOOOO
>0 0 000000000000000
OOOOOOOOOOOOOOOOOO
>00000000000000000
oooooooooooooooooo
ooooooooooooooooo
oooooooooooooooooo
>000000000000000000
oooooooooooooooooo
ooooooooooooooo
oooooooooooooooooo
oooooooooooooooooo
OO OOOOOOOOOOOOOOOOC
OOOOOOOOOOOOOOOOOO
OOOOOOOOOOOOOOOOOOO
oooooooooooooooooo
>0000000090000000
OOOOOOOOOOOOOOOOOOO
>OOOOOOOOOOOOOOOC
OOOOOOOOOOOOOOOOC
OOOOOOOOOOOOOOOOOO
ooooooooooooooooooc
OOOOOOOOOOOOOOOOOOO
00000000000000004
OOOOOOOOOOOOOOOOOOO
OOOOOOOOOOOOOOOOOO
3000000000000000000
oooooooooooooooooo
OOOOOOOOOOOOOOOOOOO
ooooooooooooooooo
ooooooooooooooooo
oooooooooooooooooo
ooooooooooooooooooc
oooooooooooooooooo
oe'fifiaooisoftooaoooaai
oooooooooooooooooc
>000000 0000000000
ooooooooooooooooooc
oooooooooooooooooo
oooooooooooooooooo
ooooooooooooooooooc
oooooooooooooooooo
OOOO&OOOOOOOOOi
OOOOOOOOOOOOOOOOOO
OOOOOOOOOOOOOOOOOOt
O O O O o o c
>0000000000
oooooooooooooooooc
oooooooooooooooooo
oooooooooooooooc
oooooooooooooooooo
>000000000000000001
oooooooooooooooooc
Minimum VMT [km]
Figure 3-39: FUFs with various data filtering sensitivities.
As shown in Figure 3-40, the "Sample-size Weighted Fit" curve using sample sizes greater
than or equal to 2, lies on top of the "Sample-size Weighted Fit" curve using sample sizes greater
than or equal to 10. Both of the "Sample-Weighted Fit" curves lie a little higher than the
"Equally Weighted Fit" curve when using sample sizes greater than or equal to 2. Therefore, the
sample-size weighted fitting (Aaron Isenstadt, Zifei Yang, Stephanie Searle, John German 2022)
with sample sizes greater than or equal to 2 are used for fitting FUF data with 6-coefficient non-
linear regression and a 399.9-mile normalized distance (SAE J2841 2010).
3-94
-------
0.9
0.8
0.7
O
S
£ 0.6
!*
§ 0.5
at
©
t. 0.4
IJ.
D
LL
0.3
0.2
0.1
0
0 20 40 60 80 100 120 140 160 180
CD Ranges [miles], GHG Emission Certified
Figure 3-40: BAR Regression FUF Curve fits with Sample-size Weighted.
3.4.3 Analyses of FUF based on BAR Dataset
While EPA used BAR data from October 2022 (California Air Resource Board [OBD data
records] 2022) for the NPRM, an additional year of data was available to inform this FRM. In
November 2023 OBD datasets (California Air Resource Board [OBD data records] 2023) were
made available for EPA to analyze. EPA found that the expanded data set confirms that, on
average, there are more charge-sustaining miles traveled and more gasoline miles traveled than
are predicted by the current SAE J2841 FUF (Fleet Utility Factor) curves. The BAR OBD data
enables the evaluation of real-world PHEV distances travelled in various operational modes;
these include charge-depleting engine-off distance, charge-depleting engine-on distance, charge-
sustaining engine-on distance, total distance traveled, odometer readings, total fuel consumed,
and total grid energy inputs and outputs of the battery pack. These fields of data allow us to use
the BAR OBD data to filter the data and calculate real-world driving FUFs (ratios of charge
depleting distance to total distance) and to then compare to the existing SAE J2841 FUFs as
calculated and applied in EPA's GHG emissions certification using the 2-cycle charge-depleting
range values.29 Although we have reached a similar conclusion to other studies that have been
conducted to evaluate PHEV utility, the BAR data has allowed EPA to analyze PHEV utility
specifically on distance traveled in each mode as recorded by the vehicle itself. Other studies
(Patrick Plotz 2023) regarding PHEV utility have attempted to calculate distance traveled in each
mode using energy and fuel consumption or the labeled values. Because energy and fuel
29 The existing regulatory FUFs are separate city and highway curves, and the charge depleting ranges that are used
with the city and highway FUF curves are 2-cycle range.
3-95
-------
consumption can vary greatly based on operating and environmental conditions, and distance
calculations can also vary, EPA did not rely on these types of analyses to inform this rulemaking.
3.4.3.1 Basis for EPA's Final Utility Factor
A fleet utility factor provides a means of accounting for a PHEV's operation using electricity,
known as the Charge Depleting Mode with respect to the total mileage that a PHEV travels. The
Charge Depleting mode is dependent on two significant factors. The first is the size or capacity
of the battery. Typically, a PHEV with a larger battery will have more all-electric range, all other
vehicle attributes equal. The second factor is an owner's propensity to charge the battery. SAE
J2841 states explicitly that the UF represented in SAE standard assumes that a PHEV is fully
charged at least once per day. Recent literature (Aaron Isenstadt, Zifei Yang, Stephanie Searle,
John German 2022) and data (California Air Resource Board [OBD data records] 2023) have
identified that the current utility factor curves overestimate the fraction of driving that occurs in
charge depleting operation. This literature (Seshadri Srinivasa Raghavan & Gil Tal 2022) also
concludes that vehicles with lower charge depleting ranges have even greater discrepancy in CO2
emissions.
The results of EPA's data analysis on 2-cycle combined GHG emission-certified CD ranges
are shown in Figure 3-50. The FUF applied in the current GHG regulations is labeled as "SAE
J2841 FUF" and EPA's data analysis of the November 2023 BAR OBD data is labeled as "BAR
Regression Fit".
The finalized FUF curve represents a modest change of about 11 percent from SAE J2841
FUF curve while the averages of the SAE J2841 FUF curve are approximately 55 percent higher
than those calculated using the "BAR Regression Fit" curve between 0 and 180-mile 2-cycle
combined GHG emission-certified CD ranges using November 2023 CARB dataset.
EPA is finalizing a revised light-duty vehicle PHEV Fleet Utility Factor curve for use in the
CO2 compliance calculation for PHEVs, beginning in MY 2031. The agency believes the current
LD vehicle PHEV compliance methodology significantly underestimates PHEV CO2 emissions.
The mechanism that is used to apportion the benefit of a PHEV's electric operation for purposes
of determining the PHEVs contribution towards the fleet average GHG requirements is the fleet
utility factor (FUF), further explained below.
The finalized FUF curve was developed using the best available public real-world PHEV
dataset, which consists of about 8,800 PHEV vehicles, 43 PHEV models and 90 PHEV model
variants in the November 2023 BAR dataset (California Air Resource Board [OBD data records]
2023). The current SAE J2841 FUF curve and the finalized FUF curve are shown in Figure 3-37,
shown above. The grey-dashed FUF curve labeled as "SAE J2841 FUF" in Figure 3-37 was
developed in SAE 2841 (SAE J2841 2010) and are used to estimate the percentage of operation
that is expected to be in charge depleting mode (vehicle operation that occurs while the battery
charge is being depleted, sometimes referred to as electric range). The measurement of the
charge depleting (CD) range is performed over the EPA city and highway test cycles, also called
the 2-cycle tests. The tested cycle specific charge depleting range is used as an input to the FUF
curves (or lookup tables, as shown in Tables 1 and 2 in 40 CFR 600.116-12) to determine the
specific city and highway FUFs. The resulting FUFs are used to calculate a composite CO2 value
for the city and highway CO2 results, by weighting the charge depleting CO2 by the FUF and
weighting the charge sustaining (CS) CO2 by one minus the FUF.
3-96
-------
The FUFs developed in SAE J2841 rely on a few important assumptions and underlying data:
(1) trip data from the 2001 National Household Travel Survey,30 used to establish daily driving
distance assumptions, and (2) the assumption that the vehicle is fully charged before each day's
operation. These assumptions are important because they affect the shape of the utility factor
curves, and therefore affect the weighting of CD (primarily electric operation)31 CO2 and CS
(primarily internal combustion engine operation)32 CO2 in the compliance value calculation. SAE
J2841 was developed more than ten years ago during the early introduction of light-duty PHEVs
and at the time was a reasonable approach for weighting the CD and CS vehicle performance for
a vehicle manufacturer's compliance calculation given the available information. The PHEV
market has since grown, and there is significantly more real-world data available to EPA on
which to design an appropriate compliance program for PHEVs. The agency believes that the use
of an FUF is still an appropriate and reasonable means of calculating the contribution of PHEVs
to GHG emissions and compliance, but the real-world data available today no longer supports
the FUF established in SAE J2841 more than a decade ago.
Because the tailpipe CO2 produced from PHEVs varies significantly between CD and CS
operation, both the charge depleting range and the utility factor curves play an important role in
determining the magnitude of CO2 that is calculated for compliance. In charge depleting mode
EPA is proposing to maintain a zero gram per mile contribution when the internal combustion
engine is not running. The significant difference noted above is the difference between,
potentially, zero grams per mile in CD mode versus CO2 grams per mile that are likely to be
similar to a hybrid (non-plug-in) vehicle. The charge depleting range for a PHEV is determined
by performing single cycle city and highway charge depleting tests according to SAE Standard
J1711 (SAE J1711 2023), Recommended Practice for Measuring the Exhaust Emissions and
Fuel Economy of Hybrid-Electric Vehicles, Including Plug-In Hybrid Vehicles. The charge
depleting range is determined by arithmetically averaging the city and highway range values
weighted 55-percent and 45-percent, respectively, as noted in 40 CFR 600.311-12(j)(4)(i) (40
CFR 600.311-12 2021).
3.4.3.2 FUF Comparisons with Real World Data
Recent literature (Aaron Isenstadt, Zifei Yang, Stephanie Searle, John German 2022),
(Seshadri Srinivasa Raghavan & Gil Tal 2022) and data (California Air Resource Board [OBD
data records] 2023) have identified that the SAE J2841 utility factor curves may overestimate the
fraction of driving that occurs in charge depleting operation. This literature (Seshadri Srinivasa
30 We used the latest NHTS data (2017) and executed the utility factor code that is in SAE J2841, Appendix C, and
found that the latest NHTS data did not significantly change the utility factor curves. NHTS data can be found at
U.S. Department of Transportation, Federal Highway Administration, 2017 National Household Travel Survey.
URL: https://nhts.ornl.gov/
31 The complexity of PHEV designs is such that not all PHEVs operate solely on the electric portion of the
propulsion system even when the battery has energy available. Engine operation during these scenarios may be
required because of such design aspects as blended operation when both the electric power and the engine are being
utilized, or during conditions such as when heat or air conditioning is needed for the cabin and can only be obtained
with engine operation.
32 Because most CD operation occurs without engine operation, the CO2 value for CD operation is often 0 or near 0
g/mi. This means that a high utility factor results in a CO2 compliance value that is heavily weighted with 0 or near
0 g/mi.
3-97
-------
Raghavan & Gil Tal 2022) also concludes that vehicles with lower charge depleting ranges have
even greater discrepancy in CO2 emissions.
EPA (Aaron Isenstadt, Zifei Yang, Stephanie Searle, John German 2022) has evaluated
recently available California Bureau of Automotive Repair (BAR) OBD data, (California Air
Resource Board [OBD data records] 2022) that has been collected through the California Bureau
of Automotive Repair and found that the data shows that, on average, there is more charge
sustaining operation and more gasoline operation than is predicted by the SAE J2841 fleet utility
factor curves. The BAR OBD data enable the evaluation of real-world PHEV distances travelled
in various operational modes; these include charge-depleting engine-off distance, charge-
sustaining engine-on distance, total distance traveled, odometer readings, total fuel consumed,
and total grid energy inputs and outputs of the battery pack. These fields of data allow us to use
the BAR OBD data to filter the data and calculate 2-cycle combined GHG emission-certified CD
ranges and 5-cycle comparable real-world driving ratios of charge depleting distance to total
distance and to then compare to the existing FUFs on 2-cycle combined GHG emission-certified
CD ranges.33
There are some limitations to the PHEV data collected through the BAR OBD data. Data
collection occurs through the California Bureau of Automotive Repair and is limited to vehicles
with ownership changes, vehicles entering the state, or vehicles that are at least 8-years old
(California Air Resource Board [OBD data records] 2023). In addition, the PHEV BAR OBD
data requirements are recent; they began in model year 2019 and were not fully phased in until
model year 2021 (California Air Resource Board [OBD data records] 2023). The dataset also
contains some reporting errors and some very low mileage data.
The results of EPA's data analysis of the BAR OBD data on 2-cycle combined GHG
emission-certified CD ranges are shown below in Figure 3-50. The combined city and highway
FUF in SAE J2841 (corresponding to the 55-percent city/45-percent highway weighing of the
city and highway FUFs) in the current regulations is labeled as "SAE J2841 FUF" and EPA's
data analysis of the CARB OBD data is labeled as "BAR Regression Fit".
3.4.3.2.1 Influence of Geographic Origin
About 79 percent of PHEV vehicles that traveled approximately 140.4 million miles are from
in-state ownership transfers, and around 21 percent of PHEV vehicles that traveled around 23.1
million miles are from out-of-state vehicle registrations. The "BAR Regression Fit" curve is
more influenced by about a factor of six by in-state ownership transfer PHEV vehicles since the
FUF values are calculated by a distance-weighted basis. Additionally, there is no reason to
expect one-time long-distance moving miles to make up more than a small portion of the 23.1
million miles from out-of-state vehicle registration vehicles.
The averaged FUF values calculated by the regression fit curves of the in-state ownership
transfers and out-of-state registrations are within about 2.3 percent and -4.8 percent of the
averaged FUF values represented by the "BAR Regression Fit" curve when using a minimum
33 Because the data collected is real-world data, we used the combined city and highway 5-cycle label range as an
input to the FUF curve described in SAE J2841, to create an apples-to-apples comparison. The existing regulatory
FUFs are separate city and highway curves, and the charge depleting ranges that are used with the city and highway
FUF curves are 2-cycle range.
3-98
-------
sample size of 2. When using larger sample sizes greater than or equal to 10, the averaged FUF
values of in-state ownership transfers and out-of-state PHEV vehicle registrations in the filtered
model data are within 1.5 percent.
As shown in Figure 3-41, the FUF differences between the blue-dashed FUF fit curve of 79
percent In-State ownership transfers and the magenta-dashed FUF fit curve of 21 percent Out-of-
State registrations are not significant. The FUF values by predicted by the "SAE J2841 FUF"
curve are still approximately 51 percent higher than those calculated by the best-case, blue-
dashed regression fit curve of In-State ownership transfers.
Figure 3-41: In-State and Out-of-State FUF Comparisons.
3.4.3.2.2 Influence of Gasoline Price
About 4.7 percent of PHEV vehicles that traveled approximately 5.6 million miles are from a
period of low gasoline prices (the pandemic period from March 1, 2020, to March 30, 2021), and
around 95.3 percent of PHEV vehicles that traveled around 157.6 million miles are from the
normal gasoline price period (mostly pre- and post-pandemic, as shown in Figure 3-42 (U.S. EIA
2023)). The distance weighted "BAR Regression Fit" curve fitting, as shown in Figure 3-43, is
more influenced by about a factor of 28 by the higher total distance travelled when gasoline
prices were normal. Furthermore, the averaged FUF values calculated by the regression fit
curves of the normal gasoline price period and the low gasoline price period are within 0.1
percent and -8.1 percent of the averaged FUF values represented by the "BAR Regression Fit"
curve when using a minimum sample size of 2. When using larger sample sizes greater than or
equal to 10, the averaged FUF values of the normal gasoline prices and the low gasoline prices in
the filtered model data are within 1.6 percent.
3-99
-------
Weekly U.S. All Grades All Formulations Retail Gasoline Prices
Dollars per Gallon
6
A, DOWNLOAD
1994 1996 1998 2000 2002 2004 2MKJ 2008 2010 2012 2014 20T6 2018 2020 2022 2024
— Weekly U.S. All Grades All Formulations Retail Gasoline Pnces
Figure 3-42: U.S. Retail Gasoline Prices.
Figure 3-43: FUF Finalized, SAE FUF, and BAR Regression Fits at low gasoline prices.
As shown in Figure 3-43, the effects of the red-dashed FUF fit curve of 5.6 million miles
traveled at low gasoline prices are minimal compared to the blue-solid FUF fit curve of 157.6-
million miles traveled distance at normal gasoline prices since FUF are calculated by a travelled
distance weighted of fleet vehicles. The FUF values by the grey-dashed "SAE J2841 FUF" Fit
curve are still approximately 55 percent higher than those by the best-case, blue-solid FUF fit
curve at normal gasoline prices.
3-100
-------
3.4.3.2.3 Influence of Aggressive Driving Behaviors
The high-volume sales "Car A" and the "SUV B" PHEV datasets were used to investigate
Grid Energy (GE) Consumptions at various driving behaviors. The Positive Kinetic Energy
(PKE) data field (California Air Resource Board [OBD data records] 2023) was used to
characterize aggressive driving behaviors. The PKE is calculated by using Equation 3-1
(Edwards 2022).
Equation 3-1. PKE = ^ccelerationX^-V;2)
distance
where Vf and Vt are the final and initial vehicle speeds, respectively.
O BAR Data /w Car A
~ BAR Data /w SUV B
— BAR Data Fit /w Car A
— -BAR Data Fit/wSUV B
US06PKE
~
~rf3
- rft
~
ff n °a
tnirf ~
~
? o_5a
c
R]
Positive Kinetic Energy (m/s )
Figure 3-44. Grid Energy Consumptions during CDO vs PKE
PKE values of about 0.14, 0.35 and 0.42 m/s2 were calculated over Highway, UDDS, and US06
dynamometer test cycle vehicle speed profiles. The "Car A" PHEV has 1.5-liter four-cylinder
gasoline engine, a 17.5-kWh lithium-ion battery, an 87-kW electric motor and about 3500
pounds vehicle weight. The "SUV B" PHEV has an 2.0L inline 4-cylinder engine, a 17.3 kWh
lithium-ion battery pack, a 100-kW electric traction motor and about 5,200-pound vehicle
weight. Most of the "Car A" PHEVs were operated at below the label value electric energy
consumption rate (cyan dashed line). As shown in the green-solid regression data fit curve in
Figure 3-44, the grid energy consumptions over the entire PKE range are almost flat. The heavier
"SUV B" PHEV requires higher grid energy consumptions to propel about 5200-pound vehicle,
and the "SUV B" grid energy data points are more scattered. The magenta-colored regression
curve fitted using the "SUV B" PHEV data points has a slightly negative slope; however, the
slope of the "SUV B" PHEV is also nearly flat over the entire PKE range.
3-101
-------
The motor peak power to vehicle weight ratio of the "Car A" PHEV is about 29 percent
higher than that of the much heavier "SUV B" PHEV. Therefore, the "SUV B" PHEV engine is
more frequently started when driving aggressively since the battery capacity to vehicle weight
ratios of the "SUV B" PHEV are about 33 percent less than those of the "Car A" PHEV while the
vehicle weights of the "SUV B" PHEV are about 49 percent heavier than those of the "Car A"
PHEV. Overall, the grid energy consumptions are flat over the entire PKE ranges, and therefore
the driving behaviors are not a major factor for grid energy consumptions and deteriorating the
FUF values compared to battery charging frequencies.
O BAR Data /w Car A
~ BAR Data /w SUV B
— 310 Wh/mile Labeled/w Car A
680 Wh/mile Labeled/w SUV B
500 1000
Grid Energy Consumption (Wh/mile) @ CD Eng OFF
1500
Figure 3-45. FUF vs Grid Energy Consumption at CD Engine Off
Figure 3-45 shows that grid energy of the "Car A" PHEV are mostly consumed at or below
the label value electric energy consumption rate while that of the "SUV B" PHEV are consumed
within reasonable percentages of its label value electric energy consumption rate. The label value
electric energy consumption rates are calculated by dividing the 2-cycle city and highway
combined grid energy consumptions by a 0.7 5-cycle adjustment factor.
3-102
-------
o
o
<0
a
o
D n. *
- oor
0
o
B
o
c
o
o
o
O BAR Data /w Car A
O ~ BAR Data /w SUV B
°°o s
» o°
O
o
^OcfcO n O o o
o . o
O O Q
o o
ftoV
b°%
o 0
®o°
o
p^D°o
i i i
3%S° 0°
»VB°C
0° O °o
o °o ° 0
o
O ° OO
o
O o
0 °
o °
oO O
fjjffri ~
O 0
o
Ey'j o
Ro
£¦¦¦¦
o
0.5
1.5
2.5
Grid Energy into Battery (kWh) x104
Figure 3-46. FUF vs Grid Energy into Battery (kWh)
° ~ o o°o
o
BAR Data /w Car A
~
BAR Data /w SUV B
Highway PKE
UDDS PKE
-
US06 PKE
o8o#><* °o
Positive Kinetic Energy {mis )
Figure 3-47. FUF vs. Positive Kinetic Energy
Figure 3-46 shows that lower FUF values are directly related to the amount of Grid Energy
charging to battery packs as shown in the poorly charged data points at below 0.2 FUF values.
As shown in Figure 3-47, there are a lot of higher BAR data points at about 0.42 m/s2 PKE
line, and therefore higher FUFs at the aggressive driving behaviors can be achieved as long as
fully charging the battery packs daily. Figure 3-47 shows that the driving behaviors are not the
most dominant factor for higher grid energy consumptions and deteriorating the FUF values.
3-103
-------
3.4.3.2.4 Influence of Data Filtering
EPA created an aggressively filtered data set, similar to the ICCT data filtering, by filtering
out vehicles with greater than 20 percent difference between odometer readings and total lifetime
traveled distance, and vehicles with greater than 20 percent difference between the total grid
energy into the battery pack and total grid energy consumption during CD operation.
The aggressively filtered dataset, the "FUF /w ICCT filter", has 36 PHEV models and 71
model variants, and approximately 4,000 individual vehicles that travelled approximately 94.9
million miles. The aggressively filtered dataset excluded about 3 percent of the PHEV data
which were charged less than 1-kWh and about 24.5 percent of the PHEV data which were
charged less than 200-kWh.
As shown in Figure 3-48, small FUF differences between the "BAR Regression Fit" curve and
the aggressively filtered "BAR Regression Fit /w ICCT filter" curve are not significant compared
to the FUF values represented by the grey-dashed "SAE J2841 FUF" Fit curve. The FUF values
by the grey-dashed "SAE J2841 FUF" Fit curve are still approximately 46 percent higher than
those by the best-case, red-dashed aggressively filtered "BAR Regression Fit /w ICCT filter"
curve. Therefore, the data filtering criteria are not a dominant factor since FUF values are mostly
determined by battery charging frequencies.
0.9
0.8
0.7
"w"
O
o
£ 0.6
£
j§ 0.5
0)
o
0.4
u.
D
Ll_
0.3
0.2
0.1
0
0 20 40 60 80 100 120 140 160 180
CD Ranges [miles], GHG Emission Certified
Figure 3-48. FUF Regression Fits with Different Filtering Criteria
3.4.3.2.5 Influence of CD Blended Vehicle Miles Traveled on Electricity (eVMT)
The CD engine-on blended mode eVMT distance is calculated by multiplying the traveled
VMT during CD engine-on blended operations by the ratio of displaced gasoline to total gasoline
consumed (displaced + CD engine on blended mode gasoline consumed). The total eVMT
distance is the sum of all electric range (AER) in EV mode and CD engine-on blended mode
eVMT distance. The displaced gasoline volumes and the CD engine-on blended mode eVMT
3-104
-------
distance are calculated using equation 5 in reference (Seshadri Srinivasa Raghavan & Gil Tal
2022).
As shown in Figure 3-49, there is about 2 percent difference between averaged FUF values
calculated using the "BAR Regression Fit" curve and the "BAR Regression /w eVMT" fit curve,
which is not significant compared to the FUF values represented by the green-solid "FUF
Finalized" curve. Therefore, the eVMT distance during CD engine on blended mode operations
is insignificant when using highly efficient electric drivetrains and high-capacity battery packs.
Figure 3-49. FUF Regression Fits with CD engine-on blended mode eVMT distance
3.4.3.2.6 FUF Curves on different CD Ranges
The "BAR Regression Fit" shown in Figure 3-50, is constructed using about 8,600 valid
PHEV data points, with 2-cycle (UDDS and Highway) combined GHG emission-certified CD
ranges from 10-miles to 180-miles, contained in the November 2023 BAR OBD dataset. This
regression fit line lies substantially lower than the SAE J2841 FUF curve. The finalized FUF
curve was adjusted modestly to be about 11% from the SAE J2841 FUF curve and
approximately 38 percent higher than the curve calculated using the "BAR Regression Fit"
curve.
The label values for CD ranges are reduced by 30 percent from 2-cycle combined CD ranges
for CAFE standards compliance and labelling the all-electric ranges and CD ranges of PHEV
vehicle stickers. The differences between the labeled CD ranges and 2-cycle GHG emission-
certified CD ranges are more than 30 percent when including the voluntary adjustments. For
example, the approximately 47-mile 2-cycle GHG emission-certified CD ranges of the 2022
Audi A7E PHEV were significantly reduced to the 26-mile labeled CD ranges by including a 20
percent voluntary adjustment. Similarly, the 47-mile 2-cycle combined GHG emission-certified
3-105
-------
CD range for a 2022 Audi A7E PHEV was reduced a total of about 45 percent by the inclusion
of a voluntary adjustment.
The BAR OBD data is a recent and relatively large dataset that includes the charge depleting
distance (or electric operating distance) and total distance, which makes it a reasonable source
for evaluating the real-world utility factors for recent PHEV usage. However, we recognize that
the curve developed from this data is a departure from the SAE J2841 FUF curves, that the BAR
OBD data has some limitations (described above), and that the original SAE J2841 FUF
methodology was also a reasonable approach at the time it was adopted. While the BAR data
suggests that a lower curve than we are finalizing might more appropriately reflect current real-
world usage, EPA recognizes that PHEV technology has the potential to provide significant
GHG reductions when operating in charge depleting mode and charged regularly. In addition,
anticipated longer all-electric range and greater all-electric performance, partially driven by
CARB's ACC II program, as well as increased consumer technology familiarity and available
infrastructure could encourage greater charge depleting operation than is evident today. EPA will
continue to monitor real-world data as it becomes available.
Table 3-26: Curve Fitting Coefficients in the FUF Finalized and BAR Regression Fit.
Curve Fitting Label
Norm
Distance
(mile)
Curve Fitting Coefficients
CI
C2
C3
C4
C5
C6
SAE J2841 FUF
399.9
10.52
-7.282
-26.37
79.08
-77.36
26.07
FUF Finalized (FRM)
583.0
10.52
-7.282
-26.37
79.08
-77.36
26.07
Illustrative Final, using
SAE Normalized Distance
399.9
7.216
-3.426
-8.511
17.506
11.747
2.715
BAR Regression Fit
399.9
3.605
-0.855
-1.061
1.090
-0.366
0.042
ICCT-2 cycle range
399.9
6.674
-4.857
0.726
0.418
0.350
-2.049
The Finalized curve (see Figure 3-37 and Figure 3-50, "FUF Finalized") is based on Equation
3-2, (40 CFR 600.116-12 2022) using a normalized distance (ND) and 6-coefficients shown in
the "FUF Finalized (FRM)" of Table 3-26. Other UF curves shown include the current SAE
J2841 FUF, which uses the combined city /highway FUF coefficients, and a ND of 399.9 miles.
The "ICCT-2 cycle range" curve in Table 3-26 was translated from the "ICCT-Labeled range"
curve using the "0.7" 5-cycle adjustment factor.
The FUF and the curve fitting coefficients (Cj) of the "FUF Finalized", "FUF Illustrative
Final", "BAR Regression Fit", and "ICCT-2 cycle range" curves are listed in Table 3-26. The
SAE J2841 FUF normalized distance and 6-curve fitting coefficients are listed in Table 2 of the
SAE J2841 standard (SAE J2841 2010). The "ICCT-2 cycle range" fit curve was translated from
the blue dash-dotted "ICCT-Labeled range" curve on 2-cycle combined GHG emission-certified
CD ranges by using the same "SAE J2841 FUF" normalized distance and 6-new curve fitting
coefficients in the Table 3-26.
3-106
-------
UF
Equation 3-2: Utility Factor (UF) Exponential Equation Fits
j '
q
where:
CD = charge depleting range in miles
ND = normalized distance
Cj = the weighting coefficient for term j
k = number of coefficients (6 for the FUF Fit and 10 for the MDIUF Fit)
The five hundred eighty-three (583)-mile normalized distance in the "FUF Finalized (FRM)"
of Table 3-26 was calculated by minimizing the sum of the squared residual norm in Equation
3-2 when using SAE J2841 FUF fitting coefficients. The red solid "FUF Illustrative Final" fit
curve, which is created using 6-new curve fitting coefficients and 399.9-mile normalized
distances in the "FUF Illustrative Final, using SAE Normalized Distance" of Table 3-26, lies on
top of the green-dashed "FUF Finalized" curve in Figure 3-50.
ICCT developed the "ICCT-Labeled range" curve for the BAR OBD, which uses the
MDIUF34 coefficients, and a ND of nine hundred eighty-five (985) miles (Aaron Isenstadt, Zifei
Yang, Stephanie Searle, John German 2022) by adjusting the normalized distances in the UF
Equation 3-2 for the BAR OBD data and using sample-size weighted nonlinear least squares
regression. As shown in Figure 3-50, the FUF values calculated using the blue-dashed "BAR
Regression Fit" curve are substantially lower than those by the blue dash-dotted ICCT-Labeled
range curve, which was fitted using the labeled CD ranges. The averaged FUF values for the
ICCT-Labeled range fitted curve are about 20 percent higher than those calculated by the "BAR
Regression Fit" curve from 0 to 180-mile CD ranges.
34 The SAE J2841, the FUF is recommended for fleet vehicle fuel consumption calculations, and the MDIUF is
recommended to estimate of an individual vehicle's fuel economy. EPA has incorporated the FUF for compliance
calculations, and the MDIUF for fuel economy labeling calculations. Among other differences, the MDIUF is a
vehicle-weighted calculation, and the FUFs are VMT distance-weighted calculations.
3-107
-------
40 60 80 100 120 140
CD Ranges [miles], GHG Emission Certified
Figure 3-50: The Finalized FUF, SAE FUF, and ICCT Curves on 2-cycle combined GHG
emission-certified CD range
SAE J2841 FUF
FUF Illustrative Final
- FUF Finalized
ICCT-Labeled range
- — BAR-Labeled range
ICCT-2 cycle range
- — BAR Regression Fit
O BAR Data
i i
The "BAR Regression Fit" curve shown in Figure 3-50 lies substantially lower than the "SAE
J2841 FUF" fit curve. The "BAR Regression Fit" curve also lies on top of the "ICCT-2 cycle
range" curve which is translated from the "ICCT-Labeled range" by dividing the label CD range
by the "0.7" 5-cycle adjustment factor. The finalized FUF curve represents a modest change of
about 11 percent from SAE J2841 FUF curve while the averages of the SAE J2841 FUF curve
are approximately 55 percent higher than those calculated using the "BAR Regression Fit" curve
between 0 and 180-mile 2-cycle combined GHG emission-certified CD ranges using November
2023 CARB dataset.
As noted above, while the BAR data suggests that an even lower curve than we are finalizing
might more appropriately reflect current real-world usage, EPA recognizes that PHEV
technology has the potential to provide significant GHG reductions when operating in charge
depleting mode and charged regularly. In addition, anticipated longer all-electric range and
greater all-electric performance, partially driven by CARB's ACC II program, as well as
increased consumer technology familiarity and available infrastructure could encourage greater
charge depleting operation than is evident today. EPA will continue to monitor real-world data as
a new BAR OBD dataset, or an OEM dataset becomes available.
Table 3-27 shows PHEV vehicles that had sample sizes from 27 to 1968 in the CARB OBD
dataset (California Air Resource Board [OBD data records] 2023) and also includes several
additional high-volume PHEVs. The compliance CO2 results range from a 19.5% to 47.8%
(median = 31%) increase in CO2 g/mi, for the example vehicles below, when using the finalized
FUF compared to the existing SAE J2841 FUF.
3-108
-------
Table 3-27: CO2 Emissions [g/mi] Calculated using Existing FUF and Finalized FUF.
Model Year
Manufacturer
PHEV Model
Existing:
Compliance CO2
using Existing FUF
Finalized:
Estimated Compliance
CO2 using Finalized FUF
2022
AUDI
Q5E
116.3
165.7
2022
BMW
330E
100.2
132.5
2021
BMW
530E
114.4
147.1
2020
BMW
18
126.7
156.8
2021
BMW
X3 xDrive
136.6
168.5
2022
BMW
X5
108.5
154.3
2019
CHEVROLET
VOLT
29.9
44.2
2022
FORD
ESCAPE
47.9
63.8
2021
HONDA
CLARITY
33.4
47.9
2022
HYUNDAI
IONIQ
47.3
62.0
2019
HYUNDAI
SONATA PHEV
63.7
86.0
2021
JEEP
WRANGLER 4XE
161.0
202.7
2022
KIA
NIRO
59.4
75.9
2020
KIA
OPTIMA PHEV
59.8
80.1
2022
KIA
SORENTO SX
68.7
90.5
2020
MERCEDES-BENZ
GLC 350E
122.5
160.4
2022
MINI
COOPER
116.7
142.3
2022
SUBARU
CROSSTREK
99.0
118.3
2022
TOYOTA
PRIUS PRIME
57.5
70.6
2022
TOYOTA
RAV4 PRIME
55.1
71.7
We believe that it is important for PHEV compliance utility factors to accurately reflect the
apportionment of charge depleting operation, for weighting the 2-cycle CO2 test results;
therefore, we are updating the city and highway fleet utility factor curves with a new, single
curve that is shown in Figure 3-37 above. We are using a single curve to better reflect real world
performance where the underlying real-world data is not parsed into city and highway data.
Since the fleet average calculations are based on a combined city and highway CO2 value, a
single FUF curve can be used for these calculations.
3.4.3.3 Statistical Evaluation of FUF based on Real-World Data.
In making comments on the proposal, several stakeholders raised issues relating to statistical
aspects of the analyses described in 3.4.3.2 above. Specific questions include:
- The minimum number of tests necessary to adequately represent a defined group of interest,
such as a vehicle model,
- whether the estimates of utility factor based on the real-world data appear significantly
lower than the SAE J2841 UF trend.
- whether the price of gasoline could exert an influence on driver behavior strong enough to
influence the utility factor, and
- whether tests primarily represented routine commuting behavior as opposed to long-distance
travel and whether this difference could influence the utility factor.
3-109
-------
3.4.3.3.1 Definitions of Utility Factor
In analyses designed to target these questions, we calculated the utility factor based on the
charge-depleting operation, with engine off and on, but with the engine-on value adjusted by
"displaced fuel" to give an estimate of the mean 'blended' distance (Seshadri Srinivasa Raghavan
& Gil Tal 2022):
UF
Equation 3-3:
^CDO,Off ^Blended
Blended 7
^Total
where the "blended distance" is calculated as
Equation 3-4:
f foispl
dBlended ~ ^CDO.Rim
\foispl fcDO/
where :
/Dispi = mean "displaced" fuel (gal),
/cdo = mean fuel consumed during charge-depleting operation, and
and where:
Equation 3-5:
f t? ( FEgas ^
foispl ~ IX).Rim I -p-p I
V ^electric/
where :
£cdo,ruii Mean Total Grid Energy consumed during engine-running charge-depleting
operation, and
FEgas = mean fuel economy over 100 mi (mpg), and
FEeiectric = mean equivalent 'electric' fuel economy over 100 mi (mpg).
The mean values in the equations were calculated by groups, where group is defined as a
model.
3.4.3.3.2 Estimation of Standard Error for the UF
To develop a basis for evaluation of statistical questions, the first task is to estimate the
variance and standard deviation for the UF (,v2i t, .vi t).
We elected to estimate the variance by bootstrap simulation. Through this approach we could
avoid the need to propagate the uncertainties for the various parameters used in calculating the
blended UF. As the UF is calculated as a ratio of means, we simulated the sampling distributions
of the UF through repeated sampling of subsets of the dataset for selected models represented by
large numbers of vehicles.
3-110
-------
For the selected groups, we performed sampling with replacement (unrestricted random
sampling) for sample sizes ranging from 5 to 275 vehicles and with 5,000 replicate samples for
each sample size. For each replicate sample, we calculated the blended UF, as described above.
After compiling the sets of replicate samples, we estimated the variance of the UF as the
variance of the set of 5,000 replicate UFs. Figure 3-51 shows the sampling distributions for the
Toyota Prius for the two smallest sample sizes. The figures show that the sampling distributions
begin to approximate normal distributions for sample sizes as low as five vehicles but more
clearly for samples of 10 vehicles or more.
Figure 3-51: Toyota Prius: Sampling Distributions for the Blended UF for two Sample
Sizes (5,000 replicates).
02 0.4 0.6
Blended UF
Normal Curve Mu=0.2564 Sigma=0.1062
0.1 0.2 0.3 0.4 0.5
Blended UF
Normal Curve Mu=0 2535 Sigma=0.0743
Table 3-28 shows summary results for the simulations for two vehicle models, the BMW
530E and the Toyota Prius. The table shows means and standard deviations for the engine-off
and blended UFs for nine sample sizes ranging from n=5 to n=275. As the UF as defined above,
represents a mean (or ratios of means), we refer to its standard deviation as a 'standard error.' In
addition, the tables show the standard error as a fraction of the mean, commonly known as the
'relative standard error' (RSE). For each sample size, the mean UF values remain stable, but as
expected, the standard errors and RSE decline as n increases.
3-111
-------
Table 3-28: Mean Utility Factors, Standard Errors and Relative Standard Errors for
Bootstrap Sampling with selected Sample Sizes for two Models (5,000 replicates drawn for
each sample size).
n
Mean
Standard Error
RSE
BMW 530E
5
0.20597
0.08291
0.40254
10
0.20604
0.05960
0.28927
15
0.20640
0.04895
0.23716
25
0.20515
0.03901
0.19018
40
0.20562
0.03062
0.14893
65
0.20536
0.02378
0.11580
105
0.20484
0.01871
0.09134
170
0.20471
0.01490
0.07277
275
0.20501
0.01158
0.05649
Toyota Prius
5
0.25637
0.10624
0.41440
10
0.25350
0.07428
0.29302
15
0.25328
0.06071
0.23970
25
0.25120
0.04770
0.18988
40
0.24977
0.03790
0.15174
65
0.24970
0.02943
0.11786
105
0.25041
0.02302
0.09192
170
0.24944
0.01827
0.07324
275
0.24980
0.01446
0.05787
The decline in the RSE with increasing sample size is shown in Figure 3-52 for one model.
The plots show that the RSE follows a power-law relationship with sample size of the form.
b
RSE = —
Vn
I
y = 0.9135x
-0.49
L
R2 = 0.9996
1 1 . . 1 1
1 .... 1 .... 1
1 . . i . 1
1 . . . . 1
•
0 50 100 150 200 250 300
Sample Size
Figure 3-52: BMW 530E: Relative Standard Error vs. Sample Size for the Utility Factor
based on Bootstrap Sampling.
Developing an approach to the estimation of standard errors enabled the calculation of
confidence intervals.
The confidence intervals were calculated to give an overall "family" a level of 0.05 (95%
confidence) for the entire dataset (i.e., all variants as an aggregate). The comparisons for
3-112
-------
individual groups (i.e., models) were thus treated as analogous to multiple post-hoc comparisons
of means following an ANOVA. For this purpose, we concluded that a one-sided Dunnett
procedure would be most appropriate.
The lower and upper limits of the confidence interval (LCL, UCL) for the UF for a group
were calculated as:
Equation 3-6:
LCL = UF ^1 - qi_a,n,c J, UCL = UF ^1 + qi_«,n,c j
where:
UF = the utility factor for a model, based on one of the three definitions given above,
e/i-iui.c = the q statistic for a one-sided Dunnett procedure for an overall a level of 0.05, a
group sample size of n vehicles, for c comparisons,
b/\n = the RSE for a sample size of n, as defined above.
3.4.3.3.3 Comparison to the SAE J2841 Trend
Figure 3-53 shows the blended utility factors plotted against the mean CD range for each
model. The confidence intervals represent one-sided Dunnett comparisons, as described above,
to represent a set of one-sided comparisons against the SAE J2841 UF trend. For this set of
comparisons, models were dropped if the lower end of the confidence interval spanned 0.0 (LCL
< 0). This result occurs for small samples where the RSE is greater than 0.40. This requirement
effectively excludes sample sizes of fewer than about 10 vehicles for a model.
0 20 40 60 80 100
2-Cycle Range (miles)
O utility Factor SAE J2841
Figure 3-53: Blended Utility Factor (UF) vs CD range, by model, with one-sided Dunnett
confidence intervals.
The confidence intervals for the vast majority of models do not span the SAE J2841 UF trend,
suggesting that the measured UFs are lower than the trend. Results vary by method. For the
blended UFs, confidence intervals for three models span the SAE trend. For an overall alpha
level of 0.05, we would expect about two models (0.05 > 40) to appear significant based on
3-113
-------
random chance. For a sample of about 40 models, though, up to 4 models could appear
significant by chance. Based merely on several examples, it would not be appropriate to reject a
conclusion that, on the whole, UF is significantly lower than the SAE trend.
3.4.3.3.4 Influence of Gasoline Price
One or more commenters noted that much of the driving reflected in the dataset under
consideration took place during the pandemic, and that gasoline prices were lower during this
period than after the pandemic ended. The question is whether reduced gasoline prices could
have influenced driver behavior during this period, ostensibly by reducing the frequency of
charging. To examine this question, we partitioned the dataset to distinguish tests occuring
during periods of'low' and 'high' gasoline prices. Specifically, we distinguished data collection
points occuring before and after April 1, 2021, with times before and after this date taken as
"low" and "high" gasoline-price conditions, respectively.
This comparison is restricted to a set of 10 models for which the samples sizes were adequate
for both price levels. Accordingly, we calculated two-sided Dunnett confidence intervals for an
overall alpha level of 0.05 and 10 comparisons. Confidence intervals are consistently narrower
for the 'high price' condition, indicating that most trips took place after the reference date. For
model years 2022 and 2023, all driving would have taken place after the reference date.
\ \ \ V \ X X \ \ \
% * V\ V \\\
<*> 0 X %
\ %
Model
gas_price ¦ High Gas Price ¦ Low Gas Price
Figure 3-54: Blended Utility Factor by Gasoline-price Level for Selected Models.
Figure 3-54 shows the comparisons for the blended UF. Directionally, the UF is higher for the
'high' price for six of the 11 models. For most models, the comparison does not appear
significant. Flowever, for several models, it does look significant. For two models in particular,
the BMW 530E and the Ford Fusion, the comparison looks significant and in the same direction,
with the UF lower for the 'low' price condition. For an overall alpha level of 0.05 we might
expect a maximum of about two comparisons to appear significant out of a sample of 10.
Comparisons for the UF for the other two definitions showed similar results.
3-114
-------
3.4.3.3.5 Influence of Geographic Origin
The fact that tests in the dataset are non-routine tests for vehicles within the exemption period
raised an additional question. Given that non-negligible fractions of these vehicles represent
vehicle owners moving into California from out of state, they may have potentially driven the
vehicles over long distances, perhaps hundreds to thousands of miles, to get to California. A
stakeholder suggested that driving behavior could differ on long trips, with vehicle owners
experiencing less need to plug in than in routine commuting behavior, and that reduced charging
frequency could effectively reduce UF.
w^pgst
%
Model
origin ¦ In-State ¦ Out-of-State
Figure 3-55: Blended Utility Factor by Origin for Selected Models.
Based on information received from CARB, we were able to distinguish in-state and out-ot-
state vehicles. To approach the question, we calculated UF separately for in- and out-of-state
groups. Out of the total, 24 models have adequate sub-samples for both groups to allow
comparisons. As with gas price, we calculated two-sided Dunnett confidence intervals, for an
overall alpha level of 0.05 for a family of 24 comparisons.
Figure 3-55 shows the comparisons in the UF by origin. The difference is not large for most
models. The in-state UF is higher for nine models and lower or very close for the remainder.
Examinations suggest that none of the individual comparisons look significant. Overall, it
appears that this factor is not a major or consistent influence in the UF for PHEVs.
3.4.4 Other studies of FUF
We have analyzed available data and compiled literature (Aaron Isenstadt, Zifei Yang,
Stephanie Searle, John German 2022), (Krajinska, Poliscanova, Mathieu, & Ambel, Transport &
Environment 2020), (Plotz, P., Moll, C., Bieker, G., Mock, P., Li, Y. 2020), (Seshadri Srinivasa
3-115
-------
Raghavan & Gil Tal 2022) showing that the SAE J2841 utility factors at the labeled CD ranges
are overestimating the operation of PHEVs on electricity, and therefore would underestimate the
CO2 g/mile compliance result. European PHEV data were not used since the mean UF values for
German company-owned PHEV vehicles were significantly lower than those for German
privately owned PHEV vehicles (Plotz, P., Moll, C., Bieker, G., Mock, P., Li, Y. 2020). All the
observed UF values (Seshadri Srinivasa Raghavan & Gil Tal 2022), except those driven by early
adopters of the discontinued Chevrolet Volt Genl Extended Range Electric Vehicles (EREVs),
were lower than the referenced SAE J2841 UFs on the labeled CD ranges when using small
volunteer private PHEV data, which consists of 3 PHEV models and four model variants of 153
PHEV owners (out of 19000 recruited owners) in USA.
Much of the literature on PHEV utility factor is presented using label CD range. However,
this is not directly comparable to the FUF used for GHG certification purposes, which is based
on the 2-cycle combined CD range. The label CD ranges are calculated by multiplying the 2-
cycle combined GHG emission-certified CD ranges by the "0.7" 5-cycle adjustment factor. The
UF values in the referenced literature were analyzed at the labeled CD ranges, which were
reduced by at least 30-percent from 2-cycle combined GHG emission certified CD ranges. The
differences between the labeled CD ranges and 2-cycle GHG emission-certified CD ranges are
more than 30 percent when including OEM voluntary adjustments. For example, the 47-mile 2-
cycle GHG emission-certified CD range of the 2022 Audi A7E PHEV was significantly reduced
to the 26-mile label CD range by the inclusion of an additional 20 percent voluntary adjustment
after applying the "0.7" 5-cycle adjustment factor. The 47-mile 2-cycle combined GHG
emission-certified CD range of the 2022 Audi A7E is about 81 percent higher than the 26-mile
labeled CD range. Therefore, those UF values in the literature are significantly lower when
translating the same UF values from the labeled CD ranges to the 43 percent higher 2-cycle
combined GHG emission-certified CD ranges. In fact, the UF values at the labeled CD ranges,
which are called the MDIUF (Multi-Day Individual Utility Factor), are not directly relevant to
the FUF Finalized at 2-cycle combined CD ranges for GHG emission certifications. We do not
propose to update the MDIUF curve, and the labeled values of the EV/CD ranges and electric
energy consumption rates which are already adjusted using the "0.7" 5-cycle adjustment factor
used for CAFE standards compliance and labelling the all-electric ranges and CD ranges of
PHEV vehicle stickers. When including OEM's voluntary labeled CD range reductions, the FUF
values on 2-cycle combined GHG emission-certified CD ranges are not directly related to the
MDIUF (Multi-Day Individual Utility Factor) on the labeled CD ranges. Therefore, the UF
values on the labeled CD ranges in the literature cannot be directly compared to the FUF values
on 2-cycle combined GHG emission-certified CD ranges.
3.4.5 Consideration of CARB ACC II PHEV Provisions
CARB recently set minimum performance requirements for PHEVs in their ACC II program.
These requirements include performance over the US06 test cycle and a minimum range and are
meant to set qualifications for PHEV's to be included in a manufacturer's ZEV compliance. EPA
received comments that it should adopt ACC II for PHEVs. ACC II is a suite of emissions
standards that includes a ZEV mandate and other tools EPA is not using in this rule and it would
not be appropriate to take only the PHEV portions of ACC II. EPA is not adopting the range and
US06 performance requirements or fleet penetration limits that are included in the CARB ACC
II ZEV provisions. EPA agrees that PHEVs meeting the performance provisions required by
CARB in ACC II have the potential to provide greater environmental benefits as compared to
3-116
-------
other PHEVs that are less capable. However, unlike the ACC II program, the GHG program in
this rulemaking is performance-based and not a ZEV mandate. In that regard, EPA believes that
it is appropriate to have a robust GHG compliance program for PHEVs that properly accounts
for their GHG emissions independent of a PHEV's range or capability over the US06 test cycle.
We are addressing the issue of ensuring appropriate GHG compliance values for PHEVs through
the revised PHEV fleet utility factor as described in section III.C.8 of the preamble; EPA is not
adopting design requirements for PHEVs, that is, we are not adopting minimum range
requirements or specifying minimum capability over any prescribed test cycles.
3.5 GHG Emissions Control Technologies
3.5.1 Engine Technologies
The following is detailed information about the ALPHA inputs for internal combustion
engines used to create ALPHA Outputs for Response Surface Equations (RSE's) used by
OMEGA. These were first discussed and listed in Table 2-2. Specific details about each engine
are contained in the engine's test data package available on EPA's webpage (U.S. EPA 2022b).
These engine data packages are combined into a set of complete engine maps suitable for use in
vehicle simulation models which are contained in a .zip file identified using the engine name
mentioned in the caption of the associated ALPHA efficiency map shown below (U.S. EPA
2023b).
3.5.1.1 2013 Chevrolet 2.5L Ecotec LCV Engine Reg E10 Fuel
This naturally aspirated engine features continuously variable valve timing, high-pressure
direct injection, electronic throttle control, coil-on-plugs and has an 11.3:1 compression ratio.
Testing was conducted in a test cell operated by FEV Engine Technologies and purchased to
support the Mid Term Evaluation (MTE) Engine Benchmarking project. (Newman, K., Kargul,
J., and Barba, D. 2015).
3-117
-------
1000 1500 2000 2500 3000 3500 4000 4500 5000 5500 6000
Speed (RPM )
Figure 3-56: 2013 Chevrolet 2.5L Ecotec LCV Engine Reg E10 Fuel BTE (%) (U.S. EPA
2023b).
165 kW
150 kW
135 kW
120 kW
105 kW
90 kW
75 kW
60 kW
45 kW
30 kW
15 kW
7.5 kW
3.5.1.2 GT Power Baseline 2020 Ford 7.3L Engine from
Argonne Report Tier 3 Fuel
This medium-duty naturally aspirated engine included port fuel injection, a 2-valve head, and
a 10.5 compression ratio. The engine was modeled in GT-Power® and then calibrated and
validated against test data available at Southwest Research Institute or provided by the Original
Equipment Manufacturers (OEMs). (Thomas E. Reinhart 2021) The provided baseline model
was configured to simulate wide-open throttle operation with power enrichment and used a
Wiebe function for describing combustion. Once the model achieved satisfactory results, the
engine performance was mapped over the speed and load range.
3-118
-------
Speed ( RPM )
Figure 3-57: GT Power Baseline 2020 Ford 7.3L Engine from Argonne Report Tier 3 Fuel
BTE (%) (U.S. EPA 2023b).
3.5.1.3 2013 Ford 1.6L EcoBoost Engine LEV III Fuel
The selected feature of this turbocharged gasoline engine was the inclusion of spray-guided
direct-injection. The testing was performed by the EPA at the National Center for Advanced
Technology (NCAT) with the engine installed in a dynamometer test cell tethered as though the
engine were operating in the vehicle. (Mark Stuhldreher, Charles Schenk, Jessica Brakora, David
Hawkins, Andrew Moskalik, and Paul DeKraker 2015)
3-119
-------
CO
CD
CL
UJ
CQ
E
z
0)
3
CT
20
18
16
14
12 -
10
8
6
4
2
250 -
200 "
15 kW
7.5 kW
1000 1500 2000 2500 3000 3500 4000 4500 5000 5500
Speed ( RPM )
Figure 3-58: 2013 Ford 1.6L EcoBoost Engine LEV III Fuel BTE (%) (U.S. EPA 2023b).
3-120
-------
3.5.1.4 2015 Ford 2.7L EcoBoost Engine Tier 3 Fuel
This turbocharged engine features intake and exhaust cam phasing, direct injection, and
integrated exhaust manifolds. The testing was performed by the EPA at the National Center for
Advanced Technology (NCAT) with the engine installed in a dynamometer test cell tethered as
though the engine were operating in the vehicle. This testing provided thorough test data for
constructing the main operating portion of the engine map. There was also subsequent testing in
a heavy-duty test cell to generate additional data for the high speed and high load mapping
needed to construct a more complete engine map.
20
CO
CD
qT 15
UU
5
CO
E 10
CD
cr
25 kW
12.5 kW
1000 1500 2000 2500 3000 3500 4000 4500 5000 5500
Speed ( RPM )
Figure 3-59: 2015 Ford 2.7L EeoBoost V6 Engine Tier 3 Fuel BTE (%) (U.S. EPA 2023b).
3-121
-------
3.5.1.5 2016 Honda 1.5L L15B7 Engine Tier 3 Fuel
o
1000 1500 2000 2500 3000 3500 4000 4500 5000 5500 6000
Speed ( RPM)
Figure 3-60: 2016 Honda 1.5L L15B7 Engine Tier 3 Fuel BTE (%) (U.S. EPA 2023b).
Features of this engine include direct-injection, single-scroll turbocharger, and dual variable
valve timing control (VTC). The testing was performed by the EPA at the National Center for
Advanced Technology (NCAT) with the engine installed in a dynamometer test cell tethered as
though the engine were operating in the vehicle. (Stuhldreher, Mark; Kargul, John; Barba,
Daniel; McDonald, Joseph; Bohac, Stanislav; Dekraker, Paul; Moskalik, Andrew 2018) The
engine was coupled to the dynamometer using a modified manual transmission and clutch with a
torsional spring assembly and rubber isolated driveshaft to allow for stable torque measurements.
Both steady-state and transient engine test data were collected during the benchmark testing.
Two different test procedures were needed to appropriately replicate steady-state engine
operation at low/mid loads and transient engine operation at high loads when the engine is
protecting itself.
250
20 h
18
16
14
12
.10
100
3-122
-------
3.5.1.6 Volvo VEP 2.0L LP Gen3 Miller Engine from 2020 Aachen Paper Octane
Modified for Tier 3 Fuel
This Miller cycle engine includes an increased compression ratio, a short intake valve opening
duration, an integrated exhaust manifold, a new intake port and piston design together with a
VGT turbo as described in Dahl et al (2020), "The New Volvo Mild Hybrid Miller Engine"
presented in Aachen Colloquium Automobile and Engine Technology. (Daniel Dahl, Ayolt
Helmantel, Fredrik Wemmert, Mats Moren, Staffan Rengmyr, and Ali Sahraeian 2020). The
image provided in this paper was digitized by loading the image into MATLAB and manually
tracing the efficiency contours. NCAT used the peak BSFC and BTE values referenced in the
paper to calculate the lower heating value for the test fuel having a reported RON of 98 and
because the authors did not provide any test data for this engine using Tier 3 fuel, the decision
was made to use ALPHA'S Octane Modifier to develop an estimated Tier 3 fuel map.
Speed ( RPM )
Figure 3-61: Volvo 2.0L VEP LP Gen3 Miller Engine from 2020 Aachen Paper
Octane Modified for Tier 3 Fuel BTE (%) (U.S. EPA 2023b).
3-123
-------
3.5.1.7 Geely 1.5L Miller GHE from 2020 Aachen Paper Octane Modified for Tier
3 Fuel
Zhang et al (2020), "Geely Hybrid Engine: World Class Efficiency for Hybrid Vehicles"
presented in the 29th Aachen Colloquium (GuiQiang Zhang, Qian Wang, Guang Chen, et al.
2020) reported this engine has a high efficiency Miller-cycle combustion system with high
tumble and turbulence kinetic energy, low friction, optimized mixture formation using a new 350
bar fuel injection system and a 13:1 compression ratio. These features are then combined with a
fully matched turbocharger with highly cooled low pressure EGR and a water-charge air cooler.
The image provided in this paper was digitized by loading the image into MATLAB and
manually tracing the efficiency contours. Since the fuel used to map this engine had a relatively
high lower heating value there was an assumption of a likely corresponding high RON value of
98 and because the authors did not provide any test data for this engine using Tier 3 fuel, the
decision was made to use ALPHA'S Octane Modifier to develop an estimated Tier 3 fuel map.
aj
CO
CL
LU
5
CO
¦ 12
5500
1000 1500 2000 2500 3000 3500 4000 4500 5000
Speed ( RPM )
Figure 3-62: Geely 1.5L Miller GIIE from 2020 Aachen Paper
Octane Modified for Tier 3 Fuel BTE (%) (U.S. EPA 2023b).
3-124
-------
3.5.1.8 2018 Toyota 2.5L A25A-FKS Engine Tier 3 Fuel
This 4-cylinder, naturally aspirated, Atkinson Cycle gasoline engine with cooled-EGR also
includes direct & port injection, VVT electric intake & hydraulic exhaust, high induction
turbulence/high speed combustion, high energy ignition, friction reduction, a variable capacity
oil pump, and an electric water pump. The testing was performed by the EPA at the National
Center for Advanced Technology (NCAT) with the engine installed in a dynamometer test cell
tethered as though the engine were operating in the vehicle. The engine was coupled to the
dynamometer using an automatic transmission and torque converter to allow for an accurate
gathering of test data where the torque measurement is very sensitive to the engine's torsional
accelerations. (John Kargul, Mark Stuhldreher, Dan Barba, Charles Schenk, Stani Bohac, Joseph
McDonald, and Paul Dekraker 2019).
Speed ( RPM )
Figure 3-63: 2018 Toyota 2.5L A25A-FKS Engine Tier3 Fuel BTE (%) (U.S. EPA 2023b).
3-125
-------
3.5.1.9 Toyota 2.5L TNGA Prototype Hybrid Engine from 2017 Vienna Paper
Octane Modified for Tier 3 Fuel
This inline 4 cylinder 2.5L gasoline naturally aspirated (NA) engine is thoroughly described
in Tadashi Toda et al (2017), "The New Inline 4 Cylinder 2.5L Gasoline Engine with Toyota New
Global Architecture Concept" presented at Internationales Wiener Motorensymposium. (T.
Toda, M. Sakai, M. Hakariya, and T. Kato 2017). Features include high energy ignition coil
rnotor-driven VVT for Atkinson cycle, a D-4S system (direct and port injection) with new multi
hole injectors, cooled EGR, and a variable oil-pressure pump system. The image provided in this
paper was digitized by loading the image into MATLAB and manually tracing the efficiency
contours. There was no information presented regarding the fuel used for the map, so the
decision was made to assume the data in the paper was based on a Tier 2 fuel and use ALPFtA's
Octane Modifier to develop an estimated Tier 3 fuel map.
7.5 kW
0
1000 1500 2000 2500 3000 3500 4000 4500 5000 5500
Speed ( RPM )
Figure 3-64: Toyota 2.5L TNGA Prototype Hybrid Engine from 2017 Vienna Paper
Octane Modified for Tier 3 Fuel BTE (%) (U.S. EPA 2023b).
135 kW
10 - 200
120 kW
105 kW
90 kW
75 kW
60 kW
45 kW
15 kW
30 kW
3-126
-------
3.5.1.10 GT Power 2020 GM 3.0L Duramax Engine from Argonne Report Diesel
Fuel
This in-line six-cylinder four-stroke engine features cooled low-pressure-loop and un-cooled
high-pressure-loop EGR systems with a variable geometry turbine, water-to-air charge air
cooling that uses a separate low temperature coolant loop, and a variable intake manifold with
dual air intake paths. The light duty diesel combines high-pressure loop EGR with low pressure
loop EGR to reduce the pumping work required to flow EGR, uses very little EGR at high load
and has very low friction for a diesel engine. The engine was modeled in GT-Power® and then
calibrated and validated against test data available at Southwest Research Institute or provided
by the Original Equipment Manufacturers (OEMs). Once the model achieved satisfactory results,
the engine performance was mapped over the speed and load range. The image and any
supporting data available were digitized by loading the image into MATLAB and manually
tracing the efficiency contours.
-600
275 kW
250 kW
| 300
a>
§-10 -
o
K 200
225 kW
200 kW
175 kW
150 kW
125 kW
100 kW
75 kW
50 kW
Figure 3-65:
25 kW
12.5 kW
1000 1500 2000 2500 3000 3500 4000
Speed ( RPM )
GT Power 2020 GM 3.0L Duramax Engine from Argonne Report Diesel Fuel
BTE (%) (U.S. EPA 2023b).
3-127
-------
3.5.1.11 Future 3.6L H LA Hybrid Concept Engine Tier 3 Fuel
Future high load application (HLA) vehicles such as hybrid light and medium duty body-on-
frame vehicles with towing capabilities as discussed in the RETT report (Bhattacharjya,S., et al.
2023) will likely require dedicated hybrid engines such as the concept for a 3.6L V6 HLA hybrid
engine.
ALPHA'S "VW 1.5L TSI evo Hybrid Concept 4 Engine from 2019 Aachen Paper Octane
Modified for Tier 3 Fuel" (U.S. EPA 2023b) engine map was used to derive an upsized concept
for a future V6 3.6L HLA hybrid engine map. There is a likelihood design changes may be
necessary to increase the durability of the original light-duty engine to be suitable for a work
truck with heavy sustained towing applications. Therefore, when deriving the scaled map using
the original 1.5L engine, surface-to-volume heat transfer effects were ignored, and the efficiency
of the original engine was maintained. Additionally, adjustments were made to the wide-open
throttle (WOT) line to account for tendencies of larger cylinders to knock under heavy loads.
0
1000 1500 2000 2500 3000 3500 4000
Speed ( RPM )
Figure 3-66: Future 3.6L HLA Hybrid Concept Engine Tier 3 Fuel BTE (%) (adapted
from VW 1.5L TSI evo Hybrid Concept 4 engine from 2019 Aachen Paper Octane
Modified for Tier 3 Fuel).
400
250 kW
225 kW
200 kW
175 kW
150 kW
125 kW
100 kW
75 kW
50 kW
25 kW
12.5 kW
3-128
-------
3.5.1.12 Future 6.0L H LA Hybrid Concept Engine Tier 3 Fuel
Future HLA vehicles such as hybrid light and medium duty body-on-frame vehicles with
towing capabilities as discussed in the REET report (Bhattachaijya,S., et al. 2023) will likely
require dedicated hybrid engines such as the concept for a 6.0L V6 HLA hybrid engine.
ALPHA'S "VW 1.5L TSI evo Hybrid Concept 4 engine from 2019 Aachen Paper Octane
Modified for Tier 3 Fuel" (U.S. EPA 2023b) engine map was used to derive an upsized concept
for a future V6 6.0L HLA hybrid engine map. There is a likelihood design changes may be
necessary to increase the durability of the original light-duty engine to one suitable for a work
truck with heavy sustained towing applications. Therefore, when deriving the scaled map using
the original 1.5L engine, surface-to-volume heat transfer effects were ignored, and the efficiency
of the original engine was maintained. Additionally, adjustments were made to the wide-open
throttle (WOT) line to account for tendencies of larger cylinders to knock under heavy loads.
Speed ( RPM )
Figure 3-67: Future 6.0L HLA Hybrid Concept Engine Tier 3 Fuel BTE (%) (adapted
from VW 1.5L TSI evo Hybrid Concept 4 engine from 2019 Aachen Paper Octane
Modified for Tier 3 Fuel).
3-129
-------
3.5.1.13 2014 Chevrolet 4.3L EcoTec3 LV3 Engine LEVIII Fuel
Features of this engine include side mount direct-injection, cylinder deactivation,
continuously variable valve timing, pushrod, single cam, and active fuel management. The
engine uses cylinder deactivation to improve thermal efficiency by reducing pumping losses
during low-load operation. This testing was performed by the EPA at the National Center for
Advanced Technology (NCAT) with the engine installed in a dynamometer test cell tethered as
though the engine were operating in the vehicle. (Mark Stuhldreher 2016) Two methods of
coupling the engine to the dynamometer were needed to gather data where the torque
measurement was very sensitive to the engine's torsional accelerations. Direct drive shaft engine
to dynamometer coupling worked best to gather most of the data but where needed, the engine
was coupled to the dynamometer through its transmission and torque converter.
250 kW
225 kW
200 kW
175 kW
150 kW
125 kW
100 kW
75 kW
50 kW
25 kW
12.5 kW
1000 1500 2000 2500 3000 3500 4000 4500 5000 5500
Speed (RPM)
Figure 3-68: 2014 Chevrolet 4.3L EcoTec3 LV3 Engine LEVIII Fuel - Cylinder Deac
Enabled BTE (%) (U.S. EPA 2023b).
3-130
-------
250 kW
225 kW
200 kW
175 kW
150 kW
Figure 3
125 kW
100 kW
75 kW
50 kW
25 kW
12.5 kW
10D0 1500 2000 2500 3000 3500 4000 4500 5000 5500
Speed ( RPM)
2014 Chevrolet 4.3L EcoTec3 LV3 Engine LEVXII Fuel - Cylinder Deac
Disabled BTE (%) (U.S. EPA 2023b).
-69:
3.5.2 Electrification Technologies
The following is detailed information about the ALPHA inputs for electric inverters, motors,
and generators used to create ALPHA Outputs for Response Surface Equations (RSE's) used by
OMEGA. These were first discussed and listed in Table 2-3. Specific details about each electric
motor are contained in its ALPHA test package available on EPA's webpage (U.S. EPA 2023a).
Each emotor package is contained in a .zip file identified using the electric motor name
mentioned in the caption of the associated ALPHA efficiency map shown below.
3.5.2.1 2010 Toyota Prius 60kW 650V MG2 EMOT
The 60kW 650V MG2 electric motor paired with an inverter and a 36hp (27kW) nickel-metal
hydride battery pack was combined with a 1.8L 4-cylinder Atkinson cycle engine. The
component benchmarking testing for this program was conducted by Oak Ridge National
Laboratory's (ORNL) Power Electronics and Electric Machinery Research Center (PEEMRC), a
broad-based research center for power electronics and electric machinery (e-motor)
development. (Olszewski, Mitch 2011). The resulting measurements were used to create a
combined efficiency map of the main drive e-motor and inverter without including any gearing,
categorized together as an EMOT.
3-131
-------
-250 kW
-200 , . , , , ,
-300 kW
0 2000 4000 6000 8000 10000 12000
Motor Output Speed ( RPM )
Figure 3-70: 2010 Toyota Prius 60kW 650V MG2 EMOT Efficiency (%) (U.S. EPA 2023a).
¦5UU KW
250 kW
200 kW
150 kW
100 kW
50 kW
25 kW
-25 kW
-50 kW
-100 kW
-150 kW
-200 kW
3.5.2.2 EST 2010 Toyota Prius 60kW 650V MG1 EMOT
The Toyota Prius uses a secondary electric motor called an MG1, which functions as a
generator to transfer power from the ICE to recharge the battery and functions as a motor to start
the ICE. Oak Ridge National Laboratory (ORNL) did not specifically benchmark this electric
generator motor, presumably because of its similarity to the MG2 electric drive motor discussed
in the previous section; MG2 functions as a motor when propelling the vehicle and acts as a
generator during regenerative braking function. (Olszewski, Mitch 2011). However, chassis test
data provided by Southwest Research Institute (SwRI) indicated the maximum operating power
for the MG1 generator motor is different than the MG2 drive motor. The maximum power curve
for the MG1 is a constant value rather than variable as MG2's power curve. Consequently, the
MG2 ORNL benchmark data was used along with the max power data provided from the SwRI
chassis test data to create a constant power version for the MG1. The MG1 efficiency map
estimates the combined efficiency of the main generator e-motor and its inverter, categorized
together as an EMOT. The "EST" in the front of the e-motor's name indicates this is an estimated
map.
3-132
-------
0 2000 4000 6000 8000 10000 12000
Motor Output Speed ( RPM )
Figure 3-71: Est 2010 Toyota Prius 60kW 650V MG1 EMOT Efficiency (%) (U.S. EPA
2023a).
30U KW
250 kW
200 kW
150 kW
100 kW
50 kW
25 kW
-25 kW
-50 kW
-100 kW
-150 kW
-200 kW
-250 kW
-300 kW
3.5.2.3 2011 Hyundai Sonata 30kW 270V EMOT
This 30 kW 270V electric motor was paired with an inverter, categorized together as an
EMOT, and powered by a 270-volt lithium polymer battery. The map was created using
benchmarked data that measured the efficiency of the combination of the main drive e-motor and
its inverter without including any gearing. The component testing for this program was
conducted by Oak Ridge National Laboratory's (ORNL) Power Electronics and Electric
Machinery Research Center (PEEMRC), a broad-based research center for power electronics and
electric machinery (e-motor) development. (Rogers, Susan 2012).
3-133
-------
0
Figure 3-72:
1000 2000 3000 4000 5000 6000
Motor Output Speed ( RPM )
2011 Hyundai Sonata 30kW 270V EMOT Efficiency (%) (U.S. EPA 2023a).
-150
75 kW
50 kW
25 kW
12.5 kW
-12.5 kW
-25 kW
-75 kW
- -100 kW
-125 kW
3.5.2.4 2012 Hyundai Sonata 8.5kW 270V BISG
Hyundai's Hybrid Starter Generator (HSG) electric motor with published specifications listed
as 43 Nm, 8.5kW, and 15,750 rpm was paired with an inverter and powered by a 270-volt
lithium polymer battery. The application of this type of electric motor is normally found in mild
hybrid electric vehicles (MHEV), often called P0 mild hybrids. The goal was to create a map
representing the combined efficiency of the starter/generator motor, its inverter, and the drive
belt, categorized together as a BISG (belt-inverter-starter/generator). The component testing for
this program was conducted by Oak Ridge National Laboratory's (ORNL) Power Electronics
and Electric Machinery Research Center (PEEMRC), a broad-based research center for power
electronics and electric machinery (e-motor) development. (Rogers, Susan 2013).
3-134
-------
-50 kW
0 1000 2000 3000 4000 5000 6000
Motor Output Speed ( RPM )
Figure 3-73: 2012 Hyundai Sonata 8.5kW 270V BISG Efficiency (%) (U.S. EPA 2023a).
70 kW
60 kW
50 kW
40 kW
30 kW
20 kW
10 kW
5 kW
-5 kW
-10 kW
-20 kW
-30 kW
-40 kW
3.5.2.5 Generic IPM 150kW 350V EDU
The Generic IPM 150kW 350V Electric Drive Unit (EDU) efficiency map was generated
using confidential benchmarking test data from several state-of-the-art internal permanent
magnet synchronous reluctance (IPMSRM) e-motors used in current production battery electric
vehicles. Transformation functions whose coefficients represent the averaged power
consumption data were utilized to blend and transform the confidential test data. The final map
was then scaled to 150kW to represent a generic EDU suitable for use in a BEV. The generated
efficiency map represents the combined operating boundaries and electrical power consumption
of the electric motor, inverter, and gearing, categorized together as an EDU. The gear ratio for
this EDU is 9.5:1.
3-135
-------
3000
2000
E
z
0)
13
cr
3
Ql
¦4—1'
o
3
Q
UJ
1000
1000
-2000
-3000
0 200
Figure 3-74:
400 600 800 1000 1200 1400 1600
EDU Output Speed ( RPM )
Generic IPM 150kW EDU Efficiency (%) (U.S. EPA 2023a).
500 kW
400 kW
300 kW
200 kW
100 kW
50 kW
-50 kW
-100 kW
-200 kW
-300 kW
-400 kW
-500 kW
3.5.2.6 Three IPM Electric Motor/Inverters (EMOTs) used to Simulate Future
LD PHEVs and MD PHEVs with Towing Capability
ALPHA has input data representing three modern IPM electric motor/inverters (EMOTs)
suitable to simulate electric drive systems for future LD and MD PHEVs. EPA received
confidential supplier data for these advanced motors and inverters which was evaluated and
compared with other data available to EPA from other advanced electric motor and inverter
systems. The evaluation determined these EMOTs were suitable to simulate electric drive
systems for future light- and medium-duty towing hybrids, which were studied and documented
in the REET report. (Bhattacharjya,S., et al. 2023). The input data used to create these ALPHA
inputs is confidential, therefore we are unable to include copies of their efficiency maps in this
document. For this rulemaking these ALPHA inputs are known as CBI-A, CBI-B, and CBI-C.
3.5.3 Vehicle Architectures
A summary of the vehicle architectures used in ALPHA 3.0 is provided in Section 2.4.4.
Figure 2-6 summarizes the seven vehicle models used to simulate vehicle efficiency, including
the one conventional (ICE) model used in previous versions of ALPHA, the five new hybrid
electric models (including a mild hybrid and four strong hybrids), and the one new battery
electric vehicle model added for ALPHA 3.0. Three of the hybrid models were available in the
3-136
-------
previous version of ALPHA. The P2-P4 hybrid (REET) and the SP-P4 Hybrid (REET) models
are new to ALPHA 3.0. They are described in the next section.
3.5.3.1 Heavy-light-duty and Medium-duty Range-extended Electric Truck
(REET) Study
In 2022, EPA conducted a study of LDT4 and MDV pickup trucks under contract with
Southwest Research Institute (SwRI) and with assistance from Argonne National Laboratory to
determine how plug-in hybrid electric vehicles (PHEVs) with significant all-electric capability,
or "range extended electric trucks" (REET), may assist in the transition to electrification of these
vehicle classes. (U.S. EPA 2022). Two pickup trucks representing high-volume examples of
LDT4 and MDV pickup trucks were selected to serve as baseline vehicles. Simulations were
conducted using Gamma Technologies GT-SUITE to model the baseline vehicles and potential
REET vehicle designs over regulatory drive cycles and simulation of the SAE J2807
"Performance Requirements for Determining Tow-Vehicle Gross Combination Weight Rating
and Trailer Weight Rating" (SAE 2016).
The base LDT4 pickup selected for the study was a MY2021 Ford F150 equipped with a 3.5L
turbocharged, direct- and port-injection gasoline engine with a rated power of 298 kW. The F150
was equipped with a "Max Trailer Tow Package" and had a GCWR of approximately 19,500
pounds. The base MDV pickup selected was a MY2021 Stellantis RAM 2500 equipped with a
6.7L Cummins diesel engine with a rated power of 275 kW and a GCWR of approximately
30,000 pounds. Please refer to the final report for this study for further details regarding the
design of the study, procedures used, and results. (Bhattacharjya,S., et al. 2023).
3.5.3.1.1 LDT4 Range-extended Electric Truck (REET)
The following targeted design criteria for the LDT4 REET simulation were established:
• Battery sizing consistent with CARB ACC II Light-duty PHEV Requirements (State
of California, Air Resources Board 2022)
• 50-mile electric vehicle label range
• 40-mile all-electric-range on the US06 drive cycle
• 0-60 mph at ETW (6,500 pounds) equivalent to baseline 2021 Ford F15035
• 0-60 mph acceleration during towing at approximately 19,500 pounds GCWR of less
than 30 seconds as per SAE J2807, 4.3.1 (SAE 2016); and SAE J1491, 4.3.2 and 4.3.3
(SAE 2006).
• 0-30 mph acceleration during towing at approximately 19,500 pounds GCWR of less
than 12 seconds as per SAE J2807, 4.3.1; and SAE J1491, 4.3.2 and 4.3.3.
35 Note that the Ford F150 had an ETW of 5,500 pounds. The LDT4 REET configurations was approximately 1,000
pounds heavier due to the battery pack weight, electric drive system weight, and other components and was rounded
to the nearest ETW weight increment (6,500 pounds) from 40 CFR 86.129-94 (40 CFR § 86.129-94 2022).
3-137
-------
• 40-60 mph acceleration during towing at approximately 19,500 pounds GCWR of less
than 18 seconds as per SAE J2807, 4.3.1; and SAE J1491, 4.3.2 and 4.3.3.
• Launch on 12% grade at approximately 19,500 pounds GCWR as per SAE J2807,
4.3.4.
• Highway gradeability (commonly referred to as the Davis Dam Grade) during towing
at approximately 19,500 pounds GCWR that maintained a minimum speed of 45 mph
in sections with a 45-mph speed limit and 55 mph in sections with speed limits of 55
mph or 65 mph sections as per SAE J2807, 4.3.5.
Based on initial vehicle simulations, two REET architectures were selected for further
simulation runs and analysis. One was an all-wheel-drive P2-P4 configuration. The other was an
all-wheel-drive series-parallel hybrid with two P4 machines and a separate motor-generator.
Both configurations were modeled using a dedicated-hybrid 3.6L Miller cycle engine with 240
kW rated power. For a summary of the hybrid drive-system and battery specifications please
refer to Table 3-29 and Table 3-30. For more details regarding the drive systems and the
dedicated hybrid engine, please refer to the final report. (Bhattacharjya,S., et al. 2023). The
electric machines used in both system configurations were internal-permanent-magnet
synchronous machines using SiC inverters. CO2 emissions and fuel economy (FE) simulation
results over the UDDS, HWFET, combined cycle results, and US06 are summarized in
Table 3-31 and Table 3-32.
Table 3-29: Hybrid drive system specifications
used for LDT4 REET simulations.
Vehicle
Location
Scaled Max.
Scaled Max.
Max.
Base
Gear Ratio
Power (kW)
Torque
Speed
Speed
(Nm)
(rpm)
(RPM)
P2-P4
P2
120
828
6,000
1.500
3.73:1
Front c-a\lc / P4
150
311
18.000
4.200
17:1
Scries -
Generator
240
619
13.500
3.500
0.75:1
Parallel
Rear series-
150
276
18.000
4.200
17:1
parallel drive
Front e-axle / P4
150
276
18.000
4.200
17:1
*P2 gear ratio is for the final drive. Note that the P2 also incorporates a 10-speed automatic transmission prior to the final drive.
¦f Gear reduction from engine to generator.
JP4 and series parallel use a single final-drive gear ratio.
3-138
-------
Table 3-30: Battery specifications used for LDT4 REET simulations.
Series-Parallel
P2-P4
Cathode:
NMC811
NMC811
Anode:
95% graphite. 5% Si
95% graphite. 5% Si
Cells in Series:
86
89
Cells in Parallel:
3
3
Cell Capacity | Ah|:
40
40
Cell Nominal Voltage |V|:
3.7
3.7
Pack Capacity | Ah|:
120
120
Pack Nominal Voltage |V|:
318.2
329.3
Pack Energy |kWh|:
38.2
39.5
Pack Weight | kg |:
208.7
215.9
Max. SOC |%|:
90
90
Min. SOC |%|:
15
15
Useable Energy |kWh|:
28.6
29.6
Table 3-31: Modeled Fuel Economy and CO2 emissions comparison between the LDT4
Series-Parallel REET and 2021 Ford F-150 using Tier 3 regular-grade fuel.
Drive ( Yele 2021 l ord 1150 (liuseline)
Class 2u RI'.I'.T Series-Parallel Results
l'ereentajie eliange from baseline**
Simulation Results*
UDDS
ETW
[lbs.]
: 5.500 :
FE
[mpg]
18.8 i
C< >2
[g/mi]
455
ETW"
[lbs.]
; 6.500 :
FE^
[mpg]
34.0
C02^
[g mi]
: 252
; EPA Compliance i
C02n
[g/mi]
n/a
FE^ ;
[°o]
81°o
C< >2V
[°°]
-45° o
EPA Compliance
C02n
[°o]
n/a
HWFET
: 5.500
24.8
345
6.500
31.2
274
n/a
26° o
-21°o
n/a
Combined
: 5.500 :
21.5
398
6.500
32.7
261
73.3
52° o
-34° o
-82%
US06
5.500 i
15.1
566
6.500
22.4
382
n/a
48° o
-32° o
n/a
* Baseline simulation results were validated via chassis dynamometer testing. Simulation results are shown to provide a comparable basis of
comparison between the baseline configuration and REET configuration. Please refer to the final report for validation results
(Bhattacharjya,S., et al. 2023).
** A negative % C02 change means less emissions with respect to the baseline 2021 Ford F-150.
¦f Charge-sustaining operation only.
Takes into account a fiilly-phased-in (2031) fleet utility factor (FUF) 0.719 based on modeled all-electric range.
Table 3-32: Modeled Fuel Economy and CO2 emissions comparison between the LDT4 P2-
P4 REET and 2021 Ford F-150 using Tier 3 regular-grade fuel.
Drive C Yele
2021 l ord K150 (ISaseline)
Simulation Results
C ,'lsiss 2si Ri l l 1'2/lM Results
Percentile eliange from
baseline**
UDDS
ETW"
[lbs.]
; 5.500 :
FE
[mpg]
18.8
C< >2
[g/mi]
455
ETW"
[lbs.]
: 6.500
FE^
[mpg]
32.3
! C02^
[g mi]
265
EPA Compliance
C02n
[g/mi]
n/a
: FE^
[°o]
72° o
C< >2V
[°°]
-42° o
; EPA Compliance
C02n
[°o]
n/a
HWFET
; 5.500
24.8
345
i 6.500
29.1
294
n/a
18°o
-15°o
n/a
Combined
5.500
21.5
398
6.500
30.9
277
84.4
44° o
-30° o
-79%
US06
5.500
15.1
566
6.500
21.5
399
n/a
42° o
-30° o
n/a
* Baseline simulation results were validated via chassis dynamometer testing. Simulation results are shown to provide a comparable basis of
comparison between the baseline configuration and REET configuration. Please refer to the final report for validation results (Bhattachaijva,S.,
etal. 2023).
** A negative % C02 change means less emissions with respect to the baseline 2021 Ford F-150.
¦f Charge-sustaining operation onlv.
•j"f Takes into account a fiilh pi d n (2031) fleet utility factor (FUF) 0.695 based on modeled all-electric range.
3-139
-------
Modeling results for the LDT4 REET with a series-parallel architecture during charge
sustaining operation showed CO2 emissions reductions of approximately 45% on the city cycle
and 21% on the highway cycle relative to a 2021 Ford F-150 equipped with a 3.5L turbocharged
GDI engine. Modeling results for a P2-P4 REET showed CO2 emissions reductions of
approximately 42% on the city cycle and 15% on the highway cycle. Both REET architectures
reduced CO2 emissions by 30% or greater during the high speed, aggressive driving represented
by the US06 cycle. When accounting for all-electric driving using the fully-phased-in MY 2031
fleet utility factor (see Section III.C.8.i of the preamble) over combined urban and highway
operation, modeling results showed the potential for both REET architectures to reduce CO2
emissions by approximately 80%. When taking into account the fleet utility factor, CO2
emissions for the series-parallel REET were approximately 13% lower than that of the P2-P4
REET.
Simulated performance at ETW and during towing at GCWR (approximately 19,500 pounds)
are summarized in Table 3-33 and Table 3-34. Modeling of the SAE J2807 towing criteria,
which includes towing at GCWR up the Davis Dam grade, indicates that both LDT4 REET
versions would complete all of the SAE J2807 tests at approximately 19,500 pounds GCWR
provided that the engine is on and under either blended or charge sustaining operation.
Table 3-33: LDT4 REET 0-60 mph acceleration performance at ETW compared to 2021
Ford F150.
Drive cycle F150 LDT4 REET LDT4 REET
Scries- P2-P4
Parallel
ETW | lbs. | 5.500 6.500 6.500
0-60 mph 5.3 4.7 6.7
Charge Sustain |see|
0-60 mph - 5.3 6.7
All-Electric |sec|
0 - 60 mph - 4.4 6.7
Blended |sec|
Top Speed |mph| 105 100 100
Table 3-34: SAE J2807 modeling results for LDT4 REET at GCWR.
Performance
l'crformanc
Minimum
LI)T4 Ri l l Series-Parallel
LI)T4 Ri l l P2-P4 1)111.1.limine
Attribute
c Metric a
.12807
Vehicle: 6592 Ihs. Trailer: 13,000 Ihs.
Vehicle: 6476 Ihs. Frailer: 13,000 Ihs.
(JCWU
Requirement
Total: 19,592 Ihs.
Total: 19,476 Ihs.
Blended
Charge Sustaining
Blended
Charge Sustaining
Level Road
0-60 mph
30
14.0
" 16.5
20.9
" 21.1
Acceleration
Time (Sees.)
Level Road
0-30 mph
12
5.2
5.2
5.0
5.0
Acceleration
i Time (Sees.)
Level Road
: 40-60 mph
18
6.4
9.1
12.6
12.6
Acceleration
i Time (Sees.)
Launch on Grade
12° o Grade.
Fwd.
-
Pass
Pass
Pass
Pass
Highway
Speed on
40
Pass
Pass
Pass
Pass
Gradeability
grade (mph)*
(Davis Dam grade)
* This study adopted highway gradeability requirements exceeding those of the J2807 minimum requirements. Thus, achieving a "pass" rating
for highway gradeabilit\ d' t s maintaining a minimum speed of 45 mph in sections with a 45-mph speed limit and 55 mph in sections with
speed limits of 55 mph or 65 mph sections for the Davis Dam grade route.
3-140
-------
The results demonstrate that significant CO2 reductions are possible for vehicles with very
high tow capability using REET powertrain architectures. Within this study, the series-parallel
architecture had lower CO2 and improved performance characteristics when compared to the
P2/P4 architecture, with modeled GHG compliance of approximately 73 g/mi CO2 for a LDT4
pickup with towing capability competitive within its weight class. Series-parallel PHEV
configurations from this study were thus used to inform component costs and emissions in
OMEGA and were scaled for high tow capacity vehicles from the LDT3 through MDPV vehicle
categories.
3.5.3.1.2 MDV Range-extended Electric Truck REET
The following targeted design criteria for the MDV REET simulation were established:
• Battery sizing sufficient for a 75-mile UDDS all-electric range consistent with
California ACT NZEV PHEV definition (State of California, Air Resources Board
2021)
• 0-60 mph at ETW (6,500 pounds) equivalent to baseline 2021 RAM 250036
• 0-60 mph acceleration during towing at approximately 29,500 pounds GCWR of less
than 30 seconds as per SAE J2807, 4.3.1; and SAE J1491, 4.3.2 and 4.3.3.
• 0-30 mph acceleration during towing at approximately 29,500 pounds GCWR of less
than 12 seconds as per SAE J2807, 4.3.1; and SAE J1491, 4.3.2 and 4.3.3.
• 40-60 mph acceleration during towing at approximately 29,500 pounds GCWR of less
than 18 seconds as per SAE J2807, 4.3.1; and SAE J1491, 4.3.2 and 4.3.3.
• Launch on 12% grade at 29,500 pounds GCWR as per SAE J2807, 4.3.4.
• Highway gradeability (commonly referred to as the Davis Dam Grade) during towing
at approximately 29,500 pounds GCWR that maintained a minimum speed of 45 mph
in sections with a 45-mph speed limit and 55 mph in sections with speed limits of 55
mph or 65 mph sections as per SAE J2807, 4.3.5.
Only the P2-P4 architecture was considered for MDV application. The other architectures,
namely P0, PI, P2, and power-split were ruled out due to sizing constraints and power/torque
limitations. The series-parallel architectures (with the exception of power-split) could potentially
be used for MDV applications but would require significantly upsized traction motors and
generator. At very high loads, e.g., during towing acceleration, the series mode energy
conversion would also be less efficient than a conventional vehicle with a high efficiency
gearbox, or the P2-P4 architecture considered selected within the study. Simulations were
conducted with both gasoline and diesel dedicated hybrid engines. The gasoline version was
36 Note that the 2021 RAM 2500 baseline vehicle had an ALVW of 9,000 pounds. The MDV REET configurations
both had ALVW of 9,500 pounds.
3-141
-------
modeled using a dedicated-hybrid 6.0L Miller cycle engine with 300 kW rated power. The diesel
version was modeled using a 4.0L dedicated hybrid engine with 284 kW rated power.
For a summary of the hybrid drive-system and battery specifications, please refer toTable
3-35 and Table 3-36. For more details regarding the drive systems and the dedicated hybrid
engines, please refer to the final report. (Bhattacharjya,S., et al. 2023). The electric machines
used in both system configurations were internal-permanent-magnet synchronous machines
using SiC inverters. Summaries of CO2 emissions and fuel economy (FE) simulation results over
the UDDS, HWFET, combined cycle results, and US06 are in Table 3-37 and Table 3-38.
Table 3-35: Hybrid drive system specifications used for MDV REET simulations.
Electric Motors Max. Power Max. Max. Speed Base Speed Gear Ratio*,
(kW) Torque (N- (rpm) (rpm) f,
m)
P2 150 1036 6.000 1.500 3.73:1
Front E-a\lc / P4 195 405 18.000 4.200 17:1
*P2 gear ratio is for the final drive. Note that the P2 also incorporates an 8-speed automatic transmission prior to the final
drive.
¦f P4 and series parallel use a single final-drive gear ratio.
Table 3-36: Battery specifications used for MDV REET simulations.
Parameter P2-P4 Gasoline and
Diesel versions
Cells in Series 82
Cells in Parallel 4
Cell Capacity | Ah| 40
Cell Nominal Voltage |V| 3.7
Pack Capacity | Ah| 160
Pack Nominal Voltage | V| 303.4
Pack Energy |kWh| 48.5
Pack Weight |kg| 265.3
Max. SOC |%| 90
Min. SOC |%| 15
Useable Energy |kWh| 36.37
3-142
-------
Table 3-37: Modeled Fuel Economy and CO2 emissions comparison between the MDV P2-
P4 REET with a gasoline DHE and a 2021 RAM 2500 Diesel.
Drive 2021 RAM 2500 (buseline) (,'lsiss 2b 1'2-1'4 w/Giis»line DIM. Results % ('hunge from ISuseline*
C ,'ycle Simulation Results
j ALMV
FE
C< >2
[g/mi]
ALMV
FE
a >2
[g/mi]
EPA
1 FE :
C< >2
[%]
EPA
[pounds]
[mpg]
[pounds]
: [mPg]
Compliance?
: [%]
Compliance
C02
[g/mi]
C02
l""l
UDDS
r 9,000
: 17.5
581
9,500
; 25.6 :
334
n/a
T 46% ;
-42%
n/a
HWFET
9,000
; 25.9
393
9,500
! 23.7'" 1
361
n/a
V -8.6%
-8.1%
n/a
Combined
9,000
; 21.3
478
9,500
; 24.7 i
346
107
16%
-28%
1 j -78%" ^
US06
9,000
I 16.7
609
9,500
1 17.8 i
480
n/a
1 " 7% '
""-21%
n/a
* Baseline simulation results were validated via chassis dynamometer testing. Simulation results are shown to provide a comparable basis of
comparison between the baseline configuration and REET configuration. Please refer to the final report for validation results
(Bhattachaijya,S., et al. 2023).
** A negative % C02 change means less emissions with respect to the baseline 2021 RAM 2500.
¦f Charge-sustaining operation onlv.
Takes into account a folly-phased-in (2031) fl t 11 ts factor (FUF) 0.691 based on modeled all-electric range.
Table 3-38: Modeled Fuel Economy and CO2 emissions comparison between the MDV P2-
P4 REET with a diesel DHE and a 2021 RAM 2500 Diesel.
Drive C Vcle 2021 RAM 2500 (buseline) (,'lsiss 2b 1'2-1'4 w/Diesel DIM. Results % ( 'hunge from ISiiseline*
Simulation Results
ALMV
FE
C< >2
[g mi]
ALVW
FE
C< >2
[g mi]
EPA
; FE
C< >2
[°o]
EPA
[pounds]
[mpg]
[pounds]
[mpg]
Compliance?
C()2
[g/mi]
[°o]
Complian
C< >2
r°]
UDDS
9.000
17.5
581
9.500
i 31.2
327
n/a
78%
-44%
n/a
HWFET
9.000
25.9
393
9.500
! 28.9
352
n/a
12%
-11%
n/a
Combined
9.000
21.3
478
9.500
30.2
337
104
42%
-29%
-78%
US06
9.000
16.7
609
9.500
: 21.5
473
n/a
29%
-22%
n/a
* Baseline simulation results were validated via chassis dynamometer testing. Simulation results are shown to provide a comparable basis of
: comparison between the baseline configuration and REET configuration. Please refer to the final report for validation results
(Bhattacharjya,S., et al. 2023).
i ** A negative % C02 change means less emissions with respect to the baseline 2021 RAM 2500.
; f Charge-sustaining operation only.
Takes into account a folly-phased-in (2031) fleet utility factor (FUF) 0.691 based on modeled all-electric range.
Both MDV REET versions offered significant potential for CO2 emissions reduction relative
to the baseline 2021 RAM 2500 diesel. Modeling of charge sustaining operation of the MDV
REET using the gasoline dedicated hybrid engine showed CO2 emissions reductions of
approximately 42 percent on the city cycle and 7 percent on the highway cycle. Modeling of
charge sustaining operation of the MDV REET using the diesel dedicated hybrid engine showed
CO2 emissions reductions of approximately 43 percent on the city cycle and 10 percent on the
highway cycle. Both versions reduced CO2 emissions by approximately 20 percent during charge
sustaining operation and aggressive driving represented by the US06 cycle. When accounting for
all-electric driving using the fully-phased-in MY 2031 fleet utility factor (see Section III.C.8.i of
the preamble) over combined urban and highway operation, modeling results showed the
potential for REET with either version of dedicated hybrid engine to reduce CO2 emissions by
approximately 78%.
Simulated performance at ALVW and during towing at GCWR (approximately 29,500
pounds) are summarized in Table 3-39 and Table 3-40. During modeling of SAE J2807 towing
conditions, which include towing at GCWR on the Davis Dam grade, results indicate that both
MDV REET versions would successfully complete all the J2807 tests at approximately 29,500
3-143
-------
pounds GCWR provided the engine is on and under either blended or charge sustaining
operation.
Table 3-39: Modeled 0-60 mph performance results at ALVW for the 2021 RAM 2500 and
both Class 2b P2-P4 REET configurations.
Drive cycle
RAM 2500
Class 2b
Class 2b
Diesel
REET
REET w/diescl
(baseline)
w/jjasolinc
DHE
DHE
ALVW (pounds)
9,000
9.500
9.500
-60 mph Charge Sustaining |scc|
8*
7.3
7.4
0-60 mph
-
7.2
7.2
All-Elcclric |see|
0-60 mph
-
7.3
7.3
Blended | sec |
Top Speed |mph|
120*
100
100
* Simulation results
Table 3-40: SAE J2807 modeling results for MDV REET.
Performance Attribute
Performance
J2807
Gasoline DHE Engine
Diesel DHE Engine
Metric
Requirement
Vehicle: 9,169 pounds
Vehicle: 9096 pounds
Trailer: 19,980 pounds
Trailer: 19,980 pounds
Total: 29,149 pounds
Total: 29,076 pounds
lilendcd (,'liargc
lilendcd Charge
Sustaining
Sustaining
Level Road Acceleration
0-60 mph
30
22.3 22.3
' 22.1 22.1
; Time (Sees.)
Level Road Acceleration
0-30 mph
12
5.5 j 5.5 ;
5.5 5.5
j Time (Sees.)
Level Road Acceleration
: 40-60 mph
18
13 13
12.2 12.3
' Time (Sees.)
Launch on Grade
] 12% Grade
Yes
Pass Pass
Pass Pass
Highway Gradeabilitv* (Davis
Maintain 55
; Min. 40 mph i
Pass Pass
Pass Pass
Dam)
mph
i * This study used a more stringent criteria of maintaining 55 mph over Davis Dam instead of the SAE J2807 minimum of 40 mph.
The modeling results demonstrate that significant CO2 reductions are possible for MDV with
very high tow capability using REET powertrain architectures. The REET modeled with the
gasoline dedicated hybrid engine obtained CO2 emissions during charge sustaining operation that
were within approximately 1.5 to 2.7 percent of that with the diesel dedicated hybrid engine.
When accounting for the fleet utility factor, the modeled GHG compliance of the REET with the
gasoline dedicated hybrid engine was approximately 107 g/mi and thus comparable to the
version with the diesel dedicated hybrid engine. This was due to the higher carbon content of
diesel fuel relative to gasoline and the ability of the gasoline dedicated hybrid engine to approach
diesel efficiency over the drive cycles when used as part of an electric hybrid drive system. Due
to the comparable CO2 emissions between the diesel and gasoline versions and the lower cost of
the gasoline dedicated hybrid engine and its associated exhaust emissions control system
compared to the diesel, the gasoline dedicated hybrid engine version was used for informing
component costs and emissions in OMEGA and was scaled to represent high tow capacity PHEV
MDVs.
3-144
-------
3.5.4 Other Vehicle Technologies
Depending on vehicle design, other vehicle technologies such as transmissions, non-hybrid
stop-start, electrified power steering, accessories, secondary axle disconnect, low drag brakes,
and air conditioning may have been used in the creation of ALPHA outputs for the Response
Surface Equations (RSE's) used by OMEGA and described in chapter 2.4.10 of this RIA. These
other technologies were first discussed in the previous version of ALPHA used for the 2017
Final Determination (U.S. EPA 2017) and the modeling has not changed.
3.6 Vehicle Air Conditioning System Related Provisions
EPA has included air conditioning (A/C) system credits in its light-duty GHG program since
the initial program adopted in the 2010 rule. Although the use of A/C credits has been voluntary,
EPA in past rules has adjusted the level of the CO2 standards downward, making them more
stringent, to reflect the availability of the credits. Manufacturers opting not to use the A/C credits
meet the standards through additional CO2 reductions. EPA is revising the A/C credits program
for light-duty vehicles in two ways. First, for A/C system efficiency credits, as proposed, EPA is
limiting the eligibility for voluntary credits for tailpipe CO2 emissions control to ICE vehicles
starting in MY 2027 (i.e., BEVs would not earn A/C efficiency credits). Second, for A/C
refrigerant leakage control, EPA is phasing down the credit from MYs 2027-2030 and retaining a
small credit for MYs 2031 and later. EPA is retaining the refrigerant-related provisions
applicable to MDV standards.
3.6.1 A/C Leakage Credit
The level to which each technology can reduce leakage can be estimated using the September
2023 version of SAE J2727. While this standard was developed for leakage of HFC-134a
refrigerant, it is also applicable to the alternative refrigerant HFO-1234yf, and may be applicable
to other low-GWP refrigerants as well. To convert J2727 chart emission (leak) rates from HFC-
134a to HFO-1234yf leakage rates, the result is multiplied by 1.03. This conversion factor for
HFO-1234yf is derived by multiplying the ratio of the molecular weights of the two refrigerants
(114 kg/kmol for HFO-1234yf and 102 kg/kmol for HFC-134a) by the inverse ratio of the
dynamic viscosities of the two refrigerants (11.1 x 10-6 Pa s for HFC-134a and 12.0 x 10-6 Pa s
for HFO-1234yf).
The J2727 standard was developed by SAE and the cooperative industry and government
EVLAC (Improved Mobile Air Conditioning) program using industry experience, laboratory
testing of components and systems, and field data to establish a method for calculating leakage.
With refrigerant leakage rates as low as 10 g/yr, it would be exceedingly difficult to measure
such low levels in a test chamber (or shed). Since the J2727 method has been correlated to "mini-
shed," or SAE J2763, results (where A/C components are tested for leakage in a small chamber,
simulating real-world driving cycles), the EPA considers this method to be an appropriate
surrogate for vehicle testing of leakage. (SAE J2727 2023). It is also referenced by the California
3-145
-------
Air Resources Board in their Environmental Performance Label regulation and the State of
Minnesota in their GHG reporting regulation.37'38
3.6.2 How Will Leakage Credits Be Calculated?
For model years 2027 through 2030, the A/C credit available to manufacturers will be
calculated based on how much a particular vehicle's annual leakage value is reduced compared
to an average MY 2008 vintage vehicle with baseline levels of A/C leakage technology and will
be calculated using a method drawn directly from the September 2023 version of SAE J2727
(SAE J2727 2023) approach. By scoring the minimum leakage rate possible on the J2727
components enumerated in the rule (expressed as a measure of annual leakage), a manufacturer
can generate the maximum A/C credit (on a gram per mile basis). To avoid backsliding on
leakage rates when using low-GWP refrigerants, where manufacturers could choose less costly
sealing technologies and/or materials, EPA is finalizing the proposed disincentive credit for
"high leak" on alternative refrigerant systems. The maximum value for this high leak
disincentive credit (or HiLeakDisincentive) is 1.8 g/mile for cars and 2.1 g/mile for trucks, with
lower amounts possible for leakage rates between the minimum leakage score (MinScore) and
the average impact (Avglmpact). The terms used for calculating the A/C Leakage Credit as well
as the HiLeakDisincentive are discussed later in this section.
The A/C credit available to manufacturers will be calculated based on the reduction to a
vehicle's yearly leakage rate, using the following equation for a baseline refrigerant which has a
GWP of 150 for MY 2031 and later, or the HFC-134a refrigerant in the previous MY2017-2025
rule and earlier:
Equation 3-7: Credit Equation for a Baseline Refrigerant
A/C Leakage Credit = (MaxCredit) * [ 1 - (§86.166-12 Score / Avglmpact39) * (GWPRefrigerant / 1430))
and the following equation for low-GWP, alternative refrigerants:
Equation 3-8: Credit Equation for Alternative Refrigerants
A/C Leakage Credit = (MaxCredit) * [ 1 - (§86.166-12 Score / Avglmpact39) * (GWPRefrigerant / 1430)) -
HiLeakDisincentive
where the HiLeakDisincentive is determined in accordance with one of the following three
conditions, depending on the refrigerant capacity (RefrigCapacity), or charge level, of the A/C
system:
For A/C systems with a refrigerant capacity <= 733g:
HileakDis = 0.0, if Score <11.0 g/yr
37 State of California, Manufacturers Advisory Correspondence MAC #2009-01, "Implementation of the New Environmental Performance
Label," This document is available in Docket EPA-HQ-OAR-2009-0175.
38 State of Minnesota, "Reporting Leakage Rates of HFC-134a from Mobile Air Conditioners," This document is available in Docket EPA-
HQ-OAR-2009-0472-0178.
39 Section 86.166-12 sets out the individual component leakage values based on the SAE value
3-146
-------
HileakDis = Max HiLeakDisincentive * [(Score - 11) / 3.3], if 11.0 < Score < 14.3,
HileakDis = Max HiLeakDisincentive, if Score > 14.3
For A/C systems with a refrigerant capacity > 733g:
HileakDis = 0.0, if Score < RefrigCapacity * 0.015
HileakDis = Max HiLeakDisincentive * (Score - (RefrigCapacity * 0.015)/3.3), if RefrigCapacity * 0.015
< Score < RefrigCapacity * 0.015 + 3.3
HileakDis = Max HiLeakDisincentive, if Score > RefrigCapacity *0.015 + 3.3
For MY 2026 and later, Equation 3-9 is used to calculate A/C leakage credits of an alternative
refrigerant with GWP (GWPRefrigerant) at or below 150. The MaxCredit for MY 2031 and later
in Equation 3-9 is 1.6 g/mile cars and 2.0 g/mile trucks shown in Table 3-41.
Equation 3-9: Leakage Credit Equation for an Alternative Refrigerant
A/C Leakage Credit = (MaxCredit40) * (1 — GWPRefrigerant / 150) — HileakDis
There are four significant terms to the credit equation. Each is briefly summarized below and
is then explained more thoroughly in the following sections. Please note that the values of any of
these terms change depending on whether a 150-GWP refrigerant or an alternative refrigerant is
used. The values are shown in Table 3-42, and are documented in the following sections.
• "MaxCredit" is a term for the maximum amount of credit entered into the equation
before constraints are applied to terms. The maximum credits that could be generated
by a manufacturer is limited by the choice of refrigerant and by assumptions regarding
maximum achievable leakage reductions.
• "Score" is the leakage score (LeakScore) of the A/C system as measured and
calculated according to the 40 CFR 86.166-12 calculation in units of g/year, where the
minimum score which is deemed feasible is fixed.
• "Avglmpact" is a term which represents the annual average impact of A/C leakage.
• "MinScore" is the lowest leak score that EPA projects is possible, when starting from
a baseline, or Avglmpact, system. The MinScore represents a 50% reduction in
leakage from the baseline levels based on the feasibility analysis detailed below.
• "GWPRefrigerant" is the global warming potential for direct radiative forcing of the
refrigerant as defined by EPA (or IPCC). The GWP values of a legally usable
refrigerant by the AIM act must be less than 150.
• "HiLeakDisincentive" is a "HiLeakDis" term for the disincentive credit deducted for
low-GWP alternative refrigerant systems which have a leakage rate greater than the
40 A/C MaxCredits of model year 2026 and later in Equation 3-9 are shown in Table 3-41.
3-147
-------
minimum leakage score of 11.0 g/year for cars and trucks. The maximum Disincentive
(Max HiLeakDisincentive) is 1.8 g/mile for cars and 2.1 g/mile for trucks.
• Detailed descriptions of Max Credit Term, § 86.166-12 implementing the J2727 Score
Term, Avglmpact Term, and GWPRefrigerant Term are presented in MY2017-2025
final rulemaking. (U.S. EPA 2012).
Table 3-41: A/C maximum leakage credits (MaxCredit) available to manufacturers, final
program (CO2 g/mile).
MY
Car
Truck
2026
13.8
17.2
2027
11.0
13.8
2028
8.3
10.3
2029
5.5
6.9
2030
2.8
3.4
2031
1.6
2.0
2032 and later
1.6
2.0
Table 3-42: A/C Component Credits /w SAE J2727-2023 default parameter settings.
Baseline Refrigerant (GWP
= 150)
Lowest-GWP
Refrigerant (GWP=1)
Cars
Trucks
Cars
Trucks
MaxCredit equation input (grams/mile CO2 EQ)
12.6
15.6
13.8
17.2
A/C credit maximum (grams/mile CO2 EQ)41
11.9
14.8
13.8
17.2
§86.166-12 MinScore (grams HFC/year)42
8.3
10.4
8.3
10.4
Avg Impact (grams HFC/year)
16.6
20.7
16.6
20.7
A small A/C leakage credit for MY 2031 and later is described in detail below. The standard
default parameter settings in the "Examples & Instructions" worksheet of the September 2023
version of SAE J2727 standard were used to calculate the "§86.166-12 Score" for the lowest-
GWP HFO-1234yf alternative refrigerant. The 10.4 g/year of the "§86.166-12" Score was
calculated using the "HFO-1234yf Belt Driven Compressor" worksheet in the September 2023
version of SAE J2727.
As shown in Table 3-42, 14.8 g/mile of trucks A/C credit maximum for the 150-GWP
baseline refrigerant are calculated by plugging 15.6 g/mile MaxCredit, the calculated 10.4 g/year
"§86.166-12" Scores, 20.7 g/year Avglmpact, and 150 GWPRefrigerant values into Equation
3-7. The 11.9 g/mile cars A/C credit maximum are multiplied the 14.8 g/mile trucks A/C credit
maximum by the cars/trucks Avglmpact ratios.
41 With a hermetically-sealed electric compressor, value increases to 12.4 and 15.4 for cars and trucks, respectively.
42 With a hermetically-sealed electric compressor, threshold value decreases to 2.4 and 3.1 for cars and trucks,
respectively.
3-148
-------
The "HiLeakDisincentive" values by the 10.4 g/year "§86.166-12" Scores become zero since
the calculated "§86.166-12" Scores are less than 11 g/year of the "HiLeakDisincentive"
threshold values.
The 17.2 g/mile MaxCredit, the calculated 10.4 g/year "§86.166-12" Scores, 20.7 g/year
Avglmpact, and 1 GWPRefrigerant values were plugged into Equation 3-8 for calculating the
trucks A/C credit maximum of the lowest-GWP HFO-1234yf refrigerant. 17.2 g/mile trucks A/C
credit maximum of a lowest-GWP HFO-1234yf alternative refrigerant are calculated using either
a belt-driven or an electric compressor since the "1/1430" term values are so small.
A/C leakage credit incentives of the baseline 150-GWP refrigerant are reset to the "0" since
the GWP values less than 150 are required by the AIM Act. As shown in Table 3-43, 2.4 g/mile
truck leakage credits of the lowest-GWP HFO-1234yf refrigerant are the trucks A/C credit
maximum relative differences between 17.2 g/mile of the lowest-GWP HFO-1234yf refrigerant
and 14.8 g/mile of the baseline 150-GWP refrigerant. The 1.9 g/mile cars leakage credits of the
lowest-GWP HFO-1234yf refrigerant are multiplied the 2.4 g/mile trucks leakage credits by the
"16.6/20.7" cars/trucks Avglmpact ratios. The A/C component credit calculation details, SAE
J2727 standard default parameter settings, the "HFO-1234yf Belt Driven Compressor", and the
"HFO-1234yf Electric Compressor" worksheet screenshots are presented (SAE J2727 Worksheet
Screenshots 2023).
Similarly, 1.4 g/year cars and 1.8 g/year trucks leakage credits for the lowest-GWP HFO-
1234yf refrigerant electric compressor are calculated using the "R-1234yf Electric Compressor"
worksheet in the September 2023 version of SAE J2727 standard.
About 1.4 g/mile cars and 1.8 g/mile trucks leakage credits of the lowest-GWP HFO-1234yf
refrigerant are calculated by simply scaling 13.8 g/mile cars and 17.2 g/mile trucks leakage
credits multiplied by the "150/1430" GWP ratios since the GWP values of the baseline
refrigerant were decreased from 1430 to 150.
Table 3-43: A/C Leakage Credits (MaxCredit) of the lowest-GWP refrigerant.
Baseline Refrigerant (GWP = 150),
Highest GWP limit
Lowest-GWP Refrigerant
(GWP=1)
Cars
Trucks
Cars
Trucks
A Simple Scaling
0
0
1.4
1.8
SAE J2727 /w an electric compressor
0
0
1.4
1.8
SAE J2727 /w a belt-driven compressor
0
0
1.9
2.4
Averaged A/C Leakage Credits
0
0
1.6
2.0
The 1.6 g/mile cars and 2.0 g/mile trucks leakage credits are also calculated by averaging the
lowest-GWP HFO-1234yf refrigerant A/C leakage credit values in the simple scaling, the
electric-compressor, and the belt-driven compressor cells in Table 3-43. Hermetically-sealed
electric compressors (Denso 2022) are widely used in the vehicle A/C systems of electrified
vehicles like various HEVs, PHEVs, and BEVs.
The HFC-152a refrigerant is the only currently available alternative refrigerant with 124-
GWP values, closer to the maximum lawful GWP limit of 150. But the flammable HFC-152a
refrigerant requires about 12 pounds of additional system components (chiller, pump, reservoir,
and plumbing for secondary loop), a new challenging packaging design, and engineering cost. As
3-149
-------
of MY2023, the lowest-GWP HFO-1234yf refrigerant already used in 97 percent of light-duty
vehicles (U.S. EPA 2023) is likely prevalent even without extra leakage credit incentives.
Therefore, 1.6 g/mile cars and 2.0 g/mile trucks leakage credits are appropriate since about
149 GWP value differences are insignificant compared to 1429 GWP differences by the HFC-
134a refrigerant, and the MaxCredit values of the baseline 150-GWP refrigerant are likely
increased above 15.6 g/mile, which were used for the HFC-134a refrigerant in the previous
MY2012-2016 and MY2017-2025 rules.
3.7 Fuel Economy Test Procedure Adjustments for Tier 3 Test Fuel
In order to respond to the need for test procedure adjustments due to the change to Tier 3
certification test fuel, EPA conducted a test program at EPA's National Vehicle and Fuel
Emissions Laboratory to quantify the differences in GHG emissions and fuel economy between
Tier 2 and Tier 3 certification test fuels. The peer-reviewed Technical Report titled "Tier 3
Certification Fuel Impacts Program" (U.S. EPA 2018) contains the details of the study design,
how we conducted the testing, and our analysis of the results, and is available in the docket for
this rule.
This section will first summarize the study design and data analysis, then the determination of
the adjusted R-factor and revised fuel economy equation.
3.7.1 Summary of EPA Test Program and Results
EPA designed the study to test vehicles that incorporated a variety of advanced powertrain
technologies that already have a significant and increasing presence in the market today and are
expected to be among the primary technologies applied by manufacturers to meet future GHG
and fuel economy standards. Our selection of vehicles for the test program was designed to
address the narrow purpose of this rule: quantifying appropriate CO2 and CAFE adjustments that
on average would prevent the change in the stringency of those standards that would otherwise
occur as the certification test fuel changed. We note that because it was necessary in this case for
EPA to estimate test fuel effects into future years, we were not able to base our vehicle selection
solely on the vehicle fleet as it currently exists. In other words, it was critical that the agency
select vehicles equipped with technologies that represent how the fleet will look in the future
(rather than how the fleet looks today).
To capture the emission and fuel economy effects with the technologies that are becoming
widespread in the fleet, we concluded that it was important to cover a wide range of engine
configurations and cylinder displacements, and related technologies. We intentionally focused on
specific technologies that we expect manufacturers to widely use in future vehicles, instead of on
specific vehicles, for two reasons: 1) Fuel effects on GHG emissions and fuel economy relate
primarily to combustion characteristics of the engine, rather than to vehicle characteristics (e.g.,
mass and aerodynamics); and 2) while we are reasonably certain that the technologies we
selected and tested will dominate the light-duty fleet in coming years, the distribution of specific
vehicles in which they will be used over the 2025 and later time period is much more difficult to
anticipate. EPA believes that the appropriateness of focusing our test vehicle selection on key
engine and powertrain technologies is further reinforced by the long-standing practice by most
manufacturers of using a single engine type in several different models of passenger cars, cross-
overs, SUVs, minivans, and/or pick-up trucks.
3-150
-------
Table 3-45 below lists the powertrain technologies that EPA selected, after a series of
technical consultation meetings with the Alliance and Global Automakers.43 The selected
vehicles cover 4-, 6-, and 8-cylinder engines, and a wide range of displacements per cylinder
(ranging from 0.375 to 0.75 liters of displacement per cylinder). In addition, EPA's selected
engines included both naturally aspirated and turbocharged engines and both direct-injection and
port-injection fuel delivery systems.44 Because these engine characteristics largely determine the
dynamics of fuel combustion, they are closely related to emissions and efficiency when test fuel
changes. We also included newer transmission technologies to reveal any potential effects
beyond the engine. Several of these engine and transmission technologies are in widespread use
today, and we expect the others to become more prevalent as future GHG, CAFE, and Tier 3
standards take effect.
As illustrated in the 2018 EPA Automotive Trends Report, the use of the key technologies
incorporated in the EPA test program is growing in a wide range of vehicle applications across
the industry, at the same time that earlier competing technologies are generally declining (U.S.
EPA 2023) 45
We chose eleven vehicles that incorporated one or more of these relevant technologies,
including the following: gasoline direct injection (GDI) (which enables higher compression
ratios for improved fuel efficiency and emissions reductions); engine turbocharging (generally in
conjunction with smaller, more efficient engines, another growing approach to improved fuel
efficiency and reduced emissions); naturally aspirated high compression engines (featuring a
high degree of valve timing authority to allow operation as Atkinson-Cycle engines when
required); cylinder deactivation technology (to allow one or more cylinders to be deactivated
while the vehicle is cruising, reducing fuel consumption and emissions); and automatic
transmissions with higher numbers of gears, as well as Continuously Variable Transmissions
(CVTs), to allow engines to stay in the most efficient engine speed range as much as possible,
improving fuel use and emissions. The test program also included a large pickup truck, a Class
2b MDV, to assess whether larger gasoline trucks with engine technology that is common today
and is likely to continue into the future show similar effects to LDVs and LDTs.46
The use of these technologies has been growing, and we expect them to continue to grow. For
example, between 2008 and 2018, in the new model year fleet:
Gasoline direct injection (GDI) penetration has grown from 2% to 51%.
43 See EPA Memorandum to Docket EPA-HQ-OAR-2016-0604: "Listing of Technical Consultation Meetings
between EPA Staff and Automobile Industry Technical Representatives Supporting the Vehicle Test Procedure
Adjustments for Tier 3 Certification Test Fuel," NPRM. Among other topics, these meetings included detailed
discussions of vehicle selection and test methodology issues for the EPA vehicle test program underway at the time.
44 EPA did not include electric hybrid powertrains in the test program because the additional test variability caused
by differences in battery state of charge and engine on/off operation would likely confound the small fuel effects.
45 The 2018 EPA Automotive Trends Report describes in detail the most recent trends among powertrain
technologies, beginning at P. 37: https://www.epa.gov/automotive-trends/download-automotive-trends-
report#Full%20Report
46 As discussed above, EPA regulates Class 2b (and Class 3) heavy-duty vehicles, which have gross vehicle weight
ratings greater than 14,000 pounds, separately from light-duty vehicles, but the 2014 Tier 3 certification test fuel
changes applied to testing for both of these vehicle categories.
3-151
-------
Gasoline engine turbocharging has grown from 3% to 31%.
Cylinder deactivation has grown from 7% to 12%.
8-speed transmissions have grown from 0.2% to 19%.
Continuously Variable Transmissions (CVTs) have grown from 6% to 20%.
The vehicles we selected for the test program were production vehicles that had emission
levels that were compliant or nearly compliant with the Tier 3 emission standards. All of the
vehicles we tested for this program were certified by the manufacturers to operate appropriately
on regular grade fuel to avoid any potential octane effects from the test fuel change (i.e., from
higher-octane Tier 2 test fuel to lower-octane Tier 3 test fuel).
Some stakeholders have asked EPA to consider using the manufacturer-generated test data
that they submit to the EPA vehicle certification database as an alternative data source for
estimating the impact of the change in CO2 and fuel economy performance due to the test fuel
change, rather than the data from the separate EPA vehicle test program.47 In fact, early in the
development of this adjustment, EPA considered the potential value of using available
manufacturer certification data for this purpose of quantifying the impact of the test fuel change.
However, EPA concluded that the manufacturer certification data submitted to EPA could not be
used for the purpose of the technical analysis needed for this rule. As shown in Table 3-44
below, EPA recognizes that there are many sources of vehicle test-to-test variability, and we
have developed methodologies to control for these sources of variability for this test program.
EPA's testing methodologies were informed by our experience with the challenges of measuring
fuel effects on vehicle emission performance. EPA concluded that it is not possible to use
manufacturer certification data, as submitted to EPA, to quantify the effects of the Tier 3 fuel
change on CO2 and fuel economy. This is why EPA instead designed a targeted, controlled test
program for the particular purposes of this rule.
In performing the testing of the selected vehicles, we took additional steps beyond those
specified in the existing compliance testing regulations in order to reduce test-to-test variability
to very low levels. This was necessary because we were working to discern very small changes
in emissions and fuel economy between tests on the two fuels, requiring lower test-to-test
variability than has been historically accepted for such testing, including compliance testing.48
We accomplished this goal in several ways, in general by reducing or eliminating potential
sources of variability. These steps included completing testing of one vehicle on one fuel in a
single work week; maintaining the same test site and vehicle driver throughout the program
across all fuels and vehicles; thorough removal of the previous test fuel from the fuel system,
with enough driving to allow for the engine to adapt to the new fuel properties; maintaining the
same number and type of test, and the same sequence, during each day of testing; and ensuring a
fully-charged battery by using a trickle-charger overnight, over weekends, and over extended
47 See briefing document provided by the Alliance of Automobile Manufacturers for E.O. 12866 meeting May 28,
2019, EPA Docket EPA-HQ-OAR-2016-0604.
48 For example, EPA historically allows up to a three percent difference in fuel economy from test to test when
performing engineering evaluations. Guidance document VPCD-97-01 for testing vehicles with knock sensors
highlights this existing variability allowance.
3-152
-------
periods between tests. By taking these kinds of actions, we were able to reduce test-to-test
variability significantly as compared to most routine testing on these test cycles.
Table 3-44 lists several of the key features of vehicle testing that affect the variability of test
results and that we specifically incorporated into the EPA vehicle test program. As shown, these
methodological features are typically not present during manufacturer certification testing (nor
are necessary for the accuracy required for that purpose).
Table 3-44: Test Variables Requiring Control for Accurate Fuel Effects Measurement.
Methodological Failures EPA Test Program Available Manufacturer
Certification Data
Identical test fuels across all test vehicles Yes No
Appropriate methods for measuring Tier 3 Yes Rarely
(oxygenated) test fuel properties
Multiple measurements of test fuel properties Yes No
across several labs/samples
Comparative testing done in same test cell (to Yes Rarely
minimize impacts from vehicle loading and
coast-down simulation, etc.)
Testing using same driver Yes No
Testing using exact same test vehicle for all Yes Rarely
testing of a vehicle model
Careful control of vehicle preparation to Yes No
reduce variability (beyond CFR requirements)
Statistical assessment of number of test Yes No
replicates needed
Monitoring driver performance metrics for Yes No
consistency with comparative tests
Highly controlled sequencing of test types Yes No
(FTP. HFET. US06)
Fuel sequence order switched to avoid vehicle Yes No
"learning bias"
Repeat of test sequences when necessary for Yes No
statistical confidence
Table 3-45 lists the test vehicles EPA used in this test program and the key technologies they
incorporated. We note that the EPA test program and the associated Technical Report only
evaluated the change in carbon-balance fuel economy between the two test fuels, not changes in
CAFE calculations. However, these data serve as a basis for developing the CAFE fuel economy
adjustment factor described below.
3-153
-------
Table 3-45: Test Program Vehicles.
Model Year
Vehicle Make/Model
Engine
Technologies
2014
Ram 1500
3.6L V6 PFI
8 speed automatic transmission.
start-stop disabled
2016
Acura ILX
2.4L 14 GDI
8 speed DCT with a torque
converter
2013
Nissan Allima
2.5L 14 PFI
CVT
2016
Honda Civic
1.5LI4GDI
CVT. downsized turbocharged
engine
2015
Ford F150 Eco-Boosl
2.7L V6 GDI
Downsized turbocharged engine.
start-stop disabled
2013
Chevrolet Malibu ("Malibu 1")
2.4L 14 GDI
Gasoline direct injection engine
2016
Chevrolet Malibu ("Malibu 2")
1.5LI4GDI
Downsized turbocharged engine
2014
Ma/da 3
2.0L 14 GDI
High compression ratio engine
2014
Chevrolet Silverado 1500
4.3L V6 GDI
Cylinder deactivation
2015
Volvo S60 T5
2.0L 14 GDI
Downsized turbocharged engine
2016
Chevrolet Silverado 2500
6.0L V8 PFI
Class 2b truck
3.7.1.1 Discussion of Results
The EPA test program described above generated a set of high-quality vehicle emissions data,
which then also served as inputs to the carbon-balance fuel-economy equation, on each of the
two fuels of interest. The associated Technical Report referenced above includes a
comprehensive summary and comparison of these data. We refer stakeholders interested in a
fuller presentation of the entire program to the Technical Report.
The Technical Report, as a comprehensive presentation of EPA test program and its results, is
independent of this rule and will likely be valuable in other contexts. Much of the data collected
in the test program and presented in the Technical Report is relevant to the development of the
adjustment factors, as described below. However, the report does not present the resulting
adjustment factors or the analyses leading to them.
In summary, Figure 3-75 shows the average percent change in CO2 emissions by vehicle,
calculated with respect to the Tier 2 fuel (or mathematically: % difference = (T3 - T2)/T2 x
100). The results indicate that for the Federal Test Procedure (FTP) and the Highway Fuel
Economy Test (HFET) cycles, going from Tier 2 fuel to Tier 3 fuel results in a reduction in CO2
per mile of 1.78 and 1.02 percent, respectively, corresponding to absolute CO2 emissions
decreases of 6.37 and 2.16 g/mi, respectively.49 Vehicles which emitted comparatively large
amounts of CO2 on Tier 2 fuel generally showed larger reductions in absolute CO2 emissions
when moving from Tier 2 fuel to Tier 3 fuel. However, these vehicles produced similar
reductions to the other vehicles in the test program when expressed as a percent reduction,
indicating a consistent effect proportional to the base vehicle performance of the test vehicle. In
our view, stringency under GHG and CAFE standards relates to this base performance, rather
than absolute CO2 emissions levels. As market representative test fuel mixes become more
efficient, it becomes comparatively easier for comparatively inefficient vehicles to comply with
these standards. Under this view of stringency, then, it is necessary to realign test results to
maintain efficiency controls at the vehicle manufacturer level.
49 The FTP and HFET are EPA's standard dynamometer driving cycles, simulating city and highway driving,
respectively.
3-154
-------
Similarly, Figure 3-76 shows the average percent change in carbon-balance fuel economy
when moving from Tier 2 to Tier 3 fuels, calculated in the same way as the CO2 differences. We
used the fuel-economy values on each fuel calculated from measured CO2 and other carbon-
containing emissions to generate the actual carbon-balance fuel economy, before the final
conversion to CAFE compliance values. The results indicate that for the FTP and the FIFET
cycles, the average reduction in fuel economy when moving from Tier 2 fuel to Tier 3 fuel are
2.29 percent and 2.98 percent, respectively, corresponding to average reductions in fuel economy
of 0.66 and 1.34 miles per gallon.
1.0%
0.5%
ra
¦3
u
<
ra
E
o
>
U
o
in
H
CM
~o
3
3
03
ro
jQ
XI
Tj
O
* ~~
ISI
Ram
4)
>
ru
fD
2
_>
Cn
O
>
XI
fN
o
T3
ro
O)
>
tTi
"5 0.0%
£ -0.5%
k-
CD
p -1.0%
X
% -1.5%
c -2.0%
ro
5 -2.5%
-3.0%
1
I
FTP ~ HFET
Figure 3-75: Percent Change in CO2 Emissions from Tier 2 to Tier 3 Test Fuel (%).
3-155
-------
Figure 3-76: Percent Change in Carbon-Balance Fuel Economy from Tier 2 to Tier 3 Test
Fuel (%).
The Acura showed a noticeably larger fuel economy difference than other vehicles on the
highway cycle (HFET). To investigate this behavior, we performed a limited number of
additional tests of this vehicle on both regular grade Tier 3 fuel and premium grade (higher
octane) Tier 3 fuel. The results showed an unexpected level of fuel economy sensitivity to the
test fuel's octane rating.50 Although we present the results for this vehicle here and in the
Technical Report, we have excluded it from the analysis we used to determine the test procedure
adjustments. Because this vehicle is not labeled by the manufacturer as requiring premium fuel,
30 Emission certification fuel, including Tier 2 test fuel, has historically been high-octane grade as a matter of
convenience to avoid having to maintain separate octane levels of test fuels for different vehicle requirements.
Later, with the implementation of electronic ignition and knock sensors in the 1990s, it became possible for the
engine controls to optimize combustion for a number of factors including the fuel octane level, with vary ing effects
on emissions and fuel economy. Thus. EPA issued guidance to manufacturers in 1997 (VPCD-97-01) clarifying that,
in order to ensure representativeness of FE test results to real-world driving, any difference in emissions or FE
between high octane and regular octane market fuel must be declared if it exceeds a 3% allowance for normal test-
to-test variability. This requirement did not apply if the vehicle was marketed as requiring higher octane fuel. Note
that under the Tier 3 program, the default test fuel is now regular octane, which obviates the situation of undeclared
octane impacts between certification tests and in-use driving on market gasoline.
3-156
-------
this behavior was unexpected on the recommended (lower octane) fuel. We thus did not want
these results to inappropriately affect the adjustments to CO2 and fuel economy.
3.7.2 Development of Adjustment Factors
In this section, we describe how we used relevant data from the EPA test program
summarized in the previous section to develop the test fuel related adjustment factors. We
present below the separate analyses we conducted to determine these adjustment factors for CO2
and for CAFE fuel economy.
We note that the EPA test program results described in the Technical Report and summarized
above differ in perspective from our development of the adjustment factors discussed in this
section. The Technical Report described the change in emissions and fuel economy with the
transition from the current Tier 2 fuel to Tier 3 fuel, so those comparisons were formed as Tier 3
relative to Tier 2 fuel. In contrast, this section describes how we used the test program results to
determine adjustment factors that would maintain the stringency of the existing standards when
testing is performed on Tier 3 test fuel. Thus, the comparison in this section is formed as Tier 2
relative to Tier 3 fuel. Another difference is the ASTM method51 used to determine the carbon
mass fraction of the test fuel for calculation of fuel economy. In the Technical Report we used
the average D5291 result from five laboratories, whereas here we use the D3343 method
modified for ethanol as appropriate, consistent with the adjusted regulatory CAFE equation.52
Most individual vehicle and powertrain combinations will react slightly differently to a
change in test fuel. As a result, an approach to test-fuel-related adjustment that attempts to
recognize the unique responses of every vehicle would be very complicated and, we believe,
difficult to implement in a practical manner for manufacturer testing. Therefore we derive the
adjustments based on average values. Such an averaging approach is not new. Historically, when
EPA has corrected new test results back to the results on a previous test fuel, EPA required that
differing vehicle responses be accounted for on average. We believe this approach continues to
be sufficient and appropriate for compliance with fleet-average requirements for fuel economy
and CO2.
We developed the CO2 and CAFE adjustment factors based on the Federal Test Procedure
(FTP) and Highway Fuel Economy Test (HFET) results from the EPA test program, as described
below for each of the two adjustment factors. For consistency with the historical FTP/HFET
weighting of 55 percent and 45 percent, respectively, which is used in the current regulations for
compliance and other testing, we believe that this same 55 percent/45 percent weighting for FTP
and HFET test results is appropriate for the adjustment factors described in this action.53
3.7.2.1 CO2 Adjustment Factor
We analyzed the data from the EPA test program (excluding the data from the Acura because
of the octane sensitivity issue discussed above). Table 3-46 presents our calculation process. The
51 ASTM International (previously known as American Society for Testing and Materials).
52 See 40 CFR 600.113-12 as amended in this rule and memo "Distillation adjustment for ethanol blending in Tier 3
and LEVIII test fuels" submitted by Aron Butler to docket EPA-HQ-OAR-2016-0604.
53 The test procedure adjustments would apply to testing on all federal Tier 3 gasoline certification fuels, including
premium certification fuel and LEV III fuels.
3-157
-------
data show that the impact of the fuel change varies slightly among the vehicles, but it is
consistently in the same direction and in the range of 1-2.5 percent, with a mean value of 1.66
percent.
Table 3-46: CO2: Results of the EPA Test Program for the FTP and HFET Cycles, With
Weighted Values for the Two Cycles, and Corresponding Percent Differences.
Vehicle FTP HFET Weighted 1 DiITcrcncc2
Tier 3
Tier 2
Tier 3
Tier 2
Tier 3
Tier 2
(g/mi)
%
(g/mi)
(g/mi)
(g/mi)
(g/mi)
(g/mi)
(g/mi)
Alii ma
270.60
276.19
163.37
165.49
222.35
226.38
4.03
1.81
Civic
213.37
216.98
143.16
144.75
181.77
184.47
2.70
1.49
F150
376.87
380.61
241.92
244.79
316.14
319.49
3.35 '
1.06
Malibu 1
, 307.37
314.53
184.01
189.15
251.86
258.11
6.25
2.48
Malibu 2
268.64
274.00
163.58
166.02
221.36
225.41
4.05
1.83
Ma/da
238.57
242.12
160.32
161.87
203.36
206.01
2.65
1.30
Ram
414.49
423.94
260.67
262.76
345.27
351.41
6.14
1.78
Silverado
419.88
427.69
281.05
281.37
357.41
361.84
4.44
1.24
Volvo
299.83
305.98
173.22
175.61
242.86
247.3 1
4.46
1.84
Silverado (2b)
706.83
721.57
443.11
447.66
588.16
598.3 1
10.15
1.73
Mean
1.66
1 As 0.55FTP + 0.45HFET.
2As T2 - T3. and as 1()()(T2 - T3)/T3.
The formula for combining and weighting CO2 test results is straightforward:
C02= 0.55>C02a,+ 0.45>C02^w
Where: CO2 = weighted CO2 in grams per mile
CChcity = CO2 as measured on the FTP test cycle
CChhighway = CO2 as measured on the HFET test cycle
Based on the results of the analysis of test data in Table 3-46, measured CO2 from FTP and
HFET testing on Tier 3 test fuel, weighted as discussed above (55/45 percent), is adjusted by
multiplying by a factor of 1.0166 to produce the expected CO2 performance had the vehicle been
tested over the same test cycles while operating on Tier 2 fuel. In other words, the CO2 emissions
test results from a vehicle being tested for GHG compliance using Tier 3 test fuel would be
multiplied by this factor to arrive at the CO2 value used for compliance.54 For example, the
compliance CO2 value would be computed as 1.0166 x (0.55 x CC>2,FTP + 0.45 x C02,HFET).
54 Compliance for the LD GHG standards is based on all carbon-related exhaust emissions (CREE). The adjustment
factor applies only to the CO2 emission aspect of the CREE equation. For discussion of CREE impacts in the EPA
test program, see memo "Carbon-related Exhaust Emissions (CREE) Measured on Current and Proposed
Certification Gasolines," submitted by Jim Warila to docket EPA-HQ-OAR-2016-0604.
3-158
-------
3.7.2.2 Analysis of Fuel Economy Data and Development of Adjusted Equation
As we did with the CO2 test data above, we used the EPA test program results (again,
excluding the Acura) to determine an adjustment factor that would be applied to the FTP and
HFET results for test vehicles operating on Tier 3 test fuel to produce CAFE fuel economy
results equivalent to those from testing on Tier 2 test fuel. Tier 2 test fuel is the result of EPA's
1986 test fuel changes and the associated adjustment, designed to produce results that represent
the CAFE fuel economy that would have been observed under 1975 test conditions (as required
by the statutes governing the CAFE program). The CAFE fuel economy adjustment described
here would align Tier 3 test fuel testing with Tier 2 test fuel results, and, by extension, with
results that would have been observed using 1975 test fuel.
Note that the adjustment factor would also be used for all other test cycles required for fuel
economy labeling. This current section summarizes EPA's analysis and the resulting value for
the CAFE fuel economy adjustment factor. As discussed above, a vehicle's CAFE fuel economy
is based primarily on the same measured CO2 emissions that determine its compliance with the
GHG standards. For the reasons discussed in that section, the CAFE calculation is necessarily
more complex than the direct CO2 emissions measurement, and adjusting the calculation carries
these complexities.
To provide NHTSA with the fuel economy data it uses for CAFE compliance, EPA uses
calculations that account for the difference in volumetric energy density (VED, e.g., Btu/gal) of
the test fuel relative to the baseline test fuel on which NHTSA based the original CAFE
standards in 1975. In the mid-1980s, when EPA last made such a test-fuel related adjustment,
empirical data available to the Agency suggested that there was not a direct, 1-to-l response of
fuel economy to changes in test fuel VED. Because of this, EPA proposed and took final action
to insert an additional factor, called the "R-factor," into the equation. EPA defined this R-factor,
established in the regulations with a value of 0.6, as the percent change in fuel economy per
percent change in test fuel VED. For example, for R = 0.6, a 10 percent decrease in test fuel
VED would only produce a 6 percent decrease in fuel economy.
Table 3-47 shows this R=0.6 adjusted fuel economy value alongside the carbon-balance fuel
economy for both test fuels. The VED of the Tier 2 fuel was higher than the 1975 CAFE
reference fuel, so the R-factor adjustment reduces the fuel economy result slightly relative to the
carbon-balance value. For Tier 3 test fuel, which has lower VED, the R-factor adjustment
increases the fuel economy result slightly If the adjustment were functioning optimally (i.e., if R
= 0.6 were exactly the right adjustment for both fuels), we'd expect the corrected value in the R
= 0.6 columns in the table to be the same value for both test fuels. However, there is still a
directionally consistent offset, with the Tier 3 test fuel values slightly lower than the Tier 2
values for all but one vehicle, suggesting that an R-factor of 0.6 is not optimal and should be
higher for this test fleet operating on Tier 3 fuel. A higher value is also supported by analyses of
other recent datasets.55
55 Sluder, C., West, B., Butler, A., Mitcham, A. et al., "Determination of the R Factor for Fuel Economy
Calculations Using Ethanol-Blended Fuels over Two Test Cycles," SAE Int. J. Fuels Lubr. 7(2):551-562, 2014.
3-159
-------
Table 3-47: Carbon-Balance and R-Adjusted Fuel Economy Results by Vehicle and Fuel
(City/Highway-Weighted Values, mpg).
Tier 2 lest fuel;) Tier 3 lest fuclb
C-balancc
R=0.6 equation
C-balancc equation
R=0.6 equation
equation
Allima
39.40
39.26
38.51
39.10
Civic
48.43
48.26
47.16
47.88
F150
27.97
27.87
27.12
27.53
Malibu 1
34.49
34.37
34.00
34.52
Malibu 2
39.61
39.48
38.72
39.3 1
Ma/.da
43.38
43.23
42.16
42.81
Ram
25.42
25.34
24.83
25.22
Silverado
24.66
24.58
23.96
24.32
Volvo
36.08
35.95
35.24
35.78
Silverado (2b)
14.90
14.85
14.56
14.79
a For the Tier 2 fuel, we calculated the adjusted fuel economy using ASTM methods D3343 and D3338, and lumped THC emission term,
consistent with how fuel economy is calculated and reported under the current requirements.
b For the Tier 3 fuel, we used modified methods D3343 and D3338, and separate NMOG and CF14 emission terms. The reason for the change
in emission terms is explain in more detail below.
Because of the remaining offset seen above, we are adopting an updated fuel economy
equation for use with Tier 3 test fuel where the R-factor is replaced by a new factor (Ra),
determined empirically so as to make the fleet-average fuel economy result using Tier 3 test fuel
numerically equivalent to the fleet-average result using Tier 2 test fuel and R=0.6. The goal is to
have no change in stringency for compliance with fuel economy standards with the new test fuel.
Note that this new factor not only updates the sensitivity of fuel economy to VED (the main
purpose of the original R-factor) but also accommodates other changes to the calculation
discussed in more detail below. For reference, we show the current equation for Tier 2 test fuel
(which we described above) here:56
F^CAFE
^-MF-r.fuei • SGTiue| • Pwater
CMFexh • THC+ 0.429-CO + 0.273-C02
NHC,fue, • SGBiue|
(R--INHCtj„„ • SGtj„„) + ((1 -R). NHC„Juel.SG„Juel)_
One of these changes to the equation is an update from using THC emissions in the Tier 2
carbon-balance denominator to using NMOG and CH4 with Tier 3 test fuel, where NMOG is
determined as specified in 40 CFR 1066.635. The inclusion of NMOG better accounts for the
oxygenated emission products resulting from ethanol in the test fuel, and is consistent with the
56 We present the equations below in a form that highlights the changes between the existing and proposed CAFE
equations. These equations are functionally equivalent to those in the regulatory language associated with this notice
(40 CFR 600.113-12), with the latter equations structured in form conventionally used for CAFE compliance
purposes. This regulatory language also defines each of the terms in these CAFE equations.
3-160
-------
use of NMOG in the Tier 3 emission standards. With the very low emission levels of Tier 3
vehicles, we expect the difference between THC and the sum of NMOG + CH4 to be negligible.
FEcafE
CMFTiuel
^^T.fuel P water
CMFexh • NMOG + 0.749-CH4 + 0.429-C0 + 0.273-C02
NHCBiue,-SGB,ue,
. (R, • IN HCT fuel • SGT fuel) + ((1 - Ra) • N H CB fuel • SGB fuel)
A second change to the fuel economy calculation is to update the test methods used in
determining specific gravity (SG), carbon mass fraction (CMF), and net heat of combustion
(NHC). As indicated earlier, EPA designed the existing CAFE equation around the use of EO test
fuel, and specified that these fuel parameters be determined using ASTM methods D1298,
D3343, and D3338, respectively. The latter two methods determine the unknown fuel property
by mathematical correlation to other known properties, and these correlations are not suitable for
ethanol blends as published. Therefore, we are adopting additional calculations to be used with
D3343 and D3338 to determine CMF and NHC of E10 test fuel. These modified methods have
been previously described in EPA guidance and other technical literature, and are specified in
detail in the regulations included as part of this action.57 We are also adopting method D4052 as
equivalent to D1298 for determining SG.
In deriving the appropriate value for Ra, i.e., the value that produces the equivalent fuel
economy with Tier 3 E10 test fuel, we used the current Tier 2 methods and R=0.6 when
calculating the fuel economy using Tier 2 test fuel, and the updated methods when using Tier 3
test fuel. Because of the changes to the measurement methods discussed in the previous
paragraph and the new Ra factor being specific to Tier 3 test fuel, this new equation would not
be valid for reporting fuel economy when testing using Tier 2 fuel. We are incorporating the
small impacts of these calculation formula changes within the single new Ra factor.
As with the CO2 adjustment factor, for the CAFE adjustment factor we weighted the results
from city (FTP) and highway (HFET) testing in the EPA test program as follows:
MPG =
1
0.55
- +
0.45
MPGMPG,
city
highway
Our analysis of the study data as described shows that a value of Ra=0.81 produces a fleet
average fuel economy difference very close to zero between the two test fuels. Table 3-48
compares the adjusted city /highway weighted fuel economy for each study vehicle as it is
currently calculated with Tier 2 fuel to the adjusted fuel economy on Tier 3 fuel using the
updated calculations and an Ra value of 0.81. At the right-hand side of the table is the percent
difference by vehicle, with the fleet average difference of near zero shown at the bottom.
57 EPA Guidance Letter CD-95-09 and SAE technical paper 930138 describe adjustment of ASTM D3338 and
D3343 results for oxygenates. More detail on accommodation of ethanol's volatility impact in the ASTM methods
can be found in the memo "Distillation adjustment for ethanol blending in Tier 3 and LEVIII test fuels," May 2,
2018, submitted by Aron Butler to docket EPA-HQ-OAR-2016-0604.
3-161
-------
Table 3-48: Adjusted Fuel Economy Results by Vehicle and Fuel Showing Impact of Ra
Factor (City/Highway-Weighted Values).
Tier 2 lest fuel
Tier 3 lest fuel
R=0.6
Ra=0.81
Tier 3 vs. Tier 2
Alii ma
39.26
39.32
0.16%
Civic
48.26
48.15
-0.23%
F150
27.87
27.69
-0.65%
Malibu 1
34.37
34.72
1.02%
Malibu 2
39.48
39.54
0.15%
Ma/da
43.23
43.05
-0.41%
Ram
25.34
25.36
0.09%
Silverado
24.58
24.46
-0.46%
Volvo
35.95
35.98
0.08%
Silverado (2b)
14.85
14.87
0.14%
Average difference
-0.01%
Figure 3-77 shows the percent change in city/highway weighted fuel economy when moving
from Tier 2 to Tier 3 test fuel using three computation methods. The bottom series (with square
markers) shows the difference using the carbon-balance calculation, which makes no adjustment
for VED and therefore is the best estimate of the actual, real-world effect. The middle series
(with round markers) shows the difference calculated using the appropriate CAFE formula and
fuel property measurements for each test fuel and R=0.6 for both (from Table 3-47). Finally, the
top series (dashed with triangular markers) shows the effect of adjusting the R-factor in the Tier
3 equation to a value of 0.81. The difference of approximately 0.6 percent between the top and
middle lines is the fuel economy reduction due to the test fuel change that would be mitigated by
the R-factor update. The top line in this figure corresponds to the right-hand column in the
previous table.
3-162
-------
Figure 3-77: City/Highway Weighted Fuel Economy Difference Between Test Fuels for
Different Calculation Methods, Shown by Vehicle (Fleet Average at Right).
3-163
-------
Chapter 3 References
40 CFR 600.116-12. 2022. https://www.ecfr.gov/current/title-40/chapter-I/subchapter-Q/part-
600/subpart-B/section-600.116-12.
40 CFR 600.311-12. 2021. https://www.ecfr.gov/current/title-40/chapter-I/subchapter-C/part-
86/subpart-N/section-86.1360.
40 CFR 86.1803-01. 2023.
77 FR 62624. 2012. (October 15). https://www.govinfo.gov/content/pkg/FR-2012-10-
15/pdf/2012-21972.pdf.
86 FR 74434. 2021. (December 30).
https://www.federalregister.gov/documents/2021/12/30/2021-27854/revised-2023-and-later-
model-year-light-duty-vehicle-greenhouse-gas-emissions-standards.
Aaron Isenstadt, Zifei Yang, Stephanie Searle, John German. 2022. Real world usage of plug-in
hybrid vehicles in the United States,, https://theicct.org/publication/real-world-phev-us-dec22/,
ICCT.
Alikin, Evgeny, and Aleksey Vedyagin. 2016. "High Temperature Interaction of Rhodium with
Oxygen Storage Component in Three-Way Catalysts." Topics in Catalysis 59: 1033-1038.
doi: 10.1007/sl 1244-016-0585-z.
ANL. 2022. "Assessment of Light-Duty Plug-in Electric Vehicles in the United States, 2010-
2021." November.
ANL. 2024a. Quantification of Commercially Planned Battery Component Supply in North
America through 2035. Argonne National Laboratory.
ANL. 2024b. Securing Critical Materials for the U.S. Electric Vehicle Industry. Argonne
National Laboratory.
Asako, Tsuyoshi, Daiki Saito, Tetsuhiro Hirao, Emily Popp, Mitsuyoshi Okada, Atsushi Kaneda,
Akifumi Kawakami, Etsuji Ohara, and Kentaro Sugimoto. 2022. "Achieving SULEV30
Regulation Requirement with Three-Way Catalyst on High Porosity Substrate while Reducing
Platinum Group Metal Loading." SAE Technical Paper 2022-01-0543. SAE. doi: 10.4271/2022-
01-0543.
Bargman, Ben, SooHyun Jang, Jan Kramer, Ihab Soliman, and Abhid Akram Abubeckar
Mohamed Sahul. 2021. "Effects of Electrically Preheating Catalysts on Reducing High-Power
Cold-Start Emissions." SAE Technical Paper 2021-01-0572. SAE. doi: 10.4271/2021-01-0572.
Battery Design. 2022. GMUltium. June 8. https://www.batterydesign.net/gm-ultium/.
Benjamini, Yoav. 1988. "Opening the box of a boxplot." The American Statistician 42 (4): 257-
262.
Bhattachaijya,S., et al. 2023. Heavy-light-duty and Medium-duty Range-extended Electric Truck
Study. Ann Arbor, Michigan: US EPA.
3-164
-------
Boger, T., and W. Cutler. 2019. Reducing Particulate Emissions in Gasoline Engines.
Warrendale, PA: SAE International.
Bohac, S. V., and S. Ludlam. 2023. "Characterization of a Lightly Loaded Underfloor Catalyzed
Gasoline Particulate Filter in a Turbocharged Light Duty Truck." Journal of Engineering for Gas
Turbines and Power 145(2): 021024. doi:10.1115/1.4056047.
Bohac, S. V., S. Ludlam, J. Martin, W. T. Preston, and M. D. Hays. 2022. "PAH Emissions from
a GPF-Equipped Light Duty Truck during Soot Accumulation and Regeneration." 31st CRC
Real World Emissions Workshop, San Diego, CA, March 14-16, 2022.
Bohac, Stanislav, and Dennis Assanis. 2004. "Effect of Exhaust Valve Timing on Gasoline
Engine Performance and Hydrocarbon Emissions." SAE Technical Paper No. 2004-01-3058.
doi: 10.4271/2004-01-3058.
Bond, T.C., S. J. Doherty, and D. W., et al. Fahey. 2013. "Bounding the Role of Black Carbon in
the Climate System, A Scientific Assessment." J. Geophys. Res. Atmos. 118, pp. 5380-5552.
Borger, T., D. Rose, P. Nicolin, B. Coulet, and A. Bachurina. 2018. "Severe Soot Oxidations in
Gasoline Particulate Filter Applications." SAE Papaer 2018-01-1699.
BrightDrop. 2022. "2024 BrightDrop Zevo 400 and Zevo 600 Order Guide." November 4.
https://www.gobrightdrop.eom/ca/_assets/files/2024-brightdrop-en-ca-zevo-400-and-zevo-600-
order-guide-12-23 -22.pdf.
—. 2023. BrightDrop - Moving things forward. Accessed December 18, 2023.
https://www.gobrightdrop.com/.
Briick, R, and K Konieczny. 2022. "Innovative Catalyst System to Achieve EU7 Legislation."
Proceedings of the 43rd International Vienna Motor Symposium, Vienna, Austria.
California Air Resource Board [OBD data records] . 2022. October.
https://www.bar.ca.gov/records-requests.
California Air Resource Board [OBD data records], 2023. November.
https://www.bar.ca.gov/records-requests.
California Environmental Protection Agency, Air Resources Board. 2022. "art 1, Section 1.4.
California Provisions: Certification and In-Use testing requirements for chassis certified
Medium-Duty Vehicles (MDV) with a Gross Combination Weight Rating (GCWR) greater than
14,000 pounds, using the Moving Average Window (MAW)." Final California 2026 and
Subsequent Model Year Criteria Pollutant Exhaust Emission Standards and Test Procedures for
Passenger Cars, Light Duty Trucks, and Medium-Duty Vehicles. August 25.
Chan, T. W., Meloche, E., Brezny, R. 2014. "Black Carbon Emissions in Gasoline Exhaust and a
Reduction Alternative with a Gasoline Particulate Filter." Environ. Sci. and Technol. 48 (10):
6027-6034. doi:10.1021/es501791b.
Chan, T. W., Meloche, E., Kubsch, J., Brezny, R., Rosenblatt, D., Rideout, G. 2013. "Impact of
Ambient Temperature on Gaseous and Particle Emissions from a Direct Injection Gasoline
Vehicle and its Implications on Particle Filtration." SAE Paper 2013-01-0527. doi: 10.4271/2013-
01-0527.
3-165
-------
Chan, T. W., Saffaripour, M., Liu, F., Hendren, J., Thomson, K. A., Kubsh, J., Brezny, R.,
Rideout, G. 2016. "Characterization of Real-Time Particle Emissions from a Gasoline Direct
Injection Vehicle Equipped with a Catalyzed Gasoline Particulate Filter During Filter
Regeneration." Emiss. Control Sci. Technol. 2 (2): 75-88. doi:10.1007/s40825-016-0033-3.
Chan, Tak, Meghdad Saffaripour, Fengshan Lui, Jill Hendren, Kevin Thomson, Joseph Kubsh,
Rasto Brezny, and Greg Rideout. 2016. "Characterization of Real-Time Particle Emissions from
a Gasoline Direct Injection Vehicle Equipped with a Catalyzed Gasoline Particulate Filte
rDuring Filter Regeneration." Emission Control Science and Technology 2: 75-88.
doi: 10.1007/s40825-016-0033-3.
Christianson, M. G., E. Bardasz, and W. Nahumck. 2010. "Impact of Lubricating Oil Condition
on Exhaust Particulate Matter Emissions from Light Duty Vehicles." SAE International Journal
of Fuels and Lubricants Vol. 3, No. 2, pp. 476-488.
Clenci, Adrian, Julien Berquez, Robert Stoica, Rodica Niculescu, Bogdan Cioc, C'at'alin
Zaharia, and Victor Iorga-Sim'an, 2022. "Experimental investigation of the effect of an
afterburner on the light-off performance of an exhaust after-treatment system." Energy Reports
8: 406-418. doi:10.1016/j.egyr.2022.07.025.
Continental. 2015. Continental Electrically Heated Catalyst Supports L0W-CO2 48 Volt Hybrid
Strategies. 9 10. Accessed 12 7, 2023. https://www.continental.com/en/press/press-
releases/2015-08-20-volt-e-kat/.
Cooper, Jonathan, and Joel Beecham. 2013. "A Study of Platinum Group Metals in Three-Way
Autocatalysts." Platinum Metals Review 57 (4): 281-288. doi: 10.1595/147106713X671457.
Corning. 2023. Corning® FLORA® Substrates. Accessed 12 15, 2023.
https://www.corning.com/worldwide/en/products/environmental-technologies/products/ceramic-
sub strates/corning-flora-sub strates .html.
Daniel Dahl, Ayolt Helmantel, Fredrik Wemmert, Mats Moren, Staffan Rengmyr, and Ali
Sahraeian. 2020. "The New Volvo Mild Hybrid Miller Engine." 29th Aachen Colloquium
Sustainable Mobility. Aachen.
Denso. 2022. How do the compressors differ in hybrid and electric vehicles? June 22. Accessed
January 4, 2024. https://www.denso-am.eu/news/how-do-the-compressors-differ-in-hybrid-and-
electric-vehicles.
Edwards, Adam Jacob. 2022. "Aggressive-dynamics metrics for drive-cycle characterization."
Transportation Research Interdisciplinary Perspectives, March 30:
https://doi.Org/10.1016/j.trip.2022.100592.
Faurecia. 2023. Ultra-low emissions solutions. Accessed 12 7, 2023.
https://www.faurecia.com/en/technologies/solutions-sustainable-mobility/Aftertreatment-
technologies.
Fortune Business Insights. 2022. Automotive E-commerce Market Size, Share & COVID-19
Impact Analysis. July. Accessed December 13, 2022.
https://www.fortunebusinessinsights.com/automotive-e-commerce-market-105728.
3-166
-------
Frigge, Michael, David C. Hoaglin, and Boris Iglewicz. 1989. "Some implementations of the
boxplot." The American Statistician 43 (1): 50-54.
Gao, Mi-Young Kim, Jae-Soon Choi, C Stuart Daw, James E Parks, and David E Smith. 2012.
"Cold-start emissions control in hybrid vehicles equipped with a passive adsorber for
hydrocarbons and nitrogen oxides." Proceedings of the Institution of Mechanical Engineers Part
D: Journal of Automobile Engineering (1396-1407) 226 (10). doi: 10.1177/0954407012443764.
Green Car Congress. 2021. Vitesco Technologies reduces emissions in high-voltage hybrid
vehicles with electrical catalyst heating; new DC/DC converter. 2 24. Accessed 12 7, 2023.
https://www.greencarcongress.com/2021/02/20210224-vitesco.html.
GuiQiang Zhang, Qian Wang, Guang Chen, et al. 2020. "Geely Hybrid Engine: World Class
Efficiency for Hybrid Vehicles." 29th Aachen Colloquium Sustainable Mobility. Aachen: 29th
Aachen Colloquium Sustainable Mobility.
IEA. 2022. "The Role of Critical Minerals in Clean Energy Transitions, World Energy Outlook
Special Report, Revised version." International Energy Agency, March.
Isenstadt, A. 2023. "What EPA's New Multi-Pollutant Emissions Proposal Means for PM
Emissions and GPFs." ICCT Fact Sheet, www.theicct.org.
Jean, Emmanuel, and Maxime Goncalves. 2023. "Electrically Heated Catalyst: A Powerful Tool
for Aftertreatment Optimization." SAE Technical Paper 2023-01-0351. SAE. doi: 10.4271/2023-
01-0351.
John Kargul, Mark Stuhldreher, Dan Barba, Charles Schenk, Stani Bohac, Joseph McDonald,
and Paul Dekraker. 2019. "Benchmarking a 2018 Toyota Camry 2.5-liter Atkinson cycle engine
with cooled-EGR." SAE International Journal of Advances and Current Practices in Mobility 1
(2): 601-638. doi: 10.4271/2019-01-0249.
Joshi, A., and T. V. Johnson. 2018. "Gasoline Particulate Filters - a Review." Emission Control
Science and Technology (4)219-239. doi:10.1007/s40825-018-0101-y.
Kawaguchi, Bungo, Kazuhiro Umemoto, Seitaro Misawa, Shigemasa Hirooka, and Takashi
Kawai. 2019. "ICE Vehicle Challenge toward Zero Emissions: Future Technology
Harmonization in Electrified Powertrain System." SAE Technical Paper No. 2019-01-2217.
SAE. doi: 10.4271/2019-01-2217.
Khameneian, Amir, Paul Dice, Joel Duncan, Jeffrey D. Naber, Mahdi Shahbakhti, Chad Archer,
Peter Moilanen, Chris Glugla, and Garlan Huberts. 2022. "Dynamic individual-cylinder analysis
of a Gasoline Direct Injection engine emissions for cold crank-start at elevated cranking speed
conditions of a Hybrid Electric Vehicle." Control Engineering Practice 129: 105349.
doi: 10.1016/j. conengprac .2022.105349.
Kim, Chang Hwan, Hyokyoung Lee, Chun Yong Kang, and Jin Woo Chung. 2018. "For a New
Paradigm in Aftertreatment: The." 27th Aachen Colloquium Automobile and Engine
Technology. 1013-1030.
Konstandopoulos, A. G., and Papaiononnou, E. 2008. "Update on teh Science and Technology of
Diesel Particulate Filters." KONA Power Part J, 26(0), pp. 36-65.
3-167
-------
Krajinska, Poliscanova, Mathieu, & Ambel, Transport & Environment. 2020. "A new Dieselgate
in the making." November: https://www.transportenvironment.org/discover/plug-hybrids-europe-
heading-new-dieselgate/.
Kunath, Florian, Dominik Rose, Bertrand Coulet, Per Nicolin, Thorsten Boger, Thomas Collins,
James Grochocinski, and Mrinmoy Dam. 2022. "Technology options to reduce cold start
emissions - electric catalyst heaters." 31st Aachen Colloquium Sustainable Mobility. Aachen.
Paper No. 58.
Lambert, C., T. Chanko, D. Dobson, X. Liu, and J. Pakko. 2017. "Gasoline Particle Filter
Development." Emiss. Control Sci. Technol. 3:105-111. doi:10.1007/s40825-016-0055-x.
Lambert, Christine. 2018. "Next Generation Three-Way Catalysts for Future, Highly Efficient
Gasoline Engines." U.S. Department of Energy Award DE-EE0006845.
Laurell, Mats, Lorenzo Pace, Fredrik Ekstrom, and Katrin Konieczny. 2019. "Strive for Zero
Emissions Impact from Hybrids." SAE Technical Paper 2019-24-0146. SAE. doi:10.4271/2019-
24-0146.
Luo, Xi, Sam Hashemi, Ram Subramanian, Anastasios Arvanitis, and Matthew Younkins. 2020.
"Fast Catalyst Light-Off with Dynamic Skip Fire." SAE Int. J. Advances & Curr. Prac. in
Mobility 2 (4): 1849-1861. doi: 10.4271/2020-01-0313.
Mark Stuhldreher . 2016. "Fuel efficiency mapping of a 2014 6-cylinder GM EcoTec 4.3 L
engine with cylinder deactivation." SAE Technical Paper. No. 2016-01-0662.
Mark Stuhldreher, Charles Schenk, Jessica Brakora, David Hawkins, Andrew Moskalik, and
Paul DeKraker. 2015. "Downsized Boosted Engine Benchmarking and Results." SAE Technical
paper no. 2015-01-1266. doi: 10.4271/2015-01-1266.
Maurer, Robert, Surya Kiran Yadla, Andreas Balazs, Matthias Thewes, Vitali Walter, and Tolga
Uhlmann. 2020. "Designing Zero Impact Emission Vehicle Concepts." Experten-Forum
Powertrain: Ladungswechsel und Emissionierung 2020. 75-116.
Maurer, Robert, Theodoros Kossioris, Stefan Sterlepper, Marco Giinther, and Stefan Pischinger.
2023. "Achieving Zero-Impact Emissions with a Gasoline Passenger Car." Atmosphere 14: 313.
doi: 10.3 3 90/atmos 14020313.
Menne, Christoph, Patrick Recker, Andreas Balazs, Joschka Schaub, Markus Ehrly, Thomas
Wilkes, Oliver Simons, et al. 2022. "Advanced Hardware, Controls and Calibration Solutions to
Meet the Upcoming EURO 7 Legislation." 31st Aachen Colloquium Sustainable Mobility.
Aachen. 1-23.
Minjares, R. J., and F. P. Sanchez. 2011. "Estimated Cost of Gasoline Particulate Filters." ICCT
Working Paper 2011-8.
Moser, David, Sanket Nipunage, JohnNunan, Ryan Day, and Chad Alltizer. 2021. "An
Unconventional Application of a HC Trap to Meet SULEV20." SAE Technical Paper 2021-01-
0574,. SAE. doi: 10.4271/2021-01-0574.
3-168
-------
Moses-DeBusk, Melanie. 2021. "Chemistry of Cold-Start Emissions and Impact of Emissions
Control." DOE Annual Merit Review, acel53. https://www.energy.gov/eere/vehicles/2021-
vehicle-technologies-office-amr-presentations-program.
Moses-DeBusk, Melanie, John M. E. Storey, Samuel A. Lewis, R. Maggie Connatser, Shannon
M. Mahurin, Shean Huff, Cyril V. Thompson, and Yensil Park. 2023. "Detailed hydrocarbon
speciation and particulate matter emissions during cold-start from turbocharged and naturally
aspirated trucks." Fuel 350: 128804. doi:10.1016/j.fuel.2023.128804.
Nakanishi, Yui, Yukio Suehiro, Masanori Hashimoto, Takeshi Narishige, Ryo Takeuchi,
Yasuhiro Shimamoto, Shogo Sano, and Hironori Iwakura. 2019. "Development of Low
Temperature Active Three Way Catalyst." SAE Technical Paper 2019-01-1293. SAE.
doi:l 0.4271/2019-01-1293.
Nakasumi, Hayaki, Akifumi Kawakami, Etsuji Ohara, Kentaro Sugimoto, Noriyuki Hibi,
Tsuyoshi Asako, Kyohei Kato, et al. 2023. "High Cell Density Flow Through Substrate for New
Regulations." SAE Technical Paper 2023-01-0359. SAE. doi: 10.4271/2023-01-0359.
Newman, K., Kargul, J., and Barba, D. 2015. "Benchmarking and Modeling of a Conventional
Mid-Size Car Using ALPHA." SAE Technical Paper 2015-01-1140. doi: 10.4271/2015-01-1140.
North American Council for Freight Efficiency. 2021. "Vans & Step Vans: Market Segment &
Fleet Profile Fact Sheet." Accessed December 13, 2022. https://nacfe.org/wp-
content/uploads/2022/01/Vans-Step-Vans-RoL-E-Fact-Sheet.pdf.
Olszewski, Mitch. 2011. EVALUATION OF THE 2010 TOYOTA PRIUS HYBRID
SYNERGY DRIVE SYSTEM. Oak Ridge National Laboratory (ORNL).
Patrick Plotz, Cornelius Moll, Georg Bieker and Peter Mock. 2023. "From lab-to-road: real-
world fuel consumption and CO2 emissions of plug-in hybrid electric vehicles (2021 Environ.
Res. Lett. 16054078)." Environ. Res. Lett. 18 099502.
Pham, A, and M Jeftic. 2018. "Characterization of Gaseous Emissions from Blended Plug-In
Hybrid Electric Vehicles during High-Power Cold-Starts." SAE Technical Paper 2018-01-0428.
SAE. doi: 10.4271/2018-01-0428.
Plotz, P., Moll, C., Bieker, G., Mock, P., Li, Y. 2020. "Real-world usage of plug-in hybrid
electric vehicles: fuel consumption, electric driving, and CO2 emissions. ICCT, September
2020." https://theicct.org/publication/real-world-usage-of-plug-in-hybrid-electric-vehicles-fuel-
consumption-electric-driving-and-C02-emissions/.
—. 2020. "real-world-usage-of-plug-in-hybrid-electric-vehicles-fuel-consumption-electric-
driving-and-C02-emissions." ICCT. September, https://theicct.org/publication/real-world-usage-
of-plug-in-hybrid-electric-vehicles-fuel-consumption-electric-driving-and-C02-emissions/.
Public Law 117-169. 2022. "The Inflation Reduction Act of 2022." August 16.
https://www.govinfo.gov/content/pkg/PLAW -117publ 169/pdf/PLAW-117publ 169.pdf.
Reuters. 2023. "China's graphite curbs will accelerate plans around alternatives." October 23.
https://www.reuters.com/business/autos-transportation/chinas-graphite-curbs-will-accelerate-
plans-around-alternatives-2023 -10-20/.
3-169
-------
Rivian. 2023. Rivian Fleet - A commercial van that gets work done. Accessed December 18,
2023. https://rivian.com/fleet.
Rodriguez, J. Felipe, and Wai K. Cheng. 2017. "Analysis of NOx Emissions during Crank-Start
and Cold Fast-Idle in a GDI Engine." SAE International Journal of Engines 10 (2): 646-655.
doi: 10.4271/2017-01-0796.
Rodriguez, J. Felipe, and Wai K. Cheng. 2016. "Reduction of Cold-Start Emissions through
Valve Timing in a GDI Engine." SAE International Journal of Engines 9 (2): 1220-1229.
doi: 10.4271/2016-01-0827.
Rogers, Susan. 2012. "FY2012 Annual Progress Report - Advanced Power Electronics and
Electric Motors Program."
Rogers, Susan. 2013. FY2013 Annual Progress Report for the Advanced Power Electronics and
Electric Motors Program. Oak Ridge National Laboratory (ORNL).
Roy, Jean, Ahmed Ghoniem, Robert Panora, and Joseph Gehret. 2018. "NOx Reduction Using a
Dual-Stage Catalyst System with Intercooling in Vehicle Gasoline Engines under Real Driving
Conditions." SAE International Journal of Engines 11 (6): 893-901. doi:10.4271/2018-01-0335.
S&P Global. 2022b. "Growth of Li-ion battery manufacturing capacity in key EV markets." May
20. Accessed November 22, 2022. https://www.spglobal.com/mobility/en/research-
analysis/growth-of-liion-battery-manufacturing-capacity.html.
—. 2022a. "US ready for a battery factory boom, but now it needs to hold the charge." October 3.
Accessed November 22, 2022. https://www.spglobal.com/marketintelligence/en/news-
insights/latest-news-headlines/us-ready-for-a-battery-factory-boom-but-now-it-needs-to-hold-
the-charge-72262329.
SAE . 2006. "Vehicle Acceleration Measurement J1491_200607."
SAE J1711. 2023. "Recommended Practice for Measuring the Exhaust Emissions and Fuel
Economy of Hybrid-Electric Vehicles, Including Plug-in Hybrid Vehicles." Issued 1999-03,
Revised 2010-06, Revised 2023-02, February.
SAE J2727. 2023. "Mobile Air Conditioning System Refrigerant Emissions Estimate for Mobile
Air Conditioning Refrigerants." Issued in 2005-06, Revised in 2023-09, September 6.
SAE J2727 Worksheet Screenshots. 2023. This document is available in Docket EPA-HQ-OAR-
2023-xxx-xxx. (December).
SAE J2841. 2010. "Utility Factor Definitions for Plug-In Hybrid Electric Vehicles Using Travel
Survey Data." Issued 2009-03, Revised 2010-09, September.
SAE. 2016. "Performance Requirements for Determining Tow-Vehicle Gross Combination
Weight Rating and Trailer Weight Rating J2807_201602."
https ://www. sae.org/ standards/content/j 2807_201602/.
Saito, C., T. Nakatani, Y. Miyairi, and et al. 2011. "New Particulate Filter Concept to Reduce
Particle Number Emissions." SAE Paper 2011-01-0814. doi: 10.4271/2011-01-0814.
3-170
-------
Seabaugh, Christian. 2022. "Take a Look Inside Amazon's New Rivian Delivery Van." Motor
Trend, November 24. Accessed March 9, 2023. https://www.motortrend.com/news/checking-in-
on-the-rivian-amazon-edv-electric-van/.
Seshadri Srinivasa Raghavan & Gil Tal. 2022. "Plug-in hybrid electric vehicle observed utility
factor: Why the observed electrification performance differ from expectation." International
Journal of Sustainable Transportation, 16:2 105-136. doi:1748-9326/abef8c.
State Department. 2023. "The Minerals Security Partnership Continues to Expand with Norway,
Italy, and India, Media Note." September 16.
State of California, Air Resources Board. 2022. "Advanced Clean Cars II Proposed Amendments
to the Low Emission, Zero Emission, and Associated Vehicle Regulations - Standardized
Regulatory Impact Assessment."
—. 2021. "Advanced Clean Trucks Regulation, Final Approval / OAL Action, Final Regulation
Order." March 15. https://ww2.arb.ca.gov/sites/default/files/barcu/regact/2019/act2019/fro2.pdf.
Stellantis. 2023. All-new 2025 Ram 1500 Ramcharger Unveiled With Class-shattering Unlimited
Battery-electric Range. November 7. Accessed December 13, 2023.
https://media. stellantisnorthamerica.com/newsrelease. do?id=25436.
Stuhldreher, Mark; Kargul, John; Barba, Daniel; McDonald, Joseph; Bohac, Stanislav; Dekraker,
Paul; Moskalik, Andrew. 2018. "Benchmarking a 2016 Honda Civic 1.5-liter L15B7
Turbocharged Engine and Evaluating the Future Efficiency Potential of Turbocharged Engines."
SAE International Journal of Engines 116: 1273-1305. doi: 10.4271/2018-01-0319.
Sun, Zin, Minggao Ouyang, and Han Hao. 2022. "Surging lithium price will not impede the
electric vehicle boom." Joule 6 (8): 1738-1742. https://doi.org/10.1016/jjoule.2022.06.028.
T. Toda, M. Sakai, M. Hakariya, and T. Kato. 2017. "The New Inline 4 Cylinder 2.5L Gasoline
Engine with Toyota New Global Architecture Concept." Internationales Wiener
Motorensymposium. Vienna.
The White House. 2021. "Building Resilient Supply Chains, Revitalizing American
Manufacturing, and Fostering Broad-Based Growth, 100-Day Reviews under Executive Order
14017, p. 121." June.
Theis, Joseph, Andrew (Bean) Getsoian, and Christine K. Lambert. 2018. "The Development of
Low Temperature Three-Way Catalysts for High Efficiency Gasoline Engines of the Future: Part
II." SAE Technical Paper 2018-01-0939. doi: 10.4271/2018-01-0939.
Thomas E. Reinhart. 2021. Engine Efficiency Technology Study: Final Report, SwRI Project
No. 03.26457. San Antonio: Southwest Research Institute.
Title 13 § 1968.2 California Code of Regulations. 2022. (November 22).
Tukey, J. W. 1977. Exploratory data analysis. Vol. 2. Addison-Wesley.
U.S. Census Bureau. 2022. February 18. Accessed December 13, 2022.
https://www.census.gov/retail/index.html.
3-171
-------
U.S. EPA. 2023. Annual Certification Data for Vehicles, Engines, and Equipment. August 30.
Accessed November 15, 2023. https://www.epa.gov/compliance-and-fuel-economy-data/annual-
certification-data-vehicles-engines-and-equipment.
—. 2023b. Combining Data into Complete Engine ALPHA Maps. Accessed February 2, 2023.
https://www.epa.gov/vehicle-and-fuel-emissions-testing/combining-data-complete-engine-alpha-
maps.
—. 2023a. Complete Electric Motor ALPHA Maps, https://www.epa.gov/vehicle-and-fuel-
emissions-testing/combining-data-complete-emotor-alpha-maps.
—. 2022. "Contract No. 68HERC20D0014, Order No. 68HERC21F0317. Powertrain and
Vehicle Modeling, Sumulation, and Validation."
U.S. EPA. 2022. "Control of Air Pollution from New Motor Vehicles: Heavy-Duty Engine and
Vehicle Standards - Regulatory Impact Analysis, EPA-420-R-22-035."
https://nepis.epa.gov/Exe/ZyPDF.cgi?Dockey=Pl 016A9N.pdf.
U.S. EPA. 2017. "Final Determination on the Appropriateness of the Model Year 2022-2025
Light-Duty Vehicle Greenhouse Gas Emissions Standards under the Midterm Evaluation, EPA-
420-R-17-001." https://nepis.epa.gov/Exe/ZyPDF.cgi?Dockey=P 100QQ91.pdf.
—. 2012. "Final Rulemaking for 2017-2025 Light-Duty Vehicle Greenhouse Gas Emission
Standards and Corporate Average Fuel Economy Standards." August:
https://www.regulations.gOv/document/EPA-HQ-OAR-2021-0208-0100.
https://www.regulations.gOv/document/EPA-HQ-OAR-2021-0208-0100.
U.S. EPA. 2018. "Office of Air Quality Planning and Standards 2014 National Air Toxics
Assessment (NATA) Technical Support Document (TSD)."
https://www.epa.gov/sites/default/files/2018-
09/documents/2014_nata_technical_support_document.pdf.
U.S. EPA. 2016. "Proposed Determination on the Appropriateness of the Model Year 2022-2025
Light-duty Vehicle Greenhouse Gas Emissions Standards under the Midterm Evaluation, EPA-
420-R-16-020." https://nepis.epa.gov/Exe/ZyPDF.cgi?Dockey=P 100Q3DO.pdf.
—. 2022b. Technical Publications and Presentations Concerning Benchmarking. Accessed 11 17,
2022. https://www.epa.gov/vehicle-and-fuel-emissions-testing/benchmarking-advanced-low-
emi s si on-li ght-duty-vehicl e-technol ogy.
—. 2023. The 2023 EPA Automotive Trends Report: Greenhouse Gas Emissions, Fuel
Economy, and Technology since 1975, EPA-420-R-23-033. Accessed January 16, 2024.
https://www.epa.gov/system/files/documents/2023-12/420r23033.pdf.
U.S. EPA. 2018. "Tier 3 Certification Fuel Impacts Test Program, EPA-420-R-18-004."
https://www.epa.gov/moves/tier-3-certification-fuel-impacts-test-program.
U.S. EPA, CARB, U.S. DOT NHTSA. 2016. "Draft Technical Assessment Report: Midterm
Evaluation of Light-Duty Vehicle Greenhouse Gas Emission Standards and Corporate Average
Fuel Economy Standards for Model Years 2022-2025, Report No. EPA-420-D-16-900."
https://nepis.epa.gov/Exe/ZyPDF. cgi/P100OXEO.PDF?Dockey=P 100OXEO.PDF.
3-172
-------
USGS. 2020. "Cobalt Deposits in the United States." U.S. Geological Survey, June 1.
—. 2022. "USGS Updates Mineral Database with Graphite Deposits in the United States." U.S.
Geological Survey, February 28.
van Nieuwstadt, M., A. Shah, E. Serban, and D. Martin. 2019. "Regeneration Strategies for
Gasoline Particulate Filters." SAE Paper 2019-01-0969. doi: 10.4271/2019-01-0969.
Warkins, Jason, Tinghong Tao, and SongLyu. 2020. "Application of Low-Mass Corning®
FLORA® Substrates for Cold-Start Emissions Reduction to Meet Upcoming LEV III SULEV30
Regulation Requirement." SAE Technical Paper 2020-01-0652. SAE. doi: 10.4271/2020-01-
0652.
Wood Mackenzie. 2022. Battery & raw materials - investment horizon outlook to 2032. Wood
Mackenzie.
—. 2023. "How will China's graphite export controls impact electric vehicle supply chain?,
subscriber material presentation." November 2.
Zhu, Shengrong, Jeffrey Hollowell, Kyoung-Pyo Ha , Nicholas Fantin, and Mark Shirley. 2013.
"A Method to Reduce Cold Start Emissions while Shortening Fast Idle Catalyst Light-off Time."
SAE Technical Paper 2023-01-0247. SAE. doi: 10.4271/2023-01-0247.
3-173
-------
Chapter 4: Consumer Impacts and Related Economic Considerations
This chapter discusses the impacts of the rule on consumers and related economic
considerations. Regarding consumer impacts, we examine the implications of the Final Standards
on consumers from three frames of reference, namely the purchase decision, the ownership
experience, and social benefits and costs. These three perspectives overlap but also differ in
important ways, which we discuss in this chapter. In addition, these three frames of reference
relate to EPA OMEGA modeling (see RIA Chapters 2, 3 and 8) and inform EPA's analysis of
costs and benefits (see RIA Chapter 9). Furthermore, the impacts of this rule on consumers affect
projections of vehicle sales and consequently inform EPA's employment analysis, which we also
discuss in this chapter.
In our representation of the purchase decision, we include costs that consumers incorporate
into their purchase decision; how consumers respond to costs; and how consumer perceptions of
technologies change or do not change over time. In the discussion of the ownership experience,
we focus on vehicle use and on consumer savings and expenses for BEVs, PHEVs, and ICE
vehicles across three body styles. Specifically, we present projected savings and expenses for
average MY 2032 vehicles at the time of purchase and averaged over the first eight years of
vehicle life. In the discussion of consumer-related costs and benefits, we include components of
social costs and benefits that are included in the benefit-cost analysis and that have direct
consumer impacts. In the discussion of vehicle sales, we explain how sales impacts are modeled,
as well as show how total vehicle sales are expected to change. We conclude with a discussion of
employment impacts in which we discuss potential impacts of the growing prevalence of electric
vehicles, present a quantitative discussion of partial employment impacts on sectors directly
impacted by this rule, and discuss potential employment impacts on other related sectors.
4.1 Modeling the Vehicle Choice within the Purchase Decision
In this section, we focus our discussion on our modeling of the consumer choice within the
purchase decision. We address the decision to buy or not buy in RIA Chapter 4.4. The LD
vehicle purchase decision is a complex process (Jackman, et al. 2023, 12-14) (Taylor and Fujita
2018). Consumers consider and value a wide array of vehicle attributes and features as they
develop and seek to satisfy their purchase criteria (Fujita, et al. 2022). Body style is a particularly
important consumer criterion. Individuals tend to consider vehicles within a body style (Fujita, et
al. 2022). Thus, we model vehicle choice within body style, namely sedan and wagons, CUVs
and SUVs, and pickups and across other vehicle attributes. The relative shares of body styles
over time are taken from the Annual Energy Outlook 2023 as described in RIA Chapter 8.1.
Value, as in "value for the money," is also among the most compelling vehicle attributes that
consumers consider (Fujita, et al. 2022, 748 & 754) (Jackman, et al. 2023). "Value" is a multi-
factor consideration; it includes factors such as purchase, fueling, maintenance, and repair costs,
wholly or in part. Thus, these costs play an important role in consumers' decision processes as
does consumer sensitivity to those costs, which we capture and quantify in our analyses. Also
important to the vehicle purchase decision, but harder to capture and quantify, are consumers'
diverse perceptions of other vehicle attributes. These more subjective assessments pertain to
vehicle attributes such as comfort, design, image, and performance (Fujita, et al. 2022, 748) as
well as to technology (e.g., ICE vehicle, PEV) where decision rules (e.g., compensatory, non-
compensatory), attitudes (e.g., technological affinity), and psychological biases (e.g., risk and
uncertainty aversion, loss aversion) may be at play (Taylor and Fujita 2018, 37).
4-1
-------
In the following discussion of consumers' decision processes, we narrow our focus and
modeling to the following key elements: costs that consumers incorporate into their purchase
decisions (i.e., purchase, fueling, maintenance, repair, and depreciation); how consumers respond
to costs (i.e., logit parameters); and how consumer perceptions of technologies change or do not
change over time (i.e., share weight parameters). In addition, as enablers of consumer acceptance
of PEVs mature and expand rapidly, we expect that electrification of the light-duty vehicle
market will also accelerate dramatically. Thus, we specifically attend to the choice consumers
will increasingly make among BEVs, PHEVs, and ICE vehicles by estimating the proportions of
new vehicle sales expected to be BEVs, PHEVs, and ICE vehicles. In our modeling, methods are
the same for all body styles and powertrains, though the inputs may differ.
During the LD vehicle purchase decision process, consumers reference a wide variety of
information during the stages of vehicle purchase. This includes what consumers believe they
already know and what they learn from other parties (e.g., friends, family) and external sources
(e.g., vehicle labels, websites). From one consumer to the next and from one purchase to another,
this information varies in type, quality, and precision. However, there are always both cost and
non-cost elements to the vehicle purchase decision. We center our representation of the purchase
decision around those two categories.
4.1.1 Cost Elements of the Purchase Decision
Regarding cost elements, in addition to initial expenses associated with purchase (e.g.,
purchase price, incentives), consumers incorporate into their purchase decisions reasonably good
estimates of the number of miles they expect to drive per year, fueling and charging expenses,
and other ownership costs. In our modeling of the consumer's purchase decision, we distill
consumer costs into cost per mile, which we also refer to as consumer generalized cost. We
discuss non-cost components of the purchase decision in section 4.1.2.
Generalized cost consists of purchase price, purchase incentive, fueling expenses, and non-
fueling ownership costs spread over time (i.e., annualized) and over miles traveled. Total cost per
mile allows consumers, as represented in the model, to compare vehicles. It is important to note
that consumer generalized cost reflects the perception and expectations of consumers (See
Consumer Module in RIA Chapter 2.2.2), not producers' expectations of consumers (See
Producer Module in RIA Chapter 2.2.2 and RIA Chapter 4.4). In our modeling and as in the real
world, consumers and producers make decisions with different information. Consumer
generalized cost also is not meant to align with costs calculated within the effects module of
OMEGA (See Effects Module in RIA Chapter 2.2.2) and therefore is not reflective of the values
used in our benefit cost analysis (See RIA Chapter 4.3 and RIA Chapter 9).
To calculate the consumer generalized cost used in modeling the consumer's purchase
decision, we annualize purchase price over 7 years using a 10% discount rate (Zabritski 2023)
(Martin 2023).58 See RIA Chapter 2.5 for the derivation of vehicle purchase price. In addition,
we assume that consumers roughly approximate that annual VMT is 15,000 miles, which is
consistent with current measures of annual VMT. In our modeling, consumers apply "rules of
thumb" when incorporating annual non-fuel ownership costs into their purchase decision. Annual
58 As points of reference, many new vehicle buyers finance their purchase with terms typically range from 24 to 84
months and average 68 months (Zabritski 2023) (Martin 2023).
4-2
-------
non-fueling ownership costs for BEVs are $1,600, annual non-fuel ownership costs for PHEVs
are $1,800, and annual non-fuel ownership costs for ICE vehicles are $2,000. These amounts are
consistent with information consumers will see on the fuel economy and environment label, and
they are consistent with OMEGA estimates of these non-fueling ownership costs. Furthermore,
via the fuel economy and environment label, consumers also have implicit information regarding
the recharging and refueling efficiency of BEVs, PHEVs, and ICE vehicles, which we capture in
our modeling of the consumer purchase decision.59 Recharging efficiency is 0.9, and refueling
efficiency for liquid fuels is 1. See Table 4-1 for a summary of operating cost inputs to consumer
generalized cost.
To determine fuel cost component of generalized cost, we note that BEV drivers consume
only off-board grid electricity as they accumulate miles; ICE vehicle drivers consume only liquid
fuels as they accumulate miles; and PHEVs drivers consume off-board grid electric and liquid
fuels as they accumulate miles. Individual PHEV drivers consume off-board grid electricity and
liquid fuels in proportions based on their own driving, fueling, and charging behaviors. The
fraction of distance that a PHEV operates in charge depleting mode (i.e., using predominantly
electricity) is called the PHEV utility factor. In our modeling and as discussed in Chapter 3, the
mechanism that is used to apportion the benefit of a PHEV's electric operation for purposes of
determining the PHEVs contribution toward the fleet average GHG requirements is the fleet
utility factor (FUF). We use the finalized FUF curve as part of the consumer purchase decision
as well. Here, we use the FUF to model the PHEV fueling costs that consumers incorporate into
their purchase decisions. FUF also appears in Table 4-1.
Table 4-1: Operating cost inputs to consumer generalized cost.
Powcrtrain
Annuali/-
Annuali/-
Annual
Annual
Refueling
Recharging
Fleet
ation
ation
VMT
Non-Fuel
Efficieney
Efficiency
Utility
Period
Rate
(miles)
Ownership
Factor
(years)
(%)
Costs
BEV
7
10
15.000
$1,600
NA
0.9
1
PHEV
7
10
15.000
$1,800
1
0.9
See
Chapter 3
ICE
7
10
15.000
$2,000
1
NA
0
Applying the information in Table 4-1, we calculate total cost per mile (aka consumer
generalized cost) with a series of related equations. We begin the series of related equations with
total cost per mile, which is separated into fuel costs per mile and non-fueling costs per mile, as
in the equation below.
Total cost per mile = fuel costs per mile + nonfueling costs per mile
Fueling cost per mile depends on fuel cost, fuel economy, refueling efficiency, recharging
efficiency, and utility factor. Calculating fuel costs per mile, though similar, clearly differs
across BEVs, PHEVs, and ICE vehicles due to the relevant energy sources (i.e., electricity and
59 The fuel economy and environment label is affixed to every new light-duty vehicle sold in the United States. The
test procedures used to determine MPGe and kWh per 100 miles for BEVs take into account charging inefficiencies.
4-3
-------
liquid fuel), units (i.e., kilowatt hours and gallons), refueling and recharging efficiency, and
utility factor.60
In general, fuel costs per mile are given by the following:
fuel costs per mile
= (fleet utility factor) * (cost per kWh * kWh per mile)
-h recharging efficiency + (1 — fleet utility factor)
* (cost per gallon * gallons per mile) -h refueling efficiency
where fleet utility factor, refueling efficiency, and recharging efficiency are populated according
to Table 4-1, specific to each powertrain. Because the FUF is effectively 1 for BEVs and 0 for
ICE vehicles, fuel costs per mile for BEV and ICE vehicles simplify to
fuel costs per mileBEV = (cost per kWh * kWh per mile) -h recharing efficiency
and
fuel costs per mileICEV
= (cost per gallon * gallons per mile) -h refueling efficiencyICEV
respectively.
Note that because recharging efficiency is 0.9, which is less than 1, dividing by recharging
efficiency increases electricity cost per mile. For liquid fuels, refueling efficiency is 1 and has no
effect on liquid fuel cost per mile.61
Non-fueling ownership costs per mile include annualized up-front capital costs and annual
non-fueling costs like maintenance and repair. Up-front capital costs include the vehicle
purchase price and purchase incentives and are generated in OMEGA (See RIA Chapter 2.6 for a
full description of purchase price and purchase incentives). To annualize capital costs over a 7-
year time period, we first calculate the annualization factor using a 10% discount rate.
annualization factor = rate * (1 + — r-r— . . .)
J V (1 + rate)ttmePertod_1/
Then, we multiply the annualization factor and capital costs to determine annualized capital
costs.
annualized capital costs = annualization factor * captial costs
The remaining non-fueling ownership costs are given as annual values in Table 4-1. Thus,
total annualized non-fueling costs are the sum of annualized capital costs and annual non-fueling
ownership costs. We then calculate non-fueling costs per mile by dividing total annualized non-
60 Note that throughout the equations in the chapter, we will be abbreviating ICE vehicle to ICEV.
61 To estimate gallons per mile in OMEGA, we divide estimated onroad grams of CO2 per mile by the estimate fuel
carbon intensity.
4-4
-------
fueling costs by estimated annual vehicle miles.62 The following equation shows these
calculations.
nonfueling costs per mile
annualized capital costs + annual nonfueling ownership costs
annual VMT
With the above equation, we have the fueling and non-fueling costs per mile used to
determine total cost per mile given by the first equation in this section. We reiterate that total
cost per mile is the cost component of the consumer decision process, which we also refer to as
consumer generalized cost. It reflects the perception and expectations of consumers during the
purchase process, not producers, and is not meant to be perfectly consistent with values used in
our benefit cost analysis.
4.1.2 Estimating Market Shares
4.1.2.1 Consumer Choice
Total sales are determined as described in Chapter 4.4.63 Here we focus on how we model
consumer choice. We model the choice of vehicle technologies (ICE, PHEV, or BEV) within
body style - namely sedan and wagons, CUVs and SUVs, and pickups - and across other vehicle
attributes. The relative shares of body styles over time are taken from the Annual Energy
Outlook 2023 as described in RIA Chapter 8.1. Most vehicle attributes are implicit in the
generalized cost, but we make the choice among vehicle technologies (i.e., BEVs, PHEVs, and
ICE vehicles) explicit. Our modeling attends to powertrain (i.e., BEVs, PHEVs, and ICE
vehicles) for several reasons, foremost among them are the "consumer facing" nature of PEV
technology and the rapid growth in PEV acceptance observed and expected (See Chapter
4.2.2.2).64 By "consumer facing," we mean that the vehicle technology (i.e., BEVs, PHEVs, and
ICE vehicles) is clearly apparent to consumers in addition to the vehicle attributes associated
with the technology (e.g., noise, acceleration, convenience).
Based on the distinct differences in refueling procedures that are readily apparent, we assume
that consumers view BEV, PHEVs, and ICE vehicles as three distinct choices. In other words,
PHEVs are not any more like a BEV than they are like an ICE vehicle.65 Thus, we calculate the
62 OMEGA also includes a dollar adjustment factor where needed to ensure that costs estimated in OMEGA are in a
consistent dollar year. Specifically, in the calculation of nonfueling costs per mile, the sum of annualized capital
costs and annual nonfueling ownership costs is also divided by a dollar adjustment factor. This ensures that costs are
estimated in 2022 dollars. The dollar adjustment factor is estimated using the GDP Implicit Price Deflator published
by the U.S. Bureau of Economic Analysis (see Chapter 2.6.7 of this RIA).
63 EPA's OMEGA model estimates total vehicle production and sales separately from BEV, PHEV, and ICE vehicle
market shares. In short, sales are based on EI A sales projections with limited revisions related specifically to this
rule. See Chapter 2 and Chapter 4.4.
64 We note that expanding the representation of choice beyond powertrain and generalized costs introduces orders of
magnitude more complexity that available data does not support.
65 Commenters Jeremy Michalek et al. from Carnegie Mellon and Yale Universities cautioned us regarding the
Independence of Irrelevant Alternatives (IIA) Property in which the introduction of a new choice draws
proportionally from existing alternatives [EPA-HQ-OAR-2022-0829-0705, p. 11], We have revised OMEGA since
these comments were made in several ways, including the introduction of PHEVs into the choice structure. We have
also re-calibrated the model and the tested its internal consistency and externally validity.
4-5
-------
proportions of BEVs, PHEVs, and ICE vehicles as one calculates weighted averages, and
proportions of BEVs, PHEVs, and ICE vehicles are given by the market share equations below.
weightBEV
market shareBEV = —— —
weightBEV + weightPHEV + weightICEV
weightPHEV
iriCLTKCt ShCLT6pfjf;y — " " "
weightBEV + weightPHEV + weightICEV
weightICEV
market shareICEV =
weightBEV + weightPHEV + weightICEV
The representation of choice appears in the weight components of the above equations, which
we explain in the weight equations below. Consumer choice includes costs and non-cost
elements, represented separately as consumer generalized costs (i.e., total costs per mile) and
consumer heterogeneous response to costs (i.e., logit parameter) and consumer perceptions of
technology overtime (i.e., shareweight parameters).
We first describe consumer choice conceptually. Consumers match vehicle attributes to
purchase criteria in their purchase decision (Fujita, et al. 2022). In our modeling, the vehicle
attributes we incorporate into consumer choice are represented by an estimate of generalized
consumer cost, as derived in Chapter 4.1.1. Generalized consumer cost creates a comparable
metric across all vehicles; this monetization of purchase price and ownership costs implicitly
includes consumer valuation of vehicle attributes. Thus, generalized consumer cost effectively
provides an ordering of vehicle alternatives within body styles.
When presented with identical products, a hypothetical consumer will select the lower priced
item. In reality, vehicle attributes and features differ, as do consumers and their purchase criteria.
Consumers purchase comparable vehicles over a range of prices. The logit parameter is specified
to allow for market penetration of lower and higher vehicle costs. We discuss the logit parameter
in more detail in Chapter 4.1.2.2. Consumers also change over time, and so do their perceptions
of vehicle technologies. Shareweight parameters represent the remaining non-cost elements of
the consumer purchase decision, in aggregate (i.e., on average for all consumers over all non-
costs elements) and over time for each technology (i.e., BEV, PHEV, and ICE vehicle). In other
words, shareweight parameters are the numerical representation of consumer acceptance
discussed in Preamble Section IV.C.6. We discuss the shareweight parameters in more detail and
present shareweight parameter values in Chapter 4.1.2.2.
Mathematically, we use a conventional logit formulation to model vehicle technology weights
resulting from the consumer decision process. This formulation employs one variable - total cost
per mile and two types of parameters - the logit and shareweights.
weightBEV = shareweightBEV * (total cost per mileBEV)logit
weightPHEV = shareweightPHEV * (total cost per milePHEV)l°ait
weightICEV = shareweightICEV * (total cost per mileICEV)logit
4-6
-------
To determine market shares, the shareweight is multiplied by the exponential term in the
weight equations above. Effectively, the shareweight mediates the effect of total cost per mile
term on the distribution of consumer purchase decisions.66 For example, a shareweight of 1 has
no mediating effect whereas shareweights greater or less than one do. Shareweights complete the
weight calculations for BEVs, PHEVs, and ICE vehicles (i.e., the second set of equations in this
section), and therefore, the calculation of BEV, PHEV, and ICE vehicle market shares (i.e., the
first pair of equations in this section). Market shares are multiplied by total vehicle sales, per
Chapter 4.4, to arrive at the OMEGA's estimate of BEV, PHEV, and ICE vehicle sales.
Before closing out this section, we note that the logit formulation that we use here reflects
well-established methods in the transportation, policy, and vehicle choice scientific literatures.
The Global Change Analysis Model (GCAM) also uses the logit formulation to represent
economic choice (JGCRI n.d.) and is among leading approaches to modeling the future of
electric vehicles (Taylor 2023). EPA has adopted the GCAM terminology of "shareweight" to
represent relative consumer acceptance of different technologies choices (JGCRI n.d.) However,
terminology varies. The logit formulation we utilize in OMEGA is also an example of random
utility discrete choice models in which shareweights are called alternative specific constants.
Commenters Jeremy Michalek et al. reviewed "over 200 automotive demand model studies in
the scientific literature and government reports [and] found random utility discrete choice models
to be the dominant paradigm for modeling automotive demand, with the logit model and its
variants most commonly used. Thus, EPA's demand model follows the dominant paradigm and
inherits the properties of random utility discrete choice models."67
4.1.2.2 Consumer Response to Costs and Acceptance of Technology
As stated above, there are always cost and non-cost elements to the vehicle purchase decision.
We center our representation of the purchase decision around those two categories. The cost
component is given by consumer generalized costs (i.e., total costs per mile) and is discussed in
Chapter 4.1.1. The non-costs components are given by the logit and shareweight parameters,
applied in the equations above and discussed in more detail below.
4.1.2.2.1 Response to Costs: More on the logit parameter
Because consumers are diverse, consumers purchase comparable vehicles over a range of
prices. The logit is specified to represent market penetration of lower and higher cost vehicle
technologies. In other words, some consumers will purchase a lower-cost vehicle technology and
others will purchase a higher-cost vehicle technology. The distributions of purchases of higher-
and lower-cost technologies overlap. Specifically, we apply a logit exponent of -8 to total cost
per mile to achieve this effect. This value is based on logit values utilized in other choice models
and consistent with the shareweight parameters we discuss below. The negative sign of the logit
parameter reflects that alternatives are represented with costs (i.e., lower cost items are generally
preferred) and the magnitude of the logit parameter regulates the degree to which relative costs
affect choice among BEVs, PHEVs, and ICE vehicle technologies.
66 For example, a shareweight of 1 has no mediating effect whereas a shareweight greater or less than one does.
Note, however, shareweight values are meaningful only relative to each other since they appear in the numerator and
denominator of the market share calculation by way of the technology weight equations.
67 [EPA-HQ-OAR-2022-0829-0705, p. 10]
4-7
-------
4.1.2.2.2 Acceptance of Technology: Background and Shareweight Estimates
We capture the evolution of consumer acceptance of vehicle technologies using parameters
called shareweights. Shareweights are the numerical representation of consumer acceptance
discussed in Section IV.C.6 of the preamble. More precisely, shareweights represent the non-cost
elements of the consumer purchase decision in aggregate (i.e., on average for all consumers over
all non-costs elements) and over time for each technology (i.e., BEV, PHEV, and ICE vehicle).
These non-cost elements include internal and external characteristics of individuals and
households (e.g., attitudes, demographics), vehicle attributes not included in generalized cost,
and conditions of the physical, social, economic, and governmental systems (e.g., charging
stations, neighborhood effects).
In the following, we first motivate our quantitative and technology-based representation of
consumer acceptance (i.e., shareweight values). Then we present the shareweight values used in
our analysis.
4.1.2.2.2.1 Research on Consumer Acceptance of Light-Duty PEVs
Our modeling separately attends to powertrain (i.e., BEVs, PHEVs, and ICE vehicles). PEV
technology is "consumer facing," meaning that the technology is clearly apparent to consumers.
EPA in coordination with the Lawrence Berkeley National Laboratory (LBNL), conducted a
comprehensive review of the scientific literature regarding consumer acceptance of PEVs. That
effort culminated in a peer-reviewed report on PEV acceptance in which EPA and LBNL
organize and summarize the enablers and obstacles of PEV acceptance evident from the
scientific literature (Jackman, et al. 2023). The review concluded that "there is no evidence to
suggest anything immutable within consumers or inherent to PEVs that irremediably obstructs
acceptance." More simply put, the enablers of PEV acceptance are external to the person. With
the evolution of the environment in which people make decisions (e.g., infrastructure,
advertising, access) and advancements in technology and vehicle attributes (e.g., range, body
style, price), widespread acceptance of PEVs is very likely to follow.
Since the conclusion of that review EPA has stayed abreast of even more recent research
regarding PEV acceptance. Foremost among those studies are the recent third-party projections
of PEV market shares. EPA reviewed several recent reports and studies containing PEV
projections, all of which include the impact of the IRA; none consider the impact of this rule.
Altogether, these studies project PEV market share in a range from 42 to 68 percent of new
vehicle sales in 2030. The mid-range projections of PEV sales from these studies, to which we
compare our No Action case, range from 48 to 58 percent in 2030. We discuss third-party
estimates in more detail in Chapter 4.1.2.2.2.5 and Figure 4-2: Moderate third party PEV market
shares with IRA.Figure 4-2.
In addition, C. Forsythe et al. (2023b), found that when consumers' basic demands for vehicle
attributes are met, they accept or prefer BEVs to combustion vehicles. They conclude that
"BEVs could constitute the majority or near-majority of cars and SUVs by 2030, given
widespread BEV availability and technology trends" and "with the assumed technological
innovations, even if all purchase incentives were entirely phased out, BEVs could still have a
market share of about 50% relative to combustion vehicles by 2030, based on consumer choice
4-8
-------
alone." Via public comments, Jeremey Michalek et al.68 describes a working paper in which they
conducted a discrete choice experiment of representative sample of 52 pickup truck owners
across the U.S. They found that 78% pickup truck owners are open to purchasing electric pickup
trucks if the technology provides sufficient performance attributes. For 43% of respondents,
vehicle choice is driven primarily by what the vehicle can do, rather than its powertrain type (C.
Forsythe, K. Gillingham, et al., Will pickup truck owners go electric? 2023a).
In a recent report, LBNL used current, publicly available data to demonstrate that a
substantial number of non-BEV owners already exhibit some of the key enabling characteristics
of current BEV adopters. They primarily examined the enablers of the "access" component of
acceptance, including characteristics of individuals and households. For example, 47 percent of
U.S. households own a single-family home with reasonable charging capabilities, making the
convenience and savings associated with residential charging feasible for nearly half of U.S.
households. Furthermore, in another recent report, LBNL challenges emergent rules of thumb
regarding PEV acceptance (e.g., wealthy, urban, male). (Taylor, Fujita and Campbell 2024 )
Their work suggests that there is untapped demand among mainstream vehicle buyers that the
conventional wisdom regarding who buys and who doesn't buy PEVs does not serve. For
example, they note that early PEVs were not well-positioned to appeal to a large segment of the
population. Specifically, most early EVs were hatchbacks in a market where hatchbacks
represent a small portion of sales generally. In addition, vehicle buyers tend to consider and
purchase vehicles with the same body style (e.g., many buyers only consider SUVs) and certain
body styles are more or less common in different locations (e.g., pickups tend to be more
common in rural areas).
4.1.2.2.2.2 Market Observations
In vehicle markets, PEV market shares have historically been in the single digits. However,
PEV sales have grown rapidly and are expected to continue to grow rapidly and robustly in
response to substantial progress in key market enablers of PEV adoption (Jackman, et al. 2023),
namely increased PEV sales, increased PEV choice, expanding infrastructure, declining PEV
prices, and production and purchase incentives. For example, annual sales of light-duty PEVs in
the U.S. have grown robustly and are expected to continue to grow. PEVs reached 9.8% of
monthly sales in January 2024 and were 9.3% of all light-duty vehicle sales in 2023, up from
6.8%) in 2022. (Argonne National Laboratory 2024) This robust growth combined with vehicle
manufacturers' plans to expand PEV production strongly suggests that PEV market share will
continue to grow rapidly. Also, the number of PEV models available to consumers is increasing,
meeting consumers demand for a variety of body styles and price points. Specifically, the
number of BEV and PHEV models available for sale in the U.S. has increased from about 24 in
MY 2015 to about 60 in MY 2021 and to over 180 in MY 2023, with offerings in a growing
range of vehicle segments.69 Data from JD Power and Associates shows that MY 2023 BEVs and
PHEVs are now available as sedans, sport utility vehicles, and pickup trucks. In addition, the
greatest offering of PEVs is in the popular crossover/SUV segment (Taylor, Fujita and Campbell
2024 ). In addition, the expansion of charging infrastructure has been keeping up with PEV
adoption as discussed in Section IV.C.4 of the preamble and RIA Chapter 5. This trend is widely
68 [EPA-HQ-OAR-2022-0829-0705, p. 8]
69 Fueleconomy.gov, 2015 Fuel Economy Guide, 2021 Fuel Economy Guide, and 2023 Fuel Economy Guide.
4-9
-------
expected to continue, particularly in light of very large public and private investments.
Furthermore, while the initial purchase price of BEVs is currently higher than for most ICE
vehicles, the price difference is likely to narrow or become insignificant as the cost of batteries
fall and PEV production rises in the coming years.70 (Slowik, Isenstadt, et al. 2022) Among the
many studies that address cost parity, an emerging consensus suggests that purchase price parity
is likely to be achievable by the mid-2020s for some vehicle segments and models.71 (Slowik,
Isenstadt, et al. 2022) (ERM 2022) Specifically, the International Council on Clean
Transportation (ICCT) projects that price parity with ICE vehicles will "occur between 2024 and
2026 for 150- to 200-mile range BEVs, between 2027 and 2029 for 250- to 300-mile range
BEVs, and between 2029 and 2033 for 350- to 400-mile range BEVs." The Environmental
Defense Fund notes that "most industry experts believe wide-spread price parity will happen
around 2025." Finally, the Inflation Reduction Act provides a purchase incentive of up to $7,500
for eligible light-duty vehicles and buyers, which is expected to increase consumer uptake of
zero emissions vehicle technology. (Slowik, Searle, et al. 2023)
4.1.2.2.2.3 EPA Analyses: Cross References and Robustness
In addition to scientific research related to consumer acceptance of PEVs, the observed and
expected acceleration of PEV adoption and key enablers in vehicle markets is further
substantiated by EPA's analysis of key enablers of PEV acceptance. These include increased
PEV production (e.g., as seen in the EPA Trends Report (U.S. EPA 2023)); increasing and
expanding private and public charging infrastructure (See RIA Chapter 5); emerging industry
standards (e.g., standardization of charging ports); advances in PEV technology (see RIA
Chapter 3), maturation of supply chains (See RIA Chapter 3.1.4 on critical minerals etc.),
reductions in battery cost (See RIA Chapter 2.5) and PEV manufacturing cost and purchase price
(see RIA Chapters and 4.2.2.1), and PEV production and purchase incentives (See RIA Chapter
2.6 on IRA). Among these interacting systems, we observe several positive feedback
mechanisms that indicate that this system of enablers is robust; in their comments, Consumer
Report called it a "virtuous cycle" in which consumer demand will continue to grow. For
example, more PEV production leads to economies of scale, maturing supply chains, and
intensified competition, which feeds cost declines resulting in more demand, looping back to
more production. More PEV options (e.g., models, body styles) and advances in PEV technology
lead to more sales among consumers whose purchase criteria can be met by PEVs. More PEV
sales leads to more experience with PEVs and more exposure, which leads to more demand and
more sales. More PEV sales leads to demand for infrastructure and charging services that, when
present, lead to more PEV demand and sales.
70 "This analysis does not consider the effect of any available state, local, or federal subsidies and tax incentives for
electric vehicles and their charging infrastructure" (page 30).
71 (ERM 2022) notes the Inflation Reduction Act (IRA), but estimates do not take into account effects of the IRA.
4-10
-------
4.1.2.2.2.4 Diffusion of Innovation
This robust network of PEV enabling systems aligns well with several well-established
models of innovation diffusion and the adoption of new technologies.72 Rogers' (2003) Diffusion
of Innovation Theory is especially well-known and characterizes well the pattern of adoption of
innovations observed through history. (Rogers 2003) Typically, sales of a new technology are
low and increase slowly and unpredictably in what is called the innovator and early adopter
stage, when consumers of the new technology are likely to share an affinity for new technology
as well as other characteristics that support trying something new. However, after the early
adopter stage, adoption increases very quickly, with rapidly accelerating demand as the
technology becomes mainstream. Mainstream adopters are often described as the early and late
majority, and as LBNL's most recent work on PEV acceptance suggests, we are seeing evidence
of the mainstream, early majority consumers entering the market or exhibiting traits consistent
with PEV adopters (Taylor, Fujita and Campbell 2024 ) (Fujita, Campbell and Taylor 2024). As
commenters from Environmental and Public Health Organizations describe, "this S-shaped pace
of technology adoption has been observed for numerous emerging technologies since the early
1900s, including the telephone, the automobile, electricity, refrigeration, clothes washers and
dryers, air conditioning, microwaves, computers, cellphones, and the internet."73
Consistent with this diffusion of innovation framework, we expect PEV adoption to follow
the S-shaped behavior so often observed in the diffusion of innovation and well established via
Diffusion of Innovation Theory. The first question is what does this S-shaped curve actually look
like? Specifically, how steep is the curve; how fast is PEV market share growing? How high is
the curve; what is the highest expected market share? When does growth in market accelerate
and slow down? The second question is what point on the curve represents the current market?
Are we in the early stages of slow and unpredictable growth? Or are we on the cusp of dramatic
growth? With OMEGA, we demonstrate multiple plausible compliance pathways for MY 2027
to MY 2032 vehicles that reflect a segment of an S-shaped adoption curve (See RIA Chapter
12.1).The Central case, which gives our estimate of the most likely future PEV market share
under the standards, shows that we are nearing the end of the early adopter phase, and by 2030,
the early majority (first wave of mainstream new vehicle consumers) will have purchased PEVs.
72 Among the most well-known are Diffusion of Innovation Theory, Theory of Planned Behavior, Theory of
Reasoned Action, and the Technology Acceptance Model. Lee at al. (2019) characterizes them well in the following:
"Diffusion of innovations theory (Rogers 2003) outlines why people adopt innovations, who adopts them, and
adoption rates over time... [including] information about the socio-demographic profile of early adopters. [Other]
consumer innovation adoption models, for example the theory of planned behaviour (Ajzen 1991), theory of
reasoned action (Fishbein and Ajzen 2009), or the technology acceptance model (Venkatesh and Davis 2000) ...
focus on motivational factors or behavioural issues."
73 Honda comments on well-established pattern on technology diffusion as well. They note, however, that these
historical transitions have typically been longer than what they perceived as roughly a decade-long transition to
roughly two-thirds of new vehicle sales being PEVs in EPA's proposal. We respond to these comments in RTC
Chapter 13. Here we note that EPA's projections are of new vehicle sales only. We also note that plug-in electric
vehicle technology has been around for decades; it became available commercially as a passenger vehicle in 2013
with the Nissan Leaf; rapid advancements in the technology and exponential growth in PEV adoption have already
been observed in the market; and the pace of technological innovation and diffusion in general in the 21st century is
much faster than in the 20th.
4-11
-------
In other words, according to our best assessment of current and future vehicle market, PEV
adoption is on the cusp of/has entered a period of very rapid and accelerating growth.
Consistent with the diffusion of innovation framework, we expect aggregate measures of PEV
acceptance to have been low and unpredictable in the early years, then grow and accelerate
rapidly until some level of saturation is reached (e.g., everyone who is ever going to accept or
prefer a PEV actually approves of PEVs, even if they do not own one). We capture the evolution
of consumer acceptance of vehicle technologies using parameters called shareweights. In the
language of models, relative acceptance of vehicle technologies is an input given by the
shareweight parameters, and projected market shares of vehicle technologies is the output that is
based on non-cost elements of the consumer's purchase decision (i.e., shareweight and logit
parameters) as well as on the consumer's estimates of the cost elements of their purchase
decision (i.e., consumer generalized costs).
4.1.2.2.2.5 Shareweights
Below we discuss the shareweights, used in the No Action case, Final Standards, and
Alternatives. Shareweights are the quantitiative representation of consumer acceptance. In
aggregate, shareweights quantitatively represent the non-cost components of vehicle technology
choice over time. As evident above, the shareweights we present here reflect the current state of
the art in terms of the scientific literature on consumer acceptance of PEVs (e.g., (Jackman, et al.
2023) (C. Forsythe, K. Gillingham, et al., Will pickup truck owners go electric? 2023a) (C. R.
Forsythe, K. T. Gillingham, et al., Technology advancement is driving electric vehicle adoption
2023b) (Gillingham, et al. 2023)), existing policy-relevant models and modeling paradigms
(e.g., (Taylor 2023)), and third party estimates (e.g., (Cole, et al. 2023) (IEA 2023) (C. R.
Forsythe, K. T. Gillingham, et al., Technology advancement is driving electric vehicle adoption
2023b) (Bloomberg NEF 2023) (U.S. DOE 2023) (Slowik, Searle, et al. 2023)) as well as
Congressional investments (e.g., BIL, IRA).
Shareweights are the numerical representation of consumer acceptance discussed in Preamble
Section IV.C.6. More precisely, shareweights represent the remaining non-cost elements of the
consumer purchase decision, in aggregate (i.e., on average for all consumers over all non-costs
elements) and over time for each technology (i.e., BEV, PHEV, and ICE vehicle).74 These non-
cost elements include internal and external characteristics of individuals and households (e.g.,
attitudes, demographics), vehicle attributes not included in generalized cost, and conditions of
the physical, social, economic and governmental systems (e.g., charging stations, neighborhood
effects). As with the concept of consumer acceptance, shareweights are meaningful only relative
to other shareweights. Therefore, shareweights should be interpreted only in comparison to the
shareweights of other vehicle technology choices. To facilitate this, we treat ICE vehicle
shareweights as the reference technology and set ICE vehicle shareweights to 1 in all years, in
every scenario. By normalizing the shareweight of ICE vehicles to 1, we can compare all other
shareweights to ICE vehicle shareweights.
Based on the motivating information provided throughout Chapter 4.1.2.2.2, we expect
consumer acceptance of PEVs initially to be lower than for ICE vehicles, meaning that all else
equal, consumers on average tend to prefer ICE vehicles to PEVs. Thus, PEV shareweights are
74 The logit parameter, discussed in Chapter 4.1.2.2.1, also represents non-cost elements of the purchase decision.
4-12
-------
less than 1. However, in 2027 we expect that PEV acceptance will already be in a period of rapid
and accelerating growth, meaning that consumer utility of PEVs, due to factors other than
generalized costs, is improving overtime (e.g., PEV awareness and access, charging
infrastructure, model variety). BEV and ICE vehicle acceptance will likely converge for sedans,
wagons, CUVs, and SUVs by 2030, signifying when consumers will be indifferent, on average
between BEV and ICE vehicles of the same generalized cost. In quantitative terms, BEV
shareweights increase to 1 for sedans, wagons, CUVs, and SUVs around 2030. In other words,
BEV and ICE vehicle acceptance converge for sedans, wagons, CUVs, and SUVs, signifying
when consumers will be indifferent, on average between BEV and ICE vehicles of the same
generalized cost. All else equal, as in C.R. Forsythe et al. (2023b), BEV could comprise 50% of
new vehicle sales by 2030 for "cars and SUVs."
Our modeling also reflects conditions in the new vehicle market specific to pickups and
PHEVs. While wagons, sedans, CUVs, and SUVs are largely, though not exclusively, used to
transport passengers and their stuff (i.e., cargo), pickups more often fulfill needs such as hauling
and towing. Though research indicates that hauling and towing in an EV reduces mileage by
about the same rate as in ICE pickup, the impact of towing and hauling is often mentioned as a
reason consumers might not want to purchase an EV pickup.75 As a result, there may be a subset
of pickup consumers who perceive BEVs as not acceptable. To capture this, BEV pickup
shareweights increase over time, though not as rapidly as for other body styles, and do not
converge with ICE vehicle shareweights; BEV pickup shareweights are always less than 1. In
other words, on average, BEV pickups will be chosen less often than ICE pickups with the same
cost to consumers. PHEV acceptance is expected to be notably lower overall than for BEV and
ICE vehicles. Based on observed market uptake of PHEVs, the complexity of PHEV technology,
the possible perception of PHEVs as stopgap or transitory technology, and the greater tendency
of PHEV buyers to convert to BEVs in subsequent purchases than BEVs to PHEVs (Lee, et al.
2023), average acceptance of PHEVs is modeled with shareweights that rise over time but are
much lower than for BEV and ICE vehicles. Under these expectations, PHEVs are expected to
be less preferred on average relative to BEV and ICE vehicles. We represent this numerically
with shareweights that increase less quickly than BEV shareweights reaching a maximum of .5,
half that of BEV and ICE vehicles.
Importantly, in this analysis, we treat shareweight parameters as exogenous to the standards.
We recognize, as some commenters also noted, that the standards themselves are likely to
motivate faster relative acceptance of PEVs by way of the multiple positive feedback
mechanisms associated with increased production of PEVs (e.g., increased exposure, economies
of scale, reduced costs and prices, expanding infrastructure, rising familiarity). However, while
we expect that increased production of PEVs will most likely be a part of automakers
compliance strategies, we have also demonstrated several compliance pathways with different
levels of PEV market share (See Preamble Section IV and RIA Chapter 12). Therefore, we apply
the same shareweight values for the No Action case, the Final Standards, and Alternative cases.
Similarly, with the exception of the Acceptance sensitivity cases (e.g., RIA Chapter 4.1.3) and
Manufacturer Compliance Pathway sensitivity cases (pathways 'B' and 'C' in preamble I.B.I,
IV.F, and IV.G), we use the same shareweight values for all remaining Sensitivity case analyses.
75 For example, see (Georgiou 2022).
4-13
-------
Table 4-2 shows shareweight values by body style for BEVs, PHEVs, and ICE vehicles. BEV
shareweights differ by body style whereas shareweights for PHEVs and ICE vehicles do not.
Figure 4-1 illustrates the shareweight values for light-duty vehicles.
Table 4-2: Central case shareweight values for light-duty vehicles.
BEV
PHEV
ICE
Calendar
Sedans/Wagons
CUV/SUV
Piekups
All body
All body styles
Year
styles
2022
0.69
0.69
0.11
0.08
1.00
2023
0.77
0.77
0.15
0.10
1.00
2024
0.83
0.83
0.21
0.13
1.00
2025
0.88
0.88
0.28
0.17
1.00
2026
0.92
0.92
0.36
0.22
1.00
2027
0.94
0.94
0.45
0.28
1.00
2028
0.96
0.96
0.54
0.34
1.00
2029
0.97
0.97
0.62
0.40
1.00
2030
0.98
0.98
0.69
0.44
1.00
2031
0.99
0.99
0.75 ""
0.47
1.00
2032
0.99
0.99
0.79
0.49
1.00
2033
0.99
0.99
0.83 "
0.49
1.00
2034
1.00
1.00
0.85
0.50
1.00
2035
1.00
1.00
0.86
0.50
1.00
2036
1.00
1.00
0.88
0.50
1.00
2037
1.00
1.00
0.88
0.50
1.00
2038
1.00
1.00
0.89
0.50
1.00
2039
1.00
1.00
0.89
0.50
1.00
2040
1.00
1.00
0.90
0.50
1.00
2041
1.00
1.00
0.90
0.50
1.00
2042
1.00
1.00
0.90
0.50
1.00
2043
1.00
1.00
0.90
0.50
1.00
2044
1.00
1.00
0.90
0.50
1.00
2045
1.00
1.00
0.90
0.50
1.00
2046
1.00
1.00
0.90
0.50
1.00
2047
1.00
1.00
0.90
0.50
1.00
2048
1.00
1.00
0.90
0.50
1.00
2049
1.00
1.00
0.90
0.50
1.00
2050
1.00
1.00
0.90
0.50
1.00
4-14
-------
/"
- — .
. — . _
/
/
/
/
/
/ ,<
2022 2027 2032 2037 2042 2047
Sedans/Wagons —• - BEV Pickups
& CUVs/SUVs
PHEVAII bodv styles » ICE All body styles
Figure 4-1: Central case shareweight values for light-duty vehicles.
The BEV and PI-IEV shareweights shown in Figure 4-1 were developed by EPA as calibrated
values using the generalized logistic form.76 By calibrating shareweight values specifically for
this analysis rather than, for example, using values directly from GCAM or other choice models,
we are ensuring consistency with EPA's other modeling assumptions such as the projected state
of ICE and PEV technologies, production constraints, consumer acceptance, charging
infrastructure, etc. Our approach to calibration involved determining the appropriate relative
position of shareweights by body style and technology, determining appropriate value bounds,
and finally, appropriate shareweight values.
For the final standards, we have revised BEV shareweights from the values used in the
Proposal. A recent study (C. R. Forsythe, K. T. Gillingham, et al., Technology advancement is
driving electric vehicle adoption 2023b) and recent BEV SUV sales following the introduction of
SUV BEV models demonstrates that acceptance of BEV CUVs, and SUVs is similar to BEV
sedans and wagons. Another recent study projects stronger consumer acceptance of BEV pickups
than estimated for the NPRM while simultaneously acknowledging the possibility that for some
locations and use cases, BEV pickups may not become acceptable (C. Forsythe, K. Gillingham,
et al., Will pickup truck owners go electric? 2023a). Thus, BEV acceptance for pickups is now
represented with larger shareweights in the near term than in the NPRM but also with maximum
76 We use the generalized logistic form in the calibration of shareweights. Specifically, Y(t) = u , where t
(C-Qe flr)
is time and Y (t) is shareweight at time t.
4-15
-------
shareweight values below 1. PHEV and therefore PHEV shareweights were added to our
analysis. As stated above, PHEV acceptance is expected to be notably lower overall than for
BEV and ICE vehicles. Thus, average acceptance of PHEVs is modeled with shareweights that
rise over time but are much lower than for BEV and ICE vehicles. In addition, and in the absence
of evidence to the contrary, shareweights for PHEVs are the same across body styles. We
examine PHEV shareweights further via alternative pathways. See preamble sections I.B.I, IV.F,
and IV. G and RIA Chapter 12.1.
In addition, for this final rulemaking analysis, consistent with the approach used for the
proposal, we calibrated shareweight values for the FRM so that the overall BEV, PHEV, and ICE
vehicle market shares produced by the OMEGA model in the No Action case are consistent with
third party estimates. EPA reviewed several recent studies to support this calibration. Unlike the
proposal, all of the third-party projections reviewed for the calibration of shareweights include
effects of the IRA. Among these third-party estimates, there is a range of assumptions that vary
across these studies such as consumer adoption, state level policies, financial incentives,
manufacturing capacity and vehicle price. We took care to account for the differing assumptions
underlying third party estimates. Altogether, these studies project PEV sales spanning a range
from 42 to 68 percent of new vehicle sales in 2030 (Cole, et al. 2023) (Wood, et al. 2023). The
mid-range third party estimates of PEV market shares range from 48 to 58 percent in 2030 (Cole,
et al. 2023) (IEA 2023) (C. R. Forsythe, K. T. Gillingham, et al., Technology advancement is
driving electric vehicle adoption 2023b) (Bloomberg NEF 2023) (U.S. Department of Energy,
Office of Policy 2023) (Slowik, Searle, et al. 2023). We show the mid-range estimates in Figure
4-2.
4-16
-------
0,6
0.5
0.4
aj
ro
to
0.3
0.2
0.1
0
*
*
*
s
* ...*
~ ~
_
,%'s
/'¦ / 1
y /
* J*
~
/ ]
ar / /
* / JT
* ' /
/ /
i/
~
~
~
m
2023 2024 2025 2026
X Bloomberg
¦ DOE - Moderate
— • — Slowik- Moderate
2027 202S
Year
~ Cole - EV -14
-•— Forsythe
> FRM Central
2029 2030 203 1 2032
a Cole - 110
¦ IEA
Figure 4-2: Moderate third party PEV market shares with IRA.
As already noted, the shareweights used in the No Action case, Final Standards, and
Alternatives reflect the current state of the art in terms of the scientific literature on consumer
acceptance of PEVs, existing policy-relevant models and modeling paradigms, and calibration to
third party estimates as well as Congressional investments (e.g., BIL, IRA). We refer to
shareweights discussed above and given in Table 4-2 and Figure 4-1 as the Central case. Below
we examine other representations of BEV acceptance via the "faster BEV acceptance" and
"slower BEV acceptance" cases presented below in Chapter 4.1.3.
4.1.3 BEV Acceptance Sensitivities
Though the Central case shareweights quantitatively estimate the most likely path of PEV
acceptance based on the research and analyses documented in RIA Chapter 4.1.2.2.2, we
4-17
-------
acknowledge that the pace of PEV acceptance could be faster or slower than the assumed in our
Central case analysis. Therefore, we also evaluated sensitivity scenarios to explore the impact of
other shareweight assumptions. We conducted sensitivities of both faster and slower consumer
acceptance of BEVs. In a Faster BEV Acceptance case, BEV acceptance could rise very quickly
and exceed acceptance of ICE vehicles by orders of magnitude. For sedans, wagons, CUVs, and
SUVs this could mean that, within just a few years, BEV acceptance will match and then surpass
that of ICE vehicles. In other words a consumer is just as willing or more likely to choose a BEV
than an ICE vehicle, all else equal. In fact, recent evidence suggests that BEVs may already be
preferred, all else equal (Gillingham, et al. 2023). Specifically, Gillingham et al. (2023)
examined all new LD vehicles sold in the U.S. between 2014 and 2020 and compared existing
electric vehicles to their most similar ICE vehicle counterpart. They found that BEVs are
preferred to the ICE counterpart in some segments. In addition, a survey from Consumer Reports
in 2022 indicates that more than 70 percent of survey respondents felt that BEVs are as good or
better than ICE vehicles, up from about 46 percent in 2017 (Bartlett 2022).
Thus, under our Faster BEV Acceptance case, we assume that BEV acceptance rises
dramatically and exceeds ICE vehicle acceptance for all body styles by 2027. Table 4-3 and
Figure 4-3 show shareweight values for faster BEV acceptance by body style. In this sensitivity,
the value of shareweights for BEVs are larger for all body styles. Values for PHEV and ICE
vehicle shareweights remain unchanged from the Central case, meaning that relative level of
acceptance for and vehicles is lower in the "faster BEV acceptance" case.
4-18
-------
Table 4-3: Faster BEV acceptance shareweight values by body style for light-duty.
BEV
Calendar Year
Sedans/Wagons
CUV/SUV
Piekups
2022
0.72
0.72
0.24
2023
0.94
0.94
0.34
2024
1.19
1.19
0.46
2025
1.45
1.45
0.62
2026
1.70
1.70
0.80
2027
1.93
1.93
1.00
2028
2.14
2.14
1.20
2029
2.32
2.32
1.38
2030
2.47
2.47
1.54
2031
2.60
2.60
1.66
2032
2.69
2.69
1.76
2033
2.77
2.77
1.83
2034
2.82
2.82
1.89
2035
2.87
2.87
1.92
2036
2.90
2.90
1.95
2037
2.93
2.93
1.96
2038
2.95
2.95
1.98
2039
2.96
2.96
1.98
2040
2.97
2.97
1.99
2041
2.98
2.98
1.99
2042
2.98
2.98
2.00
2043
2.99
2.99
2.00
2044
2.99
2.99
2.00
2045
2.99
2.99
2.00
2046
2.99
2.99
2.00
2047
3.00
3.00
2.00
2048
3.00
3.00
2.00
2049
3.00
3.00
2.00
2050
3.00
3.00
2.00
4-19
-------
0.00
2022 2027 2032 2037 2042 2047
Sedans/Wagons — ¦ — Pickups
& CUVs/SUVs
Figure 4-3: Faster BEV acceptance shareweight values by body style for
LD vehicles.
Though it appears to be very unlikely given the evidence for BEV acceptance, we
acknowledge that BEV acceptance may be slower than in the Central case as suggested by some
commenters. Jackman et al. (2023) discusses some of the issues new vehicle buyers might
encounter, such as lack of familiarity with PEVs and uncertainty about charging infrastructure.
As we discuss in Chapter 5.3.1, large investments in charging infrastructure from the private
sector and the U.S. government via the BIL and IRA are intended to address these uncertainties
over time. Nevertheless, in characterizing "slower" acceptance, we assume that ICE vehicles will
be preferred to BEVs until much later than in the Central case, all else equal. Thus, shareweights
for all body styles are less than 1 until 2044, and shareweights for pickups do not exceed 0.5. We
also assume that BEV acceptance for all body styles grows less rapidly in the early to mid-
2030's, roughly coincident with the expiration of IRA producer and consumer incentives.
Table 4-4 and Figure 4-4 show slower BEV acceptance shareweight values by body style for
light-duty vehicles. Note that the absolute value of PHEV and ICE vehicle shareweights remain
unchanged, but their relative level of acceptance compared to BEVs is higher in the "slower
BEV acceptance" case than in the Central case.
4-20
-------
Table 4-4: Slower BEV acceptance shareweight values by body style for light-duty.
BEV
Calendar Year
Sedans/Wagons
CUV/SUV
Piekups
2022
0.13
0.13
0.01
2023
0.16
0.16
0.02
2024
0.20
0.20
0.03
2025
0.24
0.24
0.04
2026
0.29
0.29
0.05
2027
0.35
0.35
0.07
2028
0.41
0.41
0.10
2029
0.48
0.48
0.13
2030
0.56
0.56
0.17
2031
0.63
0.63
0.21
2032
0.71
0.71
0.25
2033
0.77
0.77
0.29
2034
0.83
0.83
0.33
2035
0.88
0.88
0.37""
2036
0.91
0.91
0.40
2037
0.94
0.94
0.43
2038
0.96
0.96
0.45
2039
0.97
0.97
0.46
2040
0.98
0.98
0.47
2041
0.99
0.99
0.48
2042
0.99
0.99
0.49
2043
0.99
0.99
0.49
2044
1.00
1.00
0.49
2045
1.00
1.00
0.49
2046
1.00
1.00
0.50
2047
1.00
1.00
0.50
2048
1.00
1.00
0.50
2049
1.00
1.00
0.50
2050
1.00
1.00
0.50
4-21
-------
2022 2027 2032 2037 2042 2047
Sedans/Wagons — * — Pickups
& CUVs/SUVs
Figure 4-4: Slower BEV acceptance shareweight values by body style for light-duty.
4.2 Ownership Experience
Having described how we model the consumer purchase decision in Chapter 4.1, we turn to
the estimated effects of the final standards on individual consumers. In this section, we focus
specifically on the ownership experience of vehicle consumers, including vehicle miles traveled
and rebound effect, consumer savings and expenses, and other ownership considerations. A
discussion of some aspects of consumer-related social benefits and costs appears in Chapter 4.3.
4.2.1 Vehicle Miles Traveled and Rebound Effect
Critical to estimating the impacts of emissions standards is the number of vehicle miles
traveled (VMT). In the proposed rule, we acknowledged that individual vehicle miles vary. (U.S.
EPA 2021) In our analyses, aggregate vehicle miles are determined exogenously (see RIA
Chapter 8 for details). While measures and estimates of VMT for ICE vehicles is well-
established in previous EPA LD rules, and described in RIA Chapter 8, how much consumers
drive their PEVs has been changing as the technology evolves and PEVs become more common.
Thus, in the following discussion, we give particular attention to electric vehicle miles traveled
(eVMT).77
The rebound effect is the means by which aggregate VMT is influenced by the policy
alternatives. The rebound effect generally refers to the additional energy consumption that may
77 Raghavan and Tal (2021) define eVMT "as the miles driven by off-board grid electricity"; we define eVMT as the
miles produced with electricity drawn from a source external to the vehicle.
4-22
-------
arise from the introduction of a more efficient, lower cost energy service. In this and previous
rules, the rebound effect was defined as additional miles traveled in response to a lower cost of
driving. Previous rules estimated the rebound effect based on changes in fuel cost per mile,
without distinguishing between vehicles with different fuel sources. With a growing number of
PEVs, we acknowledge the importance of possible differences in rebound across vehicle
technologies and depending on the energy source or sources used to produce miles. PHEVs, for
example, produce both miles from liquid fuel and electricity in proportions that depend on
consumers' charging and use behaviors. BEVs produce only electric miles, and ICE vehicles
produce only combustion miles.
Importantly, the rebound effect offsets the energy savings benefits of efficiency
improvements to some degree. Because rebound driving consumes fuel and generates emissions,
the magnitude of the rebound effect influences actual fuel savings and emission reductions that
will result from the standards. Furthermore, rebound driving provides value to the consumer if
they choose to drive more. We discuss these costs and benefits in Chapter 4.3, and in Section
VIII of the Preamble. In this chapter, we address miles driven and rebound.
4.2.1.1 Basis for Vehicle Miles Traveled for Battery Electric Vehicles
The eVMT literature consists of a handful of studies, including the very recent studies listed
in Table 4-5. Two of the listed studies are based on California data collected by UC Davis
researchers, and both find that annual VMT for PEVs is similar to annual VMT for ICE vehicles
(Chakraborty et al. 2022; Raghavan and Tal 2021). The four other studies, using New York,
California, and national data, find that annual VMT for PEVs is less than annual VMT for ICE
vehicles (Zhao, et al. 2023) (Nehiba 2024) (Burlig, et al. 2021) (Davis 2019). These studies offer
a similar summary of the pre-existing data and research related to annual PEV miles in the U.S.
Namely, though lower cost per mile has historically been associated with more driving, this has
not been observed for PEVs, for which the cost of driving per miles is lower than for ICE
vehicles.78 Instead, average annual VMT for PEVs has historically been lower than for ICE
vehicles. This observation has been attributed to the shorter range of first generation PEVs,
typically less than 100 miles just five or six years ago, as well as to substitution across vehicles
for multiple vehicle households, and to the typical type of households who bought an electric
vehicle in the time frame of the data (Zhao, et al. 2023) (Chakraborty, Hardman and Tal 2022)
(Davis 2022) (Raghavan and Tal 2021) (Davis 2019), that is, households with characteristics
correlated with lower VMT regardless of vehicle technology (Chakraborty, Hardman and Tal
2022). This area of research continues to face several challenges including the relatively low
market penetration and uneven distribution of PEVs; the rapid evolution of PEV technology and
the PEV market; and the relative difficulty in obtaining comprehensive and up-to-date data on
how PEVs are driven (Burlig, et al. 2021) (Chakraborty, Hardman and Tal 2022) (Nehiba 2024)
(Jackman, et al. 2023). As a result, the data that are available for empirical analyses are not likely
to be representative of the current and future general population of car buyers and their driving
behavior.
78 See Chapter 11.2.3 of this RIA, which compares fueling costs for PEVs and ICE vehicles within its discussion of
energy security.
4-23
-------
Table 4-5: Recent scientific studies of eVMT.
Study
(Chakrabortv, Hardman and Tal 2022)
(Raghavan and Tal 2021)
(Zhao, et al. 2023)
(Nehiba 2024)
(Burlig, et al. 2021)
(Davis 2019)
Annual I kctric VM T Results
"Overall, we observe that factors influencing
PEV VMT are like those observed for
conventional gasoline vehicles. We find that
PEVs drive a similar amount as conventional
vehicles, not less ... as some have
suggested."
Average annual eVMT by vehicle model
ranged from 10,841 for the Nissan Leaf to
17.236 for the Tesla Model S with 80kWh
battery capacity.
"BEVs have accumulated fewer annual miles
than conventional gasoline vehicles ...
BEVs with larger ranges were driven more,
but increasing range has diminishing returns
in terms of higher annual mileage... BEV
sensitivity to operating costs was also less
than other powertrains."
"The average BEV in New York is driven
9,060 miles/year, substantially less than the
10,910 miles/year average for all passenger
cars and light truck s in New York in 2019
... However BEV mileage increased rapidly
across vehicle model years, converging
toward ICEV mileage...This convergence
could be explained by technological
improvements...Particularly, rapid increases
in battery ranges appear to play an important
role, but the relationship between range and
mileage flattens for higher range."
"... our estimates [of overall household
electricity load around EV registration
events] indicate that EV load in California is
surprisingly low. ... Given the fleet of EVs
in our sample, and correcting for the share of
out-of-home charging, our estimates
translate to approximately 1,700 electric
vehicle miles traveled (eVMT) per year for
plug-in hybrid EVs (PHEVs) and 6,700
eVMT per year for battery EVs (BEVs).
These eVMT values are substantially less
than internal combustion engine (ICE)
VMT."
"These data show that electric vehicles are
driven considerably less on average than
gasoline- and diesel-powered vehicles. In the
complete sample, electric vehicles are driven
an average of 7,000 miles per year,
compared to 10,200 for gasoline and diesel-
powered vehicles. The difference is highly
statistically significant and holds for both
all-electric and plug-in hybrid vehicles, for
both single- and multiple-vehicle
households, and both inside and outside
California."
Dsitsi Description
Location: California
Years: 2015-2019
Sources: Two surveys with reported
odometer readings and on-board recorders
Number of PEVs: 16,736 (survey) and 369
(on board recorders)
Location: California
Years: 2015 (survey and logger), 2017
(survey only), 2019 (logger only)
Sources: GPS loggers on vehicles in two-car
households and online survey
Number of households: 73
Location: United States
Years: 2016-2022 (model vears 2055 to
2020)
Source: Odometer readings from 12.5
million used cars and 11.4 million used
SUVs listed between 2016 and 2022.
Vehicle Ages: 2 to 9 years
Number of PEVs: 304.954 cars and 13.245
SUVs
Location: New York
Years: January 2017 - January 2021
Sources: Annual vehicle safety inspection
odometer readings; residential and
residential electricity prices matched by zip
code
Location: California
Year: 2014-2017
Sources: 10 percent of residential electricity
meters in the Pacific Gas and Electric
(PG&E) utility territory (362,945
households) merged with EV registration
records (63,765 vehicles)
Number of PEVs: 57,290
Name: 2017 National Household Travel
Survey
Location: United States
Year: 2017
Source: Survey with reported annual vehicle
miles
Number of PEVs: 862
4-24
-------
Based on these study results and the transparency with which they communicate data
limitations, there is no evidence that PEVs have been driven more than ICE vehicles, and study
results conflict regarding whether annual eVMT has been less for PEVs. None of these studies
accurately estimate current and future VMT for PEVs. EPA concludes that the existing empirical
evidence does not support the conclusion that current or future average annual eVMT differs or
will differ from annual VMT for ICE vehicles. Therefore, EPA uses the same annual VMT for
PEVs and ICE vehicles throughout our analyses.
4.2.1.2 Basis for the Rebound Effect for Internal Combustion Engines and PHEVs
In the 2021 rule, EPA provided a summary of the historical and recent literature on the light-
duty (LD) vehicle rebound effect, the ways it is defined (e.g., direct, indirect, economy-wide,
short- to medium-run, long-run), how it is estimated, and the basis for the rebound effect used for
internal combustion engine vehicles (ICE vehicles). Based on that review and assessment of
studies of the LD rebound effect, EPA used a single point estimate of 10 percent for the direct,
short- to medium-run rebound effect for ICE vehicles in the 2021 rule. In this current rule, EPA
is again using a value of 10 percent as an input to the agency's analyses for the direct, short- to
medium-run rebound effect for ICE vehicles. Similarly, EPA is using the point estimate of 10
percent for the direct, short- to medium-run rebound effect for PHEVs. We refer the reader to the
2021 rule RIA Chapters 3.1 and 8.3.3 and Preamble Section 1 for the full discussions of the
rebound effect and the point estimate used. (U.S. EPA 2021). See RIA Chapter 4.2.1.3 regarding
the rebound effect for BEVs.
4.2.1.3 Basis for Rebound Effect for Battery Electric Vehicles
As described briefly above, the rebound effect for BEVs is the additional miles traveled in
response to a lower cost of driving. First, we do not model increasing energy efficiency for BEVs
over time. EPA has identified three current studies that estimate a rebound effect for BEVs in the
U.S., which we list in Table 4-6. Results of the three studies are mixed. Using odometer readings
from millions of used cars and SUVs, Zhao, Ottinger, Yip, and Helveston (2023) found that
"BEV sensitivity to operating costs was less ... than other powertrains." Using data gathered
from California PEV drivers, Chakraborty, Hardman, and Tal (2022) find no evidence of an
eVMT rebound effect. Nehiba (2024) finds a rebound effect of 10 percent in an analysis of the
"entire BEV population in New York." Nehiba (2024) also finds that the responsiveness of
eVMT to electricity prices "falls as public charging stations - where prices are often decoupled
from electricity costs - become available."
4-25
-------
Study
(Zhao, cl al. 2023)
Table 4-6: Recent scientific studies of eVMT rebound.
Electric Rebound Results
" ...BEV sensitivity to operating
costs was also less than other
powertrains."
(Chakraborly. Hardman and Tal
" 2022)
(Nchiba 2024)
"Moreover, while lower electricity
price at home may lead to a higher
share of PEV VMT in total
household VMT, we do not identify
the presence of 'rebound effect'."
"A 10% increase in per mile
| residential| electricity costs
reduces mileage by 1%," but "BEV
drivers become less responsive to
residential prices when public
charging stations ... become
available."
Data
Location: United States
Years: 2016-2022 (model years
2055 to 2020)
Source: Odometer readings from
12.5 million used cars and 11.4
million used SUVs listed between
2016 and 2022.
Vehicle Ages: 2 to 9 years
Number of PEVs: 304.954 cars and
13.245 SUVs
Location: California
Years: 2015-2019
Sources: Two surveys with reported
odometer readings and on-board
recorders
Number of PEVs: 16,736 (survey)
and 369 (on board recorders)
Location: New York
Years: January 2017 - January
2021
Sources: Annual vehicle safety
inspection odometer readings;
residential and residential
electricity prices matched by zip
code
Given the mixed results of these studies and the historical evidence that BEVs are not driven
more than ICE vehicles, EPA assumes no eVMT rebound in our analyses. In other words, we
assume no additional electric vehicle miles in response to factors such as lower electricity prices,
higher BEV efficiency, or higher liquid fuel prices. See RIA Chapter 4.2.1.2 regarding the
rebound effect for ICE vehicles and PHEVs.
4.2.2 Consumer Savings and Expenses
Though manufacturers' production costs are projected to increase to meet the light-duty
standards (See Preamble Section IV.C and RIA Chapter 12), the standards are projected to save
individual consumers thousands of dollars. EPA estimates that the standards will save an average
consumer more than $6,000 over the lifetime of a light-duty vehicle, as compared to a vehicle
meeting the MY 2026 standards.79 These savings emerge as a result of modest advancements in
ICE-based vehicle technology as well as substantial increases in PEV market share. See RIA
Chapter 4.1.2.2.2.2 regarding market observations of PEV market share and PEV enablers.
In the following, we first present summaries of projected consumer savings and expenses
associated with the standards for sales-weighted average model year 2032 vehicles over a vehicle
lifetime of 24 years. Then we present consumer savings and expenses for sales-weighted average
model year 2032 PEVs and ICE vehicles over the first 8 years of vehicle life, which is the
79 This estimate excludes vehicle taxes and any investment in home charging installation.
4-26
-------
current average amount of time the first owner has possession of the vehicle (Blackley n.d.)
(Autolist 2022).80'81 For both 24-year (i.e., full vehicle lifetime) and 8-year (i.e., average duration
of first ownership) summaries, we show average consumer savings and expenses for three body
styles - sedans and wagons, CUVs and SUVs, and pickups. Specifically, we provide OMEGA
estimated national, sales-weighted average expenses associated with new model year 2032
vehicles. Consistent with OMEGA and EPA's benefit cost analysis, the EPA estimated dollar
amounts are given in 2022 dollars (2022$) with no discounting. Other dollar amounts are
consistent with original sources and noted. In addition, we group expenses based on whether they
occur with vehicle purchase (i.e., upfront), reoccur (i.e., average annual), or represent an
optional, one-time household investment (i.e., one-time optional). For upfront purchase-related
items and one-time household investments, we present the full amounts. For recurring expenses,
we present sales-weighted average annual, undiscounted amounts.
Importantly, these expenses and savings represent a subset of ownership savings, expenses,
incentives, and investments that meet the following criteria: the savings and expenses can
reasonably be expected to be experienced by buyers of MY 2032 vehicles; data and generally
accepted conventions exist to estimate reasonably precise quantities; and bounds on uncertainty
can be established. These criteria imply the exclusion of savings and expenses that vary
substantially across individuals and/or locations and can be calculated for a specific person or
place based on information in the table and other readily available information.
Furthermore, this discussion should not be interpreted as a "total cost of ownership" analysis
but as a summary of average MY2032 vehicle expenses and savings that fit the above criteria,
across body styles and powertrains, and under the final standards.82 Lastly, these consumer
ownership savings and expenses should not be confused with the societal costs and benefits that
appear in RIA Chapter 4.3 and RIA Chapter 9. That is, the presentation here is not meant to
reflect the incremental costs of the standards, nor is it meant to reflect the Benefit Cost Analysis
or net benefits of the standards. Instead, it is meant to inform the public of the average out-of-
80 The average vehicle ages at which original owners sell their cars, SUVs, and pickup trucks were 8.4, 8.3, and 8.7
years, respectively, according to a study conducted by iSeeCars.com. "iSeeCars.com analyzed more than 5 million
5-year-old or older used cars sold by their original owners between Jan. 1, 2014 and Dec. 31, 2018. Models which
were owned for less than 5 years were excluded from the analysis, to eliminate the effect of short lease terms on the
data. Models that were in production for less than 9 of the 10 most recent model years (2010 to 2019), heavy-duty
trucks and vans, and models no longer in production as of the 2019 or 2020 model years were also removed from
further analysis. The average age of each vehicle, defined as nameplate + bodystyle, was mathematically modeled
using the ages of cars when they were first listed for sale" (Blackley n.d.). In contrast, Argonne National
Laboratories (Burnham, et al. 2021, 116) state that the typical period of initial ownership is "approximately 5 years"
without citation.
81 According to S&P Global, the average age of vehicles on U.S. roadways is approximately 12 years (S&P Global
Mobility 2022). Argonne National Laboratory "analyzed survivability rates from data published by the National
Highway Traffic Safety Administration (NHTSA) of the U.S. Department of Transportation (Lu 2006) and by the
EPA (EPA 2016), finding the average lifetime of a vehicle in the United States was approximately 14 years in 2006
and just under 16 years in 2016" (Burnham, et al. 2021, 24).
82 Argonne National Laboratory provides a comprehensive and recent summary of total cost of ownership (TCO) of
vehicles, which includes a review of other TCO studies (Burnham, et al. 2021). Total costs of ownership analyses
typically aim to provide a full accounting of ownership costs rather than fitting the criteria specified here.
4-27
-------
pocket expenses that might occur for future vehicle buyers considering a new vehicle purchase
under the final standards.
The expenses and savings included in the subsequent tables are listed below:
• Purchase Price - EPA OMEGA modeled average retail price. OMEGA first estimates
the cost to the manufacturer to produce a given vehicle. Then, the model performs an
iterative search where the Producer offers different combinations of ICEV and BEV
shares and levels of cross-subsidization between BEV and ICE vehicles until the
Consumer and Producer are in agreement for vehicle shares and price. The resulting
prices are defined by the sum of the marked-up vehicle production costs and any
internal cross-subsidies applied by the model. This resulting price, representing a retail
price, is used to compute the sales-weighted average that appears in the table. See RIA
Chapter 2.5 for our cost methodology and RIA 2.6 for a description of producer
incentives. Regarding costs and producer incentives, we discuss the relationship
between incentives, cross subsidies, and purchase price in RIA 2.6.4.5.83
• Average Sales Tax - EPA OMEGA estimated national average sales tax of 5 percent
(Rearick 2023).
Federal Purchase Incentive - The maximum potential consumer purchase incentive provided via
the Inflation Reduction Act is $7,500. The actual purchase incentive any given consumer might
receive is based on several eligibility requirements for the consumer and the actual vehicle. We
included estimates of the average consumer purchase incentives as well, consistent with the
values applied within OMEGA as presented in RIA Chapter 2.6.8. As with producer incentives,
we assume consumers receive the full purchase incentive for which they are eligible. Note that
the purchase incentive is a savings for consumers and appears as a negative value in Table 4-7
and Table 4-9 and serves to reduce consumer costs relative to the purchase price discussed
above.84
• Vehicle Miles - EPA OMEGA estimated national average annual per vehicle miles
traveled. See RIA Chapter 4.2.1 and 8.3.
• Retail Fuel - EPA OMEGA estimated national average annual per vehicle fuel
expense. See RIA Chapters 2.6.6, 4.3.3, and 8.5.
83 Though partial pass through of costs and producer subsidies is possible, the net result of assuming partial pass-
through and the net result of assuming full pass-through could be quite similar. Thus, in the absence of clear
consensus in the scientific literature quantifying individual and net effects of partial pass through, the assumption of
full pass-through for technology costs and for savings associated with producer subsidies to be reasonable.
84 For new LD vehicles, the maximum potential Federal purchase incentive of $7,500 is available on cars priced up
to $55,000 and on vans, SUVs, and pickups up to $80,000 depending on the buyer's income. For used vehicles, the
maximum potential Federal purchase incentive of $4,000 is available on vehicles priced up to $25,000 depending on
the buyer's income.
4-28
-------
• Refueling Time - EPA OMEGA estimated and monetized national average annual per
vehicle refueling time. See Chapter 4.3.5 below for procedure for estimating and
monetizing refueling time.
• Maintenance - EPA OMEGA estimated national average annual maintenance
expenses. See RIA Chapter 4.3.7.
• Repair - EPA OMEGA estimated national average annual repair expenses. See RIA
Chapter 4.3.7.
• Registration - National average annual vehicle registration fee according to Burnham,
Gohkle, et al. (2021) is $68 for ICE vehicles. The national average fee is $141 for
PEVs, which includes the additional fee (e.g., excise) charged in some locations for
PEVs.
• Insurance - EPA OMEGA estimated national average annual insurance expenses. See
RIA Chapter 4.3.6.
• Residential Charging Equipment & Installation - National estimated range of
expenses associated with home charging equipment and installation. In Chapter 5.3 of
the RIA, we provide a description and summary of charging infrastructure
investments, including home charging.
Before we proceed, we note several updates to our analysis of consumer savings and expenses
completed for the FRM, in consideration of public comments and based on the best available
information in the record. Specifically, EPA has added a summary of the 24-year lifetime
expenses and savings for sales-weighted average MY 2032 vehicles, under the final standards
compared to MY 2026 standards, by body style. Table 4-7 and Table 4-8 summarize projected
average consumer expenses and savings associated with the final standards. This information is
in addition to Table 4-9 and Table 4-10, which compare savings and expenses under the
standards, across vehicle technologies, and by body style for the first eight years of vehicle life
(the average span of ownership for a new vehicle). These later two tables appeared in the NPRM
and have been updated for the FRM.
EPA has revised estimates of each line item as a result of updates to our modeling, in
consideration of public comments and based on the best available data in the public record.
Specifically, EPA added PHEVs, revised manufacturing costs and purchase price (See RIA
Chapter 2.5), revised estimates of refueling time (See RIA Chapter 4.3.5), and revised estimates
of rebound driving (See RIA Chapter 8.3.3). In addition, EPA added estimated average sales
taxes at a national average of 5 percent (Rearick 2023) and insurance estimates (See Chapter
4.3.6). EPA shows both the average estimated purchase incentive (See RIA Chapter 2.5.2.1.4 and
RIA Chapter 2.6.8) along with stating the maximum purchase incentive. Furthermore, EPA
updated the range of optional costs associated with the installation of home charging (See RIA
Chapter 5). Finally, EPA updated dollar values to 2022 dollars. We also note that the savings and
expenses summaries show registration expenses that include additional fees for PEVs (e.g.,
excise taxes).
4-29
-------
Relatedly, we emphasize that EPA assumes full pass through of manufacturing costs and of
production tax credits. Manufacturers may choose to cross-subsidize vehicles. However, because
our modeling maintains manufacturing costs across all vehicles within a policy scenario, we
effectively assume full cost pass through when averaged over all vehicles. We arrive at the
assumption of full pass through by not assuming partial pass through at any point in our
modeling. Vehicle prices are defined by the sum of the marked-up vehicle production costs and
internal cross-subsidies applied by the model (See RIA Chapter 2.5 and below in RIA Chapter
4.2.2). The 45X Advanced Manufacturing Production Tax Credit is treated within the modeling
as a reduction in direct manufacturing costs, which in turn is assumed to result in a reduction in
purchase price for the consumer after the application of the Retail Price Equivalent (RPE) factor
(See RIA Chapter 2.5.2.1.4 and RIA Chapter 2.6.8). We have conceptualized the purchase
incentive as a combination of 30D and 45W Clean Vehicle Credits. The resulting purchase
incentive is assumed to be realized entirely by the consumer and does not impact the producer
generalized cost value or the manufacturing cost (See RIA Chapter 2.6.8) or the purchase price.
4.2.2.1 Vehicle Lifetime Savings and Expenses Under the Standards Compared to
the MY2026 Standard
We now compare consumer savings and expenses under the no action case and under final
standards for the 24-year lifetime of sales-weighted average model year 2032 vehicles below. In
Table 4-7 expenses are positive values and savings are negative (i.e., in parentheses). Note also
that the maximum potential Federal Purchase incentive of $7,500 is not shown. The sales-
weighted average purchase incentive for all vehicles (PEV and ICE vehicles) is given instead
(consistent with the values applied in the OMEGA model as presented in RIA Chapter 2.6.8) and
is higher under the final standards than under to the MY 2026 standards due to the greater
number of PEVs purchased under the final standards in the central case.
4-30
-------
Table 4-7: National per vehicle ownership (savings) and expenses for new model year 2032
light-duty vehicles under the No Action case and the Final Standards (2022 dollars)
Sedan/Wagon
cuv/suv
Pickup
All Body Styles
No Action
Final
No
Action
Final
No
Action
Final
No
Action
Final
Upfront Purchase Related Expenses
Purchase Price a
(2022$)
38,100
38,900
44,400
46,900
54,000
55,600
44,900
47,000
Average Sales Tax a
2022$
1,900
2,000
2,200
2,400
2,700
2,800
2,300
2,400
Average Federal Purchase Incentive a
(2022$)
(3,300)
(4,500)
(2,700)
(4,000)
(2,700)
(4,000)
(2,800)
(4,100)
Net Purchase Price
(2022$)
36,700
36,400
43,900
45,300
54,000
54,400
44,400
45,300
Twenty-four Year Average Annual Expenses
Vehicle Milesa
(miles/year)
13,500
13,700
14,000
14,200
14,500
14,700
14,000
14,200
Retail Fuela
(2022$/year)
810
660
1,250
1,010
1,580
1,330
1,230
1,010
Re&eling Timea
(2022$/year)
40
40
60
60
60
50
60
50
Maintenancea
(2022$/year)
1,120
1,050
1,280
1,210
1,390
1,310
1,270
1,200
Repaira
(2022$/year)
640
620
740
730
690
670
710
700
Registration ^
(2022$/year)
120
140
110
130
110
130
120
130
Insurance a
(2022$/year)
490
500
450
460
460
470
460
470
Total Average Annual Expenses
(2022$/year)
3,220
3,010
3,890
3,600
4,290
3,960
3,850
3,560
Optional One-Time Investment
Residential Charging Equipment & Installation0
(2022$)
0 to 5,620
0 to 5,620
0 to 5,620
0 to 5,620
0 to 5,620
0 to 5,620
0 to 5,620
0 to 5,620
a Per OMEGA.
b Per Burnham, Gohlke, et al. (2021) adjusted from 2019$ to 2022$.
c See RIA Chapter 5
We summarize estimated savings associated with the final standards for the 24-year lifetime of
sales-weighted average model year 2032 vehicles in Table 4-8. We do not include vehicle taxes
due to the relatively small differences. In Table 4-8, savings are positive; costs are negative (i.e.,
in parentheses). Note that under the final standards, more PEVs are purchased, so the average
Federal Purchase Incentive received by consumers is larger across all vehicles.
4-31
-------
Table 4-8: Summary of estimated average savings over the 24-year lifetime of light-duty
vehicles under the Final Standards compared to vehicles meeting the MY 2026 standards
(2022 dollars)
Sedan/
Wagon
CUV/SUV
Pickup
All Body
Styles
Additional Purchase Price Expense
(900)
(2,600)
(1,700)
(2,100)
Twenty-four Year Operating Savings
5,300
7,000
7,800
6,800
Additional Average Federal Purchase Incentive
1,100
1,300
1,300
1,300
Maximum Residential Charging Expense
(5,600)
(5,600)
(5,600)
(5,600)
Total Savings with No Purchase Incentive and No Residential Charging Expense
4,400
4,400
6,100
4,700
Total Savings with Average Federal Purchase Incentive and No Residential
Charging Expense
5,500
5,700
7,400
6,000
Total Savings with Maximum Federal Purchase Incentive and No Residential
Charging Expense
11,900
11,900
13,600
12,200
Total Savings or Additional Expense with No Purchase Incentive and with
Maximum Residential Charging Expense
(1,200)
(1,200)
500
(900)
Total Savings or Additional Expense with Maximum Federal Purchase Incentive
and Maximum Residential Charging Expense
(100)
100
1,800
400
Total Savings with Average Federal Purchase Incentive and Maximum
Residential Charging Expense
6,300
6,300
8,000
6,600
In Table 4-8, we observe that estimated savings over the lifetime of the vehicle are substantial
regardless of the size of purchase incentives. Excluding purchase incentives and optional
residental charging expenses, the standards will save an average consumer $4,700 over the
lifetime of a light-duty vehicle, as compared to a vehicle meeting the MY 2026 standards.
Taking into account the estimated fleet-wide average purchase incentives, the final standards
yield average savings of $6,000 over the lifetime of the vehicle and compared to the MY 2026
standards.
4.2.2.2 Eight Year Savings and Expenses Under the Final Standards for PEVs and
ICE Vehicles
Given the projected increase in PEV market share, we narrow our focus to projected expenses
and savings associated with PEVs. Table 4-9 provides a summary of estimated consumer
expenses and savings experienced by individual new vehicle owners of BEVs, PHEVs and
ICEVs for three body styles - sedans and wagons, CUVs and SUVs, and pickups. Expenses are
positive values and savings are negative (i.e., in parentheses). Note also that the maximum
potential Federal Purchase incentive of $7,500 is not shown. As above, we provide OMEGA
estimated national, sales-weighted average expenses associated with new model year 2032
vehicles. For recurring expenses, we present sales-weighted average annual, undiscounted
amounts. EPA estimated dollar amounts are given in 2022 dollars (2022$) with no discounting.
Other dollar amounts are consistent with original sources and noted.
4-32
-------
Table 4-9: National per vehicle ownership expenses for new model year 2032 light-duty
vehicles under the Final Standards (2022 dollars).
Sedan/Wagon
cuv/suv
Pickup
BEV
(Electric)
PHEV
(Plug-in
Hybrid)
ICEV
(Gasoline)
BEV
(Electric)
PHEV
(Plug-in
Hybrid)
ICEV
(Gasoline)
BEV
(Electric)
PHEV
(Plug-in
Hybrid)
ICEV
(Gasoline)
Upfront Purchase Related
Expenses and (Savings)
Purchase Pricea
(2022$)
40,300
44,700
33,400
50,300
49,900
40,400
58,300
59,000
49,800
Average Sales Taxa
(2022$)
2,000
2,200
1,700
2,500
2,500
2,000
2,900
3,000
2,500
Average Federal Purchase
Incentivea
(2022$)
(6,000)
(6,000)
(6,000)
(6,000)
(6,000)
(6,000)
Net Purchase Price
(2022$)
36,300
40,900
35,100
46,800
46,400
42,400
55,200
56,000
52,300
Eight Year Average Annual Expenses
Vehicle Milesa
(miles/year)
15,400
15,600
15,400
16,000
16,400
16,000
17,400
17,600
17,400
Retail Fuela
(2022$/year)
450
1,230
1,340
600
1,470
1,860
840
2,200
2,530
Re&eling Timea
(2022$/year)
40
20
60
60
30
80
50
40
80
Maintenancea
(2022$/year)
530
770
820
570
850
880
670
990
1,050
Repaira
(2022$/year)
280
370
360
340
420
390
320
390
360
Registration'3
(2022$/year)
160
160
80
160
160
80
160
160
80
Insurancea
(2022$/year)
620
650
560
550
540
500
590
600
550
Total Average Annual Expenses
(2022$/year)
2,080
3,200
3,220
2,280
3,470
3,790
2,630
4,380
4,650
Optional One-Time Investment
Residential Charging Equipment
& Installation0
(2022$)
0 to 5,620
0 to
5,620
NA
0 to 5,620
0 to
5,620
NA
0 to 5,620
0 to
5,620
NA
a Per OMEGA.
b Per Burnham, Gohlke, et al. (2021) adjusted from 2019$ to 2022$.
c Per RIA Chapter 5
In the above table, when comparing new BEVs, PHEVs, and ICE vehicles within body style,
we make three general observations. First, on average, BEV owners spend less than half of what
PHEV and ICE vehicle owners spend on fuel, even after accounting for refueling time. Second,
BEV owners also save on maintenance and repair. For all operating expenses, BEV owners,
when compared to PHEV and ICE vehicle owners, save $1,100 per year for sedans and wagons,
$1,200 to $1,500 per year for CUVs and SUVs, and $1,700 to $2,000 per year for pickups. In the
above table we also show a range of investments into residential charging equipment and
installation. Importantly, home charging is not required for PEV ownership, and charging at
home is feasible via a standard 120 volt outlet (aka Level 1 which delivers 2 to 5 miles of range
per hour) or 240 volt outlet (Level 2 which delivers 10 to 20 or more miles of range per hour) (
(Borlaug, et al. 2020) citing (U.S. Department of Energy (DOE) 2020)). In some cases,
4-33
-------
additional equipment or upgrades for vehicle charging may not be needed.85 Charging at home
does deliver convenience. It very likely reduces time spent actively charging, as well as the time-
associated expense, since charging occurs when the vehicle is parked. In fact, Level 2 charging at
home has been shown to be associated with PEV continuance, that is, purchasing a PEV after
relinquishing a previous PEV (Hardman and Tal 2021). When electrical upgrades are desired,
home charging equipment and installation costs differ from one household to the next based
primarily on housing type (e.g., detached, attached, apartment) and type of upgrade required
(e.g., none, outlet upgrade, charger upgrade). Thus, the table provides a range as described in
Chapter 5 of this RIA.
Focusing on PEVs compared to ICE vehicles, we summarize the savings that consumers who
purchase a new MY 2032 LD BEV or PHEV instead of an ICE vehicle can experience over the
first 8 years of vehicle life. See
Table 4-10. On average, PEV consumers can save more than $9,000 in the first 8 years of
PEV ownership compared to an ICE vehicle.86 Those are substantial savings and would be
experienced by a PEV owner whether or not they considered those savings at the time of
purchase.
Table 4-10: Summary of estimated average savings over the first 8 years of light-duty
vehicle life when MY 2032 PEV purchased instead of ICE vehicle (2022 dollars).
Sedan/
Wagon
CUV/SUV
Pickup
All
Bodystyles
Additional Purchase Price Expense
(7,400)
(9,900)
(8,700)
(8,700)
Eight Year Operating Savings
8,100
10,200
13,300
10,600
Additional Average Estimated Federal Purchase Incentive Savings
6,000
6,000
6,000
6,000
Maximum Residential Charging Expense
(5,600)
(5,600)
(5,600)
(5,600)
Total Savings No Purchase Incentive and No Residential Charging Expense
700
300
4,600
1,900
Total Savings with Average Federal Purchase Incentive and No Residential
Charging Expense
6,700
6,300
10,600
7,900
Total Savings with Maximum Federal Purchase Incentive and No Residential
Charging Expense
8,200
7,800
12,100
9,400
Total Savings No Purchase Incentive and with Maximum Residential Charging
Expense
(4,900)
(5,300)
(1,000)
(3,700)
Total Savings Average Federal Purchase Incentive and with Maximum Residential
Charging Expense
1,100
700
5,000
2,300
Total Savings with Maximum Federal Purchase Incentive and Maximum
Residential Charging Expense
2,600
2,200
6,500
3,800
85 The ability to charge at home with at most behavior modification (i.e., electrical access with [at most] behavior
modification) varies among individual households with patterns emerging among housing types and between owners
and renters. The National Renewable Energy Laboratory (NREL) estimates home charging is currently feasible
without any upgrades (i.e., no cost) for 28 to 72% of single dwelling structures (i.e., attached and detached single
family and mobile homes) and for 11 - 40% of multiple dwelling structures (i.e., apartments) (Ge, et al. 2021).
86 This estimate includes the maximum PEV purchase incentive and excludes optional investments in residential
charging.
4-34
-------
In concluding this summary of consumer savings and expenses for new MY2032, we again
note that this is not a total costs of ownership analysis. It also is not meant to reflect the
incremental costs of the standards, nor is it meant to reflect the Benefit Cost Analysis or net
benefits of the rule. According to the criteria that we specified above, we have excluded
expenses that consumers customarily incur that may be included in some total cost of ownership
analysis. For example, we acknowledge but exclude costs associated with financing. While many
buyers finance, loan principle, interest rate, and loan period differ substantially across
individuals. We also exclude regional-, state- and local-level monetary purchase incentives as
well as other regional-, state- and local-level monetary and non-monetary, "perks'Vpolicies
associated with PEV ownership. Regional-, state-, and local-level incentives and policies take
many forms across the U.S., differing in source (e.g., governments, utilities), amount, and
eligibility (Wakefield 2023) (Bui, Slowik and Lutsey 2020) (Greschak, Kreider and Legault
2022), and some may not persist into the timeframe represented in the above.
4.2.3 Other Ownership Considerations
In addition to ownership savings and expenses experienced under the final standards,
provided above in Chapter 4.2.2, and impacts of the final standards on consumers quantified in
benefit-cost analysis, shown below in Chapter 4.3 and in Chapter 9, we also consider the effects
of the final standards on low-income households and on consumers of low-priced new vehicles
and used vehicles. These effects depend, in large part, on countervailing elements of vehicle
ownership experience under the standards, namely a) higher up front, net purchase prices,87 b)
net fuel savings,88 and c) maintenance and repair. The net effect varies across households and as
demonstrated above across vehicle types. However, net fuel savings may be especially relevant
for low-income households and consumers in the used and low-priced new vehicle markets.
First, fuel, maintenance, and repair expenditures are a larger portion of expenses for low-income
households compared to higher income households (Hardman, Fleming, et al. 2021).89 Second,
lower-priced new vehicles have historically been more fuel efficient. Third, fuel economy, and
therefore fuel savings, do not decline as vehicles age even though the price paid for vehicles
typically declines as vehicles age and are resold. Fourth, low-income households are more likely
to purchase lower-priced new vehicles and used vehicles (Hutchens, et al. 2021).
Additionally, PEV purchase incentives are available for new and used vehicles. For new
vehicles, the maximum Federal purchase incentive of $7,500 is available on cars priced up to
$55,000 and on vans, SUVs, and pickups up to $80,000 depending on the buyer's income. For
used vehicles, the maximum Federal purchase incentive of $4,000 is available on PEVs priced up
to $25,000 depending on buyer's income. Lower priced new vehicles and many used vehicles
meet the criteria for the maximum incentive and low-income buyers are more likely, by
definition, to qualify for maximum incentives. Furthermore, we discussed in Chapter 4.2.2 above
and demonstrate in Chapter 4.3 below (See also RIA Chapter 10.2.3.1), BEV maintenance and
87 Per vehicle compliance costs are $1,400 including IRA producer incentives (See Chapter 12).
88 By net fuel savings, we are referring to fuel costs and time spent refueling.
89 In the U.S., according to (Hardman, Fleming, et al. 2021), the lowest income households spend 11.2 percent of
their annual income on fuel, maintenance, and repairs of vehicles compared to all other households that spend 4.5
percent of their annual income on these expenses.
4-35
-------
repair costs are lower. For lower income buyers, lower priced vehicle buyers, and used vehicle
buyers, BEVs' lower maintenance and repair costs may be especially compelling.
Furthermore, most vehicle consumers finance, making access to credit for vehicle purchases
essential. The ability to finance may be of particular concern for low-income households. As
above, the effects of the standards on access to credit is influenced by the potentially
countervailing forces of vehicle purchase and other ownership costs. However, the degree of
influence and the net effect is not clear (See Chapter 8.4 of the 2021 rule). Increased purchase
price and presumably higher loan principal may, in some cases, discourage lending, while
reduced fuel costs may, in some cases, improve lenders' perceptions of borrowers' repayment
reliability.
Finally, while access to conventional fuels can be assumed for the most part, the number and
density of charging stations varies considerably (U.S. Department of Energy 2022). The
expansion of public and private charging infrastructure has been keeping up with PEV adoption
and is generally expected to continue to grow, particularly in light of very large public and
private investments (See RIA Chapter 5) and local level priorities (Bui, Slowik and Lutsey 2020)
(Greschak, Kreider and Legault 2022). This includes home charging events, which are likely to
continue to grow with PEV adoption but are also expected to represent a declining proportion of
charging events as PEV share increases (Ge, et al. 2021). Thus, publicly accessible charging is
an important consideration, especially among renters and residents of multi-family dwellings and
others who charge away from home (Consumer Reports 2022). Households without access to
charging at home or the workplace will likely incur additional charging costs. Thus, among
consumers who rely upon public charging, the higher price of public charging is especially
important. See Chapter 5 of this RIA for a more detailed discussion of public and private
investments in charging infrastructure, and our assessment of infrastructure needs and costs
under this rule. See also Chapter 4.2.2 for information on home charging equipment and
installation costs as well as Chapter 10.2.3.1 for a discussion of charging and home charging
installation for low-income households.
Commenters asserted that PEVs will not satisfy every consumer and urged EPA to "consider
the needs of consumers in all demographics and income levels."90 More specifically, commenters
often note the challenges associated with PEV adoption for specific use cases (e.g., towing,
hauling, long distance driving, and cold weather driving) and certain groups of consumers (e.g.,
rural households, low-income populations, and used vehicle consumers). First, EPA agrees that
consumers are heterogeneous, PEV acceptance will occur at different rates for different
consumers, and consumers will choose to satisfy their needs and preferences with PEVs and ICE
vehicles. These differences among consumers are reflected in our modeling. Based on the
evidence, we capture consumer heterogeneity in our modeling in aggregate and on average via
logit and shareweight parameters. We also note that we demonstrate several compliance
pathways. Second, PEV ownership is feasible and acceptable to some consumers for towing,
hauling, long distance driving, and cold weather driving (e.g., (C. Forsythe, K. Gillingham, et al.,
Will pickup truck owners go electric? 2023a)). Many consumer groups, including lower income
buyers and non-white consumers, purchase PEVs and indicate interest in PEVs.91 Third, we
90 Consumer Reports comment, EPA-HQ-OAR-2022-0829-0728, pp. 10-14.
91 Consumer Reports comment, EPA-HQ-OAR-2022-0829-0728, pp. 10-14.
4-36
-------
expect that PEVs will generate large savings for consumers over the lifetime of the vehicle,
whether purchased new or used. We also see a significant trend toward decreasing consumer
upfront costs for purchasing PEVs as we discuss elsewhere (e.g., See Preamble Sections IV.C.l
and IV.C.6). Lastly, our assessment projects that there will be variation in the types of
technologies that automakers adopt to meet the standards, providing consumers increased fuel
economy and associated fuel savings via PEV and ICE vehicle technologies in the new vehicle
market and eventually also in the used vehicle market. We expect that automakers recognize the
diversity of their consumers and will leverage the flexibilities built into the standards to provide
consumers with PEV and ICE vehicle choices over a wide range of utility and price points.
4.3 Consumer-Related Benefits and Costs
4.3.1 Vehicle Technology Cost Impacts
Table 4-11 shows the estimated annual vehicle technology costs of the final standards and
each alternative, estimated in OMEGA, for the indicated calendar years (CY). The table also
shows the present-values (PV) of those costs and the equivalent annualized values (AV) for the
calendar years 2027-2055 using 2, 3 and 7 percent discount rates.92
Table 4-11: Vehicle technology costs (billions of 2022 dollars)*.
Calendar
Final
Alternative
A1 tern;
Year
A
B
2027
: $2.6
$16
; $2.3
2028
"T $7.3
[ $25
$5.9
2029
$16
; $32
$15
2030
! $23
$36
$21
203 1
$29
S3 5
[ $24
2032
; $30
! $33
! $27
2035
] $55
$54
; $42
2040
; $50
$49
$40
2045
$46
$45
$39
2050
! $42
$43
! $35
2055
I $38
$39
1 $30
PV2
$870
$940
$710
PV3
: $760
$820
$610
PV7
$450
$510
: $360
AV2
$40
$43
: $32
AV3
$39
$43
I $32
AV7
1 $37
$41
1 $30
* Costs exclude consideration of IRA battery lax
credits (IRS 45X) and IRA purchase lax credits (IRS
30D and 45W).
92 For the estimation of the stream of costs and benefits, we assume that after implementation of the MY 2027 and
later standards, the MY 2032 standards apply to each year thereafter.
4-37
-------
We expect the technology costs of the program will result in a rise in the average purchase
prices for consumers, for both new and used vehicles. While we expect that vehicle
manufacturers will strategically price vehicles (e.g., subsidizing a lower price for some vehicles
with a higher price for others), we assume in our modeling that increased vehicle technology
costs will be fully reflected in higher average purchase prices paid by consumers. Note that these
technology cost increases are offset by fuel, maintenance, and repair costs, discussed in Chapter
4.3.4 and Chapter 4.3.7.
4.3.2 Value of Rebound Driving
We present the estimated rebound miles driven in Chapter 8.3.4 of this RIA. Here we discuss
the benefits associated with that rebound driving. Those benefits associated with the final
standards and each alternative are shown in Table 4-12
Table 4-12.
Table 4-12 Drive value benefits of rebound driving
(billions of 2022 dollars) *
Calendar Year
Final
Alternative A
Alternative B
2027
$0,002
$0.0052
$0.0024
2028
$0,042
$0.11
$0,043
2029
$0,081
$0.21
$0,082
2030
$0.12
! $0.32
$0.14
203 1
; $0.16
$0.42
$0.19
2032
7 $0.2
: $0.5
$0.22
2035
$1
i $1.3
"i $0.87
2040
; $2.3
1 $2.5
' $2
2045
S3.3
$3.4
i $3
2050
$4.2
$4.2
I $3.8
2055
! $4.7
i $4.7
$4.3
PV2
$46
$49
$41
PV3
: $38
$41
$34
PV7
$18
; $20
I $17
AV2
i $2.1
V $2.2
$1.9
AV3
: $2
i $2.1
$1.8
AV7
$1.5
i $1.7
$1.3
* Positive values reflect benefits: negative values reflect disbcncfits.
4-38
-------
As discussed above, the assumed rebound effect might occur when an increase in vehicle fuel
efficiency leads people to choose to drive more because of the lower cost per mile of driving.
When we estimate fuel expenditures, we multiply the number of miles driven on a given fuel by
its price per unit, i.e., dollars per gallon for liquid fuels and dollars per kWh for electricity.
Therefore, any reductions in fuel expenditures (fuel savings) associated with a policy include
additional fuel expenditures associated with rebound driving. If we ignored those rebound miles,
the fuel savings would be calculated using the same number of miles in both the policy and no-
action cases but with a lower fuel cost per mile in the policy case.
However, drivers would drive those additional rebound miles only if they find value in them.
The increase in travel associated with the rebound effect produces additional benefits to vehicle
drivers, which reflect the value of the added social and economic opportunities that become
accessible with additional travel. This analysis estimates the economic benefits from increased
rebound-effect driving as the sum of the fuel costs paid to drive those miles and the drive
surplus, which is the additional value that drivers derive from those miles.
The value of the rebound miles driven is simply the number of rebound miles multiplied by
the cost per mile of driving them.
Value of Rebound VMT = VMTrebound
action
The economic value of the increased owner/operator surplus provided by added driving, the
drive surplus, is estimated as one half of the product of the decline in vehicle operating costs per
vehicle-mile and the resulting increase in the annual number of miles driven.
\mileJNnAnti.nn \mile)AMn
VMTrebourld
DriveSurplus =
Thus, the economic benefits from increased rebound driving, called Drive Value, is then
calculated as below.
DriveValue = Value of Rebound VMT + DriveSurplus
Drive value depends on the extent of improvement in fuel consumption and fuel prices, which
depend upon vehicle model year, the calendar year, and the standards being analyzed. Thus, the
value of benefits from increased vehicle use also depends upon model year and calendar year,
and it varies among alternative standards.
4.3.3 Fuel Consumption
Overall, the Final Standards are projected to reduce liquid fuel consumption while
simultaneously increasing electricity consumption as shown in Table 4-13 and
Table 4-14, respectively. These values are generated in OMEGA and used in the benefit cost
analysis described in RIA Chapter 9.
4-39
-------
Table 4-13: Liquid-fuel consumption impacts (billion gallons)
Calendar
Liquid-Fuel Impacts.
Liquid-Fuel Impacts.
Liquid-Fuel Impacts.
Year
Final
Alternative A
Alternative B
2027
-0.07
-0.78
-0.052
2028
-0.48
-2.1
-0.4
2029
-1.5
-3.9
-1.4
2030
-3.
-6.
-2.8
203 1
-5.
-8.3
-4.5
2032
-7.2
-11
-6.6
2035
-17
-20
-15
2040
-30
-32
; -25
2045
-39
-40
f -32
2050
-43
-43
-36
2055
-43
-44
-36
sum
-780
-830
-660
Table 4-14 Electricity consumption impacts
(terawatt hours)
Calendar
Electricity Impacts.
Electricity Impacts.
Electricity Impacts.
Year
Final
Alternative A
Alternative B
2027
0.94
6.9
0.79
2028
4.1
18
3.1
2029
13
34
11
2030
r 27
r 52
23
203 1
47
[' 72
39
2032
67
92
58
2035
150
170
130
2040
260
270
200
2045
s 330
'' 330
250
2050
350
360
270
2055
360
360
270
sum
6.700
7.000
5.200
4.3.4 Monetized Fuel Savings
Table 4-15 shows the undiscounted annual monetized fuel savings associated with the final
rule and each alternative as well as the present value (PV) of those costs and equivalent
annualized value (AV) for the calendar years 2027-2055 using 2, 3, and 7 percent discount rates.
In Chapter 9, we present pretax fuel savings which are used in the benefit cost analysis. In
Chapter 9 we also present transfers, or taxes, associated with fuel expenditure changes and
battery and vehicle purchase credit incentives.
4-40
-------
Table 4-15: Retail fuel expenditure savings (billions of 2022 dollars)*
Calendar Retail Fuel Savings. Retail Fuel Savings. Retail Fuel Savings.
Year
Final
Alternative A
Alternative B
2027
$0.21
: $1.7
$0.17
2028
$1.1
$4.5
$1
2029
: $3.2
i $8.3
$3.1
2030
: $6.3
$13
$6.1
203 1
: $10
! $17
$9.5
2032
$14
i $22
$14
2035
| $35
: $42
$31
2040
! $66
T $71
$56
2045
$87
; $90
T $75
2050
i $100
$110
$91
2055
$110
$110
$96
PV2
$1,200
$1,300
$1,100
PV3
$1,000
$1,100
$890
PV7
$520
$580
$450
AV2
1 $57
$61
$49
AV3
$54
$58
$46
AV7
; $42
; $47
j "$37
* Positive values indicate savings in fuel expenditures.
4.3.5 Benefits Associated with the Time Spent Refueling
More stringent GHG standards have traditionally resulted in lower fuel consumption by liquid
fueled vehicles. Provided fuel tanks on liquid fueled vehicles retain their capacity (i.e., gas tanks
don't change volume), the lower fuel consumption would be expected to reduce the frequency of
refueling events. However, if manufacturers choose to maintain traditional range (i.e., miles
traveled on a full tank of fuel), then the possibility exists that tank capacities would become
smaller and, therefore, the frequency of refueling events would not change, although time spent
at the fuel pump may be reduced. There are indications that both outcomes are happening, with
some vehicles reducing tank sizes while others are maintaining them.
Of course, electric vehicles are not fueled in the same way. Many refueling events for electric
vehicles would be expected to occur either overnight where the vehicle is parked or during the
workday using an employer owned charge point, neither of which require extra time from the
driver, especially compared to refueling a liquid fueled vehicle. Similarly, drivers may opt to use
public charging while shopping or at other places they regularly spend time. However, some
charging events will undoubtedly require drivers to take extra time to charge, especially when
drivers are in the midst of an extended road trip. These mid-trip charging events are the focus of
this analysis. For purposes of this analysis, we have made the simplifying assumption that
PHEVs will not make use of mid-trip charging since the vehicle can continue to operate on liquid
fuel once the battery is depleted. Table 4-16 presents our estimates of the benefits associated
with the time spent refueling.
4-41
-------
Table 4-16: Benefits associated with changes to the time spent refueling
(billions of 2022 dollars) *.
Calendar Year
Final
Alternative A
Alternative B
2027
$0.0022
$0,023
$0.0015
2028
$0,026
= $0,099
$0,023
2029
-$0,012
$0.11
-$0,018
2030
-$0.11
$0,039
-$0.12
203 1
j' -$0.27
-$0,098
-$0.29
2032
-$0.47
-$0.28
-SO. 5
2035
-$0.59
-$0.43
i -$0.76
2040
-$0.86
: -$0.75
i -$1.2
2045
-$1.1
-$1
-$1.5
2050
-$1.4
-$1.4
-$1.8
2055
! -$1.7
j -$1.7
| -$2.2
PV2
; -$17
-$15
r -$23
PV3
-$15
-$13
-$19
PV7
I -$7.5
: -$6.2
-$9.8
AV2
-$0.8
1 -$0.7
-$1.1
AV3
; -$0.76
-$0.66
-$1
AV7
-$0.61
-$0.5
-$0.8
* Negative values reflect disbcncfils.
To estimate the refueling costs associated with liquid-fueled vehicles, we have borrowed
heavily from the approach used by EPA in the December 2021 GHG final rule (U.S. EPA 2021)
(U.S. DOT 2021) with updated inputs used by NHTSA in support of their 2023 CAFE proposed
rule (DOT Volpe Center 2023). The refueling costs for liquid-fueled vehicles are calculated on a
cost per gallon basis while for BEVs it is calculated on a cost per mile basis. The calculations
used are shown in the equation immediately below for liquid-fueled vehicles and in the
subsequent equation for BEVs with a discussion following.
Tank SizexShared Filled
Cost 1 Fixed Time+ -
Gallon Tank SizexShare Filled 60
X Fill Rate x f lme yalue X 0.6
Where,
Cost/Gallon = the refueling cost per gallon of fuel consumed,
Tank Size = the volume, in gallons, of the liquid fuel tank,
Share Filled = the typical share of the tank volume filled during a refill event,
Fixed Time = the fixed time, in minutes, between deciding to refill and returning to the trip,
Fill Rate = the fuel dispense rate, in gallons per minute, of liquid fuel pumps,
60 converts minutes to hours
Time Value = the value of the time for the occupants of the vehicle,
0.6 = a scalar value to count only 60 percent of refueling events
4-42
-------
We have estimated tank sizes the same way it was done in our 2021 GHG final rule, which
was based on a 2016 internal Department of Transportation (DOT) memorandum. (CAFE TSD
2021) (White September 27, 2016) The most recent data reported was for the 2016 model year
and showed that the average tank sizes of some of the most popular vehicles in the United States
were 15.7, 18.7, and 27.3 for cars, vans and SUVS, and pickup trucks, respectively, all in
gallons. We have used those values for all vehicles in each of those categories.
The share filled values are consistent for all vehicles at 0.65, meaning that the typical refill
event includes filling 65 percent of the capacity of the tank.
The fixed time value is also consistent for all vehicles at 3.5 minutes per event, while the fill
rate is held constant at 7.5 gallons per minute reflecting the legal restriction of 10 gallons per
minute and the fact that not all people refill at that maximum rate.
The time value has been extensively analyzed by DOT for use in regulatory analyses. The
values, which account for wage rates, miles driven in urban and rural settings, the different uses
of vehicles whether it be personal or commercial use, and the typical number of occupants over
the age of 5 years for different vehicles. The hourly values ($/hour) derived and which we use
are $29.36, $33.14, $28.85, and $47.28 for passenger cars, CUVs and SUVs, light-duty pickups,
and medium-duty vans and pickups, respectively, all in 2021 dollars (DOT Volpe Center 2023).
All of these values have been updated since our proposal.
As described by NHTSA, the 0.6 scaling factor is meant to capture those drivers whose
primary reason for the refueling trip was due to a low reading on the gas gauge. Such drivers
experience a cost due to added mileage driven to detour to a filling station, as well as added time
to refuel and complete the transaction at the filling station. Drivers who refuel on a regular
schedule or incidental to stops they make primarily for other reasons (e.g., using restrooms or
buying snacks) do not experience the cost associated with detouring to locate a station or paying
for the transaction, because the frequency of refueling for these reasons is unlikely to be affected
by fuel economy improvements. This restriction was imposed to exclude distortionary effects of
those who refuel on a fixed (e.g., weekly) schedule and may be unlikely to alter refueling
patterns due to increased driving range (NHTSA 2022).
To estimate the refueling costs associated with BEVs, we calculate cost per mile.
Cost ( Fixed Time 1 Share Charged\
,= — x — H—— x Time Value
Mile \Charge Frequency 60 Charge Rate )
Where,
Cost/Mile = the refueling cost per mile driven
Fixed Time = the fixed time, in minutes, between deciding to recharge and returning to the
trip,
Charge Frequency = the cumulative number of miles driven before a mid-trip charging event
is triggered,
Share Charged = the share of miles that will be charged mid-trip,
Charge Rate = the typical recharge rate, in miles per hour of charging,
Time Value = the value of the time for the occupants of the vehicle.
4-43
-------
The fixed time value is taken to be equal to that for liquid-fueled vehicles, at 3.5 minutes per
event, and the time value is equal to those stated above for liquid-fueled vehicles.
The charge rate reflects the number of miles of driving provided by a one hour charging
session. As described in Chapter 5.3, charging equipment is available at a variety of power
levels, with higher power equipment generally able to charge a vehicle more quickly. For mid-
trip charging, we assume BEV drivers will primarily use DC fast charging equipment (DCFC).
Among DCFC deployments, there is a trend toward higher power levels with more than half of
the public DCFC ports capable of power output at 250 kW or higher and about two-thirds at 150
kW or higher as of the second quarter of 2023 (Brown, et al. 2023). A combination of private
and public funds is expected to continue the buildout of the DCFC network. This includes up to
$5 billion for the National Electric Vehicle Infrastructure (NEVI) Formula Program established
by the Bipartisan Infrastructure Law. Initial funding to states under the NEVI program is
supporting station buildout along designated highway corridors, with stations required to have at
least four DCFC ports, each 150 kW or higher (U.S. DOT 2023). EPA's assessment of charging
infrastructure needs and costs under the final rule (described in Chapter 5.3.2) projects a mix of
150 kW, 250 kW, and 350 kW public DCFCs will be needed to support PEVs in future years
with the highest share of DCFC charging demand to be met by 350 kW DCFCs.93 Different
BEVs have different limits on how much energy can be delivered to the battery pack, and other
factors - ambient conditions, battery state of charge, on-vehicle accessory loads during charging
- impact the energy transfer. For this analysis, we use a value of 400 miles of driving added for
each hour of charging, selected to represent DC fast charging at 150 kW,94 and apply that value
for all BEVs. If more mid-trip charging occurs at higher-power DCFCs, this value could be
considered conservative.95
For the charge frequency and share charged parameters, we have used values developed by
NHTSA and presented in the CCEMS input files used in support of their August 2023 proposal
(U.S. DOT 2023) In their analysis, NHTSA estimated the frequency of mid-trip charging events
and the share of miles driven that require mid-trip charging as shown in Table 4-17. As Table 4-
17 shows, cars would be expected to require less frequent mid-trip charges and a smaller share of
miles driven with mid-trip charge events. Pickups and vans/SUVs have fairly similar measures,
with vans and SUVs requiring slightly more mid-trip charging than pickups.
93 The analysis assumed that among DCFCs, the highest power that a vehicle can accept (or "as fast as possible"
charging) is preferred. Residential, work, depot, and public L2 charging needs were also estimated, see Chapter
5.3.2.
94 To estimate a typical charge rate for BEVs in our analysis, we used a sales-weighted average energy consumption
rate from OMEGA for select years from 2027 to 2055, which is 0.34 kWh/mi accounting for charging losses. At
maximum power, a 150 kW EVSE port could add about 440 miles of range per hour for a vehicle with this energy
consumption rate. We rounded down to 400 miles given other factors that can impact charge rate.
95 To the extent mid-trip charging occurs at a higher charge rate or a lower charge rate, the resulting cost per mile for
time spent charging electric vehicles would be lower or higher, respectively. In the NPRM, we used a value of 100
miles per hour of charging but have updated that value based on our updated analysis.
4-44
-------
Table 4-17: BEV recharging thresholds by body style and range
Cars
Vans & SUVs
Pickups
MD Vans
MD Pickups
Miles to mid-trip charging event, BEVI50
2,000
1,500
1,600
1,200
1,200
Miles to mid-trip charging event, BEV250
3,600
2,500
2,700
1,700
Miles to mid-trip charging event, BEV300
5,200
3,500
3,800
1,700
Miles to mid-trip charging event, BEV400
10,400
7,000
7,600
Share of miles charged mid-trip, BEV150
0.06
0.09
0.08
0.125
0.125
Share of miles charged mid-trip, BEV250
0.045
0.065
0.06
0.07
Share of miles charged mid-trip, BEV300
0.03
0.04
0.04
0.07
Share of miles charged mid-trip, BEV400
0.015
0.02
0.02
* BEV150/250/300/400 refer to a BEVs having an expected 150/250/300/400 mile range.
Using the values in Table 4-17, EPA has developed curves for each body style as a function
of range. These curves, new for the final rule, are exponential curve fits as a function of BEV
range. These curves and their coefficient values are shown in Figure 4-5 and Figure 4-6.
12000
10000
8000
6000
4000
2000
0
Miles to mid-trip charge event
y = 718.62e00066x
1
.-•y = 598.26e00063x
565.31e00062x
V = 847.06eaoo23x
y = 711.67ea0035x
50
100
150
200
250
300
350
400
450
• Cars
• MD Vans
Expon. (Vans/SUVs)
• Vans/SUVs
• MD Pickups
•••• Expon. (Pickups)
Pickups
Expon. (Cars)
Expon. (MD Vans)
Expon. (MD Pickups)
Figure 4-5: Curve fits for miles driven to a mid-trip charge event.
4-45
-------
Share of miles charged mid-trip
0.14
0.12
0.1
0.08
0.06
0.04
0.02
0
V,.
y = 0.2983e-°-<,0S*.
y = 0.253e 0005x**..'.'"--..<
y = 0.209e~°005x
y = 0.1567e-'
I
y = 0.2232e~0004x
•••
50 100 150 200 250 300 350 400
450
• Cars
• MD Vans
Expon. (Vans/SUVs)
Expon. (MD Pickups)
Vans/SUVs
MD Pickups
Expon. (Pickups)
Pickups
Expon. (Cars)
Expon. (MD Vans)
Figure 4-6: Curve fits for the share of miles charged in mid-trip events.
For clarity, these curve fit equations are shown in Table 4-18.
Table 4-18 Curve fits used in calculating refueling time for BEVs *
Miles to Mid-trip Charge Event
Share of Miles Charged Mid-trip
Cars
718.62e(00066x)
0.1567e(~°'006x)
Vans/SUVs
565.3 le(00062x)
0.253e(~°'006x)
Pickups
598.26e(00063x)
0.209e(~°'°06x)
MD Vans
711.67e(0 0035x)
0.2983e(~°'006x)
MD Pickups
847.06e(00023x)
0.2232e(~°'004x)
* x is the BEV onroad range in miles.
4.3.6 Insurance Costs
Associated with the changing cost of vehicles will be a change in insurance paid by owners
and drivers of those vehicles. To estimate insurance costs, we made use of an analysis done by
Argonne National Laboratory (ANL) which focused on insurance costs associated with
comprehensive and collision coverage. (Burnham, et al. 2021) In that report, ANL presented the
data shown in 4-19 which is what we have used in OMEGA to estimate insurance costs.
4-46
-------
Table 4-19 Annual comprehensive and collision premium with $500 deductible, 2019
dollars *.
Body Style ICE. HEV. PHEV. BEV Powcrlrains
Car (Vehicle value * 0.009 + $220) * 1.19
CUV/SUV (Vehicle value * 0.005 + $240) *1.19
Pickup (Vehicle value * 0.006 + $210) * 1.19
* Vehicle value is calculated as the depreciated value of the vehicle as it ages.
To estimate the vehicle value as it ages we estimated the deterioration rate using recent data
from Black Book and Fitch Ratings which showed that the average annual depreciation rate of
two- to six-year-old vehicles fluctuated over the last decade from a high of 17.3 percent to a low
of 8.3 percent prior to the pandemic. (Black Book and Fitch Ratings 2021). Note that
depreciation largely halted during the pandemic with two- to six-year old vehicles depreciating at
only 2 percent in 2020 and projected at only 5 percent in 2021. The pandemic rates are unlikely
to be representative of future depreciation rates, so we averaged the annual rates from 2016 -
2019 to construct a more representative average depreciation rate of 14.9 percent. We estimate
that future depreciation rates will resemble pre-pandemic trends and the analysis uses the 14.9
percent depreciation rate for all future years.
We did not estimate insurance costs in the NPRM, so these costs are new and represent
increased costs relative to the proposal. As discussed, our estimated insurance rates differ
slightly by body-style, but not by powertrain type. Note that insurance costs are calculated for all
years of a vehicle's lifetime.
4.3.7 Maintenance and Repair Costs
Maintenance and repair (M&R) are large components of vehicle cost of ownership for any
vehicle. According to Edmunds, maintenance costs consist of two types of maintenance:
scheduled and unscheduled. Scheduled maintenance is the performance of factory-recommended
actions at periodic mileage or calendar intervals, like oil changes. Unscheduled maintenance
includes wheel alignment and the replacement of items such as the 12-Volt battery, brakes,
headlights, hoses, exhaust system parts, taillight/turn signal bulbs, tires, and wiper blades/inserts
(Edmunds 2023). Repairs, in contrast, are done to fix malfunctioning parts that inhibit the use of
the vehicle. The differentiation between the items that are included in unscheduled maintenance
versus repairs is likely arbitrary, but the items considered repairs seem to follow the systems that
are covered in vehicle comprehensive (i.e., "bumper-to-bumper") warranties offered by
automakers, which exclude common "wear" items like tires, brakes, and starter batteries (Muller
2017).
To estimate maintenance and repair costs, we have used the data gathered and summarized by
Argonne National Laboratory (ANL) in their look at the total cost of ownership for vehicles of
various sizes and powertrains (Burnham, et al. 2021).
4.3.7.1 Maintenance Costs
Maintenance costs, and differences between ICE vehicles and HEVs versus BEVs and
PHEVs, are an important consideration in not only the full accounting of social benefits and
costs, but also the consumer decision-making process when comparing ICE/HEV technology
versus BEV/PHEV technology. If BEVs and PHEVs are less costly to maintain, a consumer
4-47
-------
might find the potentially higher purchase price of the vehicle to be "worth it" given the possibly
lower fuel and maintenance costs over time. The reverse is also true - more costly BEV/PHEV
maintenance relative to ICE/HEV might make the potentially higher purchase price even less
appealing, even if the fueling costs are lower. Table 4-20 presents our estimated maintenance
costs for the final standards and each alternative.
Table 4-20: Maintenance costs associated with the final standards and each alternative
(billions of 2022 dollars).
Calendar Year
Final
Alternative A
Alternative B
2027
$0,042
: $0,097
$0,042
2028
$0,096
$0.14
$0,083
2029
$0,089
-$0.0079
5 $0.09
2030
-$0,027
-$0.34
I -$0.0077
203 1
j' -$0.35
-$0.91
-$0.29
2032
-$0.9
: -$1.7
-$0.79
2035
: -$3.3
-$4.9
-S3.2
2040
-$13
-$15
-$11
2045
i -$24
: -$25
-$20
2050
r -$32
! -$32
-$27
2055
"-$35 I ^
: -$35
! -$30
PV2
-$300
-$320
-$260
PV3
-$250
; -$270
-$210
PV7
-$110
-$130
-$98
AV2
-$14
-$15
; -$12
AV3
-$13
-$14
-$11
AV7
-$9.3
-$10
; -$8
* Negative values reflect lower costs (i.e..
savings).
In their study, ANL developed a generic maintenance service schedule for various powertrain
types using owner's manuals from various makes and models including the Toyota Yaris,
Camry, Camry HEV, Prius, and Prius Prime; Chevrolet Cruze, Volt, and Bolt; Nissan Sentra,
Kicks, and Leaf; Kia Optima, Optima HEV, and Optima PHEV; Kia Soul and Soul EV; Tesla
Model 3 and Model S, Ford Focus; Lincoln MKZ; BMW i3; VW Golf and e-Golf; and Fiat 500
and 500e. The analysis assumed that drivers would follow the recommended service intervals.
The authors noted that, in practice, not everyone follows the recommended service intervals but
also noted that owners likely do so at the expense of either future repair costs or the early
scrappage of the vehicle (Burnham, et al. 2021, 81). The authors also noted that estimates were
made for certain "wear items" that might not normally be included in a recommended
maintenance schedule (e.g., brake pads and rotors) for which they estimated average lifetimes
based on guidance from several experts and from automotive websites (Burnham, et al. 2021,
81).
After developing the maintenance schedules, the authors collected national average costs for
each of the preventative and unscheduled services. The authors noted that service cost varies by
several factors, including the type of mechanic (dealership vs. chain vs. independent), part
quality (OEM vs. aftermarket), and make and model cost characteristics (domestic vs. import and
mass market vs. luxury). The authors did not assume drivers would perform any of their own
maintenance services, stating a lack of data available on how often drivers do so. The authors
4-48
-------
noted that "do it yourself' maintenance would reduce costs, though depending on the service
would require investment in both tools and skill development (Burnham, et al. 2021, 81).
The authors noted that vehicle type (sedan, SUV, pickup) may influence maintenance costs as
some part sizes and fluid capacities can be larger for bigger vehicles (e.g., larger tires needed for
a pickup). However, when examining the data at their disposal, the authors found no significant
difference over 10 years of ownership. But total maintenance and repair costs of medium-duty
diesel vehicles were about 34% higher than that of their gasoline counterparts. The authors
attributed that difference to repairs rather than maintenance, since the most obvious maintenance
difference between the vehicles is that diesels do not have spark plugs which is a relatively small
cost. The authors acknowledge that their dataset had a very limited number of diesel vehicles and
there appeared to be no clear trend regarding higher or lower maintenance costs for diesel fueled
vehicles.
Specific to tires and tire replacement, an issue often cited with respect to BEVs versus ICE
vehicles, the authors noted that their analysis assumed that tire life and replacement costs are the
same for all powertrains. Some BEVs are equipped with tires that differ from those on typical
ICE vehicles to address tread wear and the instant torque of BEVs. Burnham et al (2021) further
note that advanced powertrain vehicles are often equipped with low rolling resistance (LRR)
tires, but EPA believes most new vehicles, regardless of powertrain, are equipped and sold with
LRR tires. There have also been claims that traditional tires wear 30 percent faster when installed
on BEVs (Burnham, et al. 2021, 83). Overall, while there is some evidence that BEVs and ICE
vehicles may be equipped with different tires, there is insufficient empirical evidence on the
costs of these tires or differential wear to conclude that tire and tire replacement costs vary across
powertrains.
Regarding brake-related maintenance, the authors assumed that brake pad, rotor, and caliper
replacement intervals could be extended by 33% for HEVs and by 50% for PHEVs and BEVs,
relative to ICE vehicles, due to less friction wear that would result from the use of regenerative
braking. Further, they assumed that PHEVs and BEVs would have more regenerative braking
capabilities than HEVs and, therefore, that their service intervals could be extended longer than
HEVs due to their larger battery capacity and electric motor(s) (Burnham, et al. 2021, 84). Table
4-23 shows the maintenance costs used as inputs to OMEGA.
4-49
-------
Table 4-21: Maintenance service schedule by powertrain.
Service
Miles per Event ICE
Miles per Event
Miles per
Miles per
Cost per
HEV
Event PHEV
Event BEV
Event
(2019
dollars)
Engine Oil
7,500
7,500
9,000
n/a
$65
Oil Filler
7,500
7,500
9,000
n/a
$20 """
Tire Rotation
7,500
7,500
7,500
7,500
$50
Wiper Blades
15.000
15.000
15.000
15.000
$45
Cabin Air Filter
20.000
20.000
20.000
20.000
$50
Multi-Point
20.000
20.000
20.000
20.000
$110
Inspection
Engine Air Filter
30,000
66.667
83,333
n/a
$40
Brake Fluid
37,500
37,500
37,500
37,500
$150
Tires Replaced
50.000
50.000
50.000
50.000
$525
Brake Pads
50.000
66.667
75.000
75.000
$350
Starter Battery
50.000
50.000
50.000
50.000
$175
Spark Plugs
60,000
120.000
120.000
n/a
$225
Oxygen Sensor
80.000
80.000
80.000
n/a
$350
Headlight Bulbs
80.000
80.000
80.000
80.000
$90
Transmission
90,000
1 10.000
1 10.000
n/a
$200
Service
Timing Belt
90,000
1 10.000
1 10.000
n/a
$750
Acccssorv Drive
90,000
1 10.000
1 10.000
n/a
$165
Belt
HVAC Service
100,000
100,000
100,000
100,000
$50
Brake Rotors
100.000
125.000
150.000
150.000
$500
Shocks and
100,000
100,000
100.000
100.000
$1,000
Struts
Engine Coolant
125.000
125.000
125.000
n/a
$190
EV Battery
n/a
125,000
125,000
125,000
$210
Coolant
Fuel Filter
150.000
150.000
200.000
n/a
$110
Brake Calipers
150.000
187.500
225.000
225.000
$1,000
Using the schedules and costs shown in Table 4-21, OMEGA then calculates the cumulative
maintenance costs from mile zero through mile 225,000. For example, the cumulative costs for
an ICE vehicle at 15,000 miles would be 2 x ($65+$20+$50) + $45, or $315. The cumulative
costs can then be divided by the cumulative miles to determine the average maintenance cost per
mile at any given odometer reading in a vehicle's life.
However, that average cost, while informative, suggests that the first mile incurs the same
cost as the last mile. This does not seem appropriate, especially considering that the cumulative
costs for ICE vehicles, $20,050, divided by 225,000 cumulative miles results in an average cost
per mile of $0.09. If that vehicle had a fuel economy of 35 miles per gallon, assuming $3 per
gallon of gasoline, its fuel costs would also be $0.09 per mile. Applying the average maintenance
cost per mile over 15,000 first year miles, the fuel costs and maintenance costs would both be
$1,350. This method of estimating maintenance costs vastly exceeds the above $315 estimate of
maintenance costs over the first 15,000 miles.
4-50
-------
Clearly, while the average maintenance cost per mile of $0.09 is valid and informative, it is
not the best valuation for our purpose. Instead, we have estimated the cost per mile at a constant
slope with an intercept set to $0 per mile such that the cumulative costs after 225,000 miles
would equal the $20,050 (for an ICE vehicle) included in the suggested maintenance schedule.
Following this approach, the maintenance cost per mile curves calculated within OMEGA are as
shown in Figure 4-7.
Figure 4-7: Maintenance cost per mile (2019 dollars) at various odometer readings.
Using these maintenance cost per mile curves, OMEGA then calculates the estimated
maintenance costs in any given year of a vehicle's life based on the miles traveled in that year.
Importantly, the cost curves estimate cumulative maintenance costs at any given odometer
reading. To estimate the maintenance costs in any given year, the cumulative maintenance costs
in the prior year are subtracted from the cumulative maintenance costs in the given year. This
way, the maintenance costs in each year, when summed, will equate to the cumulative
maintenance costs derived from the input data.
MaintenanceCostt = 0.5 x MaintenanceCostPerMilet x Odometert — CumulativeMaintenanceCostt_1
Where,
MaintencmceCostx = maintenance cost for a given vehicle in year t
MaintencmceCostPerMilex = maintenance cost per mile at the odometer reading reached in
year t (see Figure 4-7)
Odometerx = the odometer reading of the given vehicle in year t
4-51
-------
CumulativeMaintenanceCostt-i = the cumulative maintenance cost for the given vehicle
through year t-1
OMEGA uses these maintenance costs for light-duty and for medium-duty vehicles. The
maintenance costs are included in the benefit and cost analysis. Note that these maintenance
costs differ from those presented in Chapter 4.1 and Chapter 4.2. Chapter 4.1 costs are meant to
reflect the thought process of a potential new vehicle purchaser. Chapter 4.2 amounts are
estimated average expenses per vehicle over the first 8 years of vehicle life. Costs presented here
in Chapter 4.3 are meant to estimate the actual effects of the final rule.
4.3.7.2 Repair Costs
Table 4-22 presents our estimates of repair costs associated with the final standards and each
of the alternatives.
Table 4-22: Repair costs associated with the final standards and each alternative
(billions of 2022 dollars).
Calendar Year
Final
Alternative A
Alternative B
2027
: $0,027
$0,091
$0,026
2028
$0,081
i $0.23
$0,067
2029
$0.16
$0.36
$0.14
2030
$0.26
$0.48
$0.24
203 1
: $0.35
SO.55
$0.33
2032
$0.38
' $0.57
: $0.36
2035
; $0-7
! $0.75 ^ "
! $0.35
2040
-$0.81
-$0.88
-$1.1
2045
-$3.4
-$3.4
; -$3.3
2050
-$5.7
: -$5.7
! -$5.3
2055
! -$7.1
i -$7.3
-$6.6
PV2
-$40
-$40
-$41
PV3
| -$32
-$31
I -$32
PV7
; -$12
-$12
-$13
AV2
-$1.8
-$1.8
-$1.9
AV3
-$ 1.6
-$ 1.6
! -$i.7
AV7
-$0.99
-$0.94
-$1.1
* Negative values reflect lower costs (i.e..
. savings).
Repairs are done to fix malfunctioning parts that inhibit the use of the vehicle and are
generally considered to address problems associated with parts or systems that are covered under
typical manufacturer bumper-to-bumper type warranties. In the ANL study, the authors were
able to develop a repair cost curve for a gasoline car and a series of scalers that could be applied
to that curve to estimate repair costs for other powertrains and vehicle types. The repair cost
curve developed in the ANL study is shown in the equation below (Burnham, et al. 2021).
Repairi = vpaiebx,i = 1, ...,15
Where,
Repain = the repair cost per mile at age i,
4-52
-------
v = the appropriate vehicle type multiplier (see Car/SUV/Truck entries in Table 4-23),
p = the appropriate powertrain type multiplier (see ICE/HEV/PHEV/BEV/FCV entries in
Table 4-23),
a; = gasoline car repair cost coefficient at age i,
b = exponential constant of 0.00002,
x = the MSRP of the car when sold as new.
Table 4-23: Repair cost per mile coefficient values"
Item
Value
Car multiplier
1.0
SUV multiplier
0.91
Truck multiplier
0.7
MD Van multiplier
1.9
MD Pickup multiplier
1.6
ICE multiplier
1.0
HEV multiplier
0.91
PHEV multiplier
0.86
BEV multiplier
0.67
FCV multiplier
0.67
ao
0
ai
o
a2
0.00333
a3
0.01
a4
0.0167
3add-on
0.00333
a These coefficient values come from Burnham, Gohlke, et al.
(2021) with the exception of the medium-duty multipliers which
were added by EPA to replicate the repair cost share of the
maintenance and repair cost curve shown in Figure 3.32 of the
ANL study.
OMEGA makes use of the equation developed in the ANL study along with the coefficient
values shown in Table 4-23 to estimate repair costs per mile at any age in a given vehicle's life.
In place of the MSRP96 of the new vehicle, OMEGA uses the estimated cost to manufacture the
vehicle excluding applicable tax credits that might reduce the price paid by the purchaser.
Further, OMEGA makes use of this equation for all ages of a vehicle's life (OMEGA estimates a
30/40-year lifetime) using the aadd-on value for every age beyond the first five years. In other
words, the ax value for age 7 would be 0.0167 + 3 x 0.00333 = 0.02669 (note that, in OMEGA,
age=7 is the 8th year of a vehicle's life). The resultant repair cost per mile values at all ages are
shown in Figure 4-8. Note that the new vehicle cost (used in place of the MSRP value) is held
constant at $35,000 in Figure 4-8, regardless of vehicle type (car, van/SUV, pickup) and
96 Manufacturer suggested retail price
4-53
-------
powertrain (ICE vehicle, HEV or MHEV, PHEV, BEV) which is not likely, but is presented here
for illustration only.
OMEGA uses these repair costs for both light-duty and medium-duty. Repair costs are
included in the benefit-cost analysis.
Car
Van & SUV
Pickup
0,250
lile
%
= /W'
JU
— 0.150
^ ice
- A
.2 >£? ~ car hev
— n mn mww _ • -
^ 0-150 suv ice
5 A
~z nmn •*"
^ S ^^^truck_ice
£
J5 S* —«• truck hev
— n mn
q U.1UU
,•** — — car_phev
0 050 ** invi* cap bev
Q UtlUU
0 050 #* suv bev
o U.J.UU **
0 050 *# • truck bev
0 3 6 9 12 15 18 21 24 27 30
Vehicle Age
0 3 6 9 12 15 18 21 24 27 30
Vehicle Age
0 3 6 9 12 15 18 2124 27 30
Vehicle Age
Figure 4-8: Repair cost per mile (2019 dollars) for a $35,000 Car, Van/SUV and Pickup
with various powertrains.
4.3.8 Costs Associated with Noise and Congestion
If consumers choose to drive more, they benefit from the utility derived from those additional
miles, as described in Chapter 4.3.2. In contrast to the benefits associated with additional driving,
there are also costs. Increased vehicle use associated with a positive rebound effect also
contributes to increased traffic congestion and highway noise. Depending on how the additional
travel is distributed throughout the day and where it takes place, additional vehicle use can
contribute to traffic congestion and delays by increasing traffic volumes on facilities that are
already heavily traveled during peak periods. These added delays impose higher costs on other
road users in the form of increased travel time and operating expenses. Because drivers do not
take these external costs into account in deciding when and where to travel, we account for them
separately as a cost of additional driving associated with a positive rebound effect.
EPA relies on congestion and noise cost estimates developed by the Federal Highway
Administration (FHWA) to estimate the increased external costs caused by added driving due to
a positive rebound effect. EPA employed estimates from this source previously in the analysis
accompanying the light-duty 2010, 2012, and 2021 final rules. We continue to find them
appropriate for this analysis after reviewing the procedures used by FHWA to develop them and
considering other available estimates of these values.
FHWA's congestion cost estimates focus on freeways because non-freeway effects are less
serious due to lower traffic volumes and opportunities to re-route around the congestion. EPA
has applied the congestion cost to the overall VMT therefore the results of this analysis
potentially overestimate the congestion costs associated with increased vehicle use, and thus lead
to a conservative estimate of net benefits.
4-54
-------
EPA uses FHWA's "Middle" estimates for marginal congestion and noise costs caused by
increased travel from vehicles. This approach is consistent with the methodology used in our
prior analyses. The values used are shown in Table 4-24. These congestion costs are consistent
with those used in the 2021 final rule. These values are used as inputs to OMEGA and adjusted
within the model to the dollar basis used in the benefit and cost analysis. Table 4-25 and Table 4-
26 present our estimated congestion and noise costs, respectively, associated with the final
standards and each of the alternatives.
Table 4-24: Costs associated with congestion and noise (2018 dollars per vehicle mile)
Sedans/'Wagons CUVs/SUVs/Vans Pickups
Congestion 0.0634 0.0634 0.0566
Noise 0.0009 0.0009 0.0009
Table 4-25 Congestion costs associated with the final standards and the alternatives
(billions of 2022 dollars)
Calendar Year
Final
Alternative A
Alternative B
2027
$0,001.
1 $0.0034
; $0.0016
2028
$0,027
I $0,066
$0,025
2029
$0.05
$0.12
$0,046
2030
SO.073
$0.18
$0,078
203 1
$0,094
$0.24
$0.1
2032
$0.11
| $0.27
$0.11
2035
$0.59
: $0.73
$0.47
2040
$1.3
$1.4
$1.1
2045
$1.9
$1.9
: $1.7
2050
: $2.3
| $2.3
1 $2
2055
: $2.4
: $2.4
; $2.2
PV2
; $25
i $27
; $22
PV3
$21
r $23
$18
PV7
i $10
$11
$8.9
AV2
; $1.2
$1.2
! $1
AV3
$1.1
$1.2
i $0.96
AV7
$0.83
$0.92
1 $0.73
* Positive values reflect
increased costs.
4-55
-------
Table 4-26 Noise costs associated with the final standards and the alternatives (billions of
2022 dollars)
Calendar Year
Final
Alternative A
Alternative B
2027
$0.000015
f$0.000031
$0.000019
2028
$0.00041
: $0,001
SO.00039
2029
$0.00077
$0.0019
$0.00074
2030
$0.0011
$0.0029
$0.0012
203 1
$0.0015
$0.0038
: $0.0016
2032
$0.0017
$0.0043
$0.0018
2035
$0.0095
$0,012
$0.0077
2040
$0,021
$0,023
$0,018
2045
! $0.03
$0,032
; $0,027
2050
; $0,037
$0,038
$0,033
2055
$0.04
$0.04
! $0,036
PV2
$0.41
$0.44
: $0.36
PV3
$0.34
! $0.37
| $0.3
PV7
$0.17
$0.18
$0.15
AV2
: $0,019
$0.02
$0,017
AV3
$0,018
$0,019
! $0,016
AV7
$0,014
$0,015
$0,012
* Positive values reflect increased costs.
4.4 New Vehicle Sales
The topic of the "energy paradox" or "energy efficiency gap" has been extensively discussed
in previous analyses of vehicle GHG standards. The idea of the energy efficiency gap is that
existing fuel saving technologies were not widely adopted even though they reduced fuel
consumption enough to pay for themselves in a short period of time. Conventional economic
principles suggest that because the benefits to vehicle buyers of the new technologies would
outweigh the costs to those buyers, automakers would provide them and people would buy them.
As described in previous EPA GHG vehicle rules (most recently in the 2021 rule),
engineering analyses identified technologies (such as downsized-turbocharged engines, gasoline
direct injection, and improved aerodynamics) where the additional cost of the technology is
quickly covered by the fuel savings it provides, but they were not widely adopted until after the
issuance of EPA vehicle standards. Research also suggests that the presence of fuel-saving
technologies do not lead to adverse effects on other vehicle attributes, such as performance and
noise. Instead, research shows that there are technologies that exist that provide improved fuel
economy without hindering performance, and in some cases, while also improving performance
(Huang, Helfand, et al. 2018) (Watten, Helfand and Anderson 2021). Additionally, research
demonstrates that, in response to the standards, automakers have improved fuel economy without
adversely affecting other vehicle attributes (Helfand and Dorsey-Palmateer 2015). Lastly, while
the availability of more fuel efficient vehicles has increased steadily over time, research has
shown that the attitudes of drivers toward those vehicles with improved fuel economy has not
been affected negatively (Huang, Helfand, et al. 2018) (Huang, Helfand and Bolon 2018a). Thus,
EPA does not model tradeoffs between fuel economy and performance as a path to achieving the
standards. This "constant performance" assumption in our modeling is achieved by estimating
technology costs and emissions reductions while maintaining the performance of each vehicle
from the base year, which obviates the need to estimate potential lost consumer welfare from
4-56
-------
foregone attributes. EPA considers the assumption of constant performance to be a conservative
one, since our estimated compliance costs are higher than would be the case for manufacturers
which offer consumers lower cost vehicles with reduced performance. It appears that in the
absence of the standards, markets have not led to the adoption of fuel-efficient technologies with
short payback periods and no discernible tradeoffs, suggesting that an energy efficiency gap
appears to have existed, especially in the absence of the standards, and may still exist.
A combination of consumer-side and producer-side hypotheses in the literature may best
explain why there was limited adoption of cost-effective fuel-saving technologies before the
implementation of more stringent standards, though the literature has not settled on a single
explanation (National Academies of Science, Engineering, and Medicine 2021).97
Consumer-side hypotheses include:
• Consumers might lack information, not have a full understanding of this information
when it is presented, not have correct information, not have the ability to process the
information, or not trust the presented information.
• Consumers might weigh the present or present circumstances (e.g., current costs) more
heavily than future opportunities (e.g., long term savings, changing circumstances) in
their purchase decisions due to, for example, uncertainty about the future, a lack of
foresight, an aversion to short term losses relative to longer term gains, or a preference
for the status quo.
• Consumers might prioritize other vehicle attributes over fuel economy in their vehicle
purchase process.
• Consumers might erroneously associate higher fuel economy with lower quality
vehicles.
In addition to the research discussed above indicating that fuel-saving technologies are not
likely to be associated with adverse effects on other vehicle attributes, EPA has explored
evidence on how consumers evaluate fuel economy in their vehicle purchase decisions. Overall,
the research has not reached a consensus; results and estimates vary across a range of data types
and statistical models. Thus, it is not clear how consumers incorporate fuel economy in their
purchase decision, nor how consumer behavior might contribute to the energy efficiency gap.
Part of the uncertainty surrounding the reasons behind the energy efficiency gap is that most
of the technology applied to existing ICE vehicles may have been "invisible" to the consumer,
both literally and possibly in effect. This is for a few reasons, including that the technology itself
was not something the mainstream consumer would know about, or because it was applied to a
97 For simplicity, we present consumer- and producer-side hypotheses for the "energy efficiency gap", consistent
with traditional economic theory. Analogously but somewhat differently, we could have presented these hypotheses
organized according to individual and institutional characteristics, behaviors, and biases. Under that organization
structure, some of the hypotheses we present, such as myopia, uncertainty aversion, loss aversion, asymmetric
information, and status quo bias, could apply to both consumer and producers.
4-57
-------
vehicle at the same time as multiple other changes, therefore making it unclear to the consumer
what changes in vehicle attributes, if any, could be attributed to a specific technology.
Much less research has been conducted to evaluate the producer side of the market, though
three interrelated themes arise: market structure, business strategy, and technological innovation.
The structure of the automobile industry may inefficiently allocate car attributes, including fuel
economy, which may contribute to the existence of an energy efficiency gap. Specifically,
vehicle production involves significant fixed costs in which automakers strive to differentiate
their products from each other. In that context, fuel economy of a vehicle could be just one factor
of many in a company's product differentiation strategy. Product differentiation can lead to an
under-supply of fuel economy relative to what is cost-effective to consumers in some segments,
and an over-supply of fuel economy in other sectors (Fischer 2005). Automobile manufacturers
may adopt a "wait and see" strategy regarding the costs associated with investing in and
commercializing new technologies.
In the absence of standards, automakers have seemed willing to invest in small improvements
upon existing technologies (Helfand and Dorsey-Palmateer 2015) and more reluctant to invest in
major innovations in the absence of standards. This may be a result of first-mover disadvantages
to investing in and commercializing new technologies. The "first-mover disadvantage" occurs
when the "first-mover" pays a higher proportion of the costs of developing, implementing, or
marketing a new technology and loses the long-term advantage when other businesses move into
that market. There could also be "dynamic increasing returns" to adopting new technologies,
wherein the value of a new technology may depend on how many other companies have adopted
the technology. Additionally, there can be research and development synergies when many
companies work on the same technologies at the same time, assuming there's a reason to
innovate at the same time. Standards can create conditions under which companies invest in
major innovations. Because all companies (both auto firms and auto suppliers) have incentives to
find better, less expensive ways of meeting the standards, the possibilities for synergistic
interactions may increase. Thus, the standards, by focusing all companies on finding more
efficient ways of achieving the standards, may lead to better outcomes than if any one company
operated on its own.
At the first purchase of a PEV, the energy efficiency technology is clearly apparent to the
consumer (i.e., consumer-facing), in which case the above "invisibility" rationale does not apply.
However, as PEV technology continues to evolve, and as precedent with ICE vehicle technology
suggests, technologies that improve PEV efficiency may again become invisible to the consumer,
making the value of those improvements less apparent at the time of purchase, even if operating
savings are apparent.
Also, with the growing availability of PEV options, there may be additional risk of
information asymmetry between those selling PEVs (including manufacturers and dealerships)
and those considering purchasing one, which may be due to inexperience on one or both sides,
uncertainty in the technology, or other factors. Other reasons the energy efficiency paradox may
persist, regardless of vehicle powertrain, include uncertainty about future fuel and electricity
prices, uncertainty about charging infrastructure and availability, or perceptions of comparisons
of quality and durability of different powertrains. However, there may be factors that mitigate the
effect. Uncertainties will be resolved over time (e.g., growing familiarity with PEVs and EVSE,
durability), systems will evolve (e.g., infrastructure growth and expansion, fuel and electricity
4-58
-------
prices, supply chains), and the nature and balance of information will change. Another factor that
may reduce the magnitude of a possible energy efficiency gap are the incentives provided in the
BIL and IRA which provide support for the development, production, and purchase of PEVs and
the supporting infrastructure. Constraints on investment, either for manufacturers of the
technology or for potential purchasers of the technology, may also lead to slower adoption rates
of energy efficiency technology, even if the technology leads to reductions in operating costs.
Federal or other incentives to manufacture or purchase energy efficient technology will reduce
the impact that constraints on investment have on adoption of that technology. For PEVs, the
availability of existing incentives, including the Federal purchaser and battery manufacturing tax
credits in the IRA, is expected to lead to lower upfront costs for purchasers of PEVs than would
otherwise occur.98
There is uncertainty in the historical literature regarding consumer acceptance and adoption of
electric vehicles, as described in Chapter 4.1 and Jackman et al. (2023), however recent research
suggests that the demand for electric vehicles is robust, and adoption is constrained, at least in
part, by limited supply. Gillingham et al. (2023) examine all new LD vehicles sold in the U.S.
between 2014 and 2020, focusing on comparisons of existing electric vehicles to their most
similar ICE vehicle counterpart, finding that EVs are preferred to the ICE counterpart in some
segments (Gillingham, et al. 2023). In the paper, the authors show that, compared to ICE
counterparts, EVs have seen relative sales shares of over 30%, which indicates that the share of
PEVs in the marketplace is, at least partially, constrained due to the lack of offerings needed to
convert existing demand into market share. In addition, a survey from Consumer Reports in early
2022 shows that more than one third of Americans would either seriously consider or definitely
buy or lease a BEV today, if they were in the market for a vehicle. (Bartlett 2022). We expect the
share of Americans willing to own or lease a PEV will grow over time as exposure to, and
familiarity with, PEVs increases, as well as with infrastructure growth (Jackman, et al. 2023).
The rest of this chapter will discuss how sales effects were modeled in OMEGA, as well the
total change in sales estimated due to this rule.
4.4.1 How Sales Impacts Were Modeled
EPA has updated its OMEGA model, in part, to increase the model's useability and
transparency. In addition, the model has been updated to allow for interactions in producer and
consumer decisions in estimating total sales and the share of ICE vehicles, BEVs, and PHEVs in
the market that both meet the standard being analyzed and will be accepted by consumers. More
about the updated OMEGA model, including detailed information on the structure and
operations, can be found in RIA Chapter 2. As in previous rulemakings, the sales impacts are
based on a set of assumptions and inputs, including assumptions about the role of fuel
98 The IRA battery tax credit is also expected to reduce upfront costs for purchasers, although it is a tax credit for
battery manufacturers, not purchasers. We expect vehicle manufacturers to reduce the price of their vehicles in
accordance with their ability to take advantage of this battery tax credit in order to remain competitive in the market.
4-59
-------
consumption in vehicle purchase decisions described in Chapter 4.1, and assumptions on
consumers' demand elasticity discussed below."
At a high level, OMEGA estimates the effects of a policy on new vehicle sales volumes as a
deviation from the sales that would take place in the absence of the standards.100 This calculation
is based on applying a demand elasticity to the change in new vehicle net price, the price that
incorporates the fact that vehicle buyers are expected to take fuel consumption into consideration
in the purchase process. The modeled BEV and PHEV shares, as described in Chapter 4.1, are
then applied to the estimated total sales volumes to estimate further effects of the rule, including
costs, emissions, and benefits.
4.4.1.1 The Role of Fuel Consumption in Vehicle Sales Estimates
In the 2021 rule, as well as in this rule, EPA assumed that producers account for 2.5 years of
fuel consumption in their assessment of the consumer's purchase decision. However, as
discussed in detail in the 2021 rule, there is not a consensus around the role of fuel consumption
in vehicle purchase decisions. Greene, et al. (2018) estimates a mean willingness to pay for a one
cent per mile reduction in fuel costs over the lifetime of an average vehicle to be $1,880, with a
median of $990 and a very large standard deviation. For comparison, 2.5 years of fuel savings,
assuming one cent per mile and 15,000 vehicle miles traveled per year is about $375. This is
within the large standard deviation in Greene, et al. (2018) for the willingness to pay to reduce
fuel costs, but it is far lower than both the mean and the median of $990. On the other hand, the
2021 National Academy of Sciences (NAS) report, citing the 2015 NAS report, observed that
automakers "perceive that typical consumers would pay upfront for only one to four years of fuel
savings" (pp. 9-10), which is also within the range of values identified in Greene, et al. (2018)
for consumer response, but also well below the median or mean. (National Academies of
Science, Engineering, and Medicine 2021) Recent research shows that the results in Greene, et
al. (2018) are not unique to ICE vehicle fuel costs. Forsythe, et al. (2023) estimate a willingness
to pay for a one cent per mile reduction in operating costs for car drivers of about $1,960, and
slightly less for SUV buyers at about $1,490. (C. R. Forsythe, et al. 2023b) Based on these
results, it appears possible that automakers operate under a different perception of consumer
willingness to pay for additional fuel economy than how consumers actually behave. Consumer
response to fuel savings, and the amount of fuel savings considered in the purchase decision,
may be different with electric vehicles and in an era of high PEV sales.
Chapter 4.1 above describes how OMEGA incorporates fuel costs into the vehicle technology
choice component of consumer purchase decisions. OMEGA also incorporates fuel cost savings
into the consumers decision to buy a vehicle and in the producer assumptions. Specifically, we
assume producers account for 2.5 years of consumer fuel consumption. To do this, OMEGA
calculates an estimate of the energy consumption (gallons of fuel and/or kWh of electricity) over
a user-specified number of years (we assume 2.5), using EIA Annual Energy Outlook projections
99 The demand elasticity is the percent change in quantity associated with a one percent increase in price. For price,
we use net price, where net price is the difference in technology costs less an estimate of the change in fuel costs
over the number of years we assume fuel costs are taken into account. We also reduce BEV prices in all scenarios,
including the No Action case, due to the IRA BEV purchase and battery tax incentives.
100 We calibrate the sales in OMEGA that would take place in the absence of the standards to data from the U.S.
Energy Information Administration.
4-60
-------
of fuel and electricity cost, and the expected vehicle miles traveled per year (VMT). The same
energy costs and expected VMT are then used to calculate energy consumption in the final rule
and alternative scenarios for the same user-specified number of years.
4.4.1.2 Elasticity of Demand
By definition, the new vehicle demand elasticity relates the percent change in new vehicle
price to the percent change in new vehicle sales:
_ AQ/Q
71 AP/P
Where r| is the demand elasticity, Q is the quantity of new vehicles sold in the analysis
context,101 P is the net price in the analysis, and A refers to the change in the values between the
analysis context and the non-context sessions. Rearranging this equation produces the sales
effect:
AQ = tj * Q * AP/P
For this rulemaking, the analysis context sales (Q) are defined by EIA's AEO projections,
which include the effects of the 2021 rule's GHG standards, but not the IRA. Net price, (P) is the
sum of the average vehicle purchase price plus the fuel costs considered in the consumer's
purchase decision. Fuel costs in the price estimate account for 2.5 years of fuel consumption,
where "fuel" includes liquid fuels and/or electricity, assuming 15,000 miles of driving per year.
102 The change in net price (AP) is the difference between new vehicle net price under the EIA
projection (the analysis context) (P), and the net price under the non-context scenarios. For this
rulemaking all non-context scenarios include the effects the IRA.
For durable goods, such as vehicles, people are generally expected to have more flexibility
about when they purchase new vehicles than whether they purchase new vehicles; thus, their
behavior is more inflexible (less elastic) in the long run than in the short run. For this reason,
estimates for long-term elasticities for durable goods are expected to be smaller (in absolute
value) than short-run elasticities. At a market level, short-run responses typically focus entirely
on the new-vehicle market; longer time spans allow for adjustments between the new and used
vehicle markets, and even adjustments outside those markets, such as with public transit.
Because this rule has effects over time and could have effects related to the used vehicle market,
long-run elasticities that account for effects in the used vehicle market are more appropriate for
estimating the impacts of standards in the new vehicle market than short-run elasticities.
Continuing the approach used in the final 2021 rule, EPA is using a demand elasticity of -0.4
for LD vehicles based on an EPA peer reviewed report (U.S. EPA 2021). However, as noted in
EPA's report, -0.4 appears to be the largest estimate (in absolute value) for a long-run new
vehicle demand elasticity in recent studies. EPA's report examining the relationship between
101 For more information on the analysis context, see RIA Chapter 8. Information on the AEO projections mentioned
in this section is also included in RIA Chapter 8.
102 We note that 2.5 years of fuel consumption may be a conservative estimate. A higher value would incorporate
more fuel savings into the decision process. As a point of comparison, Tesla, use an estimate of 3 years of gas
savings in their estimates of financing a leased vehicle: https://www.tesla.eom/model3/design#overview
4-61
-------
new and used vehicle markets shows that, for plausible values reflecting that interaction, the new
vehicle demand elasticity varies from -0.15 to -0.4. A smaller elasticity does not change the
direction of sales effects, but it does reduce the magnitude of the effects. Using the value of -0.4
is conservative, as the larger estimate yields a larger change in sales.
The literature used to estimate this elasticity measure is focused on light-duty vehicles, which
are primarily purchased and used as personal vehicles by individuals and households. The
medium-duty vehicle market, in contrast, largely serves commercial applications. The
assumptions in our analysis of the LD sales response are specific to that market, and do not
necessarily carry over to the MD vehicle market. Though there are not many studies focused on
what affects purchase decisions of medium-duty, or commercial, vehicle buyers, especially in the
US, there are many articles discussing the importance of fuel efficiency, warranty considerations,
maintenance cost, and replacement part availability in choosing which commercial vehicle to
buy 103 jn addition, a working paper published by Resources for the Future reports that
commercial vehicle buyers are not sensitive to fuel price changes, likely due to specialized
vehicle needs. (Leard, McConnell and Zhou 2017) For this rule, we are assuming an elasticity of
0 for the MD vehicle sales impacts estimates and we are not projecting any differences in the
number of MD vehicles sold between the No Action and the final standards, or the more and less
stringent scenarios. This implicitly assumes that the buyers of MD vehicles are not going to
change purchase decisions if the price of the vehicle changes, all else equal. In other words, as
long as the characteristics of the vehicle do not change, commercial buyers will still purchase the
vehicle that fits their needs. The rest of this chapter focuses on the LD vehicle market.
4.4.2 New LD Vehicle Sales Estimates
For this rule, EPA is maintaining the previous assumptions of 2.5 years of fuel savings and a
new LD vehicle demand elasticity of -0.4 for its modeling. Table 4-27 shows results for total
new LD vehicle sales impacts due to the final standards. There is a very small change in total
new LD vehicle sales projected in the final standards compared to the No Action case. Sales by
between about 0.18 percent in 2027 and 0.92 percent in 2032. These impacts are generally
smaller than those estimated for the 2021 rulemaking (U.S. EPA 2021), where sales impacts
were estimated to range from a decrease of about 1 percent in MY 2027 to a decrease of 0.9
percent in MY 2032.
103 See, for example: https://www.fleetmaintenance.com/equipment/chassis-body-and-
cab/article/21136479/considerations-for-purchasing-new-and-used-trucks; https://www.automotive-
fleet. com/159336/10-factors-driving-commercial-fleet-vehicle-acquisitions ;
https://www.mwsmag.com/commercial-vehicle-demand-is-rising-and-so-are-prices/. These webpages are saved to
the docket for this rule.
4-62
-------
Table 4-27: LD sales impacts in the final rule.
Year No Action Final Standards
Total Total Sales Change from No
Sales Action (%)
2027 16.046.000 16.017.000 -29,000
2028 15.848.000 15.790.000 --">8.000
2029 15.927,000 15.840.000 -83",,,,,)
2
-------
Change in Sales From No Action
0.00%
2027 2028 2029 2030 2031 2032
-0.20%
-0.40%
-0.60%
-0.80%
-1.40%
—Final Rule — A— Alternative A * * • * * Alternative B
Figure 4-9: Total new LD vehicle sales impacts, percent change from the No Action case.
The results discussed here focus on sales of new LD vehicles, which does affect the total size
and make-up of the onroad fleet over time.104 The analysis for the effects of this rule include the
effects of the total onroad fleet. In addition to the new vehicles sold, the onroad fleet may include
vehicles from the legacy fleet (vehicles on the road before MY 2023), and re-registered vehicles
from the analysis fleet, not including the new vehicles sold in the year being analyzed. The
analysis fleet is made up of the vehicles entering the fleet starting in MY 2023, and are a result of
OMEGA estimates of the new vehicle sales impacts resulting from the policy in any given
scenario. For example, the onroad fleet for MY 2030 will include the reregistered vehicles from
the legacy fleet, as well as the reregistered analysis fleet vehicles from MY 2023 through MY
2029, and the new vehicles sold in MY 2030. Re-registered vehicles are used vehicles that
remain on the road and are registered for onroad use for that year. This is the flip side to
scrappage, which estimates the vehicles that are taken out of the total onroad fleet.105
Fleet size is normalized to AEO projections, so our fleet size does not change across
scenarios. Reregistered vehicles are aged out according to static estimates based on mileage and
vehicle age as presented in Chapter 8.3 of the RIA. As new vehicle sales change, the remaining
onroad fleet is adjusted to ensure that the fleet remains at AEO projections. Those reregistered
vehicles then accumulate miles according to static estimates based on age. As most of our effects
are estimated as a result of VMT, this process is done through normalizing the VMT of the
onroad fleet, this means that, with respect to the used vehicle market, OMEGA does not directly
model delayed scrappage but does so indirectly through estimated VMT for the reregistered fleet.
1114 The onroad fleet consists of the total count and types of vehicles on the road, and their characteristics including
transmission type and age.
1115 Note that we understand that consumers may choose between buying a new vehicle, and, for example, keeping
their current vehicle longer, buying a used vehicle or not entering the vehicle market. Our modeling accounts for
either purchasing a new vehicle or not, which lumps all other options into one category. For more information on
consumer choice, see Section 13 of the RTC.
4-64
-------
As the new vehicle sales fall, OMEGA adjusts the VMT of the existing vehicles to normalize the
fleet to the AEO projections, which in effect, leads to higher VMT for those existing vehicles.
More information on the fleet turnover response can be found in Section 12.1.3 of the RTC.
4.5 Employment
This chapter discusses potential employment impacts due to this rule and presents rough
estimates that reflect a portion of those estimates. The rule primarily affects LD and MD
vehicles, suggesting that there may be employment effects in the motor vehicle and parts sectors
due to expected effects of the standards on sales. Thus, we focus our assessment on the motor
vehicle manufacturing and the motor vehicle parts manufacturing sectors, with some assessment
of impacts on additional closely related sectors likely to be most affected by the standards.
When the U.S. economy is at full employment, even a large-scale environmental regulation is
unlikely to have a noticeable impact on aggregate net employment. Over the long run,
environmental regulation is expected to cause a shift of employment among employers rather
than affect the general employment level (Arrow et al. 1996; Hafstead and Williams, 2020). The
expectation is that labor would be reallocated from one productive use to another, as workers
transition away from jobs that are less environmentally protective and toward jobs that are more
environmentally protective. Affected sectors may nevertheless experience transitory effects as
workers move in and/or out of jobs. Some workers may retrain or relocate in anticipation of new
requirements or require time to search for new jobs, while shortages in some sectors or regions
could bid up wages to attract workers. These adjustment costs can lead to local labor disruptions.
Even if the net change in the national workforce is small, localized reductions in employment
may adversely impact individuals and communities just as localized increases may have positive
impacts. If the economy is operating at less than full employment, economic theory does not
clearly indicate the direction or magnitude of the net impact of environmental regulation on
employment; it could cause either a short-run net increase or short-run net decrease as discussed
further below.
Chapter 4.5.1 offers a brief, high-level explanation of employment impacts due to
environmental regulation and discusses a selection of the peer-reviewed literature on this topic.
Chapter 4.5.2 focuses on potential impacts from growing electrification, and Chapter 4.5.3
qualitatively discusses possible employment impacts of this rule on regulated industries. Chapter
4.5.4 presents a quantitative estimate of partial employment impacts that may occur due to this
rule. In previous rules, we have quantitatively estimated a cost effect, which should be estimated
holding vehicle sales constant. However, the cost estimates come from OMEGA, which
estimates the costs of the rule inclusive of the effects of changes in vehicles sold. Therefore, the
quantitative partial employment analysis for this rule is a combined cost and demand effect.
Chapter 4.5.5 qualitatively discusses potential impacts on related sectors.
4.5.1 Background and Literature
Economic theory of labor demand indicates that employers affected by environmental
regulation may change their demand for different types of labor in different ways. They may
increase their demand for some types, decrease demand for other types, or maintain demand for
still other types. The uncertain direction of labor impacts is due to the different channels by
which regulations affect labor demand. A variety of conditions can affect employment impacts of
environmental regulation, including baseline labor market conditions, employer and worker
4-65
-------
characteristics, industry, and region. In general, the employment effects of environmental
regulation are difficult to disentangle from other economic changes (especially the state of the
macroeconomy) and business decisions that affect employment, both over time and across
regions and industries. In light of these difficulties, we look to economic theory to provide a
constructive framework for approaching these assessments and for better understanding the
inherent complexities in such assessments.
In this chapter, we describe three ways employment at the firm level might be affected by
changes in a firm's production costs due to environmental regulation: a demand effect, caused by
higher production costs increasing market prices and decreasing demand; a cost effect, caused by
additional environmental protection costs leading regulated firms to increase their use of inputs,
including labor, to produce the same level of output; and a factor shift effect, in which post-
regulation production technologies may have different labor intensities than their pre-regulation
counterparts. These effects are outlined in a paper by Morgenstern et al., which provides the
theoretical foundation for EPA's analysis of the impacts of this regulation on labor (Morgenstern,
Pizer and Shih 2002). Due to data limitations, EPA is not quantifying the impacts of the final
regulation on firm-level employment for affected companies. Instead, we discuss demand, cost,
and factor-shift employment effects for the regulated sector at the industry level.
Additional papers approach employment effects through similar frameworks. Berman and Bui
model two components that drive changes in firm-level labor demand: output effects and
substitution effects (Berman and Bui 2001).106 Deschenes describes environmental regulations as
requiring additional capital equipment for pollution abatement that does not increase labor
productivity (Deschenes 2018). For an overview of the neoclassical theory of production and
factor demand, see Chapter 9 of Layard and Walters' Microeconomic Theory (Layard and
Walters 1978). Ehrenberg and Smith describe how at the industry level, labor demand is more
likely to be responsive to regulatory costs if: (1) the elasticity of labor demand is high relative to
the elasticity of labor supply, and (2) labor costs are a large share of total production costs
(Ehrenberg and Smith 2000).
Arrow et al. state that, in the long run, environmental regulation is expected to cause a shift of
employment among employers rather than affect the general employment level (Arrow, et al.
1996). Even if they are mitigated by long-run market adjustments to full employment, many
regulatory actions have transitional effects in the short run (Smith 2015) (U.S. OMB 2015).
These movements of workers in and out of jobs in response to environmental regulation are
potentially important distributional impacts of interest to policy makers. Of particular concern
are transitional job losses experienced by workers who might not readily find new work. This
might include workers who have skills that do not transfer easily to other industries, or who
operate in declining industries, exhibit low migration rates, or live in communities or regions
where unemployment rates are high.
Workers affected by changes in labor demand due to regulation may experience a variety of
impacts including job gains or involuntary job loss and unemployment. Compliance with
environmental regulation can result in increased demand for the inputs or factors (including
labor) used in the production of environmental protection. However, the regulated sector
106 Berman and Bui (2001) also discuss a third component, the impact of regulation on factor prices, but conclude
that this effect is unlikely to be important for large competitive factor markets, such as labor and capital.
4-66
-------
generally relies on revenues generated by their other market outputs to cover the costs of
supplying increased environmental quality, which can lead to reduced demand for labor and
other factors of production used to produce the market output. Workforce adjustments in
response to decreases in labor demand can be costly to firms as well as workers, so employers
may choose to adjust their workforce over time through natural attrition or reduced hiring, rather
than incur costs associated with job separations (see, for instance, Curtis (Curtis 2018) and
Hafstead and Williams (Hafstead and Williams III 2018)).
As suggested in this discussion, the overall employment effects of environmental regulation
are difficult to estimate. Estimation is difficult due to the multitude of small changes that occur
in different sectors related to the regulated industry, both upstream and downstream, or in sectors
producing substitute or complimentary products. Consequently, employment impacts are hard to
disentangle from other economic changes and business decisions that affect employment, over
time and across regions and industries. See Section VIII.1.1 of the preamble for more information
and background on employment effects.
4.5.2 Potential Employment Impacts from the Increasing Penetration of Electric
Vehicles
In addition to the employment effects we have discussed in previous rules (for example the
2021 rule), the increasing penetration of electric vehicles in the market is likely to affect both the
number and the nature of employment in the auto and parts sectors and related sectors, such as
providers of battery charging infrastructure. Over time, as BEVs become a greater portion of the
new vehicle fleet, the kinds of jobs in auto manufacturing are expected to change. For instance,
there will be no need for engine and exhaust system assembly for BEVs, while many assembly
tasks will involve electrical rather than mechanical fitting. In addition, batteries represent a
significant portion of the manufacturing content of an electrified vehicle, and some automakers
are likely to purchase the cells, if not pre-assembled modules or packs, from suppliers whose
employment will thereby be affected. Employment in building and maintaining battery charging
infrastructure needed to support the ever-increasing number of BEVs on the road is also expected
to affect the nature of employment in automotive and related sectors. For much of these effects,
there is considerable uncertainty in the data to quantitatively assess how employment might
change as a function of the increased electrification expected to result under the standards. Some
suggest that fewer workers will be needed because BEVs have fewer moving parts (Krisher and
Seewer 2021), while others estimate that the labor-hours involved in BEVs is almost identical to
that for ICE vehicles (Kupper, et al. 2020).
Prior analyses of potential employment effects from electrification in the auto sector
conducted outside of EPA have estimated a range of impacts. Results from California's ACC II
program analysis seem to suggest that there may be a small decrease, not exceeding 0.3 percent
of baseline California employment in any year, in total employment across all industries in CA
through 2040 (California Air Resources Board 2022). A report by the Economic Policy Institute
suggests that U.S. employment in the auto sector could increase if the share of vehicles, or
powertrains, sold in the U.S. that are produced in the U.S. increases. The BlueGreen Alliance
also states that though BEVs have fewer parts than their ICE counterparts, there is potential for
job growth in electric vehicle component manufacturing, including batteries, electric motors,
regenerative braking systems, and semiconductors, and manufacturing those components in the
U.S. can lead to an increase in jobs in that sector (BlueGreen Alliance 2021). They go on to state
4-67
-------
that if the U.S. does not become a major producer for these components, there is risk of eventual
job loss.
The UAW states that re-training programs will be needed to support auto workers in a market
with an increasing share of electric vehicles in order to prepare workers that might be displaced
by the shift to new technologies (UAW 2020). Volkswagen states that labor requirements for
ICE vehicles are about 70% higher than their electric counterparts, but these changes in
employment intensities in vehicle manufacturing may be offset by shifting to the production of
new components, for example batteries or battery cells (Herrmenn, et al. 2020).107 As discussed
in Section VIII.1.1 of the preamble, investments in the EV sector, including in battery
manufacturing and supply chains, is already happening. For example, Volkswagen announced it
will start construction of a new electric vehicle battery gigafactory supporting up to 3,000 direct
jobs in Canada, as well as supporting a new EV manufacturing plant in South Carolina. (Collins
2023) (Dr. Ernst 2023) Research from the Seattle Jobs Initiative indicates that employment in a
collection of sectors related to both BEV and ICE vehicle manufacturing is expected to grow
slightly through 2029 (Seattle Jobs Initiative 2020). Climate Nexus also indicates that
transitioning to electric vehicles will lead to a net increase in jobs in the automotive and related
sectors, a claim that is partially supported by the rising investment in batteries, vehicle
manufacturing, and charging stations (Climate Nexus 2022).
This expected investment is also supported by recent Federal actions such as the BIL, the
CHIPS Act, and the IRA, all of which will allow for increased investment along the vehicle
supply chain, including domestic critical minerals, materials processing, battery manufacturing,
charging infrastructure, and vehicle assembly and vehicle component manufacturing. The BIL
was signed in November 2021 and provides over $24 billion in investment in electric vehicle
chargers, critical minerals, and battery components needed by domestic manufacturers of EV
batteries and for clean transit and school buses. (Infrastructure Investment and Jobs Act 2021).108
The CHIPS Act, signed in August 2022, invests in expanding America's manufacturing capacity
for the semiconductors used in electric vehicles and chargers (CHIPS Act of 2022 2022).109 The
IRA provides incentives for producers to expand domestic manufacturing of BEVs and domestic
sourcing of components and critical minerals needed to produce them (117th Cong. 2022). The
IRA also provides incentives for consumers to purchase both new and used ZEVs. These laws
create domestic employment opportunities along the full automotive sector supply chain, from
components and equipment manufacturing and processing to final assembly, as well as
incentivize the development of reliable EV battery supply chains, as discussed in Section VI.1.1
107 We also note that, as discussed in Sections VI.I.l and VI.I.4 of the preamble, and RIA Chapter 4.5.5, skill sets for
ICE workers are similar to those in the EV sector and other non-automotive sectors.
108 The Bipartisan Infrastructure Law is officially titled the Infrastructure Investment and Jobs Act. More
information can be found at https://www.fhwa.dot.gov/bipartisan-infrastructure-law/
109 The CHIPS and Science Act was signed by President Biden in August 2022 to boost investment in, and
manufacturing of, semiconductors in the U.S. The fact sheet can be found at https://www.whitehouse.gov/briefing-
room/statements-releases/2022/08/09/fact-sheet-chips-and-science-act-will-lower-costs-create-jobs-strengthen-
supply-chains-and-counter-china/
4-68
-------
of the preamble.110 The BlueGreen Alliance and PERI estimate that IRA will create over 9
million jobs over the next decade, with about 400,000 of those jobs being attributed directly to
the battery and fuel cell vehicle provisions in the act (Political Economy Research Institute
2022). Additional studies find similar results: the IRA and BIL have the potential to lead to
significant job increases in transportation, electricity, and manufacturing, with some estimates of
almost 700,000 new jobs through 2030. EDF reports that more than 46,000 jobs in EV
manufacturing have already been announced since the passage of the IRA.
The U.S. Bureau of Labor Statistics (BLS) identified three key occupational areas they expect
to be affected by growth in the BEV market: the design and development of EV models, the
production of batteries, and installation and maintenance of charging infrastructure. The article
estimates changes in key occupations employed in those sectors between 2021 and 2031 (Colato
and Ice 2023). The authors note that though it is expected that the occupations outlined in the
article will be significant in BEV production and deployment, they include estimates of the total
employment change for each occupation across all sectors, not just those related to BEV
production and deployment. For example, the estimates for the change in employment of
construction laborers is the effect from all construction sectors, not just those related to the
construction of BEV charging infrastructure. In the report, BLS estimated employment changes
related to occupations employed in the design and development of electric vehicles, including
software developers, electrical engineers, electronics engineers, and chemical engineers; battery
manufacturing, including electrical, electronic and electromechanical assemblers, and
miscellaneous assemblers and fabricators; and charging network development and maintenance,
including urban and regional planners, electrical, electrical power-line installers and repairers,
and construction laborers. With the exception of the sector miscellaneous assemblers and
fabricators, BLS forecast an increase in employment in these occupational areas, with the
smallest increase, in percentages, being 1.6 percent (electrical engineers), and the largest increase
(software developers) being 26 percent.111 BLS states that though total employment in the
miscellaneous assemblers and fabricators sector is projected to fall, they do expect a number of
job openings in the sector each year in order to replace workers who transfer to different
occupations or exit the work force. Again, it is difficult to separate out the effect that the increase
in BEV production will have on these sectors from the macroeconomic effects, or the effects
from non-BEV related production activity.
4.5.3 Potential Employment Impacts of the Standards
Even with expected increases in employment in component production and new domestic jobs
related to ZEVs, shifts in production may negatively affect workers currently employed in
production of ICE vehicles. We acknowledge the possibility of geographically localized effects,
and that there may be job quality impacts associated with this rule, especially in the short term.
We note that there are Federal programs to assist workers in the transition to low or zero emitting
vehicles, including a DOE funding package which makes $2 billion in grants, and up to $10
110 More information on how these acts are expected to aid employment growth and create opportunities for growth
along the supply chain can be found in the January 2023 White House publication "Building a Clean Energy
Economy: A Guidebook to the Inflation Reduction Act's Investments in Clean Energy and Climate Action." found
online at https://www.whitehouse.gov/wp-content/uploads/2022/12/Inflation-Reduction-Act-Guidebook.pdf
111 The urban and regional planners sector is forecast to have the smaller increase in number of employees, with an
increase of about 1,600 employees between 2021 and 2031.
4-69
-------
billion in loans available to support projects converting existing automotive manufacturing
facilities to support electric vehicle production.112 The funding package is expected to result in
retention of high-quality, high-paying jobs in communities that currently host these
manufacturing facilities, and along the full supply chain for the automotive sector, from
components to assembly. The grants available give priority to refurbishing and retooling
manufacturing facilities, especially for those likely to retain collective bargaining agreements
and/or an existing higher-quality, high-wage hourly production workforce. (U.S. Department of
Energy Office of Manufacturing and Energy 2024) DOE has also announced funding to support
clean energy supply chains, with the funding going toward projects to support domestic clean
energy manufacturing (including projects supporting battery production) in, or near, nine
communities that were formerly tied to coal mining, and are expected to create almost 1,500
jobs. (EnergyTech 2023) We also note that during, and after, the comment period, several major
U.S. automakers were negotiating new labor contracts, with an emphasis on workers in facilities
that support the production of electrified vehicles, with the results including increased wages and
abilities for workers to join the union.
However, there is no data to estimate current or future job quality. Nor are we able to
determine the future location of vehicle manufacturing and supporting industries beyond the
public announcements made as of the publication of this rule. We note that, compared to the
proposal, we are finalizing standards that extend flexibilities and provide a slower increase in the
stringency of the standards in the early years of the program. The more gradual shift allows for a
more moderate pace in the industry's scale up to the battery supply chain and manufacturing,
which in turn should help to reduce any potential impacts in employment across all sectors
impacted by this rule. For example, employers may choose to reduce hiring early on to avoid the
need to reduce their workforce later. In addition, as illustrated by the alternative pathway
analyses shown in Section IV.F and IV. G of the preamble and Chapter 12 of the RIA, there are
multiple ways OEMs can choose to meet the standards, including through a wide range of BEV
and PHEV technologies, and all of these pathways continue to provide ICE vehicles to the
market. In addition, as explained above, in Section VIII.1.1 of the preamble and Section 20 of the
RTC, there are many programs and initiatives focused on training, retraining, and community-
level impacts; there is a wide range of ICE automotive jobs with similar skill sets to those in EV
automotive production and in other industries, and infrastructure work is and will continue to be
a nation-wide effort.
Shifts in PEV production associated with the final rule may lead to employment shifts with
positive impacts on affected workers. A BLS report (Hamilton 2011)113 provides detailed
descriptions of occupations employed in EV production.
Because of a variety of significant unknowns and challenges, including the state of the
macroeconomy when these standards become effective, the changes to auto manufacturing
employment due to increased production of electric vehicles, and the difficulties of modeling
impacts on employment in a complex national economy, we focus our employment impacts
analysis on the direct impacts on labor demand in closely affected sectors. In the next sections,
112 https://www.energy.gov/articles/biden-harris-administration-announces-155-billion-support-strong-and-just-
transition
113 See https://www.bls.gov/green/electric_vehicles/electric_vehicles.pdf
4-70
-------
we discuss potential impacts of the rule on industry-level demand for labor. We qualitatively
describe the employment impacts due to the factor shift, demand, and cost effects on labor
demand, following the structure of Morgenstern et al., as described above. Then we present a
quantitative estimate of partial incremental employment effects of the standards, and conclude
with a discussion of possible employment impacts on related sectors.
As discussed in RTC Section 20 and Section VIII.1.1 of the preamble, there are many existing
and planned projects focused on training new and existing employees in fields related to green
jobs, and specifically green jobs associated with electric vehicle production, maintenance and
repair, and the associated charging infrastructure. This includes work by the Joint Office of
Energy and Transportation (JOET), created by the BIL, which supports efforts related to
deploying infrastructure, chargers and zero emission vehicles.114 One example of a project from
the JOET is the Ride and Drive grant program, which targets investments in EV charging
resiliency, community-driven workforce development, and EV charging performance and
reliability. Another example is the Battery Workforce Initiative established by the Department of
Energy (DOE) in coordination with the Department of Labor (DOL), AFL-CIO, and other
organizations with the goal of accelerating the development of high-quality training. DOL has
also established the Building Pathways to Infrastructure Jobs Grant Program, which support
worker-centered sector strategy training programs. DOL also provides grants to help community
colleges provide skilled pathways to good jobs in the transportation and clean energy sectors.
DOL is also providing technical assistance to the Southeast EV Collaborative, which is made up
of a collection of state workforce agencies in the southeast region of the U.S. focused on
identifying opportunities to work together to provide equitable access to good jobs across the
region.
4.5.3.1 The Factor Shift Effect
The factor shift effect reflects employment changes due to changes in labor intensity of
production by regulated entities resulting from compliance activities. Holding vehicle sales
constant, a factor shift effect of this rule might occur if this regulation affects the labor intensity
of production of ICE vehicles, though we do not have data on how the regulation might affect
labor intensity of production within ICE vehicle production. It may also occur if PEV production
does not have the same labor intensity as ICE vehicle production and, holding vehicle sales
constant, the share of PEV and ICE vehicles changes. There is ongoing research on the different
labor intensity of production between ICE and BEV production, with inconsistent results. Some
research indicates that the labor hours needed to produce a BEV are fewer than those needed to
produce an ICE vehicle, while other research indicates there are no real differences.
EPA worked with a research group, FEV, to produce a peer-reviewed tear-down study of a
BEV (Volkswagen ID.4) to its comparable ICE vehicle counterpart (Volkswagen Tiguan).115
Peer reviewed study results were delivered in May 2023. Included in this study are estimates of
labor intensity needed to produce each vehicle under three different assumptions of vertical
integration of manufacturing scenarios ranging from a scenario where most of the assemblies and
components are sourced from outside suppliers to a scenario where most of the assemblies and
114 More information on these programs, and other programs, can be found in the memo Labor/Employment
Initiatives in the Batten /Vehicle Electrification Space located in the docket for this rule.
115 See RIA Chapter 2.5.2.2.3 for more information.
4-71
-------
components are assembled in house. Under the low and moderate levels of vertical integration,
results indicate that assembly time of the BEV at the plant is reduced compared to assembly time
of the ICE vehicle.116 Under a scenario of high vertical integration, which includes the BEV
battery assembly, results show an increase in time needed to assemble the BEV. When
powertrain systems are ignored (battery, drive units, transmission, and engine assembly), the
BEV requires more time to assemble under all three vertical integration scenarios. The results
indicate that the largest difference in assembly comes from the building of the battery pack
assembly. When the battery cells are built in-house, the BEV will require more labor hours to
build at the assembly plant. These results also indicate that if the labor input to manufacture
batteries is included in the estimated labor needs to build a BEV, regardless of the vertical
integration decisions to build batteries in-house, BEVs will require more labor to build. For more
information on this study, see Chapter 2.5.2.2.3. For information on the early indications of labor
differences, including intensity, in ICE and BEV production, see Chapter 4.5.4.
Data on the labor intensity of PHEV production compared to ICE vehicle production is also
very sparse. PHEVs share features with both ICE vehicles, including engines and exhaust
assemblies, and BEVs, including motors and batteries. If labor is a function of the number of
components, PHEVs might have a higher labor intensity of production compared to both BEV
and ICE vehicles, and if they are produced in the U.S. may provide labor demand. The labor
needs of battery production are also a factor of the total labor needs to build a PHEV.
Given the current lack of data and inconsistency in the existing literature, we are unable to
estimate a quantitative factor-shift effect of increasing relative PEV production as a function of
this rule. However, we can say, generally, that research indicates that if production of PEVs and
their power supplies are done in the U.S. at the same rates as ICE vehicles, we do not expect
employment to fall, and it may likely increase. Electric vehicle manufacturing plants and battery
plants are being announced and built in the U.S., as discussed in Section IV of the preamble. In
addition, as discussed in Section VIM.2 of the preamble, states are making efforts to support
increasing domestic production of electric vehicles and batteries, including support for the
workforce. South Carolina is focused on exploring opportunities related to electric vehicle and
automotive manufacturers,117 Ohio estimates more than 25,000 new jobs in EV manufacturing
and maintenance, battery development, and charging station installation and operations in the
state by 2030,118 California is focused on an equitable ZEV industry,119 Illinois has invested in
EV training, R&D, and workforce development and community support,120 Nevada is focused on
workforce and economic development supporting the lithium industry,121 Kentucky is providing
116 In the FEV report, "assembly time" is the time (in hours) it takes to assemble the vehicle from the component
parts.
117 SCpowersEV: State support - Driving the Future, https://scpowersev.com/state-support.
118 Accelerating Ohio's Auto & Advanced Mobility Workforce, Auto and Advanced Mobility Workforce Strategy,
2023. https://workforce.ohio.gov/wps/wcm/connect/gov/2e9f6e52-a4bc-4ef6-9080-
e6b06f067ala/Ohio%27s+Electric+Vehicle+Workforce+Strategy.pdf?MOD=AJPERES.
119 California Workforce Development Board, 2021. https://business.ca.gov/wp-
content/uploads/2021/03/CWDB_ZEV-Plan.pdf
120 Illinois Drive Electric: Abundant Workforce, https://ev.illinois.gov/grow-your-business/abundant-
workforce.html.
121 Nevada Battery Coalition: https://nevadabatterycoalition.com/about/
4-72
-------
resources toward upgrading industrial sites throughout the state,122 Tennessee is co-locating a
new Tennessee College of Applied Technology with a new EV manufacturing facility,123
Michigan is assisting with tuition and other supportive services for advanced automotive
mobility and electrification training.124'125
4.5.3.2 The Demand Effect
Demand effects on employment are due to changes in labor that result from changes in total
new vehicle sales. Compliance activities may increase the cost of production, raising market
prices and decreasing demand for the regulated industry's output and, holding labor intensity
constant, decreasing the regulated industry's demand for labor. This final rule may result in a
decrease in total new vehicle sales, suggesting that the demand effect would be a decrease in the
demand for labor.
In previous EPA LD regulations, like the 2021 rule, we have used the CAFE model to
estimate effects of a change in ICE vehicle demand on labor. The model uses a method of
estimating a demand effect on employment through the relationship of hours involved in a new
vehicle sale (for the effect on automotive dealers) or average labor hours per vehicle at a sample
of US assembly plants (for the effect on the final assembly industry) to the change in the number
of vehicles sold due to the regulation. This rule, however, uses EPA's OMEGA model. We
currently do not have the data to estimate these effects in OMEGA.
In general, if the regulation causes total sales of new vehicles to decrease, fewer workers will
be needed to assemble vehicles and manufacture their components. The demand effect may be
different for PHEV, BEV, and ICE vehicles, though, if their labor intensity of production is
different. If, for example, BEV vehicles have a lower labor intensity of production than PHEV or
ICE, and sales for BEV, PHEV, and ICE vehicles fall by the same amount, the demand effect on
labor will be smaller for BEVs than PHEVs or ICE vehicles. Due to lack of data, as discussed in
Chapter 4.5.3.1, we are unable to estimate a change in employment due to a change in demand.
We note, however, that, as explained in Chapter 4.4.2, sales effects due to this rule are small, and
therefore associated demand effects are also likely to be small.
4.5.3.3 The Cost Effect
The cost effects on employment are due to changes in labor associated with changes in costs
of production. Compliance activities may cause production costs to increase, and firms to use
more of all inputs, including labor, to produce the same level of output. This may be, in part, due
to firms producing environmental protection at the same time as industry output. In general, if a
regulation leads firms to invest in lower-emitting vehicles, we expect an increase in the labor
used to implement those technologies. In this final rule, as in previous LD and heavy-duty (HD)
122 Kentucky: Leading the Charge, https://ced.ky.gov/Newsroom/Article/20230816_Leading_th.
123 Area Development: Tennessee: A growing Capital of Electric Vehicle Production,
https://www.areadevelopment.eom/ContributedContent/Q4-2021/tennessee-growing-capital-of-electric-vehicle-
production, shtml
124 MI Labor and Economic Opportunity: Electric Vehicle Jobs Academy, https://www.michigan.gov/leo/bureaus-
agencies/wd/industry-business/mobility/electric-vehicle-jobs-academy.
125 Michigan Engineering News, $130M Electric Vehicle Center launches at U-Michigan,
https://news.engin.umich.edu/2023/04/130m-electric-vehicle-center-launches-at-u-michigan/.
4-73
-------
rules, we have estimated partial employment effects due to the change in costs of production,
where the change in costs of production are assumed to be the change in technology costs
associated with the rule. We use the historic share of labor in the cost of production to
extrapolate future estimates of impacts on labor due to compliance activities in response to the
regulation. Specifically, we multiplied estimates of the average ratio of labor to the value of
output in the regulated industry by the change in production costs estimated as an impact of the
rule. This provides a sense of the magnitude of potential impacts on employment, though as
explained in more detail in the next paragraph, this estimate is incomplete and limited. As
explained further in this chapter, the impacts estimated in this rule are a combined cost and
demand effect due to how costs are estimated in OMEGA.
The use of the average ratio of labor to value of output to estimate a cost effect on
employment has both advantages and limitations. It is often possible to estimate these ratios for
detailed sector definitions, for example, the average number of workers in the automobile and
light-duty motor vehicle manufacturing sector per $1 million spent in that sector, rather than
using ratios from more aggregated sectors, such as the motor vehicle manufacturing sector. This
avoids extrapolating employment ratios from less closely related sectors. On the other hand,
these estimates are averages, covering all the activities in these sectors, and may not be
representative of the labor effects when expenditures are required for specific activities, or when
manufacturing processes change due to compliance activities in such a way that labor intensity
changes. For instance, the ratio of workers to production cost for the motor vehicle body and
trailer manufacturing sector represents this ratio for all motor vehicle body and trailer
manufacturing activities, and not just for production processes related to emission reductions
compliance activities. Another limitation is that the ratios are averaged across firms and do not
reflect variability across individual producers. In addition, these estimates do not include changes
in industries that supply these sectors, such as steel or electronics producers. The effects
estimated with this method can be viewed as effects on employment in the sectors included in the
analysis due to the changes in expenditure in that sector, rather than as an assessment of all
employment changes due to the standards being analyzed. In addition, labor intensity is held
constant in the face of increased expenditures; this approach does not take changes in labor
intensity due to changes in the nature of production (the factor shift effect) into account, which
could either increase or decrease the employment impacts estimated using this method.
BEVs and ICE vehicles require different inputs and have different costs of production, though
there are interchangeable, or common, parts as well. We used a recent report from the Seattle
Jobs Initiative, which identified sectors most strongly associated with ICE and BEV automotive
production, to determine a list of sectors that may be directly affected by our rule (Seattle Jobs
Initiative 2020). NAICS Sectors identified by the Seattle Jobs Initiative as mainly associated
with BEV production include Electrical equipment manufacturing and Other electrical
equipment and component manufacturing. Sectors identified as related to both EV and ICE
manufacturing include Motor vehicle manufacturing, Motor vehicle body and trailer
manufacturing, and Motor vehicle parts manufacturing. And a sector identified as only
4-74
-------
associated with ICE vehicle manufacturing is Motor vehicle gasoline engine and engine parts
manufacturing.126
The Employment Requirements Matrix (ERM) provided by the U.S. Bureau of Labor
Statistics (BLS) provides direct estimates of the number of employees per $1 million in sales for
202 aggregated sectors, which roughly correspond to the 4-digit NAICS code level, from 1997
through 2022 (Bureau of Labor Statistics 2023). Figure 4-10 shows the estimates of employment
per $1 million of sales for a set of ERM sectors that generally correspond to the NAICS sectors
identified above, adjusted to 2022 dollars using the U.S. Bureau of Economic Analysis Gross
Domestic Product Implicit Price Deflator.127
The historical data show that over time, the amount of labor needed in the motor vehicle
industry has changed: automation and improved methods have led to significant productivity
increases. In Figure 4-10, we can see that the workers per $1 million in sales has, generally,
decreased over time, with the exception of Electrical equipment manufacturing and Other
electrical equipment and component manufacturing. For instance, in 1997 about 1.2 workers in
the Motor vehicle manufacturing sector were needed per $1 million, but only 0.7 workers by
2022 (in 2022$).128 Though two sectors mainly associated with BEV manufacturing, Electrical
equipment manufacturing, and Other electrical equipment and component manufacturing, show
an increase in recent years.
126 The sector Motor vehicle gasoline engine and engine parts manufacturing is a subsector of Motor vehicle parts
manufacturing.
127 The GDP IDP used can be found in the excel file "LMDV FRM EmploymentlmpactsCalculations.xlsx " in the
docket.
128 https://www.bls.gov/emp/data/emp-requirements.htm; this analysis used data for the sectors electrical equipment
and manufacturing, other electrical equipment and component manufacturing, motor vehicle manufacturing, motor
vehicle body and trailer manufacturing, and motor vehicle parts manufacturing from "Chain-weighted (2012 dollars)
real domestic employment requirements tables;" see the excel file
"LMDV FRM EmploymentlmpactsCalculations.xlsx " in the docket.
4-75
-------
7 000
0.000
cS3 cS3 $ ^ c? •0' "v3 "v3 ^ "O1 ^ •?* ^ ^ r{y
Vs -c^3 -fv3 _(S° -{V3 -cO _c\N _c\^ .c\N rg1s rKN .cO -CN1 _t\N .c\N _r\k _c\^ _r\l'
Year
rp 9 «P V
— ~ — Semiconductor and other electronic component manufacturing
— • — Electrical equipment manufacturing
— ¦ — Other electrical equipment and component manufacturing
M Motor vehicle manufacturing
* Motor vehicle body and trailer manufacturing
• Motor vehicle parts manufacturing
Figure 4-10: Workers per million dollars in sales, adjusted for domestic production.
4.5.4 Partial Employment Effects of the Standards
In previous LD rules, EPA estimated a cost effect on employment holding sales constant, and
assuming labor intensity is held constant in the face of increasing expenditures. However, the
costs of this rule are estimated in OMEGA, which works iteratively to estimate a vehicle fleet
that will meet the regulatory standards, as well as be accepted by consumers. The model
estimates this by both changing the number of new vehicles sold, as well as changing the
penetration of PIlEVs in the market between the No Action and Action cases. Therefore, though
the method used, described in Chapter 4.5.3, is the same as that in previous rules, we are unable
to estimate a cost employment effect due to this rule while holding sales constant. Therefore, the
partial employment analysis presented here is a change in employment due to the change in
costs, allowing sales to change as well. In other words, it is a combined cost and demand effect.
We estimate the partial employment effect using the historic share of labor in the cost of
production for a set of sectors affected by this rule. We use these historic shares to extrapolate
estimates of future shares of labor in the cost of producti on for each of those sectors. We then
multiply the estimated share of labor in the cost of production by the change in production costs
estimated as an impact of this rule. This provides a sense of the magnitude of potential impacts
4-76
-------
on employment. Some of the advantages and limitations of this method are described above, in
Chapter 4.5.3.3.
We rely on three different public sources to estimate a range of employment per $1 million
expenditures: the Economic Census (EC) and the Annual Survey of Manufactures (ASM), both
provided by the U.S. Census Bureau, and the Employment Requirements Matrix (ERM) provided
by the U.S. Bureau of Labor Statistics (BLS). The EC is conducted every 5 years, most recently
in 2022, however because the data is not yet publicly available, we use EC data from 2017.129
The ASM is an annual subset of the EC and is based on a sample of establishments. The latest set
of data from the ASM is from 2021. The EC and ASM have more sectoral detail than the ERM,
providing estimates out to the 6-digit North American Industry Classification System (NAICS)
code level. They provide separate estimates of the number of employees and the value of
shipments, which we convert to a ratio for this employment analysis. The total employment
across the NAICS code sectors used in this analysis (see Table 10-6) as reported in the ASM and
the EC ranges from about 1,052,500 to about 1,053,800 depending on which data source is used;
as noted above the most recent data for ASM and EC are from 2022 and 2017, respectively. The
ERM provides direct estimates of employees per $1 million in expenditures for a total of 202
aggregated sectors that roughly correspond to the 4-digit NAICS code level, and it provides data
through 2022.
We estimate cost effects on employment by separating out costs mainly associated with the
electrified portions of vehicle production (for example, batteries) and the ICE vehicle portion of
production (for example, engines), as well as the costs that are common between then (for
example, gliders130). We apply the electrified portions of cost changes ("PEV related costs") only
to sectors primarily focused on electrified portions of vehicle production, the ICE vehicle portion
of costs ("ICE related costs") only to sectors primarily focused on the ICE vehicle portions of
production, and the costs common to both the electrified portions and ICE portions of vehicle
production ("common costs") to sectors that are common to the electrified and ICE portions of
vehicle production.131 We use the sum of the estimated PEV related costs, ICE related costs, and
common costs for both LD and MD vehicles. We used a report from the Seattle Jobs Initiative to
identify sectors most strongly associated with the electrified portions of automotive production
and the ICE vehicle portions of automotive production, as well as the sectors that are common
between them (Seattle Jobs Initiative 2020).
Table 4-29 below shows the sector definitions, NAICS codes, and ERM sector numbers EPA
used to estimate employment effects in this analysis. It also provides the estimates of
129 The 2022 Economic Census was conducted starting in January 2023 and initial results will be available starting in
March 2024.
130 In this context, a glider is a vehicle without a powertrain. It includes the body, chassis, interior, and non-
propulsion related electrical components.
131 A report from the Seattle Jobs Initiative examined how electrification in the automotive industry might advance
workforce development in Oregon and Washington. As part of that study, the authors identified the sectors classified
by the North American Industry Classification System (NAICS) codes most strongly associated with automotive
production in general, those exclusive to ICE vehicles, and those primarily associated with electrified portions of
vehicle production. The report can be found at:
https://www.seattle.gov/Documents/Departments/OSE/ClimateDocs/TE/EV%20Field%20in%200R%20and%20W
A_February20.pdf.
4-77
-------
employment per $1 million of expenditure for each sector for each data source, adjusted to 2022
dollars using the U.S. Bureau of Economic Analysis Gross Domestic Product Implicit Price. The
values published in the ERM have been adjusted by BLS to remove the effects of imports. We
adjust the values for the ASM and EC to remove effects of imports through the use of a ratio of
domestic production to domestic sales of 0.81.132 While the estimated labor ratios differ across
data sources, they are fairly similar, and mainly exhibit a similar pattern across the ICE and
common sectors. The ASM and EC exhibit similar orders of most to least intensive across all
sectors, though some of the ratios in specific sectors differ across the data sets. However, the
order of most labor intensive to least intensive as estimated by the ERM differs. This may be due
to the inclusion of additional NAICS sectors within the larger ERM sectors.133 Within the ASM
and EC data, "Other electronic component manufacturing" seems to be the most labor-intensive
sector, while ERM indicates "Motor and generator manufacturing" is the most labor-intensive.
Automobile and "Light-duty motor vehicle manufacturing" is the least labor-intensive sector
across all three data sources.
Table 4-29: Sectors and associated workers per million dollars in expenditures by source.
Sector
Other electronic component
manufacturing
'S Motor and generator manufacturing
$ Battery manufacturing
H
Ph
_o
All other miscellaneous electrical
equipment and component
manufacturing
Automobile and light duty motor
^ vehicle manufacturing
^ Motor vehicle body and trailer
¦a manufacturing
^ S Motor vehicle parts manufacturing
o U
H o
U "C
HH aj
C/3
NAICS Colic
I.RM
Ratio of Workers per SI Million Kxpi-nilitun-s
Sector
ASM (2018)"
KC: (2017)"
HUM (2022)
334419
72
3.4
4.1
3.5
335312
77
1.9
2.8
5.1
33591
78
2.4
3.2
335999
78
3.8
2.3
3.2
33611
79
0.6
0.6
0.9
3362
80
2.3
2.9
4.5
3363*
81
2.1
2.2
33632
81
2.1
2.4
3.0
33631
81
(not gasoline engines)
Motor vehicle electrical and electronic
equipment manufacturing
Motor vehicle gasoline engine and
engine parts manufacturing 15 15
Values are adjusted for domestic vs. foreign production.
* In our analysis. 3363 excludes estimates for NAICS code 33631. NAICS code 33631 only includes ICE vehicle manufacturing, so we
subtract those data out from the main sector. NAICS code 3363. and apply ICE costs to that sub-sector.
Because the ERM is available annually for 1997-2022, we use these data to estimate
productivity improvements over time. We estimate a simple regression of the logged ERM
values on a year trend for each sector.134 We use this approach because the resulting coefficient
132 To estimate the proportion of domestic production affected by the change in sales, we use data from WardsAuto
for total car and truck production in the U.S. compared to total car and truck sales in the U.S. Over the period 2012-
2022, the proportion averages 84 percent. From 2017-2022, the proportion average is slightly lower, at 81 percent.
133 ERM sectors are based on the 4-digit level for NAICS code sectors. For example, ERM sector 72, consists of
results from manufacturers in NAICS code 3344.
134 Details and results are found in the file LMDV FRM EmploymentlmpactsCalculations.xlsx, which is in the
docket for this rule.
4-78
-------
describing the relationship between time and productivity is a direct measure of the average
percent change in labor productivity for a given level of output per year. The results shown in
Table 4-30 below represent the percent change in the ratio of labor to value of output per year.
These productivity changes are all negative, indicating that, for the sectors shown here, fewer
workers are needed per $1 million of output every year. For example, in Motor vehicle parts
manufacturing, the ratio falls by 2.5 percent every year. These figures coincide with the historic
decline in workers per million dollars in sales as seen in Figure 4-10.
Table 4-30 Annual change in the ratio of labor to value of output for directly impacted
sectors (%).
Sector
ERM
Annual %
Sector
Change
Number
Other electronic component
72
-5.8%
manufacturing
Electrical equipment
77
-0.4%
manufacturing
Other electrical equipment and
78
0.3%
component manufacturing
Motor vehicle manufacturing
79
-2.6%
Motor vehicle body and trailer
80
-0.6%
manufacturing
Motor vehicle parts
81
-2.5%
manufacturing
We then use those estimated percent improvements in productivity (fall in the ratio of labor to
value of output) to project the number of workers per $1 million of production expenditures
through 2032. We emphasize that the estimates provided in Table 4-31 represent an order of
magnitude effect, rather than definitive impacts. We calculate separate sets of projections
(adjusted to 2022$) for each set of data (ERM, EC, and ASM) for all sectors described above.
The ERM projections are calculated directly from the fitted regression equations used to estimate
the projected productivity growth, since the regressions themselves used ERM data. For the
ASM and EC projections of the number of workers needed per $1 million of expenditures (in
2022$), we apply ERM's ratio of projected annual productivity growth to the projected
production expenditure value in 2021 for the ASM and 2017 for the EC (the base years in our
data).135 In other words, we apply the projected productivity growth estimated using the ERM
data to the ASM and EC numbers.
To interpret the results, we compare the projected employment across data sources and report
only the maximum and minimum (in absolute value) effects in each year across all sectors.136 We
provide a range rather than a point estimate to emphasize the uncertainty in these estimates. The
135 The ERM data accounts for an adjustment to reflect domestic production. We apply an adjustment factor to the
ASM and EC values as described in footnote 132 to remove the effects of imports on the projections from those data
sources as well.
136 To see details, as well as results for all sources, see "LMDV FRM EmploymentlmpactsCalculations.xlsx" in the
docket.
4-79
-------
reported ranges provide an estimate of the expected magnitude of the effect. The employment
effect estimated here includes the costs of this rule for both LD and MD vehicles, as well as the
change in new vehicles sales for LD vehicles due to this rule. There are no estimated changes in
MD vehicle sales. See Chapter 4.4.2 for more information on the estimates of new vehicle sales
effects due to this rule.
For this analysis, we use detailed OMEGA results to get estimates of the costs of
manufacturing LD and MD vehicles separated out by costs expected to apply only to the
electrified portion of vehicle production, those expected to only apply to the ICE vehicle portion
of production, and those expected to apply to all vehicles.137 These costs (in $ million) are
multiplied by the estimates of workers per $1 million in costs for each year. Table 4-31 shows
the projected estimates of partial employment effects for each year for the three sector groups.
The effects are shown in job counts.138 We show a range of effects, from the smallest to largest
(in absolute value) effect for each sector group. Allowing for the estimated change in sales due to
the rule, increased technology costs of vehicles and parts is expected to increase employment
over the 2027-2032 timeframe for sectors focused on the electrified portion of vehicle
production, and the results show a decrease for the ICE focused sectors (except for 2027) and the
common sectors.
It should be noted that these results are exclusive of any changes in employment in related
sectors, such as charging infrastructure. In addition, while we estimate employment impacts
beginning with program implementation, some of these employment gains may occur earlier, for
example if vehicle manufacturers and parts suppliers hire staff in anticipation of compliance with
the standards, or in anticipation of ramping up PEV production.
Table 4-31: Estimated partial employment effects for sectors focused on the electrified, ICE
and common portions of vehicle production3.
Common Portions Electrified Portion ICE Portion
Year
Smallest
Largest
Smallest
Largest
Smallest
Largest
Effect
Effect
Effect
Effect
Effect
Effect
2027
-370
-3.600
3,000
6,900
2.200
2.900
2028
-900
-8.600
15.700
36.600
-800
-1.100
2029
-1.300
-13.000
36.800
89.100
-7.600
-9.800
2030
-1.900
-19.800
54.800
140.200
-13.600
-17.500
203 1
-2.100
-22.600
67,700
182.600
-18.800
-24.200
2032
-2.600
-27,700
75.100
213.900
-23,200
-29.900
a Smallest and largest effects are smallest and largest in absolute value
Table 4-32 shows the maximum combined range for the estimated change in employment
across all sectors. This represents the range from the largest employment loss (or smallest
employment gain) estimated to the largest employment gain estimated across the combination of
137 Vehicle technology cost estimates for this rule were developed in OMEGA. Chapter 9 in the RIA provides
information on the total and per-vehicle costs estimated.
138 ERM reports employment as a count of jobs, which are not based on a full-time equivalent basis.
4-80
-------
sector groups.139 The shows an expected increase in employment from 2027 through 2032.
Interpreting these results in terms of direction and relative magnitude, this estimate indicates that
possible job growth over time in PEV related sectors will be greater than possible job loss in ICE
or common sectors.
Table 4-32: Estimated maximum combined range of estimated partial employment effects
across all sectors.
Year Maximum Combined
Range
2027 i
1.600
9.400
2028
6,000
34.900
2029
14.000
80.200
2030
17.600
124.700
203 1
20.800
161.700
2032 !
17.400
188.100
As discussed in Chapter 4.5.3.1, EPA contracted with FEV to perform a detailed tear-down
study comparing two similar vehicles, a 2021 Volkswagen ID.4 (BEV) and a 2021 Volkswagen
Tiguan (ICE) (see RIA Chapter 2.5.2.2.3 for more details on this study). The results shown here
are consistent with the results of the FEV tear-down study and indicate that, even if fewer labor
hours are needed at the assembly plant, increased labor hours will be needed elsewhere in the
supply chain for the electrified portions of production, for example in building and assembling
battery packs.
While some workers may experience transitory negative impacts if they cannot readily move
into alternative employment of similar quality, these analyses suggest greater likelihood of
overall job growth over the period of these standards. In addition, as noted in Chapters 4.5.2 and
4.5.3, and throughout RTC Section 20, support for domestic employment related to electric
vehicle production, from supply chain, to production, to infrastructure, is increasing, including
through multiple programs implemented through DOE, as well as through provisions in the BIL,
IRA, and CHIPS Act.
4.5.5 Employment Impacts on Related Sectors
Economy-wide impacts on employment are generally driven by broad macroeconomic effects.
However, employment impacts, both positive and negative, in sectors upstream and downstream
from the regulated sector, or in sectors producing substitute or complementary products, may
also occur as a result of this rule.
For example, as described in RIA Chapter 8.5, we expect the rule to cause a small decline in
liquid fuel consumption and a small increase in electricity generation which may have
consequences for labor demand in those upstream industries, as well as associated industries
139 This is not a straight sum of the smallest and largest effect from Table 4-31, which are based on absolute value
(closest to and furthest from zero) and is not affected by the direction of the effect, but a sum of the minimum and
maximum effects, which include direction of the effect.
4-81
-------
such as extracting, refining, transporting, and storing of petroleum fuels. The lower per-mile fuel
costs could lead to increases in demand for ride-hailing services and cause increases in demand
for drivers in those jobs. Increased mobility related to the lower cost per mile of driving, as
discussed in Section VIII.D. 1 of the preamble, may also benefit owner/operators in MD fleets
through the fleet being able to service a greater range of customers, or benefit consumers through
increasing their feasible geographic area for employment opportunities. Firms producing
substitutes or complements to the goods produced by the regulated industry may also experience
changes in demand for labor. For example, the expected decline in gas station visits may lead to
reduced demand for labor in that sector. Although gasoline stations will sell less fuel, the fact
that many provide other goods, such as food or car washes, moderates possible losses in this
sector. Note that, as discussed in Section VIII.I.4 of the preamble, traditional gas stations and
liquid fuel providers are already incorporating EV charging into their business strategies. There
will also likely be an increase in demand for labor in sectors that build and maintain charging
stations. The magnitude of these impacts depends on a variety of factors including the labor
intensities of the related sectors as well as the nature of the linkages between them and the
regulated firms.
Expected petroleum fuel consumption reductions found in Chapter 8.5 represent fuel savings
for purchasers of fuel, however they also represent a potential loss in value of output for the
petroleum refining industry, fuel distributors, and gasoline stations. The loss of expenditures
could impact sectors throughout the petroleum fuel supply chain, including petroleum refiners,
pipeline construction, operations and maintenance, domestic oil production, and gasoline
stations, and could result in reduced employment in these sectors. In this final rule, we estimate
that the reduction in fuel consumption (see RIA Chapter 10) will be met by increasing net
exports by half of the amount of reduced domestic demand for refined product, with the other
half being met by reductions in U.S. refinery output. As discussed in RIA Chapter 8.6.4, there
have been several closures or conversions of refineries in recent years that are attributed to many
factors, including lower fuel demand due to COVID-19 or decisions to pivot away from fossil
fuels. Though the reduced domestic output may lead to future closures or conversions of
individual refineries, we are unable to estimate the future decisions of refineries to keep
operating, shut down or convert away from fossil fuels because they depend on the economics of
individual refineries, economic conditions of parent companies, long-term strategies for each
company, and on the larger macro-economic conditions of both the U.S. and the global refinery
market. Therefore, we are unable to estimate the possible effect this rule will have on
employment in the petroleum refining sector. However, because the petroleum refining industry
is material intensive and not labor intensive, and we estimate that only part of the reduction in
liquid fuel consumption will be met by reduced refinery production in the U.S., we expect that
any employment effect due to reduced petroleum demand will be small. It may also be difficult
to distinguish these effects from other trends, such as increases in petroleum sector labor
productivity that may also lower labor demand. In addition, there is uncertainty about the impact
of reduced domestic demand for petroleum fuels on the petroleum fuel supply chain. For
instance, refineries might export the volumes of gasoline and diesel fuel that would otherwise
have been consumed in light- and medium-duty vehicles, absent this rulemaking. In that scenario
there would be no impact on employment at refineries.
As discussed in Chapter 4.5.2, electrification of the vehicle fleet is likely to affect both the
number and the nature of employment in the auto and parts sectors and related sectors, such as in
4-82
-------
providers of charging infrastructure and utilities supporting grid enhancements, as well as sectors
that maintain charging infrastructure. This can happen through many avenues, including greater
demand for batteries, and therefore increased employment needs, through greater demand for
charging and fueling infrastructure to support more PEVs, leading to more private and public
charging facilities being constructed, or through greater use of existing facilities, which can lead
to increased maintenance needs for those facilities. As discussed in Section VIM.4 of the
preamble, charging infrastructure growth in the U.S. is expected to create jobs in sectors related
to constructing and maintaining these facilities, including electrical installation, maintenance and
repair, charger assembly, general construction, software maintenance and repair, planning and
design, and administration and legal. Though we received comments with concerns that there are
not enough qualified technicians to support the infrastructure needs estimated as a function of
this rule, we expect this to be a gradual increase, with more technicians being trained over time.
In addition, as described in Section VIII.1.4 of the preamble, this will also be supported by the
investments and programs focused on training for EV sector positions, and many other programs
focused on job training for positions related to EV technology, including infrastructure and EV
technicians.
In addition, the type and number of jobs related to vehicle maintenance and repair are
expected to change, though we expect this to happen over a longer time span due to the nature of
fleet turnover. Due to less need for maintenance of BEVs relative to ICE vehicles, demand for
such workers could decrease. Though we expect the sale of new PEVs to increase over the time
span of this rule, both new and used ICE vehicles will persist in the fleet for many years. As
vehicles age, they generally require greater amounts of maintenance, possibly mitigating the
expected reduction in the number of ICE vehicles in the onroad fleet over time. Over this same
time span, though we estimate less maintenance needs for BEVs compared to ICE vehicles, the
total employment related to PEV maintenance is expected to increase due to the increase in
number of PEVs in the onroad fleet. Even if the increase in PEV maintenance-related
employment is smaller than the decrease in ICE vehicle maintenance-related employment over
time, we expect opportunities for workers to retrain to other positions, for example within PEV
maintenance, charging station infrastructure, or elsewhere in the economy. Research funded by
the Department of Energy (DOE) indicates that a wide range of jobs in the ICE vehicle sector
have a relatively high similarity in needed skill sets to jobs in the EV sector, as well as in other
sectors.140 Also, as described in Section VIM.l and VIM.4 of the preamble, DOL, DOE, and
other groups are involved in existing and planned projects focused on training new and existing
employees in green energy jobs, including maintenance and repair.
Effects in the supply chain depend on where goods in the supply chain are developed.
Commenters on the proposed rule argued that developing PEVs in the U.S. is critical for
domestic employment, and for the global competitiveness of the U.S. in the future auto industry:
as other countries are moving rapidly to develop PEVs, the U.S. auto industry risks falling
behind. As discussed in Preamble Section I.A.2.iii and RIA 4.5.2, there have been several
legislative and administrative efforts, and enacted several acts, since 2021 aimed at improving
the domestic supply chain for electric vehicles, including electric vehicle chargers, critical
140 Workforce Analytic Approaches to Find Degrees of Freedom in the EV Transition;
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4699308
4-83
-------
minerals, and components needed by domestic manufacturers of PEV batteries. These actions are
also expected to provide opportunities for domestic employment in these associated sectors.
The standards may affect employment for auto dealers through a change in vehicles sold, with
increasing (decreasing) sales being associated with an increase (decrease) in labor demand.
However, vehicle sales are also affected by macroeconomic effects, and it is difficult to separate
out the effects of the standards on sales from effects due to macroeconomic conditions. In
addition, auto dealers may be affected by changes in maintenance and service costs. Increases in
those costs are likely to increase labor demand in dealerships, and reductions are likely to
decrease labor demand. Auto dealers may also be affected by changes in the maintenance needs
of the vehicles sold. For example, reduced maintenance needs of BEVs could lead to reduced
demand for maintenance labor for dealers that sell BEVs. Another factor that may affect
employment for auto dealers is if there is a change in the share of vehicles being sold under a
direct-to-consumer sales model. As of 2023, at least eight U.S. states ban the sale of direct-to-
consumer sales of vehicles.141 This would not affect many OEMs that currently sell vehicles
through dealerships across the country. However, OEMs who operate under a direct-to-consumer
sales model may affect employment levels at dealerships, either through new agreements with
existing dealers, new dealerships being built to service their vehicle sales, or increasing
alternative methods of sales. This possible effect is very uncertain, however, as it is affected by
multiple factors including state regulations, the existence of OEMs operating under direct-to-
consumer models, as well as the relative share of vehicles being sold under direct-to-consumer
models of operation.
An additional factor to consider for employment impacts across all industries that might be
affected by this rule, or by the increase in the share of PEVs in the market, is that though more
PEVs are being introduced to the market, regardless of this rule, ICE vehicles will persist in the
market for many years. Also, there are multiple pathways to compliance with this rule, and
OEMs are able to choose compliance methods that work for them. Though there may be negative
impacts on workers currently employed in ICE vehicle production, this gradual shift avoids
abrupt changes and will reduce impacts in acceptance, infrastructure availability, employment,
supply chain, and more. In addition, support through recent federal investment allows for
increased investment along the vehicle supply chain, including domestic battery manufacturing,
charging infrastructure, and vehicle manufacturing.
As discussed in Chapter 4.5.2, the BIL, CHIPS Act, and IRA are expected to create incentives
for expanding domestic manufacturing along the electric vehicle supply chain, including battery
manufacturing and infrastructure. This legislation is expected to, in turn, create incentives
opportunities for domestic employment along the full automotive sector and EV battery supply
chains, from components to final assembly.142 Importantly, domestic employment is expected to
be positively impacted due to the domestic assembly, production, and manufacturing conditions
on eligibility for purchase incentives and battery manufacturing incentives in the IRA, with
141 https://blog.onlyusedtesla.com/the-states-where-tesla-still-cant-sell-cars-and-why-it-matters-today-577c0f4e4009
142 More information on how these Acts are expected to aid employment growth and create opportunities for growth
along the supply chain can be found in the January 2023 White House publication "Building a Clean Energy
Economy: A Guidebook to the Inflation Reduction Act's Investments in Clean Energy and Climate Action." found
online at https://www.whitehouse.gov/wp-content/uploads/2022/12/Inflation-Reduction-Act-Guidebook.pdf
4-84
-------
estimates from the BlueGreen Alliance and the Political Economy Research Institute stating that
the IRA could lead to over 9 million jobs over the next decade, with about 400,000 of them
attributed directly to the IRA's battery and fuel cell vehicle provisions. (BlueGreen Alliance
2023)
As a result of these standards, consumers will likely pay higher up-front costs for the vehicles,
but they are expected to recover those costs through reduced fuel, maintenance, and repair costs,
as well as due to the IRA tax incentives for PEV purchase and battery manufacturing leading to
reduced up-front costs for BEVs. As a result, consumers are expected to have additional money
to spend on other goods and services, though the timing of access to that additional money
depends on aspects including whether the consumer borrows money to buy the vehicle. These
increased expenditures could support employment in those sectors where consumers spend their
savings. If the economy is at full employment, any change in consumer expenditures would
primarily represent a shift in employment among sectors. If, on the other hand, the economy has
substantial unemployment, these expenditures would contribute to employment through
increased consumer demand.
4-85
-------
Chapter 4 References
Argonne National Laboratory. 2024. "Light-duty Electric Dive Vehicles Monthly Sales
Updates." Energy Systems and Infrastructure Analysis. Accessed February 21, 2024.
https://www.anl.gov/esia/light-duty-electric-drive-vehicles-monthly-sales-updates.
Arrow, R.G., M.L. Cropper, G.C. Eads, R.W. Hahn, L.B. Lave, R.G. Noll, P.R. Portney, et al.
1996. "Is There a Role for Benefit-Cost Analysis in Environmental, Health, and Safety
Regulation?" Science 272 (5259): 221-222.
Autolist. 2022. How Long Should I Keep My Car. June 24. Accessed February 07, 2023.
https://www.autolist.com/guides/how-long-should-i-keep-my-car.
Bartlett, Jeff. 2022. More Americans Would Buy and Electric Vehicle, and Some Consumers
Would Use Low-Carbon Fules, Survey Shows. Consumer Reports. July 7. Accessed March 2,
2023. https://www.consumerreports.org/hybrids-evs/interest-in-electric-vehicles-and-low-
carbon-fuels-survey-a8457332578/.
Berman, E., and L.T.M. Bui. 2001. "Environmental Regulation and Labor Demand: Evidence
from the South Coast Air Basin." Journal of Public Economics 79 (2): 265-295.
Black Book and Fitch Ratings. 2021. Vehicle Depreciation Report 2021.
Blackley, Julie, n.d. How Long Do People Keep Their Cars? Accessed 02 07, 2023.
https://www.iseecars.com/how-long-people-keep-cars-study.
Bloomberg NEF. 2023. "Electric Vehicle Outlook 2023."
BlueGreen Alliance. 2023. 9 Million jobs from climate action: the ilnflation Reduction Act.
Accessed February 2024. https://www.bluegreenalliance.org/site/9-million-good-jobs-from-
climate-action-the-inflation-reduction-act/.
BlueGreen Alliance. 2021. "Backgrounder: EVs are Coming. Will They be Made in the USA?"
https://www.bluegreenalliance.org/wp-content/uploads/2021/04/Backgrounder-EVs-Are-
Coming.-Will-They-Be-Made-in-the-USA-vFINAL.pdf.
Borlaug, Brennan, Shawn Salisbury, Mindy Gerdes, and Matteo Muratori. 2020. "Levelized Cost
of Charging Electric Vehicles in the United States." Joule 4: 1470-1485.
doi:https://doi.org/10.1016/j.jopule.2020.05.013.
Brown, Abby, Jeff Cappellucci, Emily White, Alexia Heinrich, and Emma Cost. 2023. Electric
Vehicle Charging Infrastructure Trends from the Alternative Fueling Station Locator: Second
Quarter 2023. Golden, CO: National Renewable Energy Laboratory.
https://www.nrel.gov/docs/fy24osti/87033.pdf.
Bui, Anh, Peter Slowik, and Nic Lutsey. 2020. Update on electric vehicle adoption across U.S.
cities. International Council on Clean Transportation, https://theicct.org/wp-
content/uploads/2021/06/EV-cities-update-aug2020.pdf.
Bureau of Labor Statistics. 2023. Real Domestic Employment Requirements. Accessed January
2023. http://www.bls.gov/emp/ep_data_emp_requirements.htm.
4-86
-------
Burlig, Fiona, James Bushnell, David Rapson, and Catherine Wolfram. 2021. "Low Energy:
Estimating Electric Vehicle Electricity Use." AEA Papers and Proceedings 111: 430-435.
Burnham, Andrew, David Gohlke, Luke Rush, Thomas Stephens, Yan Zhou, Mark A Delucchi,
Alicia Birly, et al. 2021. Comprehensive Total Cost of Ownership Quantification for Vehicles
with Different Size Classes and Powertrains. ANL/ESD-21/4, Energy Systems Division, Lemont,
IL: Argonne National Laboratory, https://publications.anl.gov/anlpubs/2021/05/167399.pdf.
2021. "CAFE TSD." Technical Support Document: Proposed Rulemaking for Model Years
2024-2026 Light-Duty Vehicle Corporate Average Fuel Economy Standards. NHTSA,
Department of Transportation, August.
California Air Resources Board. 2022. "Public Hearing to Consider the Proposed Advanced
Clean Cars II Regulations Staff Report: Initial Statement of Reasons."
https://ww2.arb.ca.gov/sites/default/files/barcu/regact/2022/accii/isor.pdf.
Chakraborty, Debapriya, Scott Hardman, and Gil Tal. 2022. "Integrating plug-in electric vehicles
(PEVs) into household fleet- foactors influencing miles traveled by PEV owners in California."
Travel Behavior and Society 26: 67-83.
CHIPS Act of 2022. 2022. H.R. 4346 (117th Cong. Sess. 2).
Climate Nexus. 2022. "Job Impacts From the Shift to Electric Cars and Trucks." Accessed
December 2022. https://climatenexus.org/climate-issues/energy/ev-job-impacts/.
Colato, J, and L Ice. 2023. "Charging into the Future: The Transition to Electric Vehicles."
Beyond the Numbers: Employment & Unemployment (U.S. Bureau of Labor Statistics) 12 (4).
https://www.bls.gov/opub/btn/volume-12/charging-into-the-future-the-transition-to-electric-
vehicles.htm.
Cole, Cassandra, Michael Droste, Christopher Knittel, Shanjun Li, and James H. Stock. 2023.
"Policies for Electrifying the Light-Duty Fleet in the United States." AEA Papers and
Proceedings 113: 316-322. doi:https://doi.org/10.1257/pandp.20231063.
Collins, Jeffrey. 2023. South Carolina offers $1.3B to new Scout electric SUV maker. AP News.
https://apnews.com/article/scout-electric-vehicle-plant-south-carolina-
07c565669el3985738db503a86e323b0.
Consumer Reports. 2022. "Survey Says: Considerable Interest in Electric Vehicles Across
Racial, Ethnic Demographics."
Curtis, M.E. 2018. "Who Loses Under Cap-and-Trade Programs? The Labor Market Effects of
the NOx Budget Trading Program." The Review of Economics and Statistics 100 (1): 151-166.
Davis, Lucas W. 2022. "Electric vehicles in multi-vehicle households." Applied Economics
Letters. doi:https://doi.org/10.1080/13504851.2022.2083563.
Davis, Lucas W. 2019. "How much are electric vehicles driven?" Applied Economics Letters 26
(18): 1497-1502. doi:https://doi.org/10.1080/13504851.2019.1582847.
4-87
-------
Deschenes, O. 2018. "Balancing the Benefits of Environmental Regulations for Everyone and the
Costs to Workers and Firms." IZA World of Labor. Vol. 22v2.
https://wol.iza.org/uploads/articles/458/pdfs/environmental-regulations-and-labor-markets.pdf.
DOT Volpe Center. 2023. "CCEMS parameters input file." parametersref. xlsx, Economic
Values worksheet. August 17.
Dr. Ernst, Stefan. 2023. Volkswagen-backed PowerCo SE reaches significant milestone in St.
Thomas gigafactory project. Press Release, Volkswagen Group. Accessed February 2024.
https://www.volkswagen-group.com/en/press-releases/volkswagen-backed-powerco-se-reaches-
significant-milestone-in-st-thomas-gigafactory-project-17962.
Edmunds. 2023. Cost of Car Ownership. Accessed January 6, 2023. Edmunds.com/tco.html.
Ehrenberg, R.G., and R.S. Smith. 2000. Modern Labor Economics: Theory and Public Policy.
Addison Wesley Longman, Inc.
EnergyTech. 2023. DOE Announces $275M for 7 Projects to Strengthen Clean Energy Supply
Chains adn Manufacturing in Former Coal Communities. EnergyTech.
https://www.energytech.com/energy-efficiency/article/21278185/doe-announces-275m-for-7-
projects-to-strengthen-clean-energy-supply-chains-and-manufacturing-in-former-coal-
communities.
ERM. 2022. "Electric Vehicle Market Update: Manufacturer Commitments and Public Policy
Initiatives Supporting Electric Mobility in the U.S. and Worldwide. Contract Report to the
Environmental Defense Fund." Accessed 12 16, 2022.
https://blogs.edf.org/climate411/files/2022/04/electric_vehicle_market_report_v6_april2022.pdf.
Fischer, Carolyn. 2005. "On the Importance of the Supply Side in Demand-Side Management."
Energy Economics 17: 165-180.
Forsythe, C., K. Gillingham, J. J. Michalek, and K. Whitefoot. 2023a. "Will pickup truck owners
go electric?" Carnegie Mellon University and Yale University.
https://www.cmu.edu/cit/veg/publications.html.
Forsythe, Connor R, Kenneth T Gillingham, Jeremy J Michalek, and Kate S Whitefoot. 2023b.
"Technology Advancement is Driving Electric Vehicle Adoption." Sustainability Schience
(PNAS). Accessed February 2024. https://www.pnas.org/doi/epdf/10.1073/pnas.2219396120.
Forsythe, Connor R., Kenneth T. Gillingham, Jeremy J. Michalek, and Kate S. Whitefoot. 2023b.
"Technology advancement is driving electric vehicle adoption." PNAS 120 (23).
doi:https://doi.org/10.1073/pnas.2219396120.
Fujita, K Sydny, Hung-Chia Yang, Margaret Taylor, and Dana Jackman. 2022. "The Green Light
on Buying a Car: How Consumer Decision Making Interacts with Environmental Attributes in
the New Vehicle Purchase Process." Transportation Research Record 2676 (7): 743-762.
Fujita, K. Sydny, Nica Campbell, and Margaret Taylor. 2024. EVs for everyone? Identifying the
likely early majority of electric vehicle adopters. Lawrence Berkeley National Laboratory,
Energy Technologies Area.
4-88
-------
Ge, Yanbo, Christina Simeone, Andrew Duvall, and Andrew Wood. 2021. There's No Place Like
Home: Residential Parking, Electrical Access, and Implications for the Future of Electric Vehicle
Charging Infrastructure. NREL/TP-5400-81065, Golden, CO: National Renewable Energy
Laboratory, https://www.nrel.gov/docs/fy22osti/81065.pdf.
Gillingham, K, A van Benthem, S Weber, D Saafi, and X He. 2023. "Has Consumer Acceptance
of Electric Vehicles Been Increasing? Evidence from Microdata on Every New Vehicle Sale in
the United States." American Economics Association: Papers & Proceedings, forthcoming.
https://resources.environment.yale.edu/gillingham/GBWSH_ConsumerAcceptanceEVs.pdf.
Greschak, Tressa, Matilda Kreider, and Nathan Legault. 2022. "Consumer Adoption of Electric
Vehicles: An Evaluation of Local Programs in the United States." School for Environment and
Sustainability, University of Michigan, Ann Arbor, MI.
https://deepblue.lib.umich.edu/handle/2027.42/172221.
Hafstead, M.A.C., and R.C. Williams III. 2018. "Unemployment and Environmental Regulation
in General Equilibrium." Journal of Public Economics 160: 50-65.
Hardman, Scott, and Gil Tal. 2021. "Understanding Discontinuance among California's Electric
Vehicle Owners." Nature Energy 6 (5): 538-545. doi:https://doi.org/10.1038/s41560-021-00814-
9.
Hardman, Scott, Kelly L Fleming, Eesha Khare, and Mahmoud M Ramadan. 2021. "A
perspective on equity in the transition to electric vehicles." MIT Science Policy Review 2: 46-54.
doi: 10.3 8105/spr. e 1 Ordoaoup.
Helfand, G, and R Dorsey-Palmateer. 2015. "The Energy Efficiency Gap in EPA's Benefit-Cost
Analysis of Vehicle Greenhouse Gas Regulations: A Case Study." Journal of Benefit-Cost
Analysis 6 (2): 432-454.
Herrmenn, F., W. Beinhauer, D. Borrmann, M. Hertwig, J. Mack, T. Potinecke, C-P. Praeg, and
P. Rally. 2020. Employment 2030: Effect of Electric Mobility and Digitalization on the Quality
and Quantity of Employment at Volkswagen. Fraunhofer Institute for Industrial Engineering
IAO. https://www.volkswagenag.com/presence/stories/2020/12/frauenhofer-
studie/6095_EMD I_VW_Summary_um.pdf.
Huang, H, G Helfand, and K Bolon. 2018a. "Consumer Satisfaction with New Vehicles Subject
to Greenhouse Gas and Fuel Economy Standards." Society for Benefit-Cost Analysis Annual
Conference.
Huang, H, G Helfand, K Bolon, R Beach, M Sha, and A Smith. 2018. "Re-searching for hidden
costs: Evidence from the adoption of fuel-saving technologies in light-duty vehicles."
Transportation Research Part D 65: 194-212.
Hutchens, Andrew, Alecia Cassidy, Geoffrey Burmeister, and Gloria Helfand. 2021. Impacts of
Light-Duty Greenhouse Gas Emission Standards on Vehicle Affordability. U.S. Environmental
Protection Agency.
IEA. 2023. "Global EV Outlook 2023: Catching up with climate ambitions." International
Energy Agency.
Inflation Reduction Act of 2022. 2022. H.R. 5376 (117th Cong., 2nd sess.).
4-89
-------
Infrastructure Investment and Jobs Act. 2021. H.R. 3684 (117th Cong., 1st sess.).
Jackman, D K, K S Fujita, H C Yang, and M Taylor. 2023. Literature Review of U.S. Consumer
Acceptance of New Personally Owned Light Duty Plug-in Electric Vehicles. Washington, D.C.:
U.S. Environmental Protection Agency.
JGCRI. n.d. "GCAM (Global Change Analysis Model) v7 Documentation: Economic Choice in
GCAM." Joint Global Change Research Institute, https://jgcri.github.io/gcam-doc/choice.html.
Krisher, T., and J. Seewer. 2021. Autoworkers face uncertain future in an era of electric cars.
ABC News. February 11. Accessed March 1, 2023. https://www.abc27.com/news/us-
world/business/autoworkers-face-dimmer-future-in-a-new-era-of-electric-cars/.
Kupper, D., K. K. Kuhlmann, A. Tominaga, and J. Schlageter Arora. 2020. Shifting Gears in
Auto Manufacturing. September 28. Accessed October 20, 2021.
https://www.bcg.com/publications/2020/transformative-impact-of-electric-vehicles-on-auto-
manufacturing.
Layard, R., and A. Walters. 1978. "Microeconomic Theory." Chap. 9. London: McGraw-Hill.
Leard, B., V. McConnell, and Y. C. Zhou. 2017. "The Effect of Fuel Price Changes on Fleet
Demand for New Vehicle Fuel Economy." Working Paper (Resources For the Future).
https://media.rff.org/documents/RFF20WP2017-25.pdf.
Lee, Hyun Lee, Minyoung Cho, Gil Tal, and Scott Hardman. 2023. "Do plug-in hybrid adopters
switch to battery electric vehicles (and vice versa)?" Transportation Research Part D 119
(103752).
Martin, Allison. 2023. How long should your car loan be? Edited by Helen Wilbers. December 4.
Accessed December 15, 2023. https://www.bankrate.com/loans/auto-loans/how-long-should-
y our-car-loan-b e/.
Morgenstern, R.D., W.A. Pizer, and J.-S. Shih. 2002. "Jobs Versus the Environment:" An
Industry-Level Perspective." Journal of Environmental Econometrics and Management 43: 412-
436.
Muller, D. 2017. "Warranties Defined: The Truth behind the Promises." Car and Driver, May 29.
National Academies of Science, Engineering, and Medicine. 2021. Assessment of Technologies
for Improving Light-Duty Vehicle Fuel Economy 2025-2035. Washington, DC: The National
Academies Press. doi:https://doi.org/10.17226/26092.
Nehiba, Cody. 2024. "Electric vehicle usage, pollution damages, and the electricity price
elasticity of driving." Journal of Environmental Economics and Management 124 (102915).
doi:https://www.sciencedirect.com/science/article/pii/S009506962300133X.
NHTSA. 2022. Technical Support Document: Final Rulemaking for Model Years 2024-2026
LightDuty Vehicle Corporate Average Fuel Economy Standards. U.S. Department Of
Transportation, National Highway Traffic Safety Administration. Accessed March 7, 2023.
https://www.nhtsa.gov/sites/nhtsa.gov/files/2022-04/Final-TSD_CAFE-MY-2024-2026.pdf.
4-90
-------
Political Economy Research Institute. 2022. Job Creation Estimates Through Proposed Inflation
Reduction Act. University of Massachusetts Amherst, https://www.bluegreenalliance.org/site/9-
million-good-jobs-from-climate-action-the-inflation-reduction-act/.
Raghavan, Seshadri Srinivasa, and Gil Tal. 2021. "Behavioral and technology implications of
electromobility on household travel emissions." Transportation Research Part D 94 (102792).
doi:https://doi.org/10.1016/j.trd.2021.102792.
Rogers, E.M. 2003. Diffusion of Innovations. New York: Free Press.
Seattle Jobs Initiative. 2020. "Amping Up Electric Vehicle Manufacturing in the PNW:
Opportunities for Business, Workforce, and Education."
https://www.seattle.gov/Documents/Departments/OSE/ClimateDocs/TE/EV%20Field%20in%20
OR%20and%20W A_February20.pdf.
Slowik, Peter, Aaron Isenstadt, Logan Pierce, and Stephanie Searle. 2022. Assessment of Light-
Duty Electric Vehicle Costs and Consumer Benefits in the United States in the 2022-2035 Time
Frame. International Council on Clean Transportation. Accessed 12 16, 2022.
https://theicct.org/publication/ev-cost-benefits-2035-oct22/.
Slowik, Peter, Stephanie Searle, Hussein Basma, Josh Miller, Yuanrong Zhou, Felipe Rodriguez,
Claire Buysse, et al. 2023. "Analyzing the Impact of the Inflation Reduction Act on Electric
Vehicle Uptake in the United States." International Council on Clean Transportation and Energy
Innovation Policy & Technology LLC.
Smith, V.K. 2015. "Should Benefit-Cost Methods Take Account of High Unemployment?"
Review of Environmental Economics and Policy 9 (2): 165-178.
Taylor, Margaret. 2023. Draft of "Modeling the Future of Electric Vehicles to Support U.S.
Public Policy: A Comparison of Leading Approaches". Lawrence Berkeley National Laboratory.
Taylor, Margaret, and K Sydny Fujita. 2018. Consumer Behavior and the Plug-in Electric
Vehicle Purchase Decision Process: A Research Synthesis. Berkeley, CA: Lawrence Berkeley
National Laboratory.
Taylor, Margaret, K. Sydny Fujita, and Nica Campbell. 2024 . The false dichotomies of plug-in
electric vehicle markets. Lawrence Berkeley National Laboratory.
U.S. Department of Energy (DOE). 2020. Alternative fuels data center: develoing infrastructure
to charge plug-in electric vehicles. Accessed 2020.
http s ://afdc. energy. gov/fuel s/el ectri city_infrastructure. html.
U.S. Department of Energy. 2022. Electric Vehicle Charging Station Locations. Accessed March
8,2022. http s ://afdc. energy. gov/fuel s/el ectri city l ocati ons. html.
U.S. Department of Energy Office of Manufacturing and Energy. 2024. "Supply Chains Inflation
Reduction Act Domestic Manufacturing Conversion Grants Funding Opportunity Announcement
DE-FOA-0003106_FOA_Doc_Amendment_000006_IRA 50143." https://infrastructure-
exchange. energy. gov/Default.aspx#FoaIdf9eblc8a-9922-46b6-993e-78972d823cb2.
4-91
-------
U.S. Department of Energy, Office of Policy. 2023. "Investing in American Energy: Significant
Impacts of the Inflation Reduction Act and Bipartisan Infrastructure Law on the U.S. Energy
Economy and Emissions Reductions."
U.S. DOE. 2023. Investing in American Energy: Significant Impacts of the Inflation Reduction
Act and Bipartisan Infrastructure Law on the U.S. Energy Economy and Emissions Reductions.
U.S. Department of Energy, Office of Policy.
U.S. DOT. 2021. "86 FR 49602." Corporate Average Fuel Economy Standards for Model Years
2024-2026 Passenger Cars and Light Trucks. September 3.
—. 2023. "88 FR 56128." Corporate Average Fuel Economy Standards for Passenger Cars and
Light Trucks for Model Years 2027-2032 and Fuel Efficiency Standards for Heavy-Duty Pickup
Trucks and Vans for Model Years 2030-2035. August 17.
U.S. DOT, FHWA. 2023. "The National Electric Vehicle Infrastructure (NEVI) Formula
Program Guidance (Update)." June 2.
https://www.fhwa.dot.gov/environment/alternative_fuel_corridors/nominations/90d_nevi_formul
a_program_gui dance. pdf.
U.S. EPA. 2023. "2023 EPA Automotive Trends Report."
https://www.epa.gov/system/files/documents/2023-12/420r23033.pdf.
—. 2021. "86 FR 74434." Revised 2023 and Later Model Year Light-Duty Vehicle Greenhouse
Gas Emissions Standards. December 30.
U.S. EPA. 2021. The Effects of New-Vehicle Price Changes on New- and Used-Vehicle Markets
and Scrappage. EPA-420-R-21-019.
https://cfpub. epa.gov/si/si_public_record_Report.cfm?dirEntryId=352754&Lab=OTAQ.
U.S. OMB. 2015. "2015 Report to Congress on the Benefits and Costs of Federal Regulations
and Agency Compliance with the Unfunded Mandates Reform Act."
https://obamawhitehouse.archives.gov/sites/default/files/omb/inforeg/2015_cb/2015-cost-
benefit-repot.pdf/.
UAW. 2020. "Taking the High Road: Strategies for a Fair EV Future." https://uaw.org/wp-
content/uploads/2019/07/190416-EV-White-Paper-REVISED-January-2020-Final.pdf.
Wakefield, Chantel. 2023. Electric Car Rebates and Incentives: What to Know by State.
February 02. Accessed February 02, 2023. https://www.kbb.com/car-advice/electric-vehicle-
rebates-by-state/.
Watten, A, GHelfand, and S Anderson. 2021. "Attribute Production and Technocal Change:
Rethinking the Performance adn Fuel Economy Trade-off for Light-duty Vehicles." Working
Paper.
White, Vinn. September 27, 2016. "Revised Departmental Guidance on Valuation of Travel
Time in Economic Analysis."
Wood, Eric, Brennan Borlaug, Matt Moniot, D-Y Lee, Yanbo Ge, Fan Yang, and Zhaocai Liu.
2023. "The 2030 National Charging Network: Estimating U.S. Light-Duty Demand for Electric
4-92
-------
Vehicle Charging Infrastructure." National Renewable Energy Laboratory. Accessed December
18, 2023. https://www.nrel.gov/docs/fy23osti/85654.pdf.
Zabritski, Melinda. 2023. State of the Automotive Finance Market - Q3 2023. November.
Accessed December 15, 2023. https://www.experian.com/automotive/automotive-credit-
webinar? cmpid=Auto-US-24-122+Q3 +2023 + Auto+Credit+T rends.
Zhao, Lujin, Elizabeth R. Ottinger, Arthur Hong Chun Yip, and John Paul Helveston. 2023.
"Quantifying electric vehicle mileage in the United States." Joule 7: 2537-2551.
4-93
-------
Chapter 5: Electric Power Sector and Infrastructure Impacts
As plug-in electric vehicles143 (PEVs) are projected to represent a significant share of the
future U.S. light- and medium-duty vehicle fleet, EPA has developed new approaches to estimate
the power sector144 emissions of increased PEV charging. EPA combined the use of three
analytical tools to incorporate power-sector-related emissions from PEV charging demand within
the light- and medium-duty vehicle emissions inventory analysis for the final rule:
• OMEGA manufacturer compliance model
• A suite of electric vehicle infrastructure modeling tools (EVI-X) developed by the
National Renewable Energy Laboratory (NREL)
• The Integrated Planning Model (IPM)
EPA's manufacturer compliance model, OMEGA, is described in detail in Chapter 2. Chapter
5.1 below provides a summary of EVI-X and how these tools were used together with OMEGA
to estimate charge demand inputs for IPM. The power sector modeling and results are described
in 5.2. The power sector emissions results were incorporated into the emissions inventory and
cost-benefit analyses described in Chapters 6 through 9. The related retail price modeling results
were also incorporated into the analysis of costs and benefits in Chapter 9.
Chapter 5.3 describes our assessment of PEV charging infrastructure. Finally, the potential
impacts on pending changes to the power sector on grid resiliency are discussed in Chapter 5.4.
5.1 Modeling PEV Charge Demand and Regional Distribution
Under an Interagency Agreement between EPA and the U.S. Department of Energy, NREL
has continued its development of a suite of electric vehicle infrastructure modeling tools (EVI-X)
and methods for simulating PEV charging infrastructure requirements and associated electricity
loads from best available data (U.S. EPA 2022b). EVI-X tools have informed multiple national,
state, and local PEV charging infrastructure planning studies (E. Wood, C. Rames, et al. 2017)
(E. Wood, C. Rames, et al. 2018) (Alexander, et al. 2021), including a national vehicle charging
infrastructure assessment through 2030 (E. Wood, B. Borlaug, et al. 2023). Within the emissions
inventory analysis for the final rule, EVI-X models are used to translate scenario-specific
forecasts of national light-duty vehicle stock and annual energy consumption from the OMEGA
model into spatially disaggregated hourly load profiles required for subsequent power sector
modeling using the Integrated Planning Model (IPM) (see Chapter 5.2). The primary components
of the process flow from OMEGA outputs to IPM inputs is shown in Figure 5-1. IPM outputs
also flow back into inventory analyses in OMEGA as PEV emissions factors (see RIA Chapter
9).
143 Plug-in electric vehicles is defined here as both battery electric vehicles and plug-in hybrid electric vehicles
combined.
144 Power sector is defined here to include electricity generation, transmission, and the distribution system, which
typically ends at a service drop at a customer's premises.
5-1
-------
Grid impacts (emissions, energy, Spatiotemporal PEV charging load
costs, build/retirement, etc.) simulation
Figure 5-1: Modeling process flow highlighting the primary components for translating
OMEGA's national PEV stock projections and PEV attributes into hourly load profiles.
OMEGA Compliance Model
National-level PEV stock
projections, vehicle attributes, VMT,
electrical energy consumption,
mobile source inventory, program
costs
Integrated Planning Model
Hour(E5T)
12 24
Hour(EST)
PEV Likely Adopter Model
regional level
EVI-X National LDV Framework
5.1.1 PEV Disaggregation and Charging Simulation
As described in further detail in Chapter 2 of the RIA, the OMEGA model evaluates the cost
of compliance for meeting the standards and options analyzed within the final rule. Each
OMEGA run produces scenario-specific projections of national vehicle sales, stock, energy
consumption, and tailpipe emissions. For PEVs, however, tailpipe emissions are zero in the case
of battery electri c vehicles (BEVs) and are zero during the charge-depleting operation of plug-in
hybrid electric vehicles (PHEVs) with resulting emissions occurring upstream at the electricity
generation source. This expands the requisite analytical boundaries of the system with respect to
determination of emissions inventory impacts. To produce estimates of the spatiotemporal
charging loads needed for power sector emissions modeling, the national PEV stock from
OMEGA must first be disaggregated regionally.
The framework developed for PEV disaggregation leverages a likely adopter model (LAM)
adapted by NREL (Ge, et al. 2021) to rank vehicles in the private light-duty fleet for their
likelihood to be replaced by a PEV based on publicly available demographic data, including
housing type, income, tenure (rent or own), state policies (ZEV states), and population density.
The model is trained on the revealed preferences of 3,772 survey respondents (228 PEV owners)
across the United States as described in (Ge, et al. 2021). Vehicle registration data from June
2022 (Experian Automotive 2022) are used to develop a set of chassis-specific LAMs for
disaggregating PEV sedans, S/CUVs, pickups, and vans based on current regional vehicle type
preferences. This process is outlined in Figure 5-2.
5-2
-------
EPA OMEGA
(national PEVstock)
Simplified vehicle NRELPEVchassis-
representation specific LAMs
to
{-
National annual PB/stock
projections & PB/attributes
¦g—g*
''sedans
¦owo*
"pickups
PEVsedan LAM
PEV van LAM
PEV stock assigned to
IPM regions
¦ * PEVS/CUVLAM
PEVpickup LAM
demonstrative
¦
Assigns PEVs to households based Geographic distribution of OMEGA
on ranked adoption likelihood PEV stock projections
Figure 5-2: Procedure for disaggregating OMEGA national PEV stock projections to IPM
regions.
Vehicles modeled within OMEGA are first assigned to a simplified chassis type (i.e., sedan,
S/CUV, pickup, van). Next, the total number of vehicles for each chassis type are input into to
each of the four chassis-specific LAMs to disaggregate PEVs into IPM regions based on regional
vehicle type preferences and the likelihood of PEV adoption.
The OMEGA model generates vehicle adoption projections for thousands of unique PEV
models over time. Conducting detailed charging simulations for each of these models would be
computationally prohibitive and produce results that do not meaningfully differentiate from those
generated by a reduced set of representative PEV models. Thus, a clustering approach was used
to generate these representative PEV models for simulation from the complete set of OMEGA
vehicles. K-means clustering was performed over each PEV's respective battery capacity (kWh)
and energy consumption rate (kWh/mi.) parameters as specified by OMEGA. A silhouette
analysis was used to determine the appropriate number of clusters (k=6 for BEVs, k=2 for
PHEVs) and OMEGA vehicles are assigned to clusters that minimize the Euclidean distance to
the centroids of the two normalized (Z-score) parameters. These assignments are retained and
used to map OMEGA vehicles to the most similar synthetic representative PEV model. The
cluster centroids are used to produce the battery capacity and energy consumption rate
parameters for the eight representative PEVs required for subsequent charging simulations. An
additional parameter, the max DC charge acceptance, is defined as a PEV's maximum effective
charging rate over a typical 20 percent to 80 percent SOC DC fast charging (DCFC) window.
This was required to simulate DCFC for BEVs and was not directly specified by the OMEGA
model. PHEVs are assumed to be incapable of DCFC. For modeling light-duty BEV DCFC, a
simple heuristic was applied such that pre-2030 model years (Gen 1 batteries) would be capable
of 1.5C charging on average while model year 2030 and later BEVs would be capable of
charging at 3C (Gen 2 batteries).145 The key parameters for simulating charging for each of the
representative PEVs are shown in Table 5-1.
145 C-rate (or Cr) is a measure of the rate at which a battery is charged/discharged relative to its maximum energy
storage capacity. It is related to charge/discharge current in amperes (I) and maximum energy storage capacity in
amp-hours (E) by the equation I = Cr • E.
5-3
-------
Table 5-1: Representative PEV examples for charging simulations.
Sim vehicle
l'owei'ti'iiin +
ICncrgy cons, rate
Mux ACaccept.
Mux l)(nccept.
ICV Kiingre
| kWli/mi. |
|kW|
|kW|
|mi.|
((Jen 1 / (Jen 2)
((Jen 1 /
(Jen 2)
BEV1
BEV 300
0.27
9/12
134/
267
BEV2
BEV 300
0.31
9 /12 "
; 154/
308
BEV3
BEV 300
0.34
9/12
171 /
342
BEV4
| BEV 300 !
0.38
9/12
191 /
383
BEV5
BEV 300
0.42
9/12
212/
424
BEV6
BEV 300
0.47
9/12
236/
471
PHEV1
PHEV 50
0.29
9/12
-
PHEV2
PHEV 50
0.38
9/12
-
MD BEV1
BEV 150
0.54
12
300
MD BEV2
BEV 300
0.62
12
300
MDPHEV
PHEV 75 !
0.7
12
Light-duty PEV modeling in this report builds on the foundation of years of research and
collaboration at NREL and beyond, most notably the recently published 2030 National Charging
Network report (E. Wood, B. Borlaug, et al. 2023). A brief explanation of this modeling
approach is provided here; readers are directed to this previous work for more detailed
explanations of the modeling approach and assumptions.
The core tools used for modeling LDV charging demands in this study are:
• EVI-Pro: For typical daily charging needs
• EVI-RoadTrip: For fast charging along highways supporting long-distance travel
• EVI-OnDemand: For electrification of transportation network companies.
Each individual LDV model is integrated into a shared simulation pipeline (Figure 5-3).
Models are provided with a self-consistent set of exogenous inputs that prescribe the size,
composition, and geographic distribution of the national PEV fleet; technology attributes of
vehicles and charging infrastructure; assumed levels of residential/overnight charging access;
and regional environmental conditions. Each model uses these inputs in bottom-up simulations
of charging behavior by superimposing the use of a PEV over travel data from internal
combustion engine vehicles. By relying on historical travel data from conventional vehicles,
these models implicitly design infrastructure networks capable of making PEVs a one-to-one
replacement for internal combustion engine vehicles, effectively minimizing impacts to existing
driving behavior and identifying the most convenient network of charging infrastructure capable
of meeting driver needs. (E. Wood, B. Borlaug, et al. 2023).
5-4
-------
Models
Inputs
PEV Fleet
Evolution
PEV
Adoption
Weather
Conditions
Residential
EV5E Access
EVI-Pro
¦ Travel behaviors (2017 NHTS)
• Charging preferences
p EVSE availability by location
• PEV models
EVI-RoadTrip
• Long-distance travel (FHWA TAF)
¦ Land use data
¦ PEV models
EVI-OnDemand
¦ Urban TNCVMT
¦ TNC PEV models
• TNC shift behaviors
• TNC driver demographics
Intermediate
Outputs
Daily typical
PEV charging
demand
Daily
long-distance
PEV charging
demand
Daily ride-hail
PEV charging
demand
Inputs
EVSE
Utilization
Combined PEV
charging demand
EVSE port counts
by region,
locations,
and type
Final Outputs
Composite Hourly Demand
4 EVIAn Demand
EVIRoatfTrto
.1
Figure 5-3: EVI-X National light-duty vehicle framework simulation showing
spatiotemporal EV electricity demands for three separate use cases: typical daily travel
(EVI-Pro), long-distance travel (EVI-RoadTrip), and ride-hailing (EVI-OnDemand).
Adapted from Wood et al. (2023) with permission.
The independent (but coordinated) simulations produce a set of intermediate outputs
estimating daily charging demands for typical PEV use, long-distance travel, and ride-hailing
electrification. These intermediate outputs are indexed in time (hourly over a representative 24-
hour period) and space (core-based statistical area or county level) such that they can be
aggregated into a composite set of charging demands across multiple use cases. Once combined,
the peak hour for every combination of charging type (e.g., Level 1 [LI], Level 2 [L2], direct
current [DC]), location type (e.g., home, work, retail), and geography (e.g., core-based statistical
area) is identified for the purpose of network sizing. Rather than sizing the simulated charging
network to precisely meet the peak hourly demand in all situations, the simulation pipeline uses
an assumed network-wide utilization rate in the peak hour to "oversize" the network by a margin
that accounts for the fact that charging demands tend to vary seasonally and around holidays.
The simulation of MDVs (Class 2b-3, gross vehicle weight rating [GVWR] 8,500-14,000
lbs.) leverages the EVI-X LDV pipeline with some key updates, namely:
• MDVs are disaggregated from the national level to counties in a manner proportional
to existing registrations, as observed through data licensed from Experian. This
contrasts the LDV approach, which relies on a set chassis-specific LAMs to assign
PEVs to households with characteristics shown to correlate with PEV adoption.
• MDV travel patterns are derived from two sources based on chassis type: (1) Vans are
simulated based on data from NREL's FleetDNA database, and (2) pickups are
simulated based on data licensed from Wejo. This contrasts the LDV approach, which
relies on the 2017 National Household Travel Survey (NHTS).
• Because MDVs are owned by a variety of businesses, both in terms of company size
and business type, and are often used for both personal and commercial use, medium-
5-5
-------
duty PEVs in this study are assumed to be domiciled during off-shift periods at either
a commercial property (e.g., a depot) or a private residential property (e.g., a single-
family home). This study assumes that 75% of medium-duty PEVs are domiciled at
depots and 25% at single-family homes. Further research into the domicile locations of
MDVs is warranted because data on this topic are scarce, especially at the national
level.
Following the PEV charging simulations, load profiles were aggregated from counties to IPM
regions and converted from local time to Eastern Standard Time (EST) for IPM implementation.
A final corrective step was taken to ensure that the annual energy consumption estimates
supplied by OMEGA were reflected in the PEV load profiles. For a given OMEGA national
PEV stock projection file, the modeling framework produces a typical weekday and weekend 24-
hour (EST) load profile for all IPM regions and analysis years (2026, 2028, 2030, 2032, 2035,
2040, 2045, 2050, 2055).146 Load profiles were analyzed using output from two analytical cases:
• A no-action case that included modeling of electric vehicle provisions from the IRA
within the OMEGA compliance model and compliance with 2023 and later GHG
standards (86 FR 74434 2021) with the addition of heavy-duty vehicle (Class 4-8)
charge demand estimated for the California Advanced Clean Trucks (ACT) Program.
• A final rule policy case based upon Alternative 3 from the proposed rule with the
addition of heavy-duty vehicle charge demand based on an interim scenario developed
from the Greenhouse Gas Emissions Standards for Heavy-Duty Vehicles - Phase 3
Proposed Rule (HDP3).
Alternative 3 was one of the compliance scenarios modeled using OMEGA, EVI-X, and IPM
during the summer and autumn of 2023. Of the scenarios modeled in IPM after the proposal,
Alternative 3 is the closest scenario with respect to PEV charging demand to the final rule and
represents the final rule within the power sector analysis. Alternative 3 differs from the finalized
program by forecasting slightly higher PEV sales in 2027-2031 than finalized, and thus higher
PEV charging demand in earlier years and comparable PEV charging demand after 2032. Thus,
power sector impacts on emissions and cost within the final rule analysis should be considered
conservatively high estimates. Regionalized heavy-duty vehicle charge demand for both the no-
action and policy cases were based upon a combination of NREL EVI-X and LBNL HEVI-
LOAD simulations developed as part of the Multi-State Transportation Electrification Impact
Study (TEIS) (E. Wood, B. Borlaug, et al. 2024).
These analytical cases are described in more detail below. Figure 5-4 provides an example of
how specific load profiles may be used to infer annual PEV charging demands for 2030 and 2050
using the final rule policy case. The purple shading in Figure 5-4 represents the relative light-
and medium-duty vehicle charging demand in each of the 67 IPM regions. In addition to the total
hourly energy demands for PEV charging, energy demands were also broken out by the
following charging types - home Level 1 (LI), home Level 2 (L2), depot L2 (applicable to
146 Output from OMEGA and EVI-X was also generated for Hawaii, Alaska, and Puerto Rico, however the IPM
analysis only included IPM regions for the contiguous United States along with transmission dispatched across the
U. S.-Canada border.
5-6
-------
medium-duty PEVs), work L2, public L2, and public DCFC (Figure 5-5). See Chapter 5.3.1.2.
for additional discussion. Note that these have been converted to EST and reflect an unmanaged
charging scenario where drivers do not prioritize charging at certain times of the day (i.e.,
charging starts as soon as possible when vehicles are plugged in without consideration of
electricity price or other factors) in the no action case and reflects a degree of managed charging
within the analysis for the final rule based on load shifting of PEV charging that was developed
as part of the TEIS.
a) Annual light- & medium-duty PEV charging demands: 2030
b) Annual light- & medium-duty PEV charging demands: 2050
Figure 5-4: Annual light- and medium-duty vehicle PEV charging loads (2030 and 2050 are
shown) for each IPM region in the contiguous United States based on OMEGA charge
demand for the final rule in 2030 (a) and 2050 (b).
In Figure 5-5, there are clear differences in the magnitude, shape, and charger types between
the West Texas (left-ERC_WEST, containing mostly rural areas and small cities such as
Midland and Odessa) and East Texas (right-ERC REST, including multiple major population
5-7
-------
centers such as Houston, San Antonio, Austin, and Dallas-Ft. Worth) regions. The EVI-X
modeling framework conducts charging simulations that incorporate regional differences in EV
adoption, vehicle type preferences, home ownership, weather conditions, and travel patterns.
These demonstrative results reflect how in ERCWEST, EV adoption is projected to be low (due
to limited population and revealed vehicle preferences) leading to a reduced demand for home-
based charging while public DCFC demands for long-distance travel within the region (e.g., road
trips) are amplified. This leads to a disproportionate share of public DCFC charging demand
along highway corridors within the ERC WEST region. Alternatively, simulated charging
demands in the ERCREST are dominated by home and workplace charging due to the higher
EV adoption and urban travel patterns more common to the region.
The OMEGA national PEV outputs and the resulting regionalized light- and medium-duty
IPM inputs from EVI-X for both analyzed cases, for each IPM region and all analytical years
(2026, 2028, 2030, 2032, 2035, 2040, 2045, 2050, 2055) are summarized within a separate PEV
Regionalized Charge Demand Report (McDonald, PEV Regionalized Charge Demand for the
FRM - Memo to the Docket 2023).
5-8
-------
ERC WEST: MW Demand
ERC REST: MW Demand
Wkday
Wknd
2 0 5 5 437 -
wJSi
14,902
7,451
0
14,902
Wkday
Wknd
7,451 -
14,902
0
14,902
7,451
0
14,902
7,451 -
0
14,902
7,451
0
14,902
7.451 -
14,902
7,451
0
14,902
-
-
'
-
-
-
-
k.
'
'
'
"
• !
:
Public DCFC
Public L2
Private Work L2
Private Depot/Home L2
Private Home L2
Private Home LI
0 12 24 0
Hour(EST)
12 24
Hour(EST)
0 12
Hour(EST)
Hour(EST)
Figure 5-5: Yearly hourly (in EST) weekday and weekend load profiles for two IPM
regions (ERCWEST, west Texas; and ERC REST, east Texas) broken out by charger
type for an example OMEGA analytical scenario.
5-9
-------
5.2 Electric Power Sector Modeling
The analyses for the final rule used EPA's Power Sector Modeling Platform v6, which utilizes
the Integrated Planning Model (IPM). IPM is a multi-regional, dynamic, deterministic linear
programming model of the U.S. electric power sector. It provides projections of least-cost
capacity expansion, electricity dispatch, and emission control strategies for meeting energy
demand and environmental, transmission, dispatch, and reliability constraints. IPM can be used
to evaluate the cost and emissions impacts of policies to limit emissions of sulfur dioxide (SO2),
fine particulate matter (PM2.5), nitrogen oxides (NOx), carbon dioxide (CO2), hydrogen chloride
(HC1), and mercury (Hg) from the electric power sector. Post-processing IPM outputs allows for
the processing of other emissions, such as volatile organic compounds (VOC) and non-CC>2
GHGs. The power-sector modeling used for the final rule included power-sector-related
provisions of both the Bipartisan Infrastructure Law (BIL) and the Inflation Reduction Act
(IRA). Additional information regarding power-sector modeling is available via documentation
for the Post-IRA 2022 Reference Case - EPA's Power Sector Modeling Platform v6 Using IPM
(U.S. EPA 2023a). Changes made for the analysis for the final rule include updating power
sector demand using AEO2023 projections through 2055 in place of AEO2021 and substituting
electric vehicle demand from AEO with updated output from the OMEGA and EVI-X analyses
used for the final rule (see Chapter 5.1). Note that the default PEV charge demand from AEO
was replaced with charge demand from OMEGA/EVI-X for light-and medium duty (see Chapter
5.1.1) and charge demand from the TEIS (E. Wood, B. Borlaug, et al., Multi-State
Transportation Electrification Impact Study 2024) for heavy-duty charge demand for all IPM
analyses, including both no-action and policy analyses. The charge demand used for both the no
action and policy analyses for light- and medium-duty vehicles and also an estimate of charge
demand from the Heavy-duty Phase 3 GHG program have been separately docketed as part of
this rulemaking (McDonald 2024)
5.2.1 Estimating Retail Electricity Prices
The Retail Price Model (RPM) was developed to estimate retail prices of electricity using
wholesale electricity prices generated by the IPM. This model was developed by ICF under
contract with EPA (ICF 2019). The RPM provides a first-order estimate of average retail
electricity prices using information from EPA's Power Sector Modeling Platform.
IPM includes a wholesale electric power market model that projects wholesale prices paid to
generators. Electricity consumers - industrial, commercial, and residential customers - face a
retail price for electricity that is higher than the wholesale price because it includes the cost of
wholesale power and the costs of transmitting and distributing electricity to end-use consumers.
The RPM was developed to estimate retail prices of electricity based on IPM results and a range
of other assumptions, including the method of regulation and price-setting in each state.
Traditionally, cost-of-service (COS) or Rate-of-Return regulation sets rates based on the
estimated average costs of providing electricity to customers plus a "fair and equitable return" to
the utility's investors. States that impose cost-of-service regulation typically have one or more
investor-owned utilities (IOUs), which own and operate their own generation, transmission, and
distribution assets. They are also the retail service provider for their franchised service territory
in which IOUs operates. Under this regulatory structure, retail power prices are based on average
historical costs and are established for each class of service by state regulators during periodic
5-10
-------
rate case proceedings. Additional documentation on the RPM can be found at on the EPA
website (ICF 2019) (U.S. EPA 2023b). For the final rule, the RPM was updated to incorporate
distribution-level costs from the TEIS.
5.2.2 IPM emissions post-processing
Emissions of non-CC>2 GHG (methane, nitrous oxide), PM, VOC, CO, and NH3 were
calculated via post-processing of IPM power sector projections using EPA-defined emissions
factors. The EPA GHG Emissions Factors Hub was used to determine fuel-specific emissions
factors for methane and nitrous oxide emissions for the electric power sector (U.S. EPA 2022a).
Emissions factors used for post-processing of PM, VOC, CO, and NH3 were documented as part
of Post-IRA 2022 Reference Case - EPA's Power Sector Modeling Platform v6 Using IPM (U.S.
EPA 2023a).
5.2.3 IPM National-level Demand, Generation, Emissions and Costs
5.2.3.1 Power Sector Impacts of the BIL and IRA
EPA's Clean Air Markets Division (CAMD) completed an initial power sector modeling
analysis of the BIL and IRA in 2022 (U.S. EPA 2023a). The IRA provisions modeled within
IPM included:
• Clean Electricity Production and Investment Tax Credits
• Existing Nuclear Production Tax Credit
• Carbon Capture and Storage 45Q Tax Credit
This initial modeling did not include other power sector impacts, such as demand impacts
from higher levels of vehicle electrification or IRA energy efficiency provisions, however these
are likely to be included in future CAMD power sector analyses.
The initial modeling of the IRA showed a 70 percent reduction of power sector related CO2
emissions from current levels by 2055, and that the changes in CO2 emissions would be driven
primarily by increases in renewable generation and enabled by increased use of grid battery
storage capacity (see Figure 5-6).
5-11
-------
3000
2000
500
~60% reduction
from 2005 levels of
power sector C02
emissions by 2030
Power Sector Generation
~70% reduction
from 2021 levels of
power sector C02
emissions by 2055
2005 2010 2015 2020 2025 2030 2035 2040 2045 2050 2055
-Baseline with BIL —•—Baseline with BIL and IRA Historical
5000
2- 4000
5
o
~o
| 3000
3
O
-C
| 2000
2
OJ
c
<9 1000
I Other
I Hydro
Non-Hydro RE
Nuclear
Natural Gas
I Coal
Figure 5-6: Power sector modeling comparing results of the Bipartisan Infrastructure Law
(BIL) and the Inflation Reduction Act (IRA).
5.2.3.2 Power Sector Modeling Results for the Final Rule
EPA analyzed two scenarios for the final rule:
• A no-action case that included modeling of electric vehicle provisions from the IRA
within the OMEGA compliance model and compliance with 2023 and later GHG
standards (86 FR 74434 2021) with the addition of heavy-duty vehicle (Class 4-8)
charge demand estimated for the California Advanced Clean Trucks (ACT) Program.
• A final rule policy or "action" case based upon Alternative 3 from the proposed rule
with the addition of heavy-duty vehicle charge demand based on an interim scenario
developed from the Greenhouse Gas Emissions Standards for Heavy-Duty Vehicles -
Phase 3 Proposed Rule (HDP3).
Alternative 3 was one of the compliance scenarios modeled using OMEGA, EVI-X, and IPM
during the summer and autumn of 2023. Of the scenarios modeled in IPM after the proposal,
Alternative 3 is the closest scenario with respect to PEV charging demand to the final rule and
represents the final rule within the power sector analysis. Alternative 3 differs from the finalized
program by forecasting slightly higher PEV sales in 2027-2031 than finalized, and thus higher
PEV charging demand in earlier years and comparable PEV charging demand after 2032. Thus,
power sector impacts on emissions and cost within the final rule analysis should be considered
conservatively high estimates. Regionalized heavy-duty vehicle charge demand for both the no-
action and policy cases were based upon a combination of NREL EVI-X and LBNL HE VI-
LOAD simulations developed as part of the Multi-State Transportation Electrification Impact
Study (TEIS) (E. Wood, B. Borlaug, et al., Multi-State Transportation Electrification Impact
Study 2024).
One significant difference between the analysis of charge demand for the final rule compared
to the analysis for the proposal is the addition to the policy case analysis of an estimate of heavy-
5-12
-------
duty vehicle charge demand based on HDP3. The policy case estimates charge demand from this
final rule combined together with HDP3. The no-action case analysis (e.g., IRA and ACT) uses
similar assumptions to the analysis for the proposal with the exception of updated MDV charge
demand inputs and the use of EVI-X and HEVI-LOAD for regionalization of heavy-duty vehicle
charge demand. For further information on the development of regionalized light- and medium-
duty vehicle charge demand and heavy-duty vehicle charge demand, please refer to Chapter 5.1
and to the TEIS (E. Wood, B. Borlaug, et al., Multi-State Transportation Electrification Impact
Study 2024).
Emissions, demand, generation, and costs for the no-action case and for the light- and medium
duty final rule (including heavy-duty demand) are shown in Table 5-2 and Table 5-3,
respectively. Note that the total costs presented in both tables represent:
• Capital costs for building new power plants as well as retrofits
• Variable and fixed operation and maintenance costs
• Fuel costs
• Cost of capturing, transporting, and storing CO2
Table 5-2: National electric power sector emissions, demand, generation and cost for the
no-action case.
Emission
2028
2030
2035
2040
2045
2050
SO2 (million metric Ions)
0.4928
0.2838
0.1247
0.0828
0.0410
0.0165
PM2.5 (million metric tons)
0.07276
0.06149
0.04410
0.03490
: 0.02725
0.02363
NOx (million metric tons)
0.4950
0.3619
0.2128
0.1544
0.1067
0.0864
VOC (million metric tons)
i 0.03353
0.02982
0.02404
0.01996
0.01675
0.01548
CO2 (million metric tons)
1.296
1.022
638.8
479.6
406.3
347.9
CH4 (metric tons)
85.315
63.402
36.835
28.039
17.528
13.674
N2O (metric tons)
11.784
8.558
4.807
3.641
2.168
1.647
Hg (metric tons)
2.538
1.997
1.469
1.320
1.072
0.9651
HCL (million metric tons)
2.669
1.792
0.8476
0.6450
0.2286
0.1037
Total Demand (TWh)
4.457
4.593
4.924
5,230
5.546
5.893
Light- and Medium-duty
98.06
157.4
265.8
326.2
362.3
384.1
Vehicle PEV Demand (TWh)
Light- and Medium-duty
2.20%
3.43%
5.40%
6.24%
6.53%
6.52%
Vehicle PEV % of Total
Demand
Light- and Medium-duty
89.8%
i 87.2%
76.8%
70.6%
i 67.1%
64.8%
Vehicle PEV % of
Transportation Demand
Total Generation (TWh)
4.548
4.739
5.183
5.593
5.982
6.465
Total Cost (Billion $)
131.2
128.1
131.9
139.7
140.8
143.0
5-13
-------
Table 5-3: National electric power sector emissions, demand, generation and cost for the
final rule.
Emission
2028
2030
2035
2040
2045
2050
SO2 (million metric Ions)
0.4792
0.2826
0.1570
0.08755
0.04420
0.02115
PM2.5 (million metric tons)
0.07168
0.06103
0.04861
0.03635
0.02940
0.02541
NOx (million metric tons)
0.4856
0.3605
0.2445
0.1593
0.1135
0.09330
VOC (million metric tons)
0.03328
0.02987
0.02568
0.02059
0.01815
0.01592
CO2 (million metric tons)
1.267
1.014
737
513
455
394
CH4 (metric tons)
82.950
63.218
42.952
30.026
19.054
15.271
N2O (metric tons)
11.439
8.530
5.642
3.908
2.360
1.846
Hg (metric tons)
2.513
2.028
1.608
1.378
1.120
1.023
HCL (million metric tons)
2.591
1.785
1.053
0.6930
0.2604
0.1336
Total Demand (TWh)
4.475
4.646
5,222
5.734
6.173
6.578
Light- and Medium-duty
110.8
193.2
436.8
617.1
736.9
809.6
Vehicle PEV Demand (TWh)
Light- and Medium-duty
2.48%
4.16%
8.36%
i 10.8%
; 11.9%
12.3%
Vehicle PEV % of Total
Demand
Light- and Medium-duty
88.5%
84.2%
70.0%
; 66.3%
65.5%
65.8%
Vehicle PEV % of
Transportation Demand
Total Generation (TWh)
4.562
4.783
5.469
6.1 17
6.651
: 7,212
% Change in Generation from
0.293%
0.932%
5.52%
9.38%
11.2%
: 11.6%
No-action
Total Cost (Billion $)
129.8
128.5
142.4
152.3
157.0
160.4
% Change in Costs from
-1.09%
0.275%
7.97%
9.00%
11.5%
12.2%
No-action
Similar to CAMD's earlier power sector analysis and the previous analysis for the proposal,
the power sector analysis for both the final rule and a no-action case show significant year-over-
year reductions in CO2 emissions from 2028 through 2050 despite increased generation. This is
primarily due to increased use of renewables for generation as shown in Figure 5-7. The policy
case, which estimates the demand due to the light- and medium-duty final rule combined with
that of the heavy-duty Phase 3 proposed rule, shows an approximately 11.6 percent increase in
generation in 2050 relative to the no-action case. Light- and medium-duty PEV charge demand
accounted for approximately 66 percent of total demand for light, medium, and heavy-duty
vehicles combined for the policy case, thus we estimate that increased generation at a national
level due to the light-and medium-duty final rule alone relative to the no-action case is
approximately 7.6 percent in 2050. The policy case has an approximately 13.4 percent increase
in CO2 emissions in 2050 for both the LMDV final rule and HDP3 proposed rule combined
(Figure 5-8), of which an increase in CO2 emissions in 2050 of approximately 8.8 percent is
estimated to be due to light- and medium-duty vehicles. It should be noted, however, that the
increased EGU emissions are completely offset by net reductions in tailpipe emissions and that
there are significant net CO2 reductions for the final rule (see Chapter 9.6).
5-14
-------
Criteria pollutant emissions (Figure 5-9, Figure 5-10, and Figure 5-11) show similar trends of
large year-over-year reductions from 2028 through 2050 due to increased use of renewables for
generation, and similar trends of higher EGU emissions for the policy case relative to the no
action. As with CO2 emissions, the increased EGU criteria pollutant emissions are offset by net
reductions in tailpipe emissions and petroleum refining (see Chapters 9.5 and 9.6). Also note
that EPA is in the process of finalizing updated power sector modeling reflecting additional
power sector regulatory actions that will help mitigate such emission changes even further.
7,500
7,000
6,500
6,000
5,500
5,000
4,500
4,000
S
~ 3,500
0
1 3,000
c
-------
^ Policy Case
No Action Case
s
s
I
2040
Year
Figure 5-9: 2028 through 2050 power sector NOx emissions for final rule policy case (solid
gray line) and no-action case (dashed line).
5-16
-------
v
V
v
v>
v
0 Policy Case
No Action Case
->
~
0
2025
2040
Year
Figure 5-10: 2028 through 2050 power sector PM2.5 emissions for final rule policy case
(solid gray line) and no-action case (dashed line).
Year
Figure 5-11: 2028 through 2050 power sector SO2 emissions for final rule policy case (solid
gray line) and no-action case (dashed line).
5-17
-------
5.2.4 Retail Price Modeling Results
EPA estimated the change in the retail price of electricity (2022$) using the Retail Price
Model (RPM) and using the same methodology used in recent EPA power-sector rulemakings
(U.S. EPA 2022c). The RPM was developed by ICF for EPA (ICF 2019) and uses the IPM
estimates of changes in the cost of generating electricity to estimate the changes in average retail
electricity prices. The prices are average prices over consumer classes (i.e., consumer,
commercial, and industrial) and regions, weighted by the amount of electricity used by each class
and in each region. The RPM combines the IPM annual cost estimates in each of the 67 IPM
regions with EIA electricity market data for each of the 25 NERC/ISO147 electricity supply
subregions (Table 5-4 and Figure 5-12) in the electricity market module of the National Energy
Modeling System (NEMS) (U.S. Energy Information Administraton 2019). The RPM was
updated for the final rule to incorporate regional, distribution-level upgrade costs for light-
medium- and heavy-duty charging for both the no-action and policy cases based on the TEIS (E.
Wood, B. Borlaug, et al., Multi-State Transportation Electrification Impact Study 2024).
Table 5-4 summarizes the projected percentage changes in the retail price of electricity for the
final rule versus a no-action case, respectively. National level retail electricity prices from Table
5-4 were used within the analysis of costs and benefits in Chapter 9. Consistent with other
projected impacts presented above, average retail electricity price differences at the national
level are projected to be between approximately -0.4 percent (i.e., small price decrease) to 2.5
percent in 2030 and 2050, respectively. Regional average retail electricity price differences
showed small increases or decreases of approximately -4 percent to 5 percent. There is a general
trend of reduced national average retail electricity prices from 2021 through 2050, which is
largely due to reduced fuel costs from increased use of renewables for generation.
In the absence of distribution level costs, national-level average retail electricity price
differences were projected to be between approximately -0.7 percent and 0.1 percent between
2030 and 2050, reflecting approximately zero change between the no-action and policy cases on
retail price of electricity. Comparing these results to the modeling results that included
distribution-level upgrade costs, we conclude that most of the estimated 2.5 percent increase in
national average retail electricity prices in 2050 can be attributed to distribution-level upgrades.
147 NERC is the National Electricity Reliability Corporation. ISO is an Independent System Operator, sometimes
referred to as a Regional Transmission Organization.
5-18
-------
Table 5-4: Average retail electricity price by region for the no-action case and the final rule
in 2030 and 2050.
21)20* No- l'iniil No- 1'iniil No- 1'iniil Percent Percent Percent
iiction 2030 iiction 2040 iiction 2050 ('hnnge ('hnnge ('hnnge
2030 2040 2050 203o" 2040** 2050**
NK1« /ISO
Regions
2022 mil
ls/kW
TRE
98.7
sis
S3.9
76.6
77.7
68.5
69.8
y"-L02% :
L45%
1.87°/
FRCC
108.5
; 102.1
102.4
100.7
y 103.0
90.5
93.9
i' 0.31% y
2.35%
3.73°/
MISW
119.7
93.0
92.5
104.2
i 105.5
! 100.2
101.9
-0.56% :
1.21%
1.74°/
MISC
106.3
102.7
101.9
78.7
80.4
82.9
85.1
; -0.76% |
2.15%
2.66°/
MISE
128.6
: 115.0
110.7
104.9
y 107.3
97.2
100.3
-3.77% y
""2.29%
" 3.21°/
MISS
87.3
; 102.5
102.3
90.8
91.4
81.1
82.8
1 -0.22% ;
0.73%
"2.01°/
ISNE
197.0
T 169.2
168.6
176.1
177.1
| 173.8
176.2
' -0.41% y
0.59%
1.36°/
NYCW
207.1
' 239.7
239.3
248.5
250.1
' 230.5
235.7
^ -0.18% ]
0.65%
2.26°/
NYUP
130.1
T 148.0
147.7
151.6
; 152.4
I 135.6
137.0
!' -0.17% 1
0.53%
1.02°/
pjme ;
120.8
129.9
125.0
129.1
j 131.4
^ 119.9
123.3
: -3.71% ;
1.77%
2.80°/
PJMW =
114.2
; 109.2
109.2
93.7
I 95.3
: 90.1
93.2
;' -0.06% 1
1.69%
3.47°/
PJMC
106.0
T 97.9
97.4
88.9
88.9
89.8
94.5
r-0.57% 1
-0.07%
5.21°/
PJMD
94.3
82.0
"" 82.3
84.8
86.9
j 80.0
82.5
f 0.35% y
2.54%
3.08°/
SRCA V
113.4
l 108.1
109.0
134.7
'j 136.7
| 101.2
104.4
r 0.78% y
1.54%
3. IS"
SRSE T
112.2
104.0
104.3
88.9
90.4
T 86.1
87.8
1 0.30% 1
1.70%
"2.01°/
SRC 1
93.9
I 119.5
119.9
94.5
s 96.1
81.8
84.1
: 0.33%
1.64%
2.82°/
SPPS
87.5
80.3
80.6
68.2
69.9
74.4
76.5
I 0.37% !
' 2.44%
2.76°/
SPPC ;
116.0
92.4
92.9
85.4
T 85.3
y 70.7
71.8
1" 0.54%
-0.03%
1.63°/
SPPN
79.0
; 69.5
69.6
81.6
82.5
67.6
69.0
y 0.20% !
1.12%
2.15°/
SRSG ;
109.6
96.5
98.6
101.5
1 ' 102.5
94.5
96.1
I 2.15% '
0.98%
1.70°/
( AMI
166.8
7 ' 183.2
182.3
179.8
y 182.4
[ 176.7
180.1
-0.49% ;
1.45%
1.90°/
CASO
198.1
y 219.2
217.0
218.7
i'" 219.7
: 202.3
203.8
y -1.04% y
' 0.43%
0.75°/
MYl'P
96.2
T 89.2
89.9
98.4
j 102.1
91.7
95.9
1 0.80% ]
3.76%
4.55°/
RMRG ]
108.3
: 98.0
97.4
93.1
'r'" 94.2
88.4
89.3
: -0.63% ;
1.11%
1.11°/
BASN "T
101.0
y 100.5
101.7
99.4
99.5
87.4
89.7
j 1.18% y
" 0.02%
2.64°/
National
116.4
y H3.4
113.0
109.8
[ 111.4
] 101.5
104.0
-0.37% !
1.47%
2.46°/
Table Notes:
* From Post-IRA 2022 Reference Case - EPA's Power Sector Modeling Platform v6 Using IPM (U.S. EPA 2023a)
** Percentage increase in average retail electricity price for the final rule to a no-action case. Negative percentages reflect a decrease in
average retail electricity price for the final rule.
•f One mill is equal to 1/1.000 U.S. dollar, or 1/10 U.S. cent. 2022 mills per kilowatt-hour (mills/kWh) are also equivalent to 2022 dollars
per megawatt-hour ($/MWh)
5-19
-------
Figure 5-12: Electricity Market Module Regions (U.S. Energy Information Administraton
2019).
5.2.5 New Builds, Retrofits and Retirements of EGUs
The electric power sector emissions modeling undertaken in support of this rulemaking, using
IPM (described at the beginning of Chapter 5.2), also projects the anticipated mix of electric
power plants required to meet the imposed electricity demand from vehicle electrification,
subject to various constraints. These power plants are referred to here collectively as Electric
Generating Units (EGU). This definiti on includes all types of generating facilities (e.g. fossil
fuel-fired combustion, nuclear, hydroelectric, solar, wind, other renewables, etc.).
This modeling reveals anticipated EGU retirements, EGU retrofits, and new EGU
construction, which are discussed below. EGUs are retired by IPM when announced by their
owner and for economic reasons. The IRA and BIL resulted in many EGU retirements. As such,
the number and types of EGU retirements associated with the final rule when compared to a no-
action case are small in comparison to those retirements that occurred as a result of the IRA and
BIL.
New EGU capacity modelled by IPM for the no-action case is summarized in Table 5-5. New
EGU capacity modelled by IPM for the final rule is summarized in Table 5-6. EGU retirements
modelled by IPM for the no-action and for the final rule are summarized in Table 5-7 and Table
5-8.
For the no-action case, the retirement of coal-fired EGUs account for the vast majority of all
EGU retirements through 2050 (see Figure 5-8). For the final rule, the retirement of coal-fired
EGUs are very similar to the no-action case (see Table 5-8). New generation for the final rule is
similar to the no-action case. For both the no action case and the final rule, cumulative power
generation from new solar and new wind EGU builds and battery energy storage are expected to
5-20
-------
account for an increasing fraction of all new power generation through 2050 (see Chapter 5.2.3.2
and Figure 5-7). Wind-driven EGUs are expected to become the single largest new source of
EGU capacity for 2040 through 2050, and solar-powered EGUs are expected to comprise the
second largest new source of EGU capacity for 2040 through 2050.
Table 5-5: Projected EGU capacity for the no-action case.
NKW MODKLKI) CAPACITY (Cumulative (,\Y)
2028
2030
2035
2040
2045
2050
Hydro
(1
1
6
8
8
8
iNon-I Iyil ro Renewa hies
36
126
405
623
818
1,030
Biomass
0
0
0
0
0
0
Geothermal
0
0
0 "
0
i 0
0
Landfill Gas
1
1
1""
1
1
l""
Solar
5
44
| 155
; 251
310
414
Wind
29 '
81 '
248
370
; 506
615
(,'oal
(1
0
0
0
0
0
Coal without Carbon Capture & Sequestration (CCS)
o
! 0
0
: o
0
0
Integrated Gasification Combined Cycle without CCS
0
o
0
o
i 0
o
Coal with CCS
j o
0
0
0
0
0
Kiu'i'jiv Storage
23
52
84
107
118
142
Nuclear
(1
0
0
0
0
0
Natural (ins
20
25
36
76
168
270
Combined Cycle without CCS
13
f 16
20
21 '
23
24
Combined Cycle with CCS
! 0
0
0
r o
| "" o
0
Combustion Turbine
7
9
16
55
145
245
Other
0
0
0
0
0
0
(i I'll 11(1 1(1 till
80
204
531
814
1,111
1,451
Table 5-6: Projected EGU capacity for the final rule.
NKW MODKLKI) C ,'Al'AC XI'Y (Cumulative (.W)
2028
2030
2035
2040
2045
2050
Hydro
0
2
6
9
9
9
Non-I Ivd ro Renewa hies
49
140
421
741
974
1,209
Biomass
0
; 0
0
0
0
0
Geothermal
0
; o
i 0
0
0
; o
Landfill Gas
1
i
1
1
1
i
Solar
6
45
159
282
368
479
Wind
41
93
260
457
; 605
729
( ,01ll
0
0
0
0
0
0
Coal without CCS
0
0
| 0 '
0
0
0
Integrated Gasification Combined Cycle without CCS
! 0
; 0
0
0
: o
0
Coal with CCS
o""
0
: 0 ;
0
o
0
Kihtjiy Storage
20
48
82
112
130
154
Nuclear
0
0
0
0
0
0
Natural Gus
19
23
43
92
193
306
Combined Cycle without CCS
14
18
31
32
37
42
Combined Cycle with CCS
o
o
o ;
o
o
0
Combustion Turbine
4
; '5
12
60
[ 156
: 264
Other
0
0
0
0
0
0
Grand Total
88
212
552
954
1,305
1,678
5-21
-------
Table 5-7: EGU retirements for the no-action case.
RKTIRKMKNTS ((JW)
2028 2030 2035
2040
2045
2050
Combined Cycle Retirements
1 1
• 2 :
9
6
15
Coul Retirements
34 79
f 105 If.
115
126
140
Combustion Turbine Retirements
o "" 1
; 2
3
14 Y
19
Nuclen r Reti rements
0 3
10
15
29 T
48
Oil/Gns Steiim Retirements
f ' 5 ff 7
~13'"
15
15
" 17
Integrated Gnsificiition Combined Cycle Retirements
0 ; " 0
1 i
1
1 :
1
ISioinuss Retirements
: 3 i 3
1 3
3
"3
3
I' uel ( .'ell Retirements
0 ' : 0
Q ¦'
0
0 ;
0
Fossil-Other Retirements
1 o |' 0
0
0
0
0
Gcothcrmnl Retirements
0 0
; '0
o
' 0 .... j
0
Hydro Retirements
o r o
o
0
0
0
Liindfill (ins Retirements
0 0
j" o f
o
of j
0
Non-1'ossil, Other Retirements
o " ; o
0
0
0
0
Knergy Storage Retirements
0 0
o r
0
0 j
0
Grand Total
' 44 93
137
153
194 fff
fff 243
Table 5-8: EGU retirements for the final rule.
RKTIRKMKNTS ((JW)
2028
2030 2035
2040
2045
2050
Combined Cycle Retirements
1 j
1 : 2
9
6
Y" 15
Coul Retirements
37
80 i'" 102
112
"T 125
: 138
Combustion Turbine Retirements
F 0
1 2
3
14
19
Nuclcii r Reti rements
0 | "
3 10 "
15
29
| 48
Oil/Giis Stiiim Retirements
; 6
8 14
15
f 15
17
Integrated (Iiisificiition Combined Cycle Retirements
0
"o 1 ""
1
1
1 "
liiomiiss Retirements
: "" 3
3 3
3
: 3
i 3
l' uel ( .'ell Retirements
r 0 •
o f off
0
0
j 0
l'ossil-Other Retirements
f 0 f
o T o
0
0
0
Gcothcrmnl Retirements
' 0 '' T
"of off
o
Y 0
: of.
Hydro Retirements
T O f
0 0
0
0
0
Liindfill (ins Retirements
0.11"
0 0
o
f 0
0 ...
Non-1'ossil, Other Retirements
0 f
off 0
0
f f 0
of.
Knergy Storage Retirements
0 f
0 0
0
... ^
0
(i in nd To till
"" 48
97 133
150
! 192
F 242
5.2.6 Interregional Dispatch
IPM results showing international dispatch are summarized for a no-action case and for the
final rule in Table 5-9 and Table 5-10, respectively. International dispatch only occurred between
Canada and the contiguous United States represented by the IPM regions. Net international
dispatch was also very small as a percentage of total U.S. electricity demand, with electricity
imports less than 0.6 percent for all years and trending towards a very small net export by 2050
for both the no-action case and final rule.
Table 5-9: IPM results for net export of electricity into the contiguous United States for the
no-action case."'^
2028 2030 2035 2040 2045 2050
Net US Exports (TWh) -26.2 -22.9 -7.3 7.9 12.1 14.8
US Electricity Demand (TWh) 4.457 4.593 4.924 5.230 5.546 5.893
Net US Exports as a Percentage of Total Demand (%) -0.59% -0.50% -0.15% 0.15% 0.22% 0.25%
: Table Notes:
i * Negative net exports represent imports of electricity
: f International dispatch to the contiguous United States only occurred over the U.S. - Canada border.
5-22
-------
Table 5-10: IPM results for net export of electricity into the contiguous United States for
the final rule.*^
2028 2030 2035 2040 2045 2050
Net US Exports (TWh) -26.9 -23.1 -10.1 6.2 13.2 18.7
US Electricity Demand (TWh) 4.475 4.646 5.222 5.734 6.173 6.578
Net US Exports as a Percentage of Total Demand (%) -0.60% -0.50% -0.19% 0.11% 0.21% 0.28%
: Table Notes:
i * Negative net exports represent imports of electricity
: f International dispatch to the contiguous United States only occurred over the U.S. - Canada border.
International dispatch only occurred between Canada and the contiguous United States
represented by the IPM regions. To estimate interregional dispatch, IPM utilizes Total Transfer
Capabilities (TTCs), a metric that represents the capability of the power system to import or
export power reliably from one region to another.
The amount of energy and capacity transferred on a given transmission line between IPM
regions is modeled on a seasonal basis for all run years in the EPA Platform v6. All the modeled
transmission lines have the same TTCs for all seasons. The maximum values for these metrics
were obtained from public sources such as market reports and regional transmission plans,
wherever available. Where public sources were not available, the maximum values for TTCs are
based on ICF's expert view. ICF analyzes the operation of the grid under normal and
contingency conditions, using industry-standard methods, and calculates the transfer capabilities
between regions. To calculate the transfer capabilities, ICF uses standard power flow data
developed by the market operators, transmission providers, or utilities, as appropriate. Additional
information regarding power-sector modeling is available via a report submitted to the docket
(U.S. EPA 2023a).
5.2.7 Regional Comparison of No Action and Final Rule IPM Emissions and
Generation Results
For the final rule, the total number of electric vehicles in the nationwide fleet is expected to
increase after the final effective year of the rule. While reduced production of ICE vehicles will
significantly reduce on-road emissions, the burden of charging PEVs is placed on the electrical
grid. Power sector emissions vary regionally based on the mix of energy sources and the
emissions associated with each source. This section will compare the energy generation and
emissions of greenhouse gases, criteria pollutants, and key air toxic emissions between the no
action case and the final rule.
On a regional level, generation is roughly proportional to population size (Figure 5-13). From
2028 to 2050, total generation is expected to increase across most regions, and total generation
under the final rule will outpace the no action case, as expected given the difference in
nationwide total generation between the two cases (see Chapter 5.2.3.2). A handful of regions are
expected to see a decrease in power sector generation, including Arkansas, Idaho, Kentucky,
Maryland, Nebraska, Wyoming, and parts of Montana and North Dakota. In each of these
regions, coal, nuclear, and natural gas facilities are incrementally phased out over the modeled
period. This is also the case for most other regions; however, these listed areas experience a net
decrease in total generation due to slower introduction of nonhydroelectric renewable capacity.
For example, under the final rule in 2028, Wyoming will generate 25.7 TWh from coal, 0.38
TWh from solar, and 19.6 TWh from wind; in 2050, coal, solar, and wind will generate 3.4, 4.3,
5-23
-------
and 29.9 TWh, respectively. Wyoming has a long history of coal mining and until recent years
has relied almost exclusively on coal-fired power plants for generation within the region. Most of
Wyoming's oldest coal-fired EGUs are scheduled to retire by 2038, which will drastically reduce
the region's total coal capacity at a rate that new solar and wind facilities are not projected to
keep pace with by 2050.
No Action. 2028
Combined HDP3 and LMDV Final Rules, 2028
No Action, 2050
r
K
Combined HDP3 and LMDV Final Rules, 2050
450,000
Figure 5-13: Total Generation by IPM Region in 2028 and 2050 in No Action Case and
Final Rule.
Understanding the sources from which each region receives their energy is also vital to
ensuring that the reduced tailpipe emissions are not exceeded by increased regional emissions
from EGUs. From 2028 to 2050, the composition of the national energy sector will experience
significant shifts toward renewable energy sources, primarily wind and solar. For both the no
action case and the final rule, nonhydroelectric renewables account for approximately one-fourth
of nationwide generation in 2028 and by 2050 renewables account for approximately three-
fourths of generation. These projections are based on planned retirements and expansions of
existing facilities as well as plans for new builds of EGUs and transmission lines.
With respect to new transmission, the need for new transmission lines associated with the
LMDV and FIDP3 rules between now and 2050 is projected to be very small, approximately one-
percent or less of transmission. Nearly all of the projected new transmission builds appear to
overlap with pre-existing transmission line right of ways (ROW), which makes the permitting
process simpler. Approximately 41-percent of the potential new transmission line builds
projected by IPM have already been independently publicly proposed by developers. The
approximate regional distribution of the potential new transmission line builds are:
• 24 percent in the West (excluding Southern California), which are largely Federal
lands, that are more-easily permittable for new transmission builds;
5-24
-------
• 21 percent in the desert Southwest, which are largely Federal lands, that are more-
easily permittable for new transmission builds;
• 14 percent in the Midwest;
• 9 percent for each of the Northeast, Mid-Atlantic, and Southeast and Mid-Atlantic
regions; and
• 5 percent each for Southern California and New York State/City regions.
In 2028, the primaiy energy source for electric generation varies considerably by region, with
many regions still relying heavily on coal, nuclear, and natural gas. By 2050, nonrenewable
energy is the primary energy source in only ten regions, and the vast majority of the country will
generate over 50 percent of their total electricity from wind turbines. Renewables, including
wind, solar, hydroelectric, and geothermal, will compose approximately three-fourths of
nationwide energy generation (see Figure 5-14 and Figure 5-15). There are very few differences
between the primary energy sources in the no action case and the final rule by 2050.
Figure 5-14: Percentage of Total Generation from Renewable Energy Sources in
2050 in No Action Case and Final Rule.
2028
and
No Action, 2028
fy,
No Action, 2050
Combined HDP3 and LMDV Final Rules, 2028
m t
Combined HDP3 and LMDV Final Rules, 2050
5-25
-------
No Action. 2028
No Action. 2050
Combined HDP3 and LMDV Final Rules. 2028
*0*
Combined HDP3 and LMDV Final Rules. 2050
Primary Energy Sources for
Generation
¦ >50% Wind
*50% Wind
~ >50% Natural Gas
| | <50% Natural Gas
H >50% Coal
| | <=50% Coal
¦ >50% Hydroelectric
J <50% Hydroelectric
J p-50% Nuclear
~ <50% Nuclear
| | >50% Solar
~ <50% Solar
~ <50% Geothermal
Figure 5-15: Primary Energy Source by IPM Region in 2028 and 2050 in No Action Case
and Final Rule.
Emissions directly associated with the power sector decrease significantly in both the no
action case and the final rule year over year due to increased reliance on renewables for
generation, but there are a small number of IPM regions in which emissions increase by less than
0.5 percent, and one case in which SO2 increases to the same degree in both the no action and
final rule within the Midcontinent ISO in Louisiana. Figure 5-16 through Figure 5-20 show the
total emissions of CO2, NOx, PM2.5, SO2. and mercury, respectively.
5-26
-------
No Action. 2028
Combined HDP3 and LMDV Final Rules. 2028
100M
No Action. 2050 Combined HDP3 and LMDV Final Rules. 2050
Figure 5-16: Comparing CO2 Emissions between the No Action Case and Final Rule in
2028 and 2050.
5-27
-------
No Action. 2028
Combined HDP3 and LMDV Final Rules. 2028
ti*.-*
v
No Action, 2050
*
*
* >
*
400
Combined HDP3 and LMDV Final Rules, 2050 l±j
Figure 5-17: Comparing Mercury (Hg) Emissions between the No Action Case and Final
Rule in 2028 and 2050.
No Action, 2028
No Action, 2050
Combined HDP3 and LMDV Final Rules, 2028
Combined HDP3 and LMDV Final Rules, 2050
A
40,000
Figure 5-18: Comparing NQx Emissions between the No Action Case and Final Rule in
2028 and 2050.
5-28
-------
No Action, 2028 Combined HDP3 and LMDV Final Rules, 2028
Th
No Action, 2050 Combined HDP3 and LMDV Final Rules. 2050 M
50,000
Figure 5-19: Comparing SO2 Emissions between the No Action Case and Final Rule in 2028
and 2050.
5-29
-------
No Action, 2028
Combined HDP3 and LMDV Final Rules. 2028
6,000
in
c
O
in
C
O
No Action, 2050
Combined HDP3 and LMDV Final Rules. 2050
A
A
0
Figure 5-20: Comparing PM2.5 Emissions between the No Action Case and Final Rule in
2028 and 2050.
Power sector emissions are projected to be slightly higher in the final rule than in the no
action case both nationally and regionally due to the higher energy demand for the final rule
from PEV charging. Due to this higher demand, some nonrenewable energy facilities,
particularly natural gas, are projected to produce more energy in 2050 in the final rule than in the
no action case. In the southwest portion of the country, the primary energy source is projected to
be natural gas in 2050, and this is reflected by projections of elevated CO2 emissions. Under the
final rule, Arizona is projected to generate approximately 9 TWh more energy from natural gas
than the no action case. In the region, 2050 C02 emissions are projected to be 16,826 thousand
tons under the final rule and 11,981 thousand tons in the no action case. These are both
significant reductions from 2028, which have projected CO2 emissions of 23 million tons in the
no action case and 23.3 million tons in the final rule case.
Overall net emissions of criteria air pollutants and GHG due to the final rule (i.e., the net sum
of power sector emissions, tailpipe emissions, refinery emissions, etc.) are still significantly
reduced both nationally and in nearly all regions of the U.S. For summaries of net emissions,
please refer to RIA Chapters 7.4 and 8.6.
5.3 Assessment of PEV Charging Infrastructure
As PEV adoption grows, more charging infrastructure will be needed to support the fleet. This
section summarizes the status and outlook of U.S. PEV charging infrastructure, how much and
what types of charging may be needed to support the level of PEV penetration in the rulemaking,
and how we estimated the associated costs.
5-30
-------
5.3.1 Status and Outlook for PEV Charging Infrastructure
5.3.1.1 Definitions
Terminology for charging infrastructure varies in the literature with terms like "charger",
"plug", "outlet", and "port" sometimes being used interchangeably. Throughout this chapter, we
use the following definitions, consistent with DOE's Alternative Fuels Data Center.148 When
referring to public charging, a station is the physical location where charging occurs. Each
station may have one or more Electric Vehicle Supply Equipment (EVSE) ports that provide
electricity to a vehicle. The number of vehicles that can simultaneously charge at the station is
equal to the number of EVSE ports. Each port may also have multiple connectors or plugs, e.g.,
to accommodate vehicles that use different connector types, but each port can only charge one
vehicle at a time. While it is less common to refer to the place home charging occurs (e.g.,
garage or driveway) as a station, we use the term ports in the same way for residential and non-
residential charging.
We do not include electric power infrastructure—generation, transmission, and distribution
systems—in our definition of PEV charging infrastructure.149 Discussions of electric power
infrastructure can be found in Chapters 5.2 and 5.4.
5.3.1.2 Charging Types
Electric Vehicle Supply Equipment (EVSE) ports can be alternating or direct current (AC or
DC); they also vary by power level. Common AC charging types include LI (up to about 2 kW
power) and L2 (up to 19.2 kW power) (AFDC 2024a) (Schey, Chu and Smart 2022). DC fast
charging (DCFC) is available in a range of power levels today, e.g., 50 kW to 350 kW. A
standard for even higher-powered DCFC designed to serve medium- and heavy-duty PEVs, the
Megawatt Charging System (MCS), is currently in development (ANL 2023) (CharIN 2022).
Generally, the use of higher-power EVSE ports corresponds to faster charging (AFDC 2024a)
though the maximum power that vehicles can accept varies by model.150 Most vehicle models
currently use the SAE J1772 standard connector for Level 1 and 2 charging.151'152 There are
multiple connectors for DCFC, including Combined Charging System (CCS), CHAdeMO, and
the North American Charging Standard (NACS) connector or J3400 (AFDC 2024a). NACS
began as a proprietary connector developed by Tesla and is now undergoing a standardization
process by SAE (JOET 2023a) (SAE 2024).
148 Definitions are consistent with those used by the U.S. Department of Energy, Alternative Fuels Data Center
(AFDC 2024a). A diagram is available at: https://afdc.energy.gov/fuels/electricity_infrastructure.html (last accessed
January 9, 2024).
149 The electric power utility distribution system infrastructure, which includes substations, feeders, and distribution
transformers among other components, typically ends at a service drop to a customer's premises.
150 Table 5-1 shows the maximum DCFC power levels we assumed for BEV models in our infrastructure cost
analysis.
151 Tesla vehicles use the NACS connector for AC charging though a J1772 adapter is available.
152 As noted in Chapter 5.3.1.3, many auto manufacturers have announced that they will offer the NACS standard
developed by Tesla on future production models (Reuters 2023). The NACS connector is capable of both L2
charging and DCFC (JOET 2023a).
5-31
-------
Wireless or inductive charging systems have also been demonstrated and sold as aftermarket
add-ons but have not been widely deployed (AFDC 2024a). Due to the uncertainty about the
timing and uptake of wireless charging, we consider it outside the scope of this analysis.
5.3.1.2.1 PEV Charging Infrastructure Status and Trends
Charging infrastructure has grown rapidly over the last decade (AFDC 2024b). As shown in
Figure 5-21 there are about 60,000 public charging stations in the U.S. today with about 160,000
EVSE ports. This is more than double the 74,000 EVSE ports as of the end of 2019 (AFDC
2024b). About three-quarters of public EVSE ports today are L2 (AFDC 2024c), however,
DCFC deployments have generally experienced faster growth than L2 in the past few years
(Brown, Cappellucci, et al., Electric Vehicle Charging Infrastructure Trends from the Alternative
Fueling Station Locator: Third Quarter 2023 2024). Among DCFCs, there is a trend toward
higher power levels with more than half of the EVSE ports capable of power output at 250 kW or
higher and about two-thirds at 150 kW or higher as of the third quarter of 2023 (Brown,
Cappellucci, et al., Electric Vehicle Charging Infrastructure Trends from the Alternative Fueling
Station Locator: Third Quarter 2023 2024).
180,000
EVSE Ports Stations
Figure 5-21: U.S. Public PEV Charging Infrastructure from 2011—2023 (Data Source: U.S.
DOE, Alternative Fuels Data Center (AFDC 2024b) (AFDC 2024c)153.
Estimates for future infrastructure needs vary in the literature, reflecting different
assumptions about driving and charging behavior, residential charging access, and the mix of
EVSE by power levels, among other factors. A recent national assessment by NREL (E. Wood,
B. Borlaug, et al. 2023) estimated that to support 33 million PEVs in 2030, about 1.25 million
public EVSE ports (including 182,000 DCFC ports) would be needed, along with 26.8 million
153 EVSE port and station counts are for the end of the calendar year shown. Values shown for 2023 are from the
Alternative Fuels Data Center (AFDC) Station Locator as captured on January 10, 2024.
5-32
-------
private ports (most at single family homes, but also at multi-family homes and workplaces) (E.
Wood, B. Borlaug, et al. 2023). That yields a ratio of about one public EVSE port needed per 26
PEVs.154 This fits well within a range of other recent studies examining public infrastructure
needs. An ICCT report looking across a dozen studies published between 2018 to 2021 found
that two-thirds of the estimates (including its own) fell between 20 and 40 PEVs per public
EVSE port (Bauer, et al. 2021).155 A new report conducted by ICF for the Coordinating Research
Council, which assessed infrastructure needs for the level of PEV adoption in the proposed
rule,156 found one public EVSE port would be needed for every 34 light-duty PEVs (CRC 2023).
There was approximately one public EVSE port for every 26 PEVs on the road as of the third
quarter of 2023157 suggesting public charging infrastructure is generally keeping pace with PEV
adoption. See RTC Section 17 for additional discussion of infrastructure availability and recent
infrastructure needs assessments.
5.3.1.3 PEV Charging Infrastructure Investments
Investments in PEV charging infrastructure have grown rapidly in recent years and are
expected to continue to climb. According to BloombergNEF, total cumulative global investment
in PEV charging reached almost $55 billion in 2022 and was estimated to reach nearly $93
billion in 2023 (BloombergNEF 2023). U.S. infrastructure spending has also grown significantly
over the past several years with estimated public charging investments of $2.7 billion in 2023
alone (BloombergNEF 2023).
The U.S. government is making large investments in infrastructure through the Bipartisan
Infrastructure Law (Public Law 117-58 2021) and the Inflation Reduction Act (Public Law 117-
169 2022). However, we expect that private investments will also play a critical role in meeting
future infrastructure needs. Private charging companies have already attracted billions globally in
venture capital and mergers and acquisitions, indicating strong interest in the future of the
charging industry. Bain projects that by 2030, the U.S. market for electric vehicle charging will
be "large and profitable" with both revenue and profits estimated to grow by a factor of twenty
relative to 2021 (Zayer, et al. 2022).158 Domestic manufacturing capacity is also increasing.
About forty companies have announced over $500 million of investments in U.S. facilities to
construct charging equipment, with planned domestic production capacity of more than one
million chargers (including 60,000 DCFCs ) annually (U.S. DOE 2024) (U.S. DOE 2023).159
154 The number of EVSE ports needed to meet a given level of charging demand will depend on the mix of L2 and
DCFC ports (and EVSE power levels of each). A comparison of estimated charging needs in (E. Wood, B. Borlaug,
et al. 2023) to available charging infrastructure by EVSE type is discussed in (Brown, et al. 2024).
155 Note the full range of studies spanned 12 to 129 PEVs per public charger though all but two were between 20 and
56.
156 The study assessed infrastructure needs associated with ZEV adoption in the proposed rule, the proposed
Greenhouse Gas Emissions Standards for Heavy-Duty Vehicles-Phase 3, as well as California policies including
Advanced Clean Cars II rule. The EVSE port to PEV ratio discussed is for light-duty vehicles only.
157 Estimated from approximately 4.16 million EVs and 160,000 public EVSE ports as reported in (Brown, et al.
2024).
158 Estimates account for hardware and installation as well as operations and other charging services such as vehicle -
grid integration.
159 Note: investment and production capacity totals include only those available in public announcements (as
reported by DOE) and may not be comprehensive.
5-33
-------
These activities suggest that companies are positioning themselves to meet the growing demand
for PEV charging.
The NREL study discussed above (E. Wood, B. Borlaug, et al. 2023) estimated that between
$31 billion and $55 billion would be needed by 2030 for public infrastructure, noting that $24
billion in investments from public and private sources had already been announced as of March
2023. The White House estimates that as of January 2024 total investments to expand the U.S.
charging network had grown to over $25 billion (The White House 2024). This includes more
than $10 billion in private sector investments from automakers, charging companies, and
retailers, among others (The White House 2024). Considering 2030 is still six years away, and
that the standards themselves will spur additional investments, announced charging infrastructure
investments in the U.S. (described in the following sections) appear to be putting us on track to
support the PEV adoption under the finalized standards. Furthermore, these public and private
parties are already responding to the market that is developing for infrastructure, and we see no
reason to believe they won't continue to meet infrastructure demand as the PEV market grows.
5.3.1.3.1 Bipartisan Infrastructure Law
The Bipartisan Infrastructure Law (BIL)160 (Public Law 117-58 2021) provides up to $7.5
billion over five years to build out a national network of PEV chargers. Two-thirds of this
funding is for the National Electric Vehicle Infrastructure (NEVI) Formula Program (U.S. DOT,
FHWA 2022a). The remaining $2.5 billion is for the Charging and Fueling Infrastructure (CFI)
Discretionary Grant Program, which is evenly divided between funds for charging and fueling
infrastructure along corridors and in communities where fueling infrastructure can include
hydrogen, propane, or natural gas (U.S. DOT, FHWA 2022a). Both programs are administered
under the Federal Highway Administration (FHWA) with support from the Joint Office of
Energy and Transportation (JOET).
The first phase of NEVI formula funding for states was launched in 2022 and is focused on
building out Alternative Fuel Corridors (AFCs) on highways. Charging stations for AFCs are
required to have at least four DCFC ports, each 150 kW or higher (88 FR 12724 2023). Per
FHWA's guidance to states, stations generally must be located no more than 50 miles apart and
one mile from the Interstate (U.S. DOT, FHWA 2022a). Initial plans for all 50 states, DC, and
Puerto Rico covering FY22 and FY23 funds were approved in September 2022 (U.S. DOT,
FHWA 2022b). Together, the $1.5 billion in funding will help deploy or expand charging
infrastructure on about 75,000 miles of highway (U.S. DOT, FHWA 2022b). Ohio was the first
state to open a NEVI-funded station near Columbus in December 2023 (JOET 2023c). New
York and Pennsylvania followed with stations in Kingston (JOET 2023d) and Pittston (JOET
2024b), respectively. Another 30 states are soliciting proposals and making awards (JOET
2024e). An additional $885 million is available for state plans in FY24 (JOET 2023e). In
September 2023, JOET announced that up to $100 million in NEVI funding would be available
to increase reliability of the existing charging infrastructure network with funds going to repair
or replace EVSE ports (JOET 2023f). This will complement efforts of the National Charging
Experience (ChargeX) Consortium. Launched in May 2023 by JOET and led by U.S. DOE labs,
the ChargeX Consortium will develop solutions and identify best practices for common problems
related to the consumer experience, e.g., payment processing and user interface, vehicle-charger
160 Signed into law as the "Infrastructure Investment and Jobs Act"
5-34
-------
communication, and diagnostic data sharing (JOET 2023g). Relatedly, in January 2024, JOET
announced $46.5 million in federal funding to support 30 projects to increase charging access,
reliability, resiliency, and workforce development (JOET 2024c). This includes projects to
increase the commercial capacity for testing and certification of high-power electric vehicle
chargers, which will accelerate the deployment of interoperable, safe, and efficient electric
vehicle and charger systems (JOET 2024d). Also in January 2024, over $600 million in grants
under the CFI Program was announced to deploy PEV charging and alternative fueling
infrastructure in communities and along corridors in 22 states (JOET 2024a). This first round of
CFI grants is expected to fund about 7,500 EVSE ports.
In addition to NEVI and CFI, there are a variety of other Federal programs that could help
reduce State or private costs associated with deploying EVSE. For example, constructing and
installing charging infrastructure is an eligible activity for other U.S. Department of
Transportation formula programs including the Congestion Mitigation & Air Quality
Improvement Program, National Highway Performance Program, and Surface Transportation
Block Grant Program, which have a total of more than $40 billion in FY22 funds authorized
under the BIL (U.S. DOT, FHWA 2022a).161 Discretionary grant programs include the Rural
Surface Transportation Grant Program, Infrastructure for Rebuilding America Grant Program,
and the Discretionary Grant Program for Charging and Refueling Infrastructure (U.S. DOT,
FHWA 2022a).
5.3.1.3.2 Inflation Reduction Act
The Inflation Reduction Act (IRA), signed into law on August 16, 2022 (Public Law 117-169
2022), will also help reduce the cost that consumers and businesses pay toward PEV charging
infrastructure. The IRA extends the Internal Revenue Code 30C Alternative Fuel Refueling
Property Tax Credit (Section 13404) through Dec 31, 2032, with modifications. Under the new
provisions, residents in low-income or non-urban areas, representing around two-thirds of
Americans (The White House 2024), are eligible for a 30 percent credit for the cost of installing
residential charging equipment up to a $1,000 cap. Businesses are eligible for up to 30 percent of
the costs associated with purchasing and installing charging equipment in these areas (subject to
a $100,000 cap per item) if they meet prevailing wage and apprenticeship requirements.
5.3.1.3.3 Equity Considerations in BIL and IRA
The infrastructure funding in the BIL and the IRA tax credit discussed above can help to
address equity challenges for PEV charging infrastructure. One of the stated goals of the $7.5
billion in infrastructure funding under the BIL is to support equitable access to charging across
the country (U.S. DOT, FHWA 2022a). Accordingly, FHWA instructed states to incorporate
public engagement in their planning process for the NEVI Formula program, including reaching
out to Tribes, and rural, underserved, and disadvantaged communities among other stakeholders.
Both the formula funding and discretionary grant program are subject to the Justice40 Initiative
target that 40 percent of the overall benefits of certain covered federal investments go to
disadvantaged communities (U.S. DOT, FHWA 2022a). As noted above, the Internal Revenue
161 Only a portion is likely to be used to support PEV charging infrastructure, and limits and restrictions may apply.
5-35
-------
Code 30C tax credit will help residents in low-income and non-urban census tracts to install
home charging and provide an incentive for businesses to site stations in these areas.
5.3.1.3.4 Other Public and Private Investments
States, utilities, auto manufacturers, charging network providers, and others are also investing
in and supporting PEV charging infrastructure deployment. California announced plans to invest
$1.9 billion in state funds through 2027 for charging and hydrogen refueling infrastructure
serving light-, medium-, and heavy-duty vehicles (and related activities), which it estimates
could support 40,000 new EVSE ports (California Energy Commission 2024). The New York
Power Authority is investing $250 million to support up to 400 DCFC stations (NYPA 2023).
Several states including New Jersey and Utah offer partial rebates for residential, workplace, or
public charging while others such as Georgia and D.C. offer tax credits (AFDC 2023c).162 The
NC Clean Energy Technology Center identified more than 200 actions taken across 38 states and
D.C. related to providing financial incentives for electric vehicles and or charging infrastructure
in 2022, a four-fold increase over the number of actions in 2017 (Apadula, et al. 2023).163 Other
programs will increase charging access at multi-unit dwellings. For example, the municipal
utility in Burlington, Vermont, in partnership with EVmatch, offers rebates for EVSE
installations at these properties with an additional $300 incentive provided if owners make
charging equipment available for public use during the day to further extend charging access
(Oreizi 2022). The Edison Electric Institute estimates that electric companies are investing $5.2
billion in infrastructure and other transportation electrification efforts in 35 states and the District
of Columbia (EEI 2023a).164 And over 60 electric companies and cooperatives serving customers
in 48 states and the District of Columbia have joined together to advance fast charging through
the National Electric Highway Coalition (EEI 2023b).
In July 2023, seven automakers—BMW, GM, Honda, Hyundai, Kia, Mercedes-Benz, and
Stellantis—announced that they would jointly deploy 30,000 EVSE ports in North America
(Domonoske 2023). GM is also partnering with charging provider EVgo to deploy over 2,700
DCFC ports (GM 2021) and charging provider FLO to deploy as many as 40,000 Level 2 ports
(with a focus on deployments in rural areas) (Valdes-Dapena 2022). Ford plans to install publicly
accessible DCFC ports at many of its dealerships (JOET 2023b). Mercedes-Benz recently
announced that it is planning to build 2,500 charging points in North America by 2027 (Reuters
2023). Tesla has its own network with nearly 24,000 DCFC ports and more than 10,000 L2 ports
in the United States (AFDC 2024c). Tesla announced that by 2024, 7,500 or more existing and
new ports (including 3,500 DCFC) would be open to all PEVs, and that it would double the size
of its DCFC network (The White House 2023). Many auto manufacturers have announced that
they will offer the NACS standard developed by Tesla on future production models in order to
access the Tesla network (Reuters 2023).
Auto manufacturers are also providing support to customers. Volkswagen, Hyundai, and Kia
all offer customers complimentary charging at Electrify America's public charging stations
(subject to time limits or caps) in conjunction with the purchase of select new electric vehicle
162 Details on eligibility, qualifying expenses, and rebate or tax credit amounts vary by state.
163 Includes actions by states and investor-owned utilities.
164 The $5.2 billion total reflects approved filings for infrastructure deployments and other customer programs to
advance transportation electrification.
5-36
-------
models (VW 2023) (Hyundai 2023) (Kia 2023). Ford has agreements with several charging
providers to make it easier for their customers to charge and pay across different networks (Ford
2019).
Other charging networks are also expanding. Francis Energy, which has fewer than 1000
EVSE ports today (U.S. Department of Energy, Alternative Fuels Data Center 2023d), aims to
deploy over 50,000 by the end of the decade (JOET 2023b). Electrify America, a subsidiary of
VW that is implementing the $2 billion investment165 required as part of a 2016 Clean Air Act
settlement (EPA 2023), plans to more than double its network size (U.S. Department of Energy,
Alternative Fuels Data Center 2023d) to 10,000 fast charging ports across 1800 U.S. and
Canadian stations by 2026. This is supported in part by a $450 million investment from Siemens
and Volkswagen Group (Electrify America 2022). Blink plans to invest over $60 million to grow
its network over the next decade (JOET 2023b). Charging companies are also partnering with
major retailers, restaurants, and other businesses to make charging available to customers and the
public. For example, EVgo is deploying DCFC at certain Meijer locations, CBL properties, and
Wawa. Volta is installing DCFC and L2 ports at select Giant Food, Kroger, and Stop and Shop
stores, while ChargePoint and Volvo Cars are partnering with Starbucks to make charging
available at select Starbucks locations (JOET 2023b). Walmart recently announced plans to
expand their network of DCFCs from fewer than 300 locations to thousands of Walmart and
Sam's Club facilities by 2030 (Kapadia 2023). Other efforts will expand charging access along
major highways at up to 500 Pilot and Flying J travel centers (through a partnership between
Pilot, GM, and EVgo) and 200 TravelCenters of America and Petro locations (through a
partnership between TravelCenters of America and Electrify America) (JOET 2023b). BP plans
to invest $1 billion toward charging infrastructure by the end of the decade, including through a
partnership to provide charging at various Hertz locations across the country that could support
rental and ridesharing vehicles, taxis, and the public (BP 2023). All of these investments indicate
that the charging infrastructure market is rapidly expanding with market participants taking the
necessary steps to meet demand. As previously noted, we see no reason to believe that this trend
will not continue throughout the timeframe of this rulemaking.
5.3.2 PEV Charging Infrastructure Cost Analysis
To assess the infrastructure needs and associated costs for this final rule, we start with
estimates of PEV charging demand generated using the methodology described in Chapter 5.1.
These demand estimates are then used to project the number and mix of EVSE ports that may be
needed each year under the final rule and a no-action case.166 Finally, we assign costs for each
EVSE port type intended to reflect upfront hardware and installation costs based on values in the
literature. This section summarizes the methodology and assumptions used for the PEV
165 The $2 billion investment is for charging or hydrogen refueling infrastructure as well as other activities to
advance ZEVs (e.g., education and public outreach).
166 The final rule and no-action cases used throughout the PEV charging infrastructure cost analysis were based on a
preliminary analysis compared to the final compliance modeling. While annual PEV charging demand is generally
higher in the compliance scenarios relative to those in the preliminary analysis (with annual differences of between
plus and minus five percent), cumulative electricity consumption associated with PEV charging from 2027 to 2055
in the final rule compliance scenario is only four percent higher for the action case (the final standards) and one
percent higher in the no action case, compared to the preliminary analysis used to assess PEV charging
infrastructure needs and costs. (Note the scenarios used for power sector modeling are described in Chapter 5.2.)
5-37
-------
infrastructure cost analysis and presents the resulting EVSE costs under the final rule relative to
the no-action case.
5.3.2.1 Charging Demand Projections
Regionalized PEV charging demand and EVSE port needs under our final rule and a no-
action case were simulated by NREL for select years from 2026-2055 under an Interagency
Agreement between EPA and the U.S. Department of Energy (U.S. EPA 2022b). The analysis
framework utilized to estimate charging demand (and EVSE port counts described in Chapter
5.3.2.2) was adapted from the framework used in NREL's 2023 study, The 2030 National
Charging Network: Estimating U.S. Light-Duty Vehicle Demandfor Electric Vehicle Charging
Infrastructure (E. Wood, B. Borlaug, et al. 2023) though the PEV adoption scenarios are specific
to this analysis as described in Chapter 5.1.167
NREL's EVI-Pro model was used to simulate charging demand from typical daily travel, EVI-
RoadTrip was used to simulate demand from long-distance travel, and EVI-OnDemand to
simulate demand from ride-hailing applications (see (E. Wood, B. Borlaug, et al. 2023).) Eight
unique charging types and locations were considered: home LI, home L2, depot L2, work L2,
public L2, and public DCFC at 150 kW, 250 kW, and 350 kW power levels (DC-150, DC-250,
and DC-350). The following assumptions informed the respective charging shares for daily
travel modeled with EVI-Pro.
For light-duty PEVs:
• PEV drivers with access to residential charging are assumed to prefer home over either
work or public charging when home charging is sufficient to support all travel needs.
• 75 percent of BEVs and 53 percent of PHEVs are assumed to use L2 for home
charging with the remaining share using LI.168
• Workplace L2 is the next most preferred charging type after home charging.
• Remaining charging needs are met with public charging. DCFC is generally preferred
for BEVs, and among DCFC, the highest power that a vehicle can accept (or "as fast
as possible" charging) is preferred.
• Public L2 charging is used by PHEVs, which are assumed not to be DCFC-capable.
It's also used by BEVs in certain long dwell time location types such as schools or
medical facilities where it's assumed that DCFC is not available.
167 Other aspects of the modeling methodology that are specific to this analysis are described in Chapter 5.1 and
5.3.2.
168 This in part reflects assumptions about the characteristics of PEVs modeled by OMEGA, including a percentage
of low mileage PEVs for which LI meets daily charging needs.
5-38
-------
For medium-duty PEVs:169
• All medium-duty PEVs (e.g., work van or pickup truck) are assumed to have access to
L2 charging at their home base (i.e., the location they are regularly parked when not in
use.) For some PEVs, this could be at a dedicated depot for commercial fleets whereas
others could be regularly parked overnight and charged at the owner's home. For
simplicity, we refer to both options as "depot L2".
• Drivers are assumed to prefer depot L2 charging over public charging when it is
sufficient to support all travel needs.
• Remaining charging needs are met with public charging. DCFC is generally preferred
for BEVs, and among DCFC, the highest power that a vehicle can accept (or "as fast
as possible" charging) is preferred. Public L2 charging is used by medium-duty
PHEVs.
For road trips and travel by ride-hailing vehicles modeled in EVI-RoadTrip and EVI-
OnDemand,170 respectively, all public charging is assumed to be met with DCFC for BEVs.
Additionally, BEVs able to accept higher-power charging (Gen 2) are assumed to be adopted
more quickly for these applications than for daily travel needs modeled in EVI-Pro.171
As shown in Figure 5-22, the share of PEV charging demand by location and type is similar
between the final rule and no-action case, though depot charging increases more under the final
rule due to a higher relative share of medium-duty PEVs. The majority of PEV charging demand
across all years is projected to be met at homes (primarily with L2 charging) though under the
final rule, the home charging (LI and L2) share declines from over 70 percent in 2028 to 53
percent in 2055 as the share of depot, workplace, and public charging grow.172 DCFC has the
next highest share of demand after home charging. Due to the modeling assumption that BEVs
charge "as fast as possible" when using DCFC, 350 kW charging dominates.173
169 Charging infrastructure needs for medium-duty PEVs were not simulated for the NPRM due to timing
constraints, and therefore depot charging and other projected medium-duty PEV demands are new additions for this
analysis. We also note that medium-duty PEVs are out of scope for (E. Wood, B. Borlaug, et al. 2023), which
assessed charging needs for light-duty vehicles.
170 Medium-duty PEVs are not modeled within EVI-OnDemand or EVI-RoadTrip. All ride-hailing vehicles are
assumed to be light-duty PEVs. Medium-duty PEVs are assumed to be used primarily for commercial applications
and therefore less likely to be regularly used for long-distance travel.
171 For max DC fast charging rates for different vehicle types modeled in this analysis, see Table 5-1.
172 While public L2 and DC-350 kW charging share grow, lower-powered DCFC options decline between 2028 and
2055.
173 While the modeling framework allows for 50 kW DCFC, no charging demand at this power level is found in
either the final rule or no-action case since simulated BEV models are capable of higher-power charging. Therefore,
we do not include 50 kW DCFCs in the discussion or tables presented in this chapter.
5-39
-------
100%
90%
80%
70%
60%
50%
¦
¦
-
-
-
-
-
-
-
-
-
_
-
¦
¦
¦
¦
¦
¦
¦
¦
I
¦
I
¦
1
¦
1
40%
30%
20%
10%
(W
¦
¦
u
m
u
m
\J /0
No Action
Final Rule
No Action
Final Rule
No Action
_QJ
CC
~CO
c
i_i_
No Action
Final Rule
No Action
Final Rule
No Action
Final Rule
No Action
Final Rule
2028
2030
2035
2040
2045
2050
2055
¦ Home LI
Home L2
¦ Depot L2
1 Work L2 ¦ Public L2
¦ DC-150 1
¦ DC-250
¦ DC-350
Figure 5-22: Share of charging demand by location and type for the no-action case (left side
of each pair of bars) and final rule (right side of each pair of bars) for 2028—2055.174
5.3.2.2 Projected EVSE Port Needs
The number of EVSE ports needed to meet the level of PEV charging demand in our final rule
and the no-action case was estimated for all charging types described above. Home charging was
further delineated into EVSE ports at single family houses (SFHs)—including both detached and
attached houses (e.g., townhouses) — and at non-SFHs. Non-single family home ports include
those serving multi-family housing (e.g., apartments and condominiums), mobile homes, as well
as curbside or other neighborhood ports used by PEV drivers without access to dedicated off-
street parking.175 Several additional assumptions informed this network sizing. For both home
and depot charging, it was assumed that as PEV adoption increases, more charging ports would
be shared across vehicles. This could reflect SFHs with more than one PEV, residents of multi-
unit dwellings sharing L2 ports, or medium-duty PEVs sharing ports at depots or the owner's
home. Specifically, we assume that at 1 percent PEV adoption, 1 EVSE port is needed per light-
duty PEV with home charging access. This declines to 0.6 EVSE ports per light-duty PEV for
SFHs and 0.5 EVSE ports per light-duty PEV for other home types when PEVs make up the
entire light-duty fleet. For medium-duty PEVs, we assume that at 1 percent adoption, an EVSE
174 The demand shares shown and used within the PEV charging infrastructure analysis do not assume any managed
charging.
175 Curbside or neighborhood ports are modeled in this analysis as a home charging option and could be either public
or semi-private (restricted access). This differs from (E. Wood, B. Borlaug, et al. 2023), in which these ports are
classified as public "neighborhood" ports.
5-40
-------
port is needed for each vehicle, declining to 0.5 EVSE ports per PEV when PEVs make up the
entire medium-duty fleet.
Network sizing for public and workplace charging is based on the regional charging load
profiles described in Chapter 5.1.176 For each DCFC port type (DC-150, DC-250, and DC-350)
the total number of ports needed is scaled such that during the peak hour of usage 20 percent of
ports in the region are fully utilized. For work and public L2 charging, 60 percent and 55 percent
of ports, respectively, are assumed to be fully utilized during the peak hour. These percentages
are modeled after highly utilized stations today (E. Wood, B. Borlaug, et al. 2023).177 Figure
5-23 and Figure 5-24 show the growing charging network that would be needed to meet PEV
charging demand if auto manufacturers comply by using the PEV penetrations under a central
case analysis178 of the final standards and no-action case, respectively.179 We anticipate that the
highest number of ports will be needed at homes, growing from under 16 million in 2027 to over
77 million in 2055 under the final standards.180 This is followed by public charging, estimated to
grow from under 600,000 ports to over 7.8 million total EVSE ports in that timeframe. The
majority of these are public L2 ports with about 685,000 DCFCs estimated to be needed by 2055.
Depot and workplace181 charging needs also increase to over 3.7 million and about 5.8 million
EVSE ports in 2055, respectively. Notably, while DCFC at 350 kW constitutes a significant
fraction of total electricity demand (Figure 5-22), the number of ports needed is relatively small
compared to the scale shown. This is because far fewer 350 kW ports are needed to deliver the
same amount of electricity as lower-powered options (e.g., public L2 ports). Similar patterns are
observed in the no-action case though fewer total ports are needed than under the final standards
due to the lower anticipated PEV demand. Table 5-11 summarizes port counts by EVSE type for
select years182 under the final standards and in the no-action case.
176 However, as previously noted, the final rule and no-action cases used in the PEV charging infrastructure analysis
differ from those used in the power sector modeling.
177 The same method and thresholds for sizing the non-residential charging network based on peak hour of usage
was applied for all years in this analysis. If we instead assumed the percentage of L2 or DCFC ports that are fully
utilized at peak grew as a function of time or PEV penetration, we would expect higher average utilizations per port
and fewer total ports needed.
178As noted earlier, the PEV charging analysis is based on a preliminary analysis that includes minor differences
with the final central case.
179 Charging simulations were conducted for 2026, 2028, 2030, 2035, 2040, 2045, 2050, and 2055. Linear
interpolations were used to estimate the network size in intermediate years.
180 The number of EVSE ports needed to meet a given level of electricity demand will vary based on the mix of
charging ports, charging preferences, vehicle characteristics, and other factors. Estimates shown reflect assumptions
specific to this analysis, but actual needs could vary.
181 Compared to the NPRM, the relative shares of work and public L2 EVSE ports has changed with some charging
that was previously classified as work shifting to public L2. This is due to a change in the modeling assumptions.
Consistent with (E. Wood, B. Borlaug, et al. 2023), the FRM analysis assumes some workplace charging will be
done at publicly-accessible ports where employees work. Since work and public L2 are assigned the same cost per
EVSE port (as described in Chapter 5.3.2.4), this change has no impact on total costs.
182 EVSE port counts for all years from 2027—2055, along with data inputs used in this analysis, are available in the
docket, see (Burke 2024).
5-41
-------
100
90
80
70
60
50
40
30
20
10
. 11111111111 lit
..iiifTTTTllllllllll
¦ I 1 '
I I
11111
rsicsirsimmmmmmmmm
LO LO LD Ln LTJ LTJ
rsicsirsirsirsirsirsirsirsirsirsirsirsirsirsicsirsirsirsirsirvirsirsicsirsirsirvirsirvi
SFH LI
SFH L2
Non-SFH L2
Depot L2
Work L2
Public L2
DC-150
DC-250
DC-350
Figure 5-23: EVSE port counts by charging type for the final standards 2027—2055.
80
70
60
° 50
^ 40
+-<
o
Q_
I I ¦
i/i
>
30
20
10
¦ i
Mill
r--cocr)0
-------
Table 5-11: EVSE port counts (thousands) for select years under the final rule and no-
action case.
Final rule
No-action case
2027
2030
2040
2055
2027
2030
2040
2055
SFH LI
4.251
8.047
17.241
19.228
4.143
7.695
13.571
16.650
SFH L2
10.537
20.894
47.271
53.669
10.190
19.398
35,022
43.838
Non-SFH L2
727 1
1.424
3.447
4.300
705
1.332
2.443
3.129
Depot L2
62
357
j 2,472 'i
3,737 f
45
137
702
1.157
Work L2
385
777
3.417
5.793
375
754
1.839
2.948
Public L2
492
992
4.289
7.187
480
964
2.343
3.740
DC-150
37
67
80
11
35
61
54
7
DC-250
14
37
44
45
13
36
37
40
DC-350
25
68
334
629
25
63
165
301
Total183
16.530
32.662
78.594
94.599
16.011
30.440
56.175
71.809
5.3.2.3 EVSE Cost Approach
In order to estimate the costs incurred each year, we calculate how many EVSE ports of each
type would need to be procured and installed to achieve the charging network sizes shown in
Figure 5-23 and Figure 5-24.184 There is limited data on the expected lifespan and maintenance
needs of PEV charging infrastructure. We make the simplifying assumption that all EVSE ports
have a 15-year equipment lifetime (Borlaug, Salisbury, et al. 2020). After that, we assume they
must be replaced at full cost. This assumption likely overestimates costs as some EVSE
providers may opt to upgrade existing equipment rather than incur the cost of a full replacement.
Some installation costs such as trenching or electrical upgrades may also not be needed for the
replacement. We do not attempt to estimate EVSE maintenance costs due to uncertainty but note
that maintenance may be able to extend equipment lifetimes. Another simplifying assumption we
make is that EVSE ports are operational and able to meet PEV charging demand the same year
costs are incurred. The actual time to permit and install can vary widely by port type, power
level, region, site conditions, and other factors.
5.3.2.4 Hardware & Installation Costs per EVSE Port
We assign costs to each of the above infrastructure types intended to reflect the upfront capital
costs associated with procuring and installing the EVSE ports. There are many factors that can
impact equipment costs, including whether ports are wall-mounted or on a pedestal as well as
183 Due to rounding, totals may differ from the sums of port counts shown for a given year.
184 The EVSE port needs described in Chapter 5.3.2.2 were simulated independently for run years 2026, 2028, 2030,
2035, 2040, 2045, 2050, and 2055 without accounting for the mix of ports in previous years. For select years and
EVSE types (e.g., DC-150), this results in lower future port needs as demand shifts to other types (e.g., DC-350).
We estimate costs needed to achieve full network sizes in Figure 5-23 and Figure 5-24 even if this results in a
slightly 'overbuilt' network for select years.
5-43
-------
differences in equipment features and capabilities (Schey, Chu and Smart 2022). For example, an
ICCT paper found that costs more than doubled between networked and non-networked L2
hardware (Nicholas 2019). Among networked units with one or two ports per pedestal, about a
10 percent difference in per-port hardware costs was found (Nicholas 2019). The power level of
the EVSE is one of the most significant drivers of cost differences. While estimates for charging
equipment vary across the literature, higher-power charging equipment is typically more
expensive than lower-power units.
Installation costs may include labor, materials (e.g., wire or conduit), permitting, taxes, and
upgrades or modifications to the on-site electrical service. These costs—particularly labor and
permitting—can vary widely by region (Schey, Chu and Smart 2022). They also vary by site. For
example, how much trenching is needed will depend on the distance from where the charging
equipment will be located and the electrical panel. A recent study found that average L2
installation costs at condominiums and commercial locations increased by $16 or $20 for each
extra foot of distance between the EVSE and power source respectively (Schey, Chu and Smart
2022). How many EVSE ports are installed also impacts cost. ICCT estimated that on a per-port
basis, installation costs for 150 kW ports were about 2.5 times higher when only one port is
installed compared to 6-20 per site (Nicholas 2019). And, as with hardware costs, installation
costs may rise with power levels.
To reflect the diversity of hardware and installation costs, we considered a range of costs for
each charging type as shown in Table 5-12 and detailed below.
Table 5-12: Cost (hardware and installation) per EVSE port.185
Home
Depot
Work
Public
SFH LI
SFH L2
non-SFH L2
L2
L2
L2
DC-150 DC-250
DC-350
Low
$0
$870
$3,740
$1,690
$4,400
$4,400
$112,200 $146,150
$180,100
Mid
$0
$1,280
$5,620
$6,150
: $7,500
; $7,500
$154,200 $193,450
$232,700
High
$550
$1,690
$7,500
$10,600
i $10,600
i $10,600
$ 196.200 $240,750
$285,300
5.3.2.4.1 Home Charging
PEVs typically come with a charging cord that can be used for LI charging by plugging it into
a standard 120 VAC186 outlet, and, in some cases, for L2 charging by plugging into a 240 VAC
outlet.187 We include the cost for this cord as part of the vehicle costs described in Chapter 2, and
therefore don't include it here. For our "Low" and "Mid" cases, we make the simplifying
assumption that PEV owners opting for LI home charging already have access to a 120 VAC
outlet and therefore do not incur installation costs.188 To reflect that some PEV drivers opting for
185 All costs shown above and used within the cost analysis are rounded to the nearest ten and expressed in 2022
dollars.
186 Volts, alternating current.
187 Not all charging cords may be capable of Level 2 charging.
188 (Ge, et al. 2021) found that while residential charging access is expected to decline as PEV adoption grows, the
majority of PEVs are projected to have access to an outlet either where they regularly park or at another parking
location at their home even if PEVs reach 100% of the light-duty fleet.
5-44
-------
LI charging may need to install an outlet near their preferred parking spot, we assign installation
costs for our "High" case, set at the average of the high and low estimates for LI home
installation costs provided in (E. Wood, B. Borlaug, et al. 2023).189
For L2 home charging, some PEV owners may opt to simply install or upgrade to a 240 VAC
outlet for use with a provided cord while others may choose to purchase or install a wall-
mounted or other L2 charging unit, which may have additional features and capabilities. In
Table 5-12, the "Low" cost for single-family homes assumes outlet installations only, the
"High" cost assumes the purchase and installation of L2 units, and the "Mid" cost assumes an
even split.190
For other home types (non-SFHs), which include both ports at multi-family housing as well as
curbside or other neighborhood L2 ports, the "Low" cost is also assigned to reflect outlet
installations only at apartments whereas the "High" cost reflects costs for a public L2 EVSE port.
The "Mid" cost is the average of the two.
Costs vary by housing type with installation costs for SFHs typically lower than those for
apartments, condos, or mobile homes (non-SFHs). We use costs by housing type from (Nicholas
2019) for outlet upgrades for all housing types and L2 unit installations for SFHs.191 Public L2
costs are described in Chapter 5.3.2.4.3 below.
5.3.2.4.2 Depot Charging
As described in Chapter 5.3.2.1, depot L2 charging may reflect charging at commercial
depots, i.e., dedicated locations for a fleet of commercial medium-duty vehicles. Alternately, a
medium-duty vehicle (e.g., a work van or pickup truck) could be parked at the PEV owner's
home. In this case, residential L2 charging equipment would be used. In Table 5-12, the "Low"
cost assumes all medium-duty vehicles are charged at single-family homes using hardwired
residential L2 equipment, the "High" cost assumes all medium-duty vehicles charge at
commercial depots.192 The "Mid" cost assumes an even split of residential and commercial L2
charging.
5.3.2.4.3 Work and Public Charging
Cost estimates for work and public EVSE ports (both L2 and DCFC) are updated for the final
rule analysis to align with NREL's 2023 national charging network assessment (E. Wood, B.
Borlaug, et al. 2023). This study drew from various data and literature sources, including the
studies that were used as sources for work and public L2 (Nicholas 2019) and DCFC costs
189 In the NPRM, we assigned $0 costs for all LI home charging.
190 This is unchanged from the NPRM. We note that NPRM costs were expressed in 2019 dollars. We adjusted these
to 2022 dollars (starting from unrounded values).
191 We use costs from Table 5 of (Nicholas 2019), specifically "Level 2 outlet upgrade" for outlet only installations
and "Level 2 charger upgrade" for hardware and installation costs associated with a Level 2 charging unit. For
SFHs, we weight the relative share of light-duty vehicles owned by residents of detached versus attached houses,
sourced from Figure 12 of (Ge, et al. 2021). Apartment costs are used for non-SFHs.
192 We assign the high end of our public L2 cost range to reflect that medium-duty PEVs may use higher power
charging compared to light-duty PEVs or that depots may incur higher installation costs.
5-45
-------
(Borlaug, Muratori, et al. 2021) in the NPRM. We use the sum of the low unit and installation
costs in (E. Wood, B. Borlaug, et al. 2023) as the "Low" costs in
Table 5-12,193 and the sum of the high costs in (E. Wood, B. Borlaug, et al. 2023) as the
"High" costs. Our "Mid" costs are the average of "Low" and "High".
5.3.2.5 Will Costs Change Over Time?
The infrastructure costs shown above reflect present day costs (expressed in 2022 dollars).
However, both hardware and installation costs could vary over time. For example, hardware
costs could decrease due to manufacturing learning and economies of scale. Recent studies by
ICCT assumed a 3 percent annual reduction in hardware costs (Nicholas 2019) (Bauer, et al.
2021). By contrast, installation costs could increase due to growth in labor or material costs. As
noted above, installation costs also depend on site conditions, including whether sufficient
electric capacity exists to add charging infrastructure and how much trenching is required
between the EVSE port and electrical panel. If easier and, therefore, lower cost sites are selected
first, then over time installation costs could rise as charging stations start to be installed in more
challenging locations. (Bauer, et al. 2021) found that these and other countervailing factors could
result in the average cost of a 150 kW EVSE port in 2030 being similar (~3 percent lower) to
that in 2021.
Due to the uncertainty on how costs may change over time, we have made the simplifying
assumption for this analysis to keep combined hardware and installation costs per EVSE port
constant.
5.3.2.6 PEV Charging Infrastructure Cost Summary
Table 5-13 shows the estimated annual EVSE costs194 for the indicated calendar years in the
final rule relative to the no-action case using the "Low", "Mid", and "High" per port cost
estimates. Annual costs range from $0.3 billion dollars under the low scenario to $17 billion
under the high scenario. The table also shows the present value (PV) of these costs and the
equivalent annualized value (EAV) for the calendar years 2027-2055 using 2 percent, 3 percent,
and 7 percent discount rates. The "Mid" costs are included as social costs in the net benefits
estimates for this final rule, presented in Chapter 9.2.
193 We apply "L2 commercial" costs in (E. Wood, B. Borlaug, et al. 2023) for both work and public L2. We treat
costs in (E. Wood, B. Borlaug, et al. 2023) as 2022 dollars.
194 As described in Chapter 5.3.2.4 above, EVSE costs include hardware and installation costs for the EVSE. They
do not include any costs associated with distribution system upgrades. Those costs are accounted for in our FRM
analysis in the electricity price (see Chapter 5.2.4).
5-46
-------
Table 5-13: EVSE costs for the final rule relative to no-action case (billions of 2022 dollars).
Calendar Year
Low
Mid
High
2027
$0.9
$1.3
$1.9
2028
$0.3
$0.6
$0.8
2029
$1.4
$2.3
$3.2
2030
$1.4
$2.3
$3.2
203 1
$6.7
$10.0
$14.0
2032
$6.7
$10.0
$14.0
2035
$6.7
$10.0
$14.0
2040
$6.0
$9.0
$12.0
2045
$7.6
$12.0
$16.0
2050
$8.3
$13.0
$17.0
2055
$5.7
$8.6
$12.0
PV2
$120
$190
$260
PV3
$110
$160
$220
PV7
$63
$96
$130
EAV2
$5.9
$9.0
$12.0
EAV3
$5.7
$8.8
$12.0
EAV7
$5.2
$7.9
$11.0
5.4 Grid Reliability
Electric power system reliability can be determined using a variety of statistical metrics. The
generally accepted metrics by which electric utilities across the U.S. measure and report electric
power system reliability is set by the Institute of Electrical and Electronics Engineers (IEEE)
using the standard IEEE 1366-2022 (IEEE Guide for Electric Power Distribution Reliability
Indices). The formulation of overall electric power system reliability metrics includes electric
power outages associated with what is known as "loss of supply" events; these are events in
which electric power generation and/or electric power transmission is the root cause for a power
outage.
Using this approach, we observed that electric power utilities in 48 U.S. Census Division and
State regions tracked by the U.S. Energy Information Administration (EIA) had overall trends in
distribution grid reliability that were less than the national average for the years 2013 and 2021
(the most-recent years for which data is available) (EIA, 2022). Conversely, 13 U.S. Census
Division and State regions had overall trends in distribution grid reliability for the same years
that were greater than the national average for the years in question. According to the California
Public Utilities Commission, "This data alone does not fully capture the current state of
reliability of the U.S. electric power distribution system..." (Enis 2021). Given the massive size
of the electric power distribution system - with its multitude of regional, climate, and density
variations - interpreting distribution system reliability indices can be challenging. Moreover,
such reliability statistics focus on outage duration and customer counts, which may obscure
important regional variations. However, as the expected increase in electricity generation
associated with the final rule relative to the no action case is relatively small - ranging from 4
percent in 2030 to approximately 12 percent in 2050 - we do not expect the U.S. electric power
distribution system to be adversely affected by the projected additional number of charging
electric vehicles.
5-47
-------
It is not uncommon for the electric power system to have additional, unutilized generation
capacity at various times throughout a given day. Grid operators can utilize this previously
untapped generation capacity by shifting the charging of electric vehicles to times where excess
underutilized generation capacity exists and/or shift electric vehicle charging away from times
where generation capacity is less prevalent, without affecting the availability of electric vehicles.
Such benefits are also conferred to non-EV loads as well. This allows the grid operators to more
effectively use existing electric power system resources, which decreases overall operative costs
for all ratepayers. These research efforts (Kintner-Meyer, et. al., 2020; Pless et. al., 2020;
Satchwell et. al., 2023; and Lipman et. al., 2021) have capitalized on the mismatch between
electric generation capacity and demand by demonstrating the ability to shift up to 20 percent of
electric vehicle charging loads from any hour of the day to other times of the day. Conversely,
the research efforts also demonstrated the ability to increase electric vehicle charging loads by up
to 30 percent in a given hour of the day. The ability to shift and curtail electric power is a feature
that can improve grid operations and, therefore, grid reliability. Integration of electric vehicle
charging into the power grid, by means of vehicle-to-grid software and systems that allow
management of vehicle charging time and rate, has been found to create value for electric vehicle
drivers, electric grid operators, and ratepayers (Chhaya et. al., 2022). We anticipate similar
strategies could be used to shift PEV charging loads from peak times as needed to reduce grid
impacts across different regions. As the expected increase in electric power demand resulting
from PEV charging in this final rule will be well under 20 percent, we do not anticipate it to pose
grid reliability issues.
How the additional electricity demand from PEVs will impact the grid will depend on many
factors, including the time-of-day that charging occurs, the use of battery storage, and vehicle-to-
grid (V2G) or other Vehicle-Grid Integration (VGI) technology. For example, PEVs can be
scheduled to charge at off-peak hours when the electricity demand is easier to meet. Onsite
battery storage, if deployed at charging stations, could also reduce potential grid impacts by
shifting when electricity is drawn from the grid while still providing power to vehicles when
needed. Stationary battery systems, when combined with DC fast chargers, can also help to
reduce the overall impact on the grid by reducing power drawn directly from the distribution
system with supplemental power provided by the battery systems. Such stationary battery
systems can then recharge when unutilized power is available on the distribution system.
Managed charging and battery storage could also enable increasing renewable use if charging
load is shifted to times with excess solar or wind that might otherwise be curtailed. V2G
technology, which allows electricity to be drawn from vehicles when not in use, could even
allow PEVs to enhance grid reliability.
Management of PEV charging can reduce overall costs to utility ratepayers by delaying
electric utility customer rate increases associated with equipment upgrades and may allow
utilities to use electric vehicle charging as a resource to manage intermittent renewables. When
PEVs charge during hours when existing grid infrastructure is underutilized, they can put
downward pressure on all customers' electric rates by spreading fixed grid investment costs
across greater electricity sales (Satchwell et. al., 2023). The development of new electric utility
tariffs, including those for submetering for electric vehicles, will also help to facilitate the
management of electric vehicle charging and can help to reduce PEV operating costs. When
employed as distributed energy resources (DER), PEVs can help to defer and/or replace the need
for specific transmission and distribution system equipment upgrades. Recently, NREL found
5-48
-------
that a vehicle-to-grid control strategy which lowered an EV battery's average state of charge
when parked - while ensuring it was fully recharged in anticipation of the driver's next need -
could extend the life of the battery if continued over time (NREL, 2023). Similarly, a study by
Environment and Climate Change Canada, NRC Canada, and Transport Canada also found no
significant difference in usable battery energy between a vehicle that was used for bidirectional
V2G and one that was not, and identified an improved SOC profile resulting from V2G activity
as a possible factor (Lapointe et. al., 2023). Application programming interfaces have been
developed by industry in partnership with ANL to manage the exchange of energy services
contracts, enabling the dispatch of PEVs and other distributed energy resources into utility
planning and operations territory-wide or within a specific section of the distribution grid (Evoke
Systems 2023). Further, automakers including BMW, Ford, and Honda developed a joint venture
that promises to enable their EV customers to earn financial savings from managed charging and
energy-sharing services (Honda 2024). See Section IV.C.5 of the preamble for a discussion of
DERs and their potential benefits.
Many stakeholders have been engaged in Vehicle-Grid Integration (VGI) efforts. These
include most major automakers (e.g., Ford, GM, FCA, BMW, Audi, Nissan, Toyota, Honda, and
others), electric utilities (e.g., SCE, PG&E, SDG&E, etc.), the Electric Power Research Institute,
EVSE providers, researchers, and the California Energy Commission (Chhaya, et al. 2019)
(Lipman, Harrington and Langton 2021), among others.
The increasing integration of electric vehicle charging into the electric power grid has also
been found to increase grid reliability (Chhaya, et al. 2019), as the ability to shift and curtail
electric power loads improves grid operations and, therefore, grid reliability. Such integration
has been found to create value for electric vehicle drivers, electric grid operators, and ratepayers.
Management of PEV charging can reduce overall costs to utility ratepayers by delaying electric
utility customer rate increases associated with equipment upgrades and may allow utilities to use
electric vehicle charging as a resource to manage intermittent renewables or provide ancillary
services.
The Electric Power Research Institute (EPRI)195 is undertaking a three-year-long research
project to better understand the scale of commitment and investment in the electric power grid
that is required to meet the anticipated electric power loads. Thus far, the electric power sector
and its regulators have focused on incremental EV load growth and charger utilization (Electric
Power Research Institute 2022). The work of EPRI focuses on grid impacts and associated lead
times required to better prepare the grid (including transmission, substation, feeder, and
transformer) for vehicle electrification. These efforts are, in part, based upon grid reliability
research conducted by EPRI (Maitra 2013) (Electric Power Research Institute 2012), which
identified grid and charging behavior characteristics associated with grid resiliency. We also
consulted with FERC staff on distribution system reliability and related issues.
Managed EV charging provides several benefits to vehicle owners, rate payers that do not
operate electric vehicles, and the operators of the electric power system, including lower costs
and longer lifespans for electric power system assets. Managed electric vehicle charging, when
coupled with time-of-use (TOU) electric rates, can help to further reduce already low refueling
195 EPRI is an independent, nonprofit, U.S.-based organization that conducts research and development related to the
generation, delivery, and use of electricity [https://www.epri.com/].
5-49
-------
costs of EVs by allowing vehicle operators to charge when electricity rates are most
advantageous. Since low electricity costs coincide with surpluses of electricity, such charging
reduces the overall costs of electricity generation and delivery to all electricity rate payers, not
just those charging electric vehicles. Researchers at the Lawrence Berkeley National Laboratory
(LBNL) identified 136 active or approved EV-specific TOU electric utility rates for U.S.
investor-owned utilities in 37 states and the District of Columbia (Cappers et. al., 2021). Of the
136 active or approved EV-specific TOU electric utility rates, 54 rates are for residential
customers, 48 rates are for commercial customers, 27 rates are for utility-owned facilities, four
rates are for fleet operators, and the remaining three rates are for mixed facilities.
As discussed above, managed charging has demonstrated the ability to shift up to 20 percent
of EV charging loads from any hour of the day to other times of the day as well as to increase
EV charging loads by up to 30 percent in a given hour of the day (Lipman et. al., 2021). By
more-effectively utilizing existing electric power system assets, managed electric vehicle
charging can also help to further reduce overall electricity costs by allowing for the deferral of
electric power system upgrades, with deferment potential of between 5 and 15 years over the
2021-2050 period (Kintner-Meyer et. al., 2022). While such deferrals reduce immediate capital
expenditures for electric power system operators, they also extend the functional lifespan of
these assets, provide electric utility planners with additional time to consider cost-effective
planning options, and helps to mitigate supply chain shortages for electric power system
components.
New technologies and solutions exist and are emerging to connect all these new charging
stations to the grid as quickly as possible. Utility hosting capacity maps are one tool available
that developers can use to identify faster and lower cost locations to connect new EV chargers.
These maps can help charging station developers identify locations where there is excess
available grid capacity. Hosting capacity maps provide greater transparency into the ability of the
distribution grid to host additional distributed energy resources (DERs) such as BEV charging. In
addition, hosting capacity maps can identify where DERs can alleviate or aggravate grid
constraints. Hosting capacity is commonly defined as the additional injection or withdrawal of
electric power up to the limits where individual grid assets exceed their power ratings or where a
voltage violation would occur. Hosting capacity maps, analyzed and created by the utility that
owns the distribution system, are usually color-coded lines or surface diagrams overlayed on
geographic maps, representing the conditions on the grid at the time when the map is published
or updated. The analysis is based on power flow simulations of the distribution circuits given
specific customers' load profiles supplied by the electric circuit and the grid asset data as
managed by the utility. The hosting capacity is highly location specific. A DOE review found
that utilities have published 39 hosting capacity maps in 24 states and the District of Columbia
(U.S. DOE 2024).
Hosting capacity maps can help direct new EV charger deployment to less constrained
portions of the grid, giving utilities more lead time to make distribution system upgrades. In
tandem, new technologies and power control protocols are helping connect new EV loads faster
even where there are grid capacity constraints. Southern California Edison, a large electric utility
in California, proposed a pilot to allow faster connection of new EV loads in constrained areas
by deploying Power Control Systems (PCS). In addition to the anticipated build out of charging
infrastructure and electric distribution grids, innovative charging solutions implemented by
electric utilities have further reduced lead times to deploying BEVs. One approach is for utilities
5-50
-------
to make non-firm capacity available immediately as they construct distribution system upgrades.
In California, Southern California Edison (SCE) proposed a two-year Automated Load Control
Management Systems (LCMS) Pilot. The program would use third-party owned LCMS
equipment approved by SCE to accelerate the connection of new loads, including new EVSE,
while "SCE completes necessary upgrades in areas with capacity constraints." SCE would use
the LCMS to require new customers to limit consumption during periods when the system is
more constrained, while providing those customers access to the distribution system sooner than
would otherwise be possible. Once SCE completes required grid upgrades, the LCMS limits will
be removed, and participating customers will gain unrestricted distribution service. SCE hopes to
evaluate the extent to which LCMS can be used to "support distribution reliability and safety,
reduce grid upgrade costs, and reduce delays to customers obtaining interconnection and utility
power service." SCE states that prior CPUC decisions have expressed clear support for this
technology and SCE is commencing the LCMS Pilot immediately (Southern California Edison
2023).
Plans to use LCMS to connect new EV loads faster in constrained sections of the grid, like
that employed by SCE, are being bolstered by new standards for load control technologies. UL,
an organization that develops standards for the electronics industry, published the UL 3141
Outline of Investigation (OOI) for Power Control Systems (PCS) in January 2024 (UL 2024).
Manufacturers can now use this standard for developing devices that utilities can use to limit the
energy consumption of BEVs. The OOI identifies five potential functions for PCS. One of these
functions is to serve as a Power Import Limit (PIL) or Power Export Limit (PEL). In these use
cases, the PCS controls the flow of power between a local electric power system (local EPS,
most often the building wiring on a single premises) and a broader area electric power system
(area EPS, most often the utility's system). Critically, the standardized PIL function will enable
the interconnection of new BEV charging stations faster by leveraging the flexibility of BEVs to
charge in coordination with other loads at the premise. With this standard in place and
manufacturer completion of conforming products, utilities will have a clear technological
framework available to use in load control programs that accelerates charging infrastructure
deployment for their customers.
In addition to the flexible interconnection enabled by PCS, technologies including battery- or
generation-backed charging and mobile charging can facilitate rapid charging deployment, even
before utility connections can be upgraded. Mobile chargers can be deployed immediately
because they do not require an on-site grid connection. They can be used as a temporary solution
to bring additional charging infrastructure to locations before a stationary, grid-connected
charger can be deployed. Mobile chargers can also help bring charging infrastructure to locations
where traditional charger deployments can be more difficult, such as at multi-unit dwellings
(Bloomberg 2023).
Additional innovative charging solutions will further accelerate charging deployment by
optimizing the use of chargers that have already been installed. Technologies are emerging to
make the most of existing charging infrastructure. Other companies are working on facilitating
the sharing of chargers between more drivers. One company, EVMatch, developed a software
platform for sharing, reserving, and renting EV charging stations, which can allow owners of
charging stations to earn additional revenue while making their chargers available to more EV
drivers to maximize the benefit of each deployed charger. EVMatch is also rolling out a new
product called the EVMatch adapter in partnership with Argonne National Laboratory. The
5-51
-------
EVMatch adapter is a smart charging adapter that can turn any Level 1 or 2 EVSE into a smart
charger that can remotely monitor and control charging to enable even more efficient utilization
of existing chargers (Chenoweth 2023; Harper 2021). Innovative charging models like these can
be efficient ways to increase charging access for EVs with a smaller amount of physical
infrastructure.
State government plays an important role in vehicle electrification (including aspects of grid
resilience), as most electric utilities are regulated by state Public Service Commissions (PSC)
and Public Utility Commissioners (PUC) and since Federal funding for vehicle electrification is
largely distributed through state agencies. The National Association of Regulatory Utility
Commissioners (NARUC), a national association representing the state public service
commissioners who regulate essential utility services, including energy, telecommunications, and
water, produced a series of documents aimed at providing vehicle electrification-related guidance
for state regulators (National Association of Regulatory Utility Commissioners 2022a),
facilitating electric vehicle interoperability (National Association of Regulatory Utility
Commissioners 2022b), and fostering vehicle electrification equity (National Association of
Regulatory Utility Commissioners 2022c). NARUC, in conjunction with the National
Association of State Energy Officials (NASEO) and the American Association of State Highway
and Transportation Officials (AASHTO), also produced a guide for public utility commissions,
state energy offices, and departments of transportation discussing the state-level roles and their
interrelations vis-a-vis transportation electrification (National Association of Regulatory Utility
Commissioners 2022a).
As discussed in Section IV.C.3 of the preamble and as part of our upstream analysis, we
model changes to power generation due to the increased electricity demand anticipated under the
final standards. Bulk generation and transmission system impacts are felt on a larger scale, and
thus tend to reflect smoother load growth and be more predictable in nature. Further, we project
the additional generation needed to meet the projected demand of the light- and medium-duty
PEVs under the final standards to be relatively modest compared to the No Action case, ranging
from 4 percent in 2030 to approximately 12 percent in 2050. This is roughly equal to the
combined latest U.S. annual electricity consumption estimates for data centers (U.S. DOE
2023b) and cryptocurrency mining operations (EIA 2024) or slightly more than the increase in
total U.S. electricity end-use consumption between 2021 and 2022 (EIA 2023b). Electric power
consumption associated with this final rule is expected to increase by 12 percent during the 26-
year period between now and 2050. By comparison, U.S. annual electric power consumption
increased by a similar amount over a shorter 20-year period between 2002 and 2022, a period
during which the electric power sector reliability remained "relatively consistent" between 2013
and 2019 (EIA 2023c), according to the EIA, when excluding major events. While changes in
2013 to EIA's electric reliability reporting complicate direct comparisons with electric reliability
data prior to that year, researchers at Lawrence Berkeley National Laboratory did not find
evidence of a change in electric reliability metrics between 2000 and 2015, when considering the
effects of severe weather and utility spending on the long-term reliability of the U.S. power
system (Larsen et. al., 2020).
5.4.1 Factors Affecting Distribution Grid Reliability
The electric power system in the U.S. has historically been a very reliable system (NREL
2024), with utilities, system planners, and reliability coordinators working together to ensure an
5-52
-------
efficient and reliable grid with adequate resources for supply to meet demand at all times. The
power sector analysis conducted in support of this rule indicates that resource adequacy (EPA
2024a) and grid reliability can be maintained (see Section IV.C.3 of the preamble and RIA
Chapter 5). The electric power system is comprised of three subsystems - generation,
transmission, and distribution - each of which have their own unique electric reliability
attributes. These attributes are discussed below.
Power interruptions caused by extreme weather are the most-commonly reported, naturally-
occurring factors affecting grid reliability, with the frequency of these severe weather events
increasing significantly over the past twenty-years due to climate change (U.S. DOE 2023c).
Conversely, decreasing emissions of greenhouse gases can be expected to reduce future extreme
weather events relative to current severe weather events as well as future severe weather events
if GHG emissions continue unabated, which would serve to reduce the risks for electric power
sector reliability. Extreme weather events include snowstorms, hurricanes, and wildfires. These
power interruptions have significant impact on economic activity, with associated costs in the
U.S. estimated to be $44 billion annually (LaCommare et. al., 2018). The effects of extreme
weather events on grid reliability can be largely discerned using electric power reliability indices.
These indices are reported to EIA with and without inclusion of Major Event Days (MED).
MEDs are used to statistically differentiate between electric power reliability associated with
normal day-to-day operation and reliability associated with atypical operation, such as extreme
weather events. MEDs allow for "major events to be studied separately from reliability
performance that occurs during what would be considered normal operation, and, to better reveal
trends in normal operation that would be hidden by the large statistical effect of major events."
(Warren 2003). Increasingly, MEDs are associated with extreme weather events; climate change
has led to an increase in the frequency and intensity of extreme weather events (NASA 2024),
which have led to significant, widespread power interruptions (G. U. DOE 2023). For example,
six hurricanes were classified as "major" in the 2017 Atlantic hurricane season. The long-term
average number of major hurricanes since 1851 is six per decade (NOAA n.d.). The average
duration of annual electric power interruptions in the U.S., approximately two hours, decreased
slightly from 2013 to 2021, when extreme weather events associated with climate change are
excluded from reliability statistics (EIA 2023). When extreme weather events associated with
climate change are not excluded from reliability statistics, the national average length of annual
electric power interruptions increased to about seven hours (EIA 2022b).
Around 93 percent of all power interruptions in the U.S. occur at the distribution-level, with
the remaining fraction of interruptions occuring at the generation- and transmission-levels (J. L.
Eto, Distribution system versus bulk power system: identifying the source of electric service
interruptions in the US 2019) (Larsen, Severe Weather, Utility Spending, and the Long-term
Reliability of the U.S. Power System 2020). As a part of its overall national security and energy
emergency management responsibilities, DOE's Office of Cybersecurity, Energy Security, and
Emergency Response collects grid reliability information on electric incidents and emergencies
via the Electric Emergency Incident and Disturbance Report (Form DOE-417). Summaries of
these reports appear annually in Electric Disturbance Events (OE-417) Annual Summaries
(DOE, Electric Disturbance Events (OE-417) Annual Summaries 2023). In 2023, about two
percent of all power interruptions reported to DOE were attributable to unexpected transmission
losses and about two and a half percent of all reported power interruptions reported to DOE were
attributable to fuel supply emergencies and large uncontrolled losses of generation capacity
5-53
-------
(DOE, Electric Disturbance Events (DOE-417) 2023). Given differences in study methodologies,
total power interruption percentages may not sum to one hundred percent. Grid reliability is also
tracked by NERC and the EIA. Power interruption reports are submitted with standardized
reliability indices that focus upon the frequency and duration of power interruptions. These
standardized reliability indices are described by the Institute of Electrical and Electronics
Engineers (IEEE) in 1366-2022 IEEE Guide for Electric Power Distribution Reliability Indices
(IEEE 2022). A discussion of grid reliability and factors affecting it for each of the three electric
power sector subsystems appears below, starting with the distribution system.
5.4.2 Distribution Grid Reliability Continues to Improve
As discussed above, most power interruptions occur at the distribution system level.
Researchers at Lawrence Berkeley National Laboratory (LBNL) found very few publicly
available, peer-reviewed studies that evaluate long-term trends in electric reliability over broad
geographic areas that also include formal analysis to validate the statistical significance (J. L.
Eto, Distribution system versus bulk power system: identifying the source of electric service
interruptions in the US 2019) (Larsen, Severe Weather, Utility Spending, and the Long-term
Reliability of the U.S. Power System 2020). Earlier studies on grid reliability suggest that power
system reliability was decreasing over time. For instance, such studies examined publicly
available data from over 155 utilities, representing 50 percent of U.S. electricity sales, and
spanning up to 10 years and found a modest, yet statistically significant, decrease in grid
reliability at a rate of about two percent annually (J. L. Eto 2012). As of 2007, all power
interruptions reported to DOE are also required to be reported to NERC. As such, the DOE and
NERC grid reliability datasets would be expected to mirror each other.
However, further analysis by LBNL researchers of previous grid reliability studies and the
underlying grid reliability data submitted to DOE and NERC was found to reveal that the
apparent decrease in grid reliability reversed when the study timeframe was expanded beyond 10
years and data appropriately screened for errors (J. L. Eto 2012). The LBNL researchers also
found that changes in the grid reliability reporting rules, which were made mandatory during the
timeframe of some studies, may have resulted in a skewing of the data gathered after the rule
change (Fisher 2012).
The LBNL researchers also found that many previous grid reliability studies rely upon data
from DOE and NERC which have been found to be incomplete and inconsistent. For example,
nine power interruptions reported to DOE were not found in the NERC data, and power
interruptions reported to NERC were not reported to DOE, even though they meet the DOE
reporting criteria (Larsen, Severe Weather, Utility Spending, and the Long-term Reliability of
the U.S. Power System 2020). The researchers also found three events in the NERC grid
reliability dataset in which the power to more than 50,000 customers was interrupted, but these
were not found in the DOE dataset. As noted earlier, the DOE and NERC grid reliability datasets
should have been identical, since as of 2007 all DOE events are required to be reported to NERC
as well. NERC grid reliability data were inconsistent with comparable information reported to
DOE.
The researchers also found grid reliability studies in which the statistical significance of the
accompanying results were not tested yet included in the analyses. Researchers also found as
much as a 54 percent difference between power interruption statistics in a given year.
5-54
-------
After accounting for study design shortcomings and data inconsistencies observed in other
studies, the LBNL researchers concluded, with statistical significance, that the regional and U.S.
national-level grid reliability trends are, in fact, improving. Across all utilities, there are, on
average, 1.8 percent fewer interruptions for an average customer. The trend ranges from less than
a one-percent decrease in New England to more than a four-percent decrease in the Rocky
Mountain region. This study considered 16 years of data from 203 U.S. utilities, which represent
about 70 percent of electricity sales.
Other studies have found that electric vehicles can provide valuable grid reliability benefits
(Tuffner 2021). These researchers, from the Pacific Northwest National Laboratory (PNNL) and
Idaho National Laboratory (INL), found that "PEVs can be programmed to act during voltage
dips in a way that, both, is friendly to the grid and causes no significant inconvenience to the
operation of the vehicles." This research was motivated by concerns as early as the late 1980's
that some residential air conditioners could, under certain conditions, inadvertently affect electric
power system reliability (Williams 1992) (Kosterev 2009). As a result, a consortium from the
private sector, electric utilities, academia, national laboratories, NERC, air conditioning
manufacturers, and the DOE, collaborated in a series of three DOE-sponsored national
workshops in 2008, 2009, and 2015 to identify the nature of the potential residential air
conditioning issues. Through these efforts, methodology and protocols to avert such issues were
developed (CERTS 2008) (DOE, U.S. DOE-NERC Workshop on Fault-Induced Delayed
Voltage Recovery (FIDVR) 2009). While this research was expanded to include the potential
effects of electric vehicle charging, it must be noted that electric vehicles have not been
implicated in any such electric power sector disturbances. The potential characteristics of plug-in
electric vehicle charging was further discussed by EPRI at the NERC-DOE FIDVR Workshop in
2015 (Halliwell 2015).
In the study from PNNL/INL, which started in 2014, researchers at INL observed the
behavior of six different light-duty electric vehicles when charging from 240 V Level 2 chargers.
The real-life charging characteristics obtained from the electric vehicles were then used to
simulate the charging of various combinations of the six test vehicles. The researchers applied
the simulated electric vehicle charging loads to a representative distribution feeder from Phoenix,
AZ, which consisted of 1,594 single family residences, 45 percent of which were equipped with
standard air conditioning systems. The simulated charging loads for 120 electric vehicles was
superimposed upon the residential loads of the 1,594 single-family residences as well as their
associated air conditioning loads. As a result of this analysis, researchers identified electric
vehicle charging characteristics that could potentially benefit grid reliability, referred to as "grid
friendly" charging attributes (J. Eto 2023).
To address concerns about the potential of the distribution system to integrate new PEV
charging loads, we commissioned a study under an interagency agreement with DOE to assess
both the costs of potential distribution system upgrades as well as the availability of necessary
distribution system components that could be associated with the level of PEV charging demand
projected for both this final rule and Greenhouse Gas Emissions Standards for Heavy-Duty
Vehicles - Phase 3. A discussion of the study is in Section IV.C.5.ii of the preamble.
5.4.3 Transportation Electrification Impact Study
We commissioned a study as part of an interagency agreement with the U.S. Department of
Energy entitled the "Transportation Electrification Impact Study" (TEIS) to estimate the
5-55
-------
potential costs and benefits associated with electrical distribution system upgrades that may incur
as a result of this final rule in addition to those of the Greenhouse Gas Emissions Standards for
Heavy-Duty Vehicles - Phase 3 (E. Wood, B. Borlaug, et al., Multi-State Transportation
Electrification Impact Study 2024). These costs and benefits include new or replacement
substations, underground and overhead distribution feeders, and service transformers, all in rural,
suburban, and urban locations, as well as along freight corridors. To do so, our study builds upon
the methodology developed by the California Public Utilities Commission (CPUC) for their
Electrification Impacts Study Part 1. The Part 1 CPUC study, which is considered preliminary,
focuses on costs only - and not benefits - associated with preparing California's electrical
distribution system to accommodate the expected large-scale integration of distributed energy
resources (DER).
DERs are a wide variety of resources, such as electric battery storage systems, rooftop solar
panels, smart thermostats, energy efficiency measures, thermal energy storage systems, or
electric vehicles and their charging equipment that reduce power usage. DERs are considered
non-wires alternative (NWA), a non-traditional approach to defer and/or replace the need for
specific transmission and distribution system equipment upgrades. The use of DERs can help
support diverse electrification technologies while maintaining system reliability and
affordability. DERs have been shown to provide significant distribution system benefits, both
financially and in terms of their ability to defer necessary distribution system upgrades. Such
deferment allows the upgrades to be coordinated more-effectively by electric utilities and
provides greater flexibility in accommodating supply chain shortfalls.
The CPUC Part 1 (California Public Utilities Commission 2023) study upon which our study
is, in part, modeled is the first of a two-part study series and is referred to as the DER
"unmitigated" case. By design, the CPUC Part 1 study captures only distribution system costs
associated with DER implementation in California - that is, costs expected by using exclusively
traditional distribution investments. As such, the Part 1 study did not consider alternatives or
future potential mitigation strategies, such as alternative time-variant rates or dynamic rates and
flexible load management strategies. In short, the Part 1 study depicts what would happen if
DER deployment in California is disorderly and "measures were not taken to reduce costs and
manage load." As such, the Part 1 study did not consider alternatives or future potential
mitigation strategies, such as alternative time-variant rates or dynamic rates.
Currently underway, the second CPUC study (California Public Utilities Commission 2023)
("Part 2") of the two-part series is referred to as the "mitigated" case. By design, this study
captures only the distribution system benefits associated with DER deployment in California -
that is, the study is not limited exclusively to traditional distribution investments and therefore
considers alternatives or future potential mitigation strategies, such as alternative time-variant
rates or dynamic rates and flexible load management strategies. In short, the Part 2 study depicts
what would happen if DER deployment in California is orderly and measures were taken to
reduce costs and manage load. As such, the Part 2 study considers alternatives or future potential
mitigation strategies, such as alternative time-variant rates or dynamic rates.
The Part 2 CPUC benefits-only study is designed to complement the already-completed Part 1
CPUC cost-only study insofar as the Part 2 study captures the financial benefits associated with
DER deployment as well as the benefits associated with distribution system upgrade deferment -
essential factors not considered in the Part 1 study (again, by design). Only after the benefit
5-56
-------
estimates associated with DER deployment (from the Part 2 CPUC study) are combined with the
cost estimates of DER deployment (from the Part 1 CPUC study), the CPUC argues, would an
accurate portrayal of large-scale DER deployment in California emerge.
However, some commentors to this rule appear to have misunderstood the purpose of the
CPUC "unmitigated" Part 1 cost-only study as well as the inherent cost-only limitations
associated with the study and, subsequently, cite incorrectly the preliminary and incomplete
results of the Part 1 cost-only study.
In our study, aggregate distribution system-level costs (and benefits) were estimated for five
states using premise-level load profiles that were summed and applied to known utility
infrastructure elements (i.e., substations, distribution feeder lines, service transformers, etc.) and
combined with utility-specific cost information. The resulting system-level cost estimates
quantified the level of traditional grid investment required to meet the vehicle electrification load
requirements associated with the proposed rule.
Time-series data, geospatial and utility network data, socioeconomic data, and advanced
metering infrastructure (AMI) data which were collected, ingested, mapped, and analyzed for
this analysis. Using a full-scale distribution capacity expansion approach from the bottom
(premise-level) up to the substation level, the methodology identifies where and when the
distribution grid will need capacity enhancements under certain policy and charging behavior
scenarios consistent with this final rule. The estimates are developed using thermal capacity
analysis at the substation, distribution feeder, and service transformer levels. Using machine
learning, the study estimates each customer's premise-level electric load over the study period,
using actual customer data.
Premise-level load profiles are developed reflecting the expected adoption of DERs, such as
photovoltaic systems and electric vehicles. Unlike our power sector analysis for the proposal, the
load profiles used for this analysis combine, for the first time, the load profiles for a no-action
case and for the final rule for both the Light- and Medium-Duty Multipollutant Standards
(LMDV) and the Greenhouse Gas Emissions Standards for Heavy-Duty Vehicles - Phase 3
(HDP3) Standards into a single power sector analysis. The load profiles from light-, medium-,
and heavy-duty are distributed into IPM regions using NREL's EVI-X suite of models for light-
duty, MDV, and heavy-duty buses; and using LBNL's HEVI-LOAD model for all other heavy-
duty applications. The resulting premise-level load profiles were aggregated up to electric utility
service territories. These profiles included DER-specific adoption and are used to identify the
magnitude and location of high electrification and DER adoption. The system-level grid impacts
and costs of electricity service were determined based upon the profiles.
This methodology is first applied to five-states, which were selected based on their diversity
in urban/rural population, utility distribution grid composition, freight travel demands, and state
EV policies. The selected states are California, Oklahoma, Illinois, Pennsylvania, and New York.
The results from these five-states are then extrapolated to the 67 IPM regions that we use to
represent the remaining 48 contiguous states within our power sector analysis.
5.4.3.1 TEIS Results
The combined results of the five-state study are summarized in Figure 5-25 through Figure
5-27. Results were also extrapolated nationally and aggregated into the 67 IPM regions and
5-57
-------
national level results are shown in Figure 5-28 for the combined LMDV and HDP3 PEV
charging demand. The "no action" scenario represents transportation impacts due to existing
policies, examples of which include vehicle electrification provisions of the IRA and due to
California's Advanced Clean Trucks (ACT) program. The "action" scenario represents the
impacts of the Agency's LMDV final rule combined with estimated impacts projected due to
HDP3 (see Chapter 5.2.3.2).
Note that while distribution-level costs increase year-over-year for both the no-action case
and for the combined results of the LMDV final rule and projected HDP3 impacts, net costs for
generation and transmission combined decrease year-over-year primarily due to the impact of
power sector provisions within the IRA. Thus the projected retail price of electricity was
relatively insensitive to distribution-level costs. The net impact was a year-over-year decrease in
the retail price of electricity for both the no action and for LMDV and HDP3 (Figure 5-28), and a
small, incremental (0.35 cents/kWh) increase in the national average retail price of electricity.
However, as described in our summary of RPM in Chapter 5.2.4, this small incremental increase
did appear to be primarily due to distribution-level costs.
The TEIS also examines the benefits of a simple managed charging approach, wherein the
peak charging rate is reduced by spreading charging over the full dwell period. In 95 percent of
all cases (Action and No Action), managed charging was less expensive or as expensive as
unmanaged charging. In other words, when neglecting the other benefits associated with this
final rule, it would cost more to do nothing than (not manage electric vehicle charging) than it
would to manage electric vehicle charging in 95 percent of the cases. Managed charging was less
expensive than unmanaged charging by about one percent, on average. These values ranged from
-0.9 percent in Pennsylvania (the lone exception, where managed charging was more expensive
than managed charging) to 6.0 percent in Illinois (where Action Unmanaged was more expensive
than Action managed). The benefits of managed charging increases with increasing PEV
penetration. For instance, the degree to which managed charging costs were less than unmanaged
charging costs were the greatest in California, the state with the highest PEV penetration of the
five states considered, in both 2027 and 2032. Conversely, there was no difference between
managed charging costs and unmanaged charging costs in Oklahoma, the state with the lowest
PEV penetration of the five states considered, in 2027 and 2032. This effect is also apparent over
time for states that had the greatest increase in PEV penetration between the years 2027 and
2032. More advanced managed charging that accounts for the load profiles of other end uses
would be expected to result in even greater benefits.
Some other findings associated with the TEIS include: annual charging infrastructure needs
could increase by 3 percent across five states; incremental distribution grid investment needs
represent approximately 3 percent of current annual utility investments in the distribution
system; incremental distribution grid investment needs decrease by 30 percent with basic
managed charging techniques; and benefits of vehicle electrification to consumers outweigh the
estimated cost of charging infrastructure and grid upgrades.
5-58
-------
California
31.2
No Action - Unmanaged
I No Action - Managed
Action - Unmanaged
I Action - Managed
30.2
29.3
29.9
2027 2032
Figure 5-25: A comparison of the costs needed for distribution level upgrades for the no-
action and action scenarios and showing the impacts of managed vs. unmanaged charging
in California.
No Action - Unmanaged
I No Action - Managed
Action - Unmanaged
| Action - Managed
No Action - Unmanaged
I No Action - Managed
Action - Unmanaged
I Action - Managed
3.3 3.3 3.4
2.5 2.5 2.5 2.5
llll Illl
Figure 5-26: A comparison of the costs needed for distribution level upgrades for the no-
action and action scenarios and showing the impacts of managed vs. unmanaged charging
in Illinois (left) and Oklahoma (right).
5-59
-------
No Action - Unmanaged
I No Action - Managed
Action - Unmanaged
I Action - Managed
No Action - Unmanaged
No Action - Managed
Figure 5-27: A comparison of the costs needed for distribution level upgrades for the no-
action and action scenarios and showing the impacts of managed vs. unmanaged charging
in New York (left) and Pennsylvania (right).
0.140
0.120
0.100
¦ft 0.080
0.060
a.
^ 0.040
0.020
0.000
$350
$300 _
c
o
m
$250 JA
o
u
$200 sj
FRM - Managed - Distribution Cost
2055
Figure 5-28: National distribution-level cost comparison of the no action case with
unmanaged growth to the FRM with managed growth and respective national average
retail price of electricity
(see also Chapter 5.2.4)
5-60
-------
With managed charging, the final standards provide significant distribution system benefits,
both financially and in terms of their ability to defer necessary distribution system upgrades. By
more-effectively utilizing existing electric power system assets, managed electric vehicle
charging has been shown to help reduce overall electricity costs by allowing for the deferral of
electric power system upgrades, with deferment potential of between 5 and 15 years over the
2021-2050 period (Kintner-Meyer et. al., 2022). While such deferrals reduce immediate capital
expenditures for electric power system operators, they also extend the functional lifespan of
these assets, provide electric utility planners with additional time to more-effectively schedule
and coordinate needed distribution system upgrades, consider cost-effective planning options,
and help mitigate supply chain shortages for electric power system components.
Relative to the No-Action case, the cost associated with increasing distribution system
capacity under the Action case decreases in three of the five modeled states. California, Illinois,
and Pennsylvania see decreases of about -0.9 percent in California and Illinois to -1.9 percent in
Pennsylvania. New York and Oklahoma see increases in costs of about 0.8 percent to 3.1
percent, respectively. Oklahoma was selected as one of the five modeled states because it is
representative of typical long-distance freight corridors. However, the anomalously high costs in
the state appear to be attributed to the disproportionately large number of high-power charging
located along its freight corridors relative to lower power chargers, which were found to be
undercounted.
The final rate-payer impacts associated with the final rule are to be determined using the
Retail Price Model (RPM), which provides a first order estimate of average retail electricity
price. RPM is a part of the IPM suite of power sector modeling tools.
While additional capacity must be built in any case to meet the increasing electric power
demands associated with vehicle electrification, an additional 5 GW (eight percent) of generation
capacity must be added under the Action-Unmanaged Charging case, compared to the Action
case with managed charging.
The results from the five-state analysis are extrapolated across the U.S. to the remaining 48
contiguous states for 2027 and 2032. These cost estimates are extrapolated at the county-level for
each state to yield a total cost for each county by asset type (i.e., substation, distribution line,
service transformer, etc.). The net cost of the distribution-level upgrades estimated within TEIS
are included within our analysis of costs and benefits for the final rule along with other grid-
related costs modeled by IPM. These costs were aggregated into IPM regions, and the rate-payer
impacts associated with the final rule are to be determined using the Retail Price Model (RPM),
which provides a first-order estimate of average retail electricity price (see Chapter 5.2.1 and
5.2.4).
The estimated cost per kWh for noncoincident peaks occuring during EV charging was
determined by using the incremental costs for capital addition for each asset type. Extreme
outliers were removed from the dataset when they were found to exceed six standard deviations
from the mean by EVSE type, scenario, and asset type.
The weighted-average cost per scenario and EVSE type are applied to the number of ports per
county per EVSE and averages were calculated for each EVSE type by asset type for each
scenario. These values were summed to obtain 2027 cost estimates. Cost estimates for 2032 are
5-61
-------
incremental to 2027 cost estimates. The substation costs per kW are the highest across all
component types, while feeder costs are the lowest.
Initial cost estimates suggest that 28 states have less than $1B impact of total costs of lower
48, or 11 percent of total for the final rule in 2032. Distribution system upgrade costs associated
with the final rule are greatest in California, followed by New York, Texas, New Jersey, and
Florida. Maryland and Massachusetts have cost commensurate with Pennsylvania and Illinois.
These costs are distributed over the electric utility rate base on a per kW basis. Final costs
estimates are to be developed using the Retail Price Model as noted.
5.4.4 Transmission Improvements Increase Grid Reliability
The transmission system is another portion of the electric power system with unique grid
reliability attributes. The federal government has limited authority to direct transmission system
planning. Delays in the planning, permitting, and construction of transmission power lines are
common (Rand, et. al., 2023).
Existing transmission lines are often congested and may not have sufficient capacity to carry
new generation loads (DOE, 2023a), while existing generation loads often pay congestion
penalties for the right to use the overcrowded transmission lines (Millstein, et. al., 2022; DOE
2023b; Rand, et. al., 2023). While obstacles, such as transmission line delays and transmission
line congestion, exist, several innovative solutions have been developed and deployed by DOE,
including the Grid Deployment Office (DOE, 2023b). For example, two 230-kV transmission
lines used by PPL Electric Utilities, in Pennsylvania, were found to be approaching their
maximum transmission capacity in 2020. As a result, the utility paid more than $60 million in
congestion fees in the winters of 2021-2022 and 2022-2023 (Lehmann, et. al., 2023). Rather than
rebuilding or reconductoring the two transmission lines, which would have cost tens of millions
of dollars, the utility spent under $300 thousand installing dynamic line rating (DLR) sensors,
which helped the utility to rebalance each of the two transmission lines and allowed them to
reliably carry an additional 18 percent of power (FERC, 2023a; Lehmann, et. al., 2023).
New solar and wind generation, as well as fossil-fuel-fired generation, depend upon access to
transmission lines to deliver power they generate to end-users. New generation awaiting
authorization to connect to the transmission system are said to be in "interconnection queues."
The amount of new potential electric generating capacity in these queues is growing
significantly, with over 2,000 gigawatts (GW) of total generation and storage capacity now
seeking connection to the grid. Approximately 95 percent of this new potential electric
generating capacity is from renewable resources, such as solar, wind, and battery storage.
Without adequate transmission line capacity, projects in the interconnection queues are often
cancelled. Most proposed electric generating projects applying for interconnection are
withdrawn, and those that are built take longer on average to complete the required studies and
become operational. The typical interconnection wait time from connection request to
commercial operation increased from less than two years, for projects built between 2000-2007,
to nearly four years for those projects built between 2018-2022 (with a median of 5 years for
projects built in 2022).
To alleviate the interconnection queue backlog, DOE recently announced several programs
and projects. Examples of such programs and projects include DOE's Interconnection Innovation
e-Xchange (i2X), which aims to increase data access and transparency, improve process and
5-62
-------
timing, promote economic efficiency, and maintain grid reliability; FERC Order 2023, which
provides generator interconnection procedures and agreements to address interconnection queue
backlogs, improve certainty, and prevent undue discrimination for new technologies; and DOE's
Grid Resilience and Innovation Partnerships (GRIP) program, with $10.5 billion in Bipartisan
Infrastructure Law funding to develop and deploy Grid Enhancing Technologies (GET), such as
Dynamic Line Ratings (DLR) and Advanced Power Flow Controllers (APFC).
Energy storage projects can also be used to help reduce transmission line congestion and are
seen as alternatives to transmission line construction. These projects, known as Storage As
Transmission Asset (SATA), can help to reduce transmission line congestion, have smaller
footprints, have shorter development, permitting, and construction times, and can be added
incrementally, as required. Examples of SATA projects include the ERCOT Presidio Project, a 4
MW battery system that improves power quality and reduces momentary outages due to voltage
fluctuations; the APS Punkin Center, a 2 MW, 8 MWh battery system deployed in place of
upgrading 20 miles of transmission and distribution lines; the National Grid Nantucket Project, a
6 MW, 48 MWh battery system installed on Nantucket Island, MA, as a contingency to undersea
electric supply cables; and the Oakland Clean Energy Initiative Projects, a 43.25 MW, 173 MWh
energy storage project to replace fossil generation in the Bay area.
Through such efforts, the interconnection queues can be reduced in length, transmission
capacity on existing transmission lines can be increased, additional generation assets can be
brought online, and electricity generated by existing assets will be curtailed less often. These
factors help to improve overall grid reliability.
5.4.5 Electric Generation Will Continue To Be Reliable Under this Final Rule
Electricity production from coal-fired electric power plants decreased from about fifty-percent
of the U.S. generation base in 2004 to about twenty-percent in 2022 (EIA 2023). Some states and
regions, such as New England and California, have shifted almost entirely away from coal, and
several electric power utilities are already coal-free or have announced their intent to close coal-
fired generation capacity.
Given the additional electricity demand stemming from this rule, some commenters raised
concerns that the additional demands associated with the rule could impact the reliability of the
power grid. As such, we conducted an additional resource adequacy and grid reliability
assessment of the impacts of the vehicle rule and how projected outcomes under the rule
compare with projected baseline outcomes in the presence of the IRA. We used power sector
modeling (IV.C.3) to estimate emissions from electric power plants for loads associated with
vehicle electrification as well as to assess generation resource adequacy and grid reliability of the
rapidly-transitioning electric grid. The results of the additional resource adequacy assessment
appear in the associated Technical Memorandum for Multi-Pollutant Emissions Standards for
Model Years 2027 and Later Light-Duty and Medium-Duty Vehicles, and Greenhouse Gas
Emissions Standards for Heavy-Duty Vehicles - Phase 3 ("The Report").
The Report uses the same scenario and years of analysis contained in the RIA. The scenarios
include a base case and the final rule scenario. For purposes of the resource adequacy and
reliability assessment, estimates and projections are taken from those same scenarios and years
as shown in the RIA (2030, 2040, and 2050).
5-63
-------
The focus of the analysis is on comparing the illustrative final rule scenario from the RIA to a
baseline "No Action case" (absent the rule requirements). In this framework, the emphasis is on
the incremental changes in the power system that are projected to occur under the presence of the
rule in the 2030, 2040, and 2050 model run years. The results presented in the Report further
demonstrate, for the specific power sector cases illustrated in Chapter 5 of the RIA, that the
implementation of this rule combined with the HD Phase 3 proposal can be achieved without
undermining resource adequacy, which is a central element of grid reliability. The Report also
evaluates the cumulative impacts of these rules combined several recently proposed Power
Sector Rules, and it finds that these cumulative impacts are associated with changes to the
electric grid that are well within the range of fleet conditions that respect resource adequacy, as
projected by multiple, highly respected peer-reviewed models. Please refer to the docketed
Report for more detailed information on the resource adequacy analysis. Electric generation is
currently reliable (EIA 2023d) with ample resource adequacy (NERC, 2023), and power sector
modeling conducted in support of this rule projects that this rule will not adversely affect
resource adequacy.
Several independent studies, with scenarios and assumptions that bracket the electric power
loads expected with this rule, also indicate that resource adequacy will not be adversely affected
by this rule. EPA's report, Electricity Sector Emissions Impacts of the Inflation Reduction Act,
summarized results from fourteen multi-sector and power sector models under the IRA in 2030
and 2035 (EPA, 2023c). Across the models, wind and solar resources provide 22-54 percent of
generation (median of 45 percent) in 2030 and 21-80 percent (median of 50 percent) in 2035.
The North American Renewable Integration Study showed how the U.S. could accommodate
between 70-79 percent of wind and solar generation by 2050 (Brinkman, et. al., 2021). The Solar
Futures Study illustrated power systems with upwards of 80 percent of renewable energy by
2050 (DOE, 2021). Finally, a study published in the journal Joule demonstrates a 100 percent
renewable power system for the contiguous U.S. (Cole, et. al., 2021).
Power outages in the U.S. are infrequent, occurring about 1.4 times per customer annually and
typically lasting between 2-5 hours (EIA 2023). The effect of power outages on electric vehicle
owners is expected to be similar to that of non-electric vehicle drivers. Neither driver will be able
to "fuel" during power outages, as gasoline pumps are electric powered. However, electric
vehicles can provide their owners with residential power for a limited time. Moreover, electric
vehicle chargers that are attached to distributed energy resources, such as homes or businesses
with solar and/or stationary battery storage, would be unaffected by power outages and, thereby,
can continue to provide charge for electric vehicles via its independent capacity. In fact, electric
vehicles could be used to power gasoline pumps during electric power outages. Given that the
physical extent of typical power outages tends to be relatively small, electric vehicle drivers, as
well as conventional vehicle drivers, can be expected to drive out of the outage area and to
unaffected charging or refueling stations, should it become necessary.
5-64
-------
Chapter 5 References
86 FR 74434. 2021. (December 30).
https://www.federalregister.gov/documents/2021/12/30/2021-27854/revised-2023-and-later-
model-year-light-duty-vehicle-greenhouse-gas-emissions-standards.
88 FR 12724. 2023. https://www.federalregister.gov/documents/2023/02/28/2023-
03500/national-electric-vehicle-infrastructure-standards-and-requirements.
AFDC. 2024c. "Alternative Fuels Data Center, Alternative Fueling Station Locator." January 10.
Accessed January 10, 2024.
https://afdc.energy.gov/stations/#/analyze?country=US&fuel=ELEC.
—. 2024a. Developing Infrastructure to Charge Electric Vehicles,
http s ://afdc. energy. gov/fuel s/el ectri city_infrastructure. html.
—. 2023c. "State Laws and Incentives." Accessed 111, 2023. https://afdc.energy.gov/laws/state;
https://afdc.energy.gov/laws/12832; https://afdc.energy.gov/laws/12320;
https://afdc.energy.gov/laws/5182; https://afdc.energy.gov/laws/11493.
—. 2024b. U.S. Public Electric Vehicle Charging Infrastructure.
https://afdc.energy.gov/data/10972.
Alexander, Matt, Noel Crisostomo, Wendell Krell, Jeffrey Lu, and Raja Ramesh. 2021.
Assembly Bill 2127: Electric Vehicle Charging Infrastructure Assessment. California Energy
Commission, https://www.energy.ca.gov/programs-and-topics/programs/electric-vehicle-
charging-infrastructure-assessment-ab-2127.
ANL. 2023. "Charging for Heavy-Duty Electric Trucks Frequently Asked Questions about the
Megawatt Charging System and SAE J3721." Argonne National Laboratory, March.
https://www.anl.gov/sites/www/files/2023-03/MCS_FAQs_Final_3-13-23.pdf.
Apadula, Emily, Rebekah de la Mora, Justin Lindemann, Brian Lips, Vincent Potter, Autumn
Proudlove, and David Sarkisian. 2023. 50 States of Electric Vehicles Q4 2022 Quarterly Report
& 2022 Annual Review Executive Summary. NC Clean Energy Technology Center. Accessed 3
5, 2023. https://nccleantech.ncsu.edu/wp-content/uploads/2023/02/Q4-
22_EV_execsummary_Final.pdf.
Bauer, Gordon, Chih-Wei Hsu, Mike Nicholas, and Nic Lutsey. 2021. Charging Up America:
Assessing the Growing Need for U.S. Charging Infrastructure Through 2030. The International
Council on Clean Transportation. Accessed 111, 2023. https://theicct.org/wp-
content/uploads/2021/12/charging-up-ameri ca-j ul2021 .pdf.
BloombergNEF. 2023. "Zero-Emission Vehicles Factbook, A BloombergNEF special report
prepared for COP28." December. https://assets.bbhub.io/professional/sites/24/2023-COP28-
ZEV-F actbook. pdf.
Borlaug, Brennan, Matteo Muratori, Madeline Gilleran, David Woody, William Muston, Thomas
Canada, Andrew Ingram, Hal Gresham, and Charlie McQueen. 2021. "Heavy-duty truck
5-65
-------
electrification and the impacts of depot charging on electricity distribution systems." Nature
Energy 6: 673-682. Accessed 1 13, 2023. https://doi.org/10.1038/s41560-021-00855-0.
Borlaug, Brennan, Shawn Salisbury, Mindy Gerdes, and Matteo Muratori. 2020. "Levelized Cost
of Charging Electric Vehicles in the United States." Joule 4 (7): 1470-1485. Accessed 111,
2023. doi:https://doi.org/10.1016/j.joule.2020.05.013.
BP. 2023. "BP plans to invest $1 billion in EV charging across US by 2030, helping to meet
demand from Hertz's expanding EV rentals." February 15. https://www.bp.com/en_us/united-
states/home/news/press-releases/bp-plans-to-invest-l-billion-in-ev-charging-across-us-by-2030-
helping-to-meet-demand-from-hertzs-expanding-ev-rentals.html.
Brown, Abby, Jeff Cappellucci, Emily White, Alexia Heinrich, and Emma Cost. 2024. Electric
Vehicle Charging Infrastructure Trends from the Alternative Fueling Station Locator: Third
Quarter 2023. Golden, CO: National Renewable Energy Laboratory.
https://www.nrel.gov/docs/fy24osti/88223.pdf.
Burke, S. 2024. "Files supporting the PEV Charging Infrastructure Analysis. Memorandum to
Docket EPA-HQ-OAR-2022-0829." March.
California Energy Commission. 2024. "CEC Approves $1.9 Billion Plan to Expand Zero-
Emission Transportation Infrastructure." February 14. https://www.energy.ca.gov/news/2024-
02/cec-approves-19-billion-plan-expand-zero-emission-transportation-infrastructure.
California Public Utilities Commission, Electric Division. 2023. Electrification Impacts Study
Part 1. Proceeding R.21-06-017 (Order Instituting Rulemaking to Modernize the Electric Grid
for a High Distributed Energy Resources Future), San Francisco: California Public Utilities
Commission.
CERTS. 2008. 2008 US DOE Workshop on the Role of Residential AC in Contributing to Fault-
Induced Delayed Voltage Recovery (FIDVR). Accessed 2 28, 2024.
https://certs.lbl.gov/publications/2008-us-doe-workshop-role-residential.html.
CharlN. 2022. "CharIN e.V. officially launches the Megawatt Charging System (MCS) at
EVS35 in Oslo, Norway." June 14. Accessed March 8, 2023.
https://www.charin.global/media/pages/news/charin-e-v-officially-launches-the-megawatt-
charging-system-mcs-at-evs35-in-oslo-norway/2bcal0e315-
1655213 569/press_release_mcs_charin_final.pdf.
Chevrolet. 2023. Installation Made Easy: Home Charging Installation on Us. Accessed 3 1, 2023.
https://www.chevrolet.com/electric/living-electric/home-charging-installation.
Chhaya, S., V. Ananth, R. Ravikumar, G. Bellino, R. Scholer, E. Cutter, C. Botsford, et al. 2019.
"Distribution System Constrained Vehicle-to-Grid Services for Improved Grid Stability and
Reliability; Publication Number: CEC-500-2019-027; Chhaya, Sunil; McCollough, Norman;
Ananth, Viswanath; Maitra, Arindam; Ravikumar, Ramakrishnan; Dunckley, et.al." Accessed
December 13, 2022. https://www.energy.ca.gOv/sites/default/files/2021-06/CEC-500-2019-
027.pdf.
CRC. 2023. "Assessing the Battery-Recharging and Hydrogen-Refueling Infrastructure Needs,
Costs and Timelines Required to Support Regulatory Requirments for Light-, Medium-, and
5-66
-------
Heavy-Duty Zero-Emission Vehicles." Accessed December 17, 2023. https://crcao.org/wp-
content/uploads/2023/09/CRC_Infrastructure_Assessment_Report_ICF_09282023_Final-
Report.pdf.
DOE. 2023. Electric Disturbance Events (DOE-417). Accessed 2 28, 2024.
http s ://www. oe. netl. doe. gov/oe417. aspx.
—. 2023. Electric Disturbance Events (OE-417) Annual Summaries. Accessed 2 28, 2024.
http s ://www. oe. netl. doe. gov/OE417_annual_summary. aspx.
—. 2009. "U.S. DOE-NERC Workshop on Fault-Induced Delayed Voltage Recovery (FIDVR)."
Arlington: U.S. DOE, 9 29.
DOE, GDO U.S. 2023. Keeping the Lights On in Our Neighborhoods During Power Outages. 11
8. Accessed 2 28, 2024. https://www.energy.gov/gdo/articles/keeping-lights-our-neighborhoods-
during-power-
outages#:~:text=According%20to%20Climate%20Central%20%2C%20about,by%20almost%20
80%25%20since%202011.
Domonoske, Camila. 2023. "Big carmakers unite to build a charging network and reassure
reluctant EV buyers." July. https://www.npr.Org/2023/07/26/l 190188838/ev-chargers-network-
range-anxiety-bmw-gm-honda-hyundai-kia-mercedes-stellantis.
EEI. 2023a. "Electric Transportation Biannual State Regulatory Update." December.
https://www.eei.Org/-/media/Proj ect/EEI/Documents/Issues-and-Policy /Electric-
Transportation/ET-Biannual-State-Regulatory-Update.pdf.
—. 2023b. "Issues & Policy: National Electric Highway Coalition." Accessed January 11 2023.
https://www.eei.org/issues-and-policy/national-electric-highway-coalition.
EIA. 2023. Annual Energy Outlook 2023. EIA. https://www.eia.gov/outlooks/aeo/.
—. 2023. Average duration of total annual electric power interruptions, United States (2013-
2020). 11 10. Accessed 2 28, 2024.
https://www.eia.gov/todayinenergy/detail.php?id=50316#:~:text=U.S.%20electricity%20custom
ers%20experienced%20eight%20hours%20of%20power%20interruptions%20in%202020,-
Source%3 A%20U. S. %20Energy&text=On%20average%2C%20U. S. %20electricity%20custome
rs,electricit.
Electric Power Research Institute. 2022. "Supplemental Project Notice: EVs2Scale2030."
Accessed December 14, 2022. https://www.epri.com/research/products/000000003002025622.
—. 2012. "Understanding the Grid Impacts of Plug-In Electric Vehicles (PEV): Phase 1 Study —
Distribution Impact Case Studies; Maitra, Arindam; 1024101." December 31. Accessed
December 13, 2022. https://www.epri.eom/research/products/000000000001024101.
Electrify America. 2022. "Electrify America Raises $450 Million - Siemens Becomes a Minority
Shareholder; Company Intensifies Commitment to Rapid Deployment of Ultra-Fast Charging."
June 28. https://media.electrifyamerica.com/en-us/releases/190.
Enis, Julian. 2021. "Electric System Reliability (Presentation from the California Public Utilities
Commission)." February 17. https://www.cpuc.ca.gov/-/media/cpuc-
5-67
-------
website/transparency/commissioner-committees/emerging-trends/2021/2021-02-17-electric-
system-reliability-presentation—final.pdf.
EPA. 2023. "Volkswagen Clean Air Act Civil Settlement."
https://www.epa.gOv/enforcement/volkswagen-clean-air-act-civil-settlement#investment.
Eto, J., interview by U.S. EPA. 2023. Discussion of FIDVR and Electric Vehicles
Eto, J., LaCommare, K. 2012. "A quantitative assessment of utility reporting practices for
reporting electric power distribution events." 2012 IEEE Power & Energy Society General
Meeting. San Diego: IEEE.
Eto, J., LaCommare, K., Caswell, H, and Till, D. 2019. Distribution system versus bulk power
system: identifying the source of electric service interruptions in the US. Institution of
Engineering and Technology.
—. 2019. "Distribution system versus bulk power system: identifying the source of electric
service interruptions in the US." IET Gener. Transm. Distrib.
Experian Automotive. 2022. "Vehicle in Operation Market Data & Reports." Data Set. Accessed
June 2022. https://www.experian.com/automotive/vehicles-in-operation-vio-data.
Fisher, E., Eto, J., and LaCommare, K. 2012. Understanding Bulk Power Reliability: The
Importance of Good Data and A Critical Review of Existing Sources. Berkeley : Lawrence
Berkeley National Laboratory.
Ford. 2019. "Ford Introduces North America's Largest Electric Vehicle Charging Network,
Helping Customers Confidently Switch to an All-Electric Lifestyle." October 17. Accessed
January 11, 2023. https://media.ford.com/content/fordmedia/fna/us/en/news/2019/10/17/ford-
introduces-north-americas-largest-electric-vehicle-charting-network.html.
Ge, Y., C. Simeone, A. Duvall, and E. Wood. 2021. There's No Place Like Home: Residential
Parking, Electrical Access, and Implications for the Future of Electric Vehicle Charging
Infrastructure. Technical Report No. NREL/TP-5400-81065, National Renewable Energy
Laboratory, https://www.nrel.gov/docs/fy22osti/81065.pdf.
GM. 2021. "To Put 'Everybody In' an Electric Vehicle, GM introduces Ultium Charge 360."
https://media.gm.com/media/us/en/gm/home.detail.html/content/Pages/news/us/en/2021/apr/042
8 -ultium -charge-3 60. html.
Halliwell, J. 2015. "Plug-in Electric Vehicle Charging Characteristics." NERC-DOE FIDVR
Workshop. Washington, D C.: U.S. DOE, 9 30.
Hyundai. 2023. Stay powered for your drive. Accessed 111, 2023.
https://www.hyundaiusa.com/us/en/electrified/charging.
ICF. 2019. "Documentation of the Retail Price Model." ICF Contract Report to the U.S. EPA.
https://www.epa.gov/sites/default/files/2019-06/documents/rpm_documentationjune2019.pdf.
IEEE. 2022. IEEE Guide for Electric Power Distribution Reliability Indices. 11 12. Accessed 2
28, 2024. https://standards.ieee.org/ieee/1366/7243/.
5-68
-------
JOET . 2024b. "Pennsylvania Continues Shift Toward Thriving Electric Transportation Sector."
January 23. https://driveelectric.gov/news/new-pennsylvania-nevi-station.
JOET. 2024e. "2024 Q1 NEVI Progress Update." February 16.
http s ://dri veel ectric. gov/news/nevi -update-q 1.
—. 2024a. "Biden-Harris Administration Bolsters Electric Vehicle Future with More than $600
Million in New Funding." January 11. https://driveelectric.gov/news/new-cfi-funding.
—. 2023f. "Biden-Harris Administration to Invest $100 Million for EV Charger Reliability."
September 13. https://driveelectric.gov/news/ev-reliability-funding-opportunity.
—. 2023c. "First Public EV Charging Station Funded by NEVI Open in America." December 13.
https://driveelectric.gov/news/first-nevi-funded-stations-open.
—. 2024d. "FY23 Ride and Drive FOA DE-FOA-0002881." https://driveelectric.gov/files/ride-
and-drive-foa.pdf.
—. 2023g. "Joint Office Announces National Charging Experience Consortium." May 18.
https://driveelectric.gov/news/chargex-consortium.
—. 2024c. "New Funding Enhances EV Charging Resiliency, Reliability, Equity, and Workforce
Development." January 19. https://driveelectric.gov/news/workforce-development-ev-projects.
—. 2023d. "New York Continues NEVI Charging Station Momentum." December 15.
https://driveelectric.gov/news/new-york-NEVI-charging-station-momentum.
—. 2023b. "Private Sector Continues to Play Key Part in Accelerating Buildout of EV Charging
Networks." February 15. https://driveelectric.gOv/news/#private-investment.
—. 2023a. "SAE J3400 Charging Connector." https://driveelectric.gov/charging-connector.
—. 2023e. "State Plans for Electric Vehicle Charging." https://driveelectric.gov/state-plans/.
Kapadia, Vishal. 2023. "Leading the Charge: Walmart Announces Plan to Expand Electric
Vehicle Charging Network." April 6. https://corporate.walmart.com/news/2023/04/06/leading-
the-charge-walmart-announces-plan-to-expand-electric-vehicle-charging-network.
Kia. 2023. Enroll Now in the Electrify America Kia Charge Pass Program. Accessed 3 1, 2023.
http s://owners. ki a. com/content/owners/en/ki a-el ectrify. html.
Kosterev, D., Undrill, J. 2009. "Load Modeling in WECC." DOE-NERC FIDVR Workshop.
U.S. DOE, September.
Larsen, P., Lawson, M., LaCommare, K., and Eto, J. 2020. Severe Weather, Utility Spending,
and the Long-term Reliability of the U.S. Power System. Lawrence Berkeley National
Laboratory.
Larsen, P., Lawson, M., LaCommare, K., and Eto, J. 2020. Severe Weather, Utility Spending,
and the Long-term Reliability of the U.S. Power System. Berkeley : Lawrence Berkeley National
Laboratory.
Lipman, Timothy, Alissa Harrington, and Adam Langton. 2021. Total Charge Management of
Electric Vehicles, Publication Number: CEC-500-2021- 055. California Energy Commission.
5-69
-------
Accessed December 13, 2022. https://www.energy.ca.gov/sites/default/files/2021-12/CEC-500-
2021-055.pdf.
Maitra, Arindam. 2013. "Analysis of Projected Impact of Plug-in Electric Vehicles on the
Distribution Grid." In Frontiers of Engineering: Reports on Leading-Edge Engineering from the
2012 Symposium. The National Academies Press. Accessed December 13, 2022.
doi: 10.17226/18185.
McDonald, J. 2024. "IPM Input Data and Results for the LMDV Final Rule - Submission to
Docket No. EPA-HQ-OAR-2022-0829."
—. 2023. "PEV Regionalized Charge Demand for the FRM - Memo to the Docket."
NASA. 2024. Extreme Weather and Climate Change. 1 30. Accessed 2 82, 2024.
https://climate.nasa.gov/extreme-weather/.
National Association of Regulatory Utility Commissioners. 2022b. "Electric Vehicle
Interoperability Considerations for Public Utility Regulators - An addendum to the NARUC
report, Electric Vehicles: Key Trends, Issues, and Considerations for State Regulators (2019);
Villarreal, Christopher;." Accessed December 15, 2022. https://pubs.naruc.org/pub/D548E5DA-
1866-DAAC-99FB-70957246AEBE.
—. 2022a. "Electric Vehicles: Key Trends, Issues, and Considerations for State Regulators;
Harper, Charles; McAndrews, Gregory; Sass Byrnett, Danielle;." Accessed December 15, 2022.
https://pubs.naruc.org/pub/32857459-0005-B8C5-95C6-1920829CABFE.
—. 2022c. "Models for Incorporating Equity in Transportation Electrification Considerations for
Public Utility Regulators - An addendum to the NARUC report, Electric Vehicles: Key Trends,
Issues, and Considerations for State Regulators (2019); McAdams, Jasmine;." Accessed
December 15, 2022. https://pubs.naruc.org/pub/D4E4DD39-1866-DAAC-99FB-
4CE60044ACA7.
Nicholas, Michael. 2019. Estimating electric vehicle charging infrastructure costs across major
U.S. metropolitan areas. The International Council on Clean Transportation. Accessed January
11, 2023.
https://theicct.org/sites/default/files/publications/ICCT_EV_Charging_Cost_20190813.pdf.
NOAA. n.d. U.S. Hurricane Strikes by Decade. Accessed 2 28, 2024.
https://www.nhc.noaa.gov/pastdec.shtml.
NYPA. 2023. "EVolve NY's Mission: A Fast Electric Charging Station Near You."
http s://evol veny. ny pa. gov/ab out-evol ve-new-y ork.
Oreizi, Darya. 2022. "Burlington Electric Department Launches New Program with EVmatch to
Expand EV Charging at Multi-family Properties." September 30.
https://evmatch.com/blog/burlington-electric-department-launches-new-program-with-evmatch-
to-expand-ev-charging-at-multi-family-
properties/#:~:text=Burlington%20Electric%20Department%20(BED)%20recently,stations%20a
t%20multi%2Dfamily%20properties.
Public Law 117-169. 2022. (August 16). https://www.congress.gov/117/plaws/publl69/PLAW-
117publl69.pdf.
5-70
-------
Public Law 117-58. 2021. (November 15). https://www.govinfo.gov/content/pkg/PLAW-
117publ58/pdf/PLAW-l 17publ58.pdf.
Reuters. 2023. "Mercedes to launch vehicle-charging network, starting in North America."
January 6. Accessed January 10, 2023. https://www.reuters.com/business/autos-
transportation/mercedes-launch-vehicle-charging-network-starting-north-america-2023 -01 -05/.
—. 2023. "More automakers plug into Tesla's EV charging network." November 1.
https://www.reuters.com/business/autos-transportation/more-automakers-plug-into-teslas-ev-
charging-network-2023 -10-05/.
SAE. 2024. "WIP 2024-01-24 North American Charging System (NACS) for Electric Vehicles."
https ://www. sae. org/ standards/content/j 3400/.
Schey, Stephen, Kang-Ching Chu, and John Smart. 2022. Breakdown of Electric Vehicle Supply
Equipment Installation Costs. Idaho National Laboratory.
https://inldigitallibrary.inl.gov/sites/sti/sti/Sort_63124.pdf.
The White House. 2024. "FACT SHEET: Biden-Harris Administration Announces New Actions
to Cut Electric Vehicle Costs for Americans and Continue Building Out a Convenient, Reliable,
Made-in-America EV Charging Network." January 19. https://www.whitehouse.gov/briefing-
room/statements-releases/2024/01/19/fact-sheet-biden-harris-administration-announces-new-
actions-to-cut-electric-vehicle-costs-for-americans-and-continue-building-out-a-convenient-
reliable-made-in-america-ev-charging-ne.
—. 2023. "Fact Sheet: Biden-Harris Administration Announces New Standards and Major
Progress for a Made-in-Am erica National Network of EV Chargers." February 15. Accessed
February 27, 2023. https://www.whitehouse.gov/briefing-room/statements-
releases/2023/02/15/fact-sheet-biden-harris-administration-announces-new-standards-and-major-
progress-for-a-made-in-america-national-network-of-electric-vehicle-chargers/.
Tuffner, F., Undrill, J., Scoffield, D., Eto, J, Kosterev, D., and Quint, D. 2021. Distribution-
Level Impacts of Plug-in Electric Vehicle Charging on the Transmission System during Fault
Conditions. Richland: Pacific Northwest National Lab.
U.S. Department of Energy, Alternative Fuels Data Center. 2023d. "Electric Vehicle Charging
Station Locations." Accessed 2 28, 2023.
https://afdc.energy.gov/fuels/electricity_locations. html#/analyze?country=US&fuel=ELEC&ev_
networks=Tesla&ev_networks=Tesla%20Destination&ev_levels=dc_fast;
https://afdc.energy.gov/fuels/electricity_locations. html#/analyze?country=US&fuel=ELEC&ev_
networks=.
U.S. DOE. 2024. Building America's Clean Energy Future, https://www.energy.gov/invest.
—. 2023. "FOTW #1314, October 30, 2023: Manufacturers Have Announced Investments of
Over $500 million in More Than 40 American-Made Electric Vehicle Charger Plants." October
30. https://www.energy.gOv/eere/vehicles/articles/fotw-1314-october-30-2023-manufacturers-
have-announced-investments-over-500.
U.S. DOT, FHWA. 2022b. "Historic Step: All Fifty States Plus D.C. and Puerto Rico Greenlit to
Move EV Charging Networks Forward, Covering 75,000 Miles of Highway." September 27.
5-71
-------
Accessed January 10, 2023. https://highways.dot.gov/newsroom/historic-step-all-fifty-states-
plus-dc-and-puerto-rico-greenlit-move-ev-charging-networks.
—. 2022a. "The National Electric Vehicle Infrastructure (NEVI) Formula Program Guidance."
February 10. Accessed January 10, 2023.
https://www.fhwa.dot.gov/environment/alternative_fuel_corridors/nominations/90d_nevi_formul
a_program_gui dance. pdf.
U.S. Energy Information Administraton. 2019. "The National Energy Modeling System: An
Overview 2018." Independent Statistics and Analysis.
https://www.eia.gov/outlooks/aeo/nems/overview/pdf/0581(2018).pdf.
U.S. EPA. 2022a. "GHG Emissions Factors Hub, Table 1 - Stationary Combustion."
https://www.epa.gov/climateleadership/ghg-emission-factors-hub.
—. 2022b. "Interagency Agreement with U.S. DOE. OTAQ Transportation Research
Collaboration with NREL. EPA IA Identification No. DW-089-92534101-3."
—. 2023a. "Post-IRA 2022 Reference Case - EPA's Power Sector Modeling Platform v6 Using
IPM." https://www.epa.gov/power-sector-modeling/post-ira-2022-reference-case.
U.S. EPA. 2022c. "Regulatory Impact Analysis for Proposed Federal Implementation Plan
Addressing Regional Ozone Transport for the 2015 Ozone National Ambient Air Quality
Standard, EPA Docket No. EPA-HQ-OAR-2021 -0668-0151."
https://www.regulations.gov/document/EPA-HQ-OAR-2021-0668-0151.
—. 2023b. Retail Price Model. April 12. Accessed November 28, 2023.
https://www.epa.gov/power-sector-modeling/retail-price-model.
Valdes-Dapena, Peter. 2022. "GM to put thousands of electric vehicle chargers in rural
America." December 7. https://www.cnn.com/2022/12/07/business/gm-chargers/index.html.
VW. 2023. ID.4. Accessed 111, 2023. https://www.vw.com/en/models/id-4.html.
Warren, C. A., J. D. Bouford, R. D. Christie, D. Kowalewski, J. McDaniel, R. Robinson, D. J.
Schepers, J. Viglietta, and C. Williams. 2003. Classification of major event days. 2003 IEEE
Power Engineering Society General Meeting (IEEE Cat. No. 03CH37491), vol. 1, pp. 466-471.
IEEE, 2003.
Williams, B., Schmus, W., and Dawson, D. 1992. "Transmission voltage recovery delayed by
stalled air conditioner compressors." IEEE Transactions on Power Systems.
Wood, E., B. Borlaug, K. McKenna, J. Keen, B. Liu, J. Sun, D. Narang, et al. 2024. Multi-State
Transportation Electrification Impact Study. U.S. Department of Energy.
Wood, E., B. Borlaug, K. McKenna, J. Keen, B. Liu, N. Narang, L. Kiboma, et al. 2024. Multi-
State Transportation Electrification Impact Study. U.S. Department of Energy.
Wood, E., C. Rames, M. Muratori, and S. and Young, S. Raghavan. 2018. Charging Electric
Vehicles in Smart Cities: An EVI-Pro Analysis of Columbus, Ohio. Golden, CO: National
Renewable Energy Laboratory, https://www.nrel.gov/docs/fyl8osti/70367.pdf.
5-72
-------
Wood, Eric, Brennan Borlaug, Matt Moniot, Dong-Yeon Lee, Yanbo Ge, Fan Yang, and Zhaocai
Liu. 2023. The 2030 National Charging Network: Estimating U.S. Light-Duty Demand for
Electric Vehicle Charging Infrastructure. Golden, CO USA: National Renewable Energy
Laboratory, https://www.nrel.gov/docs/fy23osti/85654.pdf.
Wood, Eric, Clement Rames, Matteo. Muratori, Sesha. Raghavan, and Marc Melaina. 2017.
National Plug-In Electric Vehicle Infrastructure Analysis. U.S. Department of Energy Office of
Energy Efficiency & Renewable Energy. Accessed March 8, 2023.
https://www.nrel.gov/docs/fyl7osti/69031.pdf.
Zayer, Eric, Lucas Martin, Trent Murphey, Mary Stroncek, and Ingo Stein. 2022. EV Charging
Shifts into High Gear. Bain & Company. Accessed March 5, 2023.
https://www.bain.com/insights/electric-vehicle-charging-shifts-into-high-gear/.
5-73
-------
Chapter 6: Health and Welfare Impacts
This rule will impact emissions of GHGs, criteria pollutants, and air toxic pollutants. There
are health and welfare impacts associated with ambient concentrations of GHGs, criteria
pollutants, and air toxics which are described in this chapter.
6.1 Climate Change Impacts from GHG Emissions
Elevated concentrations of greenhouse gases (GHGs) have been warming the planet, leading
to changes in the Earth's climate that are occurring at a pace and in a way that threatens human
health, society, and the natural environment. While EPA is not making any new scientific or
factual findings with regard to the well-documented impact of GHG emissions on public health
and welfare in support of this rule, EPA is providing in this section a brief scientific background
on climate change to offer additional context for this rulemaking and to help the public
understand the environmental impacts of GHGs.
Extensive information on climate change is available in the scientific assessments and the
EPA documents that are briefly described in this section, as well as in the technical and scientific
information supporting them. One of those documents is EPA's 2009 Endangerment and Cause
or Contribute Findings for Greenhouse Gases Under section 202(a) of the CAA (74 FR 66496,
December 15, 2009). In the 2009 Endangerment Finding, the Administrator found under section
202(a) of the CAA that elevated atmospheric concentrations of six key well-mixed GHGs - CO2,
methane (CH4), nitrous oxide (N2O), HFCs, perfluorocarbons (PFCs), and sulfur hexafluoride
(SF6) - "may reasonably be anticipated to endanger the public health and welfare of current and
future generations" (74 FR at 66523). The 2009 Endangerment Finding, together with the
extensive scientific and technical evidence in the supporting record, documented that climate
change caused by human emissions of GHGs threatens the public health of the U.S. population.
It explained that by raising average temperatures, climate change increases the likelihood of heat
waves, which are associated with increased deaths and illnesses (74 FR 66497). While climate
change also increases the likelihood of reductions in cold-related mortality, evidence indicates
that the increases in heat mortality will be larger than the decreases in cold mortality in the U.S.
(74 FR 66525). The 2009 Endangerment Finding further explained that compared with a future
without climate change, climate change is expected to increase tropospheric ozone pollution over
broad areas of the U.S., including in the largest metropolitan areas with the worst tropospheric
ozone problems, and thereby increase the risk of adverse effects on public health (74 FR 66525).
Climate change is also expected to cause more intense hurricanes and more frequent and intense
storms of other types and heavy precipitation, with impacts on other areas of public health, such
as the potential for increased deaths, injuries, infectious and waterborne diseases, and stress-
related disorders (74 FR 66525). Children, the elderly, and the poor are among the most
vulnerable to these climate-related health effects (74 FR 66498).
The 2009 Endangerment Finding also documented, together with the extensive scientific and
technical evidence in the supporting record, that climate change touches nearly every aspect of
6-1
-------
public welfare196 in the U.S., including: Changes in water supply and quality due to changes in
drought and extreme rainfall events; increased risk of storm surge and flooding in coastal areas
and land loss due to inundation; increases in peak electricity demand and risks to electricity
infrastructure; and the potential for significant agricultural disruptions and crop failures (though
offset to some extent by carbon fertilization). These impacts are also global and may exacerbate
problems outside the U.S. that raise humanitarian, trade, and national security issues for the U.S.
(74 FR 66530).
In 2016, the Administrator issued a similar finding for GHG emissions from aircraft under
section 231(a)(2)(A) of the CAA. (81 FR 54422 2016) In the 2016 Endangerment Finding, the
Administrator found that the body of scientific evidence amassed in the record for the 2009
Endangerment Finding compellingly supported a similar endangerment finding under CAA
section 231(a)(2)(A), and also found that the science assessments released between the 2009 and
the 2016 Findings "strengthen and further support the judgment that GHGs in the atmosphere
may reasonably be anticipated to endanger the public health and welfare of current and future
generations" (81 FR 54424).
Since the 2016 Endangerment Finding, the climate has continued to change, with new
observational records being set for several climate indicators such as global average surface
temperatures, GHG concentrations, and sea level rise. Additionally, major scientific assessments
continue to be released that further advance our understanding of the climate system and the
impacts that GHGs have on public health and welfare both for current and future generations.
These updated observations and projections document the rapid rate of current and future climate
change both globally and in the U.S. (USGCRP 2017, USGCRP 2016, USGCRP 2018, IPCC
2018, IPCC 2019, IPCC 2019, IPCC 2013, NASEM 2016, NASEM 2017, NASEM 2019)
(Blunden and Boyer 2022, US EPA 2021, USGCRP 2023)
The most recent information demonstrates that the climate is continuing to change in response
to the human-induced buildup of GHGs in the atmosphere. These recent assessments show that
atmospheric concentrations of GHGs have risen to a level that has no precedent in human history
and that they continue to climb, primarily because of both historical and current anthropogenic
emissions, and that these elevated concentrations endanger our health by affecting our food and
water sources, the air we breathe, the weather we experience, and our interactions with the
natural and built environments. For example, atmospheric concentrations of one of these GHGs,
CO2, measured at Mauna Loa in Hawaii and at other sites around the world reached 419 parts per
million (ppm) in 2022 (nearly 50 percent higher than preindustrial levels)197 and have continued
to rise at a rapid rate. Global average temperature has increased by about 1.1 °C (2.0 °F) in the
2011-2020 decade relative to 1850-1900. (IPCC 2021) The years 2015-2021 were the warmest
7 years in the 1880-2021 record, contributing to the warmest decade on record with a decadal
temperature of 0.82 °C (1.48 °F) above the 20th century. (Blunden and Boyer 2022, NOAA
196 The CAA states in section 302(h) that "[a]ll language referring to effects on welfare includes, but is not limited
to, effects on soils, water, crops, vegetation, manmade materials, animals, wildlife, weather, visibility, and climate,
damage to and deterioration of property, and hazards to transportation, as well as effects on economic values and on
personal comfort and well-being, whether caused by transformation, conversion, or combination with other air
pollutants." 42 U.S.C. 7602(h).
197 https://gml.noaa.gov/webdata/ccgg/trends/CO2/CO2_annmean_mlo.txt.
6-2
-------
2022) The IPCC determined (with medium confidence) that this past decade was warmer than
any multi-century period in at least the past 100,000 years. (IPCC 2021) Global average sea level
has risen by about 8 inches (about 21 centimeters (cm)) from 1901 to 2018, with the rate from
2006 to 2018 (0.15 inches/year or 3.7 millimeters (mm)/year) almost twice the rate over the 1971
to 2006 period, and three times the rate of the 1901 to 2018 period. (IPCC 2021) The rate of sea
level rise over the 20th century was higher than in any other century in at least the last 2,800
years. (USGCRP 2018) Higher CO2 concentrations have led to acidification of the surface ocean
in recent decades to an extent unusual in the past 65 million years, with negative impacts on
marine organisms that use calcium carbonate to build shells or skeletons. (IPCC 2018) Arctic sea
ice extent continues to decline in all months of the year; the most rapid reductions occur in
September (very likely almost a 13 percent decrease per decade between 1979 and 2018) and are
unprecedented in at least 1,000 years. (IPCC 2021) Human-induced climate change has led to
heatwaves and heavy precipitation becoming more frequent and more intense, along with
increases in agricultural and ecological droughts198 in many regions. (IPCC 2021)
The assessment literature demonstrates that modest additional amounts of warming may lead
to a climate different from anything humans have ever experienced. The 2022 CO2 concentration
of 419 ppm is already higher than at any time in the last 2 million years.199 If concentrations
exceed 450 ppm, they would likely be higher than any time in the past 23 million years: (IPCC
2013) at the current rate of increase of more than 2 ppm a year, this would occur in about 15
years. While GHGs are not the only factor that controls climate, it is illustrative that 3 million
years ago (the last time CO2 concentrations were above 400 ppm) Greenland was not yet
completely covered by ice and still supported forests, while 23 million years ago (the last time
concentrations were above 450 ppm) the West Antarctic ice sheet was not yet developed,
indicating the possibility that high GHG concentrations could lead to a world that looks very
different from today and from the conditions in which human civilization has developed. If the
Greenland and Antarctic ice sheets were to melt substantially, sea levels would rise
dramatically—the IPCC estimated that over the next 2,000 years, sea levels will rise by 7 to 10
feet even if warming is limited to 1.5 °C (2.7 °F), from 7 to 20 feet if limited to 2 °C (3.6 °F),
and by 60 to 70 feet if warming is allowed to reach 5 °C (9 °F) above preindustrial levels. (IPCC
2021) For context, almost all of the city of Miami is less than 25 feet above sea level, and the 4th
National Climate Assessment (NCA4) stated that 13 million Americans would be at risk of
migration due to 6 feet of sea level rise.
The NCA4 found that it is very likely (greater than 90 percent likelihood) that by mid-
century, the Arctic Ocean will be almost entirely free of sea ice by late summer for the first time
in about 2 million years. (USGCRP 2018) Coral reefs will be at risk for almost complete (99
percent) losses with 1°C (1.8°F) of additional warming from today (2 °C or 3.6 °F since
preindustrial). At this temperature, between 8 and 18 percent of animal, plant, and insect species
could lose over half of the geographic area with suitable climate for their survival, and 7 to 10
percent of rangeland livestock would be projected to be lost. (IPCC 2018) The IPCC similarly
198 These are drought measures based on soil moisture.
199 Annual Mauna Loa CO2 concentration data from
https://gml.noaa.gov/webdata/ccgg/trends/CO2/CO2_annmean_mlo.txt, accessed September 9, 2023.
6-3
-------
found that climate change has caused substantial damages and increasingly irreversible losses in
terrestrial, freshwater, and coastal and open ocean marine ecosystems.
Every additional increment of temperature comes with consequences. For example, the half
degree of warming from 1.5 to 2 °C (0.9 °F of warming from 2.7 °F to 3.6 °F) above
preindustrial temperatures is projected on a global scale to expose 420 million more people to
frequent extreme heatwaves, and 62 million more people to frequent exceptional heatwaves
(where heatwaves are defined based on a heat wave magnitude index which takes into account
duration and intensity—using this index, the 2003 French heat wave that led to almost 15,000
deaths would be classified as an "extreme heatwave" and the 2010 Russian heatwave which led
to thousands of deaths and extensive wildfires would be classified as "exceptional"). It would
increase the frequency of sea-ice-free Arctic summers from once in 100 years to once in a
decade. It could lead to 4 inches of additional sea level rise by the end of the century, exposing
an additional 10 million people to risks of inundation as well as increasing the probability of
triggering instabilities in either the Greenland or Antarctic ice sheets. Between half a million and
a million additional square miles of permafrost would thaw over several centuries. Risks to food
security would increase from medium to high for several lower-income regions in the Sahel,
southern Africa, the Mediterranean, central Europe, and the Amazon. In addition to food security
issues, this temperature increase would have implications for human health in terms of increasing
ozone concentrations, heatwaves, and vector-borne diseases (for example, expanding the range
of the mosquitoes which carry dengue fever, chikungunya, yellow fever, and the Zika virus, or
the ticks which carry Lyme, babesiosis, or Rocky Mountain Spotted Fever). (IPCC 2018)
Moreover, every additional increment in warming leads to larger changes in extremes, including
the potential for events unprecedented in the observational record. Every additional degree will
intensify extreme precipitation events by about 7 percent. The peak winds of the most intense
tropical cyclones (hurricanes) are projected to increase with warming. In addition to a higher
intensity, the IPCC found that precipitation and frequency of rapid intensification of these storms
has already increased, the movement speed has decreased, and elevated sea levels have increased
coastal flooding, all of which make these tropical cyclones more damaging. (IPCC 2021)
The NCA4 also evaluated a number of impacts specific to the U.S. Severe drought and
outbreaks of insects like the mountain pine beetle have killed hundreds of millions of trees in the
western U.S. Wildfires have burned more than 3.7 million acres in 14 of the 17 years between
2000 and 2016, and Federal wildfire suppression costs were about a billion dollars annually.
(USGCRP 2018) The National Interagency Fire Center has documented U.S. wildfires since
1983, and the 10 years with the largest acreage burned have all occurred since 2004. (NIFC
2021) Wildfire smoke degrades air quality, increasing health risks, and more frequent and severe
wildfires due to climate change would further diminish air quality, increase incidences of
respiratory illness, impair visibility, and disrupt outdoor activities, sometimes thousands of miles
from the location of the fire. Meanwhile, sea level rise has amplified coastal flooding and erosion
impacts, requiring the installation of costly pump stations, flooding streets, and increasing storm
surge damages. Tens of billions of dollars of U.S. real estate could be below sea level by 2050
under some scenarios. Increased frequency and duration of drought will reduce agricultural
productivity in some regions, accelerate depletion of water supplies for irrigation, and expand the
distribution and incidence of pests and diseases for crops and livestock. The NCA4 also
recognized that climate change can increase risks to national security, both through direct
impacts on military infrastructure and by affecting factors such as food and water availability
6-4
-------
that can exacerbate conflict outside U.S. borders. Droughts, floods, storm surges, wildfires, and
other extreme events stress nations and people through loss of life, displacement of populations,
and impacts on livelihoods. (USGCRP 2018)
EPA modeling efforts can further illustrate how these impacts from climate change may be
experienced across the U.S. EPA's Framework for Evaluating Damages and Impacts (FrEDI)
uses information from over 30 peer-reviewed climate change impact studies to project the
physical and economic impacts of climate change to the U.S. resulting from future temperature
changes. (Hartin, Advancing the estimation of future climate impacts within the United States
2023, U.S. EPA 2022, U.S. Dept of State and U.S. Executive Office of the President 2021, OMB
2022) These impacts are projected for specific regions within the U.S. and for more than 20
impact categories, which span a large number of sectors of the U.S. economy. (U.S. EPA
2021)200 Using this framework, the EPA estimates that global emission projections, with no
additional mitigation, will result in significant climate-related damages to the U.S.201 These
damages to the U.S. would mainly be from increases in lives lost due to increases in
temperatures, as well as impacts to human health from increases in climate-driven changes in air
quality, dust and wildfire smoke exposure, and incidence of suicide. Additional major climate-
related damages would occur to U.S. infrastructure such as roads and rail, as well as
transportation impacts and coastal flooding from sea level rise, increases in property damage
from tropical cyclones, and reductions in labor hours worked in outdoor settings and buildings
without air conditioning. These impacts are also projected to vary from region to region with the
Southeast, for example, projected to see some of the largest damages from sea level rise, the
West Coast projected to experience damages from wildfire smoke more than other parts of the
country, and the Northern Plains states projected to see a higher proportion of damages to rail
and road infrastructure. While information on the distribution of climate impacts helps to better
understand the ways in which climate change may impact the U.S., recent analyses are still only
a partial assessment of climate impacts relevant to U.S. interests and in addition do not reflect
increased damages that occur due to interactions between different sectors impacted by climate
change or all the ways in which physical impacts of climate change occurring abroad have
spillover effects in different regions of the U.S.
Some GHGs also have impacts beyond those mediated through climate change. For example,
elevated concentrations of CO2 stimulate plant growth (which can be positive in the case of
beneficial species, but negative in terms of weeds and invasive species, and can also lead to a
reduction in plant micronutrients (Ziska, et al. 2016)) and cause ocean acidification. Nitrous
oxide depletes the levels of protective stratospheric ozone. (WMO 2018)
Transportation is the largest U.S. source of GHG emissions, representing 27 percent of total
GHG emissions. The GHG emission reductions resulting from compliance with this final rule
will significantly reduce the volume of GHG emissions from this sector. Section VI.D.2 of the
preamble discusses impacts of GHG emissions on individuals living in socially and economically
vulnerable communities. While EPA did not conduct modeling to specifically quantify changes
in climate impacts resulting from this rule in terms of avoided temperature change or sea-level
200Available at https://www.epa.gov/cira/fredi. Documentation has been subject to both a public review comment
period and an independent expert peer review, following EPA peer-review guidelines.
201 Compared to a world with no additional warming after the model baseline (1986-2005).
6-5
-------
rise, we did quantify climate benefits by monetizing the emission reductions through the
application of the social cost of greenhouse gases (SC-GHGs), as described in Section VILA of
the preamble.
These scientific assessments, the EPA analyses, and documented observed changes in the
climate of the planet and of the U.S. present clear support regarding the current and future
dangers of climate change and the importance of GHG emissions mitigation.
6.2 Climate Benefits
The EPA estimates the climate benefits of GHG emissions reductions expected from the final
rule using estimates of the social cost of greenhouse gases (SC-GHG) that reflect recent
advances in the scientific literature on climate change and its economic impacts and incorporate
recommendations made by the National Academies of Science, Engineering, and Medicine.
(National Academies 2017) The EPA published and used these estimates in the RIA for the
December 2023 Final Oil and Gas NSPS/EG Rulemaking, "Standards of Performance for New,
Reconstructed, and Modified Sources and Emissions Guidelines for Existing Sources: Oil and
Natural Gas Sector Climate Review." The EPA solicited public comment on the methodology
and use of these estimates in the RIA for the agency's December 2022 Oil and Gas NSPS/EG
Supplemental Proposal and has conducted an external peer review of these estimates, as
described further below.
The SC-GHG is the monetary value of the net harm to society associated with a marginal
increase in GHG emissions in a given year, or the benefit of avoiding that increase. In principle,
SC-GHG includes the value of all climate change impacts (both negative and positive), including
(but not limited to) changes in net agricultural productivity, human health effects, property
damage from increased flood risk and natural disasters, disruption of energy systems, risk of
conflict, environmental migration, and the value of ecosystem services. The SC-GHG, therefore,
reflects the societal value of reducing emissions of the gas in question by one metric ton and is
the theoretically appropriate value to use in conducting benefit-cost analyses of policies that
affect GHG emissions. In practice, data and modeling limitations restrain the ability of SC-GHG
estimates to include all physical, ecological, and economic impacts of climate change, implicitly
assigning a value of zero to the omitted climate damages. The estimates are, therefore, a partial
accounting of climate change impacts and likely underestimate the marginal benefits of
abatement.
Since 2008, the EPA has used estimates of the social cost of various greenhouse gases (i.e.,
SC-CO2, SC-CH4, and SC-N2O), collectively referred to as the "social cost of greenhouse gases"
(SC-GHG), in analyses of actions that affect GHG emissions. The values used by the EPA from
2009 to 2016, and since 2021 - including in the proposal for this rulemaking - have been
consistent with those developed and recommended by the IWG on the SC-GHG; and the values
used from 2017 to 2020 were consistent with those required by E.O. 13783, which disbanded the
IWG. During 2015-2017, the National Academies conducted a comprehensive review of the SC-
CO2 and issued a final report in 2017 recommending specific criteria for future updates to the
SC-CO2 estimates, a modeling framework to satisfy the specified criteria, and both near-term
updates and longer-term research needs pertaining to various components of the estimation
process. (National Academies 2017) The IWG was reconstituted in 2021 and E.O. 13990
directed it to develop a comprehensive update of its SC-GHG estimates, recommendations
regarding areas of decision-making to which SC-GHG should be applied, and a standardized
6-6
-------
review and updating process to ensure that the recommended estimates continue to be based on
the best available economics and science going forward.
The EPA is a member of the IWG and is participating in the IWG's work under E.O. 13990.
As noted in previous EPA RIAs, while that process continues, the EPA is continuously
reviewing developments in the scientific literature on the SC-GHG, including more robust
methodologies for estimating damages from emissions, and looking for opportunities to further
improve SC-GHG estimation.202 As EPA noted in the LMDV NPRM, in the December 2022 Oil
and Gas NSPS/EG Supplemental Proposal RIA the Agency included a sensitivity analysis of the
climate benefits of the Supplemental Proposal using a new set of SC-GHG estimates that
incorporates recent research addressing recommendations of the National Academies (National
Academies 2017) in addition to using the interim SC-GHG estimates presented in the Technical
Support Document: Social Cost of Carbon, Methane, and Nitrous Oxide Interim Estimates under
Executive Order 13990 (IWG 2021) that the IWG recommended for use until updated estimates
that address the National Academies' recommendations are available.
The EPA solicited public comment on the sensitivity analysis and the accompanying draft
technical report, External Review Draft of Report on the Social Cost of Greenhouse Gases:
Estimates Incorporating Recent Scientific Advances, which explains the methodology underlying
the new set of estimates, in the December 2022 Supplemental Oil and Gas Proposal. The
response to comments document can be found in the docket for that action.
To ensure that the methodological updates adopted in the technical report are consistent with
economic theory and reflect the latest science, the EPA also initiated an external peer review
panel to conduct a high-quality review of the technical report, completed in May 2023. See 88
FR at 26075/2 noting this peer review process. The peer reviewers commended the agency on its
development of the draft update, calling it a much-needed improvement in estimating the SC-
GHG and a significant step toward addressing the National Academies' recommendations with
defensible modeling choices based on current science. The peer reviewers provided numerous
recommendations for refining the presentation and for future modeling improvements, especially
with respect to climate change impacts and associated damages that are not currently included in
the analysis. Additional discussion of omitted impacts and other updates have been incorporated
in the technical report to address peer reviewer recommendations. Complete information about
the external peer review, including the peer reviewer selection process, the final report with
individual recommendations from peer reviewers, and the EPA's response to each
recommendation is available on EPA's website (EPA 2023f).
The remainder of this section provides an overview of the methodological updates
incorporated into the SC-GHG estimates used in this final RIA. A more detailed explanation of
each input and the modeling process is provided in the final technical report, EPA Report on the
Social Cost of Greenhouse Gases: Estimates Incorporating Recent Scientific Advances.
Appendix to Chapter 9 shows the benefits of the final rule using the interim SC-GHG (IWG
2021) estimates presented in the proposal.
202 EPA strives to base its analyses on the best available science and economics, consistent with its responsibilities,
for example, under the Information Quality Act.
6-7
-------
The steps necessary to estimate the SC-GHG with a climate change integrated assessment
model (IAM) can generally be grouped into four modules: socioeconomics and emissions,
climate, damages, and discounting. The emissions trajectories from the socioeconomic module
are used to project future temperatures in the climate module. The damage module then
translates the temperature and other climate endpoints (along with the projections of
socioeconomic variables) into physical impacts and associated monetized economic damages,
where the damages are calculated as the amount of money the individuals experiencing the
climate change impacts would be willing to pay to avoid them. To calculate the marginal effect
of emissions, i.e., the SC-GHG in year "t," the entire model is run twice - first as a baseline and
second with an additional pulse of emissions in year "t." After recalculating the temperature
effects and damages expected in all years beyond "t" resulting from the adjusted path of
emissions, the losses are discounted to a present value in the discounting module. Many sources
of uncertainty in the estimation process are incorporated using Monte Carlo techniques by taking
draws from probability distributions that reflect the uncertainty in parameters.
The SC-GHG estimates used by the EPA and many other federal agencies since 2009 have
relied on an ensemble of three widely used IAMs: Dynamic Integrated Climate and Economy
(DICE) (W. Nordhaus 2010) Climate Framework for Uncertainty, Negotiation, and Distribution
(FUND) (Anthoff and Tol 2013); (Anthoff and Tol 2013b) and Policy Analysis of the
Greenhouse Gas Effect. (PAGE) (Hope 2013) In 2010, the IWG harmonized key inputs across
the IAMs, but all other model features were left unchanged, relying on the model developers'
best estimates and judgments. That is, the representation of climate dynamics and damage
functions included in the default version of each IAM as used in the published literature was
retained.
The SC-GHG estimates in this RIA no longer rely on the three IAMs (i.e., DICE, FUND, and
PAGE) used in previous SC-GHG estimates. As explained previously, EPA uses a modular
approach to estimate the SC-GHG, consistent with the National Academies' (National
Academies 2017) near-term recommendations. That is, the methodology underlying each
component, or module, of the SC-GHG estimation process is developed by drawing on the latest
research and expertise from the scientific disciplines relevant to that component. Under this
approach, each step in the SC-GHG estimation improves consistency with the current state of
scientific knowledge, enhances transparency, and allows for more explicit representation of
uncertainty.
The socioeconomic and emissions module relies on a new set of probabilistic projections for
population, income, and GHG emissions developed under the Resources for the Future (RFF)
Social Cost of Carbon Initiative (K. P. Rennert 2021) (Rennert, Prest, et al. 2022a). These
socioeconomic projections (hereinafter collectively referred to as the RFF-SPs) are an internally
consistent set of probabilistic projections of population, GDP, and GHG emissions (CO2, CH4,
and N2O) to 2300. Based on a review of available sources of long-run projections necessary for
damage calculations, the RFF-SPs stand out as being most consistent with the National
Academies' recommendations. Consistent with the National Academies' recommendation, the
RFF-SPs were developed using a mix of statistical and expert elicitation techniques to capture
uncertainty in a single probabilistic approach, taking into account the likelihood of future
emissions mitigation policies and technological developments, and provide the level of
disaggregation necessary for damage calculations. Unlike other sources of projections, they
provide inputs for estimation out to 2300 without further extrapolation assumptions. Conditional
6-8
-------
on the modeling conducted for the SC-GHG estimates, this time horizon is far enough in the
future to capture the majority of discounted climate damages. Including damages beyond 2300
would increase the estimates of the SC-GHG. As discussed in U.S. EPA (EPA 2023f) the use of
the RFF-SPs allows for capturing economic growth uncertainty within the discounting module.
The climate module relies on the Finite Amplitude Impulse Response (FaIR) model, (Smith,
et al. 2018, IPCC, Climate Change 2021 - The Physical Science Basis 2021, Millar, et al. 2017) a
widely used Earth system model which captures the relationships between GHG emissions,
atmospheric GHG concentrations, and global mean surface temperature. The FaIR model was
originally developed by Richard Millar, Zeb Nicholls, and Myles Allen at Oxford University, as
a modification of the approach used in IPCC AR5 to assess the GWP and GTP (Global
Temperature Potential) of different gases. It is open source, widely used (e.g., (IPCC 2018);
(IPCC 2021a) and was highlighted by the National Academies (National Academies 2017) as a
model that satisfies their recommendations for a near-term update of the climate module in SC-
GHG estimation. Specifically, it translates GHG emissions into mean surface temperature
response and represents the current understanding of the climate and GHG cycle systems and
associated uncertainties within a probabilistic framework. The SC-GHG estimates used in this
RIA rely on FaIR version 1.6.2 as used by the IPCC (IPCC 2021). It provides, with high
confidence, an accurate representation of the latest scientific consensus on the relationship
between global emissions and global mean surface temperature and offers a code base that is
fully transparent and available online. The uncertainty capabilities in FaIR 1.6.2 have been
calibrated to the most recent assessment of the IPCC (which importantly narrowed the range of
likely climate sensitivities relative to prior assessments). See U.S. EPA (EPA 2023f) for more
details.
The socioeconomic projections and outputs of the climate module are inputs into the damage
module to estimate monetized future damages from climate change.203 The National Academies'
recommendations for the damage module, scientific literature on climate damages, updates to
models that have been developed since 2010, as well as the public comments received on
individual EPA rulemakings and the IWG's February 2021 TSD, have all helped to identify
available sources of improved damage functions. The IWG (e.g., (IWG 2010) (IWG 2016a)
(IWG 2021)), the National Academies (2017), comprehensive studies (e.g., (Rose, et al. 2014)),
and public comments have all recognized that the damages functions underlying the IWG SC-
GHG estimates used since 2013 (taken from DICE 2010 (W. Nordhaus 2010); FUND 3.8
(Anthoff and Tol 2013b); (Anthoff and Tol 2013); and PAGE 2009 (Hope 2013)) do not include
all of the important physical, ecological, and economic impacts of climate change. The climate
change literature and the science underlying the economic damage functions have evolved, and
DICE 2010, FUND 3.8, and PAGE 2009 now lag behind the most recent research.
203 In addition to temperature change, two of the three damage modules used in the SC-GHG estimation require
global mean sea level (GMSL) projections as an input to estimate coastal damages. Those two damage modules use
different models for generating estimates of GMSL. Both are based off reduced complexity models that can use the
FaIR temperature outputs as inputs to the model and generate projections of GMSL accounting for the contributions
of thermal expansion and glacial and ice sheet melting based on recent scientific research. Absent clear evidence on
a preferred model, the SC-GHG estimates presented in this RIA retain both methods used by the damage module
developers. See (EPA 2023f) for more details.
6-9
-------
The challenges involved with updating damage functions have been widely recognized.
Functional forms and calibrations are constrained by the available literature and need to
extrapolate beyond warming levels or locations studied in that literature. Research and public
resources focused on understanding how these physical changes translate into economic impacts
have been significantly less than the resources focused on modeling and improving our
understanding of climate system dynamics and the physical impacts from climate change
(Auffhammer 2018). Even so, there has been a large increase in research on climate impacts and
damages in the time since DICE 2010, FUND 3.8, and PAGE 2009 were published. Along with
this growth, there continues to be wide variation in methodologies and scope of studies, such that
care is required when synthesizing the current understanding of impacts or damages. Based on a
review of available studies and approaches to damage function estimation, the EPA uses three
separate damage functions to form the damage module: (1) a subnational-scale, sectoral damage
function (based on the Data-driven Spatial Climate Impact Model (DSCIM) developed by the
Climate Impact Lab (CIL 2023) (Carleton 2022) (Rode, et al. 2021); (2) a country-scale, sectoral
damage function (based on the Greenhouse Gas Impact Value Estimator (GIVE) model
developed under RFF's Social Cost of Carbon Initiative (Rennert, Errickson, et al. 2022); (3) and
a meta-analysis-based damage function (based on (Howard and Sterner 2017)).
The damage functions in DSCIM and GIVE represent substantial improvements relative to
the damage functions underlying the SC-GHG estimates used by the EPA to date and reflect the
forefront of scientific understanding about how temperature change and SLR lead to monetized
net (market and nonmarket) damages for several categories of climate impacts. The models'
spatially explicit and impact-specific modeling of relevant processes allow for improved
understanding and transparency about mechanisms through which climate impacts are occurring
and how each damage component contributes to the overall results, consistent with the National
Academies' recommendations. DSCIM addresses common criticisms related to the damage
functions underlying current SC-GHG estimates (e.g., (Pindyck 2017)) by developing multi-
sector, empirically grounded damage functions. The damage functions in the GIVE model offer a
direct implementation of the National Academies' near-term recommendation to develop
updated sectoral damage functions that are based on recently published work and reflective of
the current state of knowledge about damages in each sector. Specifically, the National
Academies noted that "[t]he literature on agriculture, mortality, coastal damages, and energy
demand provide immediate opportunities to update the [models]" (National Academies 2017),
which are the four damage categories currently in GIVE. A limitation of both models is that the
sectoral coverage is still limited, and even the categories that are represented are incomplete.
Neither DSCIM nor GIVE yet accommodate estimation of several categories of temperature
driven climate impacts (e.g., morbidity, conflict, migration, biodiversity loss) and only represent
a limited subset of damages from changes in precipitation. For example, while precipitation is
considered in the agriculture sectors in both DSCIM and GIVE, neither model takes into account
impacts of flooding, changes in rainfall from tropical storms, and other precipitation related
impacts. As another example, the coastal damage estimates in both models do not fully reflect
the consequences of SLR-driven salt-water intrusion and erosion, or SLR damages to coastal
tourism and recreation. Other missing elements are damages that result from other physical
impacts (e.g., ocean acidification, non-temperature-related mortality such as diarrheal disease
and malaria) and the many feedbacks and interactions across sectors and regions that can lead to
6-10
-------
additional damages.204 See U.S. EPA (EPA 2023f) for more discussion of omitted damage
categories and other modeling limitations. DSCIM and GIVE do account for the most commonly
cited benefits associated with CO2 emissions and climate change - CO2 crop fertilization and
declines in cold related mortality. As such, while the GIVE- and DSCIM-based results provide
state-of-the-science assessments of key climate change impacts, they remain partial estimates of
future climate damages resulting from incremental changes in CO2, CH4, and N2O.205
Finally, given the still relatively narrow sectoral scope of the recently developed DSCIM and
GIVE models, the damage module includes a third damage function that reflects a synthesis of
the state of knowledge in other published climate damages literature. Studies that employ meta-
analytic techniques206 offer a tractable and straightforward way to combine the results of
multiple studies into a single damage function that represents the body of evidence on climate
damages that pre-date CIL and RFF's research initiatives. The first use of meta-analysis to
combine multiple climate damage studies was done by (Tol 2009) and included 14 studies. The
studies in (Tol 2009) served as the basis for the global damage function in DICE starting in
version 2013R (W. Nordhaus 2014). The damage function in the most recent published version
of DICE, DICE 2016, is from an updated meta-analysis based on a rereview of existing damage
studies and included 26 studies published over 1994-2013. Howard and Sterner provide a more
recent published peer-reviewed meta-analysis of existing damage studies (published through
2016) and account for additional features of the underlying studies (Howard and Sterner 2017).
This study addresses differences in measurement across studies by adjusting estimates such that
the data are relative to the same base period. They also eliminate double counting by removing
duplicative estimates. Howard and Sterner's final sample is drawn from 20 studies that were
published through 2015. (Howard and Sterner 2017) They present results under several
specifications and show that the estimates are somewhat sensitive to defensible alternative
modeling choices. As discussed in detail in U.S. EPA (EPA 2023f), the damage module
underlying the SC-GHG estimates in this RIA includes the damage function specification (that
excludes duplicate studies) from (Howard and Sterner 2017) that leads to the lowest SC-GHG
estimates, all else equal.
The discounting module discounts the stream of future net climate damages to its present
value in the year when the additional unit of emissions was released. Given the long-time
horizon over which the damages are expected to occur, the discount rate has a large influence on
the present value of future damages. Consistent with the findings of (National Academies 2017),
the economic literature, OMB Circular A-4's guidance for regulatory analysis, and IWG
recommendations to date (IWG 2010) (IWG 2013) (IWG 2016a) (IWG 2016b) (IWG 2021), the
EPA continues to conclude that the consumption rate of interest is the theoretically appropriate
204 The one exception is that the agricultural damage function in DSCIM and GIVE reflects the ways that trade can
help mitigate damages arising from crop yield impacts.
205 One advantage of the modular approach used by these models is that future research on new or alternative
damage functions can be incorporated in a relatively straightforward way. DSCIM and GIVE developers have work
underway on other impact categories that may be ready for consideration in future updates (e.g., morbidity and
biodiversity loss).
206 Meta-analysis is a statistical method of pooling data and/or results from a set of comparable studies of a problem.
Pooling in this way provides a larger sample size for evaluation and allows for a stronger conclusion than can be
provided by any single study. Meta-analysis yields a quantitative summary of the combined results and current state
of the literature.
6-11
-------
discount rate to discount the future benefits of reducing GHG emissions and that discount rate
uncertainty should be accounted for in selecting future discount rates in this intergenerational
context. OMB's Circular A-4 points out that "the analytically preferred method of handling
temporal differences between benefits and costs is to adjust all the benefits and costs to reflect
their value in equivalent units of consumption and to discount them at the rate consumers and
savers would normally use in discounting future consumption benefits" (OMB 2003). The
damage module described above calculates future net damages in terms of reduced consumption
(or monetary consumption equivalents), and so an application of this guidance is to use the
consumption discount rate to calculate the SC-GHG. Thus, EPA concludes that the use of the
social rate of return on capital (7 percent under current OMB Circular A-4 guidance), which does
not reflect the consumption rate, to discount damages estimated in terms of reduced consumption
would inappropriately underestimate the impacts of climate change for the purposes of
estimating the SC-GHG.207
For the SC-GHG estimates used in this RIA, EPA relies on a dynamic discounting approach
that more fully captures the role of uncertainty in the discount rate in a manner consistent with
the other modules. Based on a review of the literature and data on consumption discount rates,
the public comments received on individual EPA rulemakings, the February 2021 TSD (IWG
2021), and the (National Academies 2017)208 recommendations for updating the discounting
module, the SC-GHG estimates rely on discount rates that reflect more recent data on the
consumption interest rate and uncertainty in future rates. Specifically, rather than using a
constant discount rate, the evolution of the discount rate over time is defined following the latest
empirical evidence on interest rate uncertainty and using a framework originally developed by
(Ramsey 1928) that connects economic growth and interest rates. The Ramsey approach
explicitly reflects (1) preferences for utility in one period relative to utility in a later period and
(2) the value of additional consumption as income changes. The dynamic discount rates used to
develop the SC-GHG estimates applied in this RIA have been calibrated following the (Newell,
Pizer and Prest 2022) approach, as applied in (Rennert, Errickson, et al. 2022) (Rennert, Prest, et
al. 2022a). This approach uses the discounting formula (Ramsey 1928) in which the parameters
are calibrated such that (1) the decline in the certainty-equivalent discount rate matches the latest
empirical evidence on interest rate uncertainty estimated by (Bauer and Rudebusch 2020) (Bauer
and Rudebusch 2023) and (2) the average of the certainty-equivalent discount rate over the first
decade matches a near-term consumption rate of interest. Uncertainty in the starting rate is
addressed by using three near-term target rates (1.5, 2.0, and 2.5 percent) based on multiple lines
of evidence on observed market interest rates.
The resulting dynamic discount rate provides a notable improvement over the constant
discount rate framework used for SC-GHG estimation in previous EPA RIAs. Specifically, it
provides internal consistency within the modeling and a more complete accounting of
uncertainty consistent with economic theory (Arrow, et al. 2013) (Cropper, et al. 2014) and the
(National Academies 2017) recommendation to employ a more structural, Ramsey-like approach
207 See also the discussion of the inappropriateness of discounting consumption-equivalent measures of benefits and
costs using a rate of return on capital in Circular A-4 (2023) (OMB 2003).
208 Similarly, OMB's Circular A-4 (2023) points out that "The analytically preferred method of handling temporal
differences between benefits and costs is to adjust all the benefits and costs to reflect their value in equivalent units
of consumption before discounting them" (OMB 2003)
6-12
-------
to discounting that explicitly recognizes the relationship between economic growth and
discounting uncertainty. This approach is also consistent with the (National Academies 2017)
recommendation to use three sets of Ramsey parameters that reflect a range of near-term
certainty-equivalent discount rates and are consistent with theory and empirical evidence on
consumption rate uncertainty. Finally, the value of aversion to risk associated with net damages
from GHG emissions is explicitly incorporated into the modeling framework following the
economic literature. See U.S. EPA (EPA 2023f) for a more detailed discussion of the entire
discounting module and methodology used to value risk aversion in the SC-GHG estimates.
Taken together, the methodologies adopted in this SC-GHG estimation process allow for a
more holistic treatment of uncertainty than past estimates used by the EPA. The updates
incorporate a quantitative consideration of uncertainty into all modules and use a Monte Carlo
approach that captures the compounding uncertainties across modules. The estimation process
generates nine separate distributions of discounted marginal damages per metric ton - the
product of using three damage modules and three near-term target discount rates - for each gas
in each emissions year. These distributions have long right tails reflecting the extensive evidence
in the scientific and economic literature that shows the potential for lower-probability but higher-
impact outcomes from climate change, which would be particularly harmful to society. The
uncertainty grows over the modeled time horizon. Therefore, under cases with a lower near-term
target discount rate - that give relatively more weight to impacts in the future - the distribution
of results is wider. To produce a range of estimates that reflects the uncertainty in the estimation
exercise while also providing a manageable number of estimates for policy analysis, the EPA
combines the multiple lines of evidence on damage modules by averaging the results across the
three damage module specifications. The full results generated from the updated methodology
for methane and other greenhouse gases (SC-CO2, SC-CH4, and SC-N2O) for emissions years
2020 through 2080 are provided in U.S. EPA (EPA 2023f).
Table 6-1 summarizes the resulting averaged certainty-equivalent SC-GHG estimates under
each near-term discount rate that are used to estimate the climate benefits of the GHG emission
reductions expected from the final rule. These estimates are reported in 2022 dollars but are
otherwise identical to those presented in U.S. EPA (EPA 2023f). The SC-GHGs increase over
time within the models — i.e., the societal harm from one metric ton emitted in 2030 is higher
than the harm caused by one metric ton emitted in 2027 — because future emissions produce
larger incremental damages as physical and economic systems become more stressed in response
to greater climatic change, and because GDP is growing over time and many damage categories
are modeled as proportional to GDP.
6-13
-------
Table 6-1: Annual Rounded SC-COi, SC-CH4, and SC-N2O Values, 2027-2055.
SC-GHG and Near-term Ramsey Discount Rate
SC-CO2 SC-CH4 SC-N2O
(2022 dollars per metric ton (2022 dollars per metric ton (2022 dollars per metric ton of
ofC02) of CH4) N2Q)
Emission
Near-term rate
Near-term rate
Near-term rate
Year
2.5%
2.0%
1.5%
2.5%
2.0%
1.5%
2.5%
2.0%
1.5%
2027
150
250
410
1900
2400
3200
47000
70000
110000
2028
160
250
420
2000
2500
3300
48000
72000
110000
2029
160
250
430
2000
2600
3400
49000
73000
110000
2030
160
260
430
2100
2600
3500
50000
74000
120000
203 1
160
260
440
2200
2700
3600
51000
76000
120000
2032
170
270
440
2300
2800
3700
52000
77000
120000
2033
170
270
450
2400
2900
3800
53000
79000
120000
2034
170
270
450
2500
3000
4000
54000
80000
120000
2035
180
280
460
2500
3100
4100
55000
81000
120000
2036
180
280
460
2600
3200
4200
57000
83000
130000
2037
180
290
470
2700
3300
4300
58000
84000
130000
2038
190
290
470
2800
3400
4400
59000
86000
130000
2039
190
290
480
2900
3500
4500
60000
87000
130000
2040
190
300
480
3000
3600
4600
61000
88000
130000
2041
200
300
490
3100
3700
4800
62000
90000
140000
2042
200
310
490
3200
3800
4900
63000
91000
140000
2043
200
310
500
3300
3900
5000
65000
93000
140000
2044
210
320
500
3400
4100
5100
66000
95000
140000
2045
210
320
510
3500
4200
5200
67000
96000
140000
2046
210
330
520
3500
4300
5400
69000
98000
150000
2047
220
330
520
3600
4400
5500
70000
99000
150000
2048
220
340
530
3700
4500
5600
70000
100000
150000
2049
230
340
530
3800
4600
5700
72000
100000
150000
2050
230
340
540
3900
4700
5800
73000
100000
150000
2051
230
350
550
4000
4800
6000
75000
100000
150000
2052
240
350
550
4100
4900
6100
76000
110000
160000
2053
240
360
560
4200
5000
6200
77000
110000
160000
2054
240
360
560
4300
5100
6300
78000
110000
160000
2055
250
360
570
4400
5200
6400
79000
110000
160000
Source: (EPA 2023f)
Note: These SC-CH4 values are identical to those reported in the technical report U.S. EPA (2023f) adjusted for inflation to 2022 dollars
using the annual GDP Implicit Price Deflator values in the U.S. Bureau of Economic Analysis' (BEA) NIPA Table 1.1.9 (Bureau of
Economic Analysis (BEA) 2021). The values are stated in $/metric ton GHG and vary depending on the year of GHG emissions. This table
displays the values rounded to two significant figures. The annual unrounded values used in the calculations in this RIA are available in
Appendix A.5 of U.S. EPA (EPA 2023f) and at: www.epa.gov/environmental-economics/scghg.
The methodological updates described above represent a major step forward in bringing SC-
GHG estimation closer to the frontier of climate science and economics and address many of the
(National Academies 2017) near-term recommendations. Nevertheless, the resulting SC-GHG
estimates presented in Table 9-1, still have several limitations, as would be expected for any
modeling exercise that covers such a broad scope of scientific and economic issues across a
complex global landscape. There are still many categories of climate impacts and associated
damages that are only partially or not reflected yet in these estimates and sources of uncertainty
that have not been fully characterized due to data and modeling limitations. For example, the
6-14
-------
modeling omits most of the consequences of changes in precipitation, damages from extreme
weather events, the potential for nongradual damages from passing critical thresholds (e.g.,
tipping elements) in natural or socioeconomic systems, and non-climate mediated effects of
GHG emissions. More specifically for methane, the SC-CH4 estimates do not account for the
direct health and welfare impacts associated with tropospheric ozone produced by methane. As
discussed further in U.S. EPA (EPA 2023f), recent studies have found the global ozone-related
respiratory mortality benefits of CH4 emissions reductions, which are not included in the SC-CH4
values presented in Table 7-1, to be, in 2022 dollars, approximately $2,700 per metric ton of
methane emissions in 2030. (McDuffie, et al. 2023). In addition, the SC-CH4 estimates do not
reflect that methane emissions lead to a reduction in atmospheric oxidants, like hydroxyl
radicals, nor do they account for impacts associated with CO2 produced from methane oxidizing
in the atmosphere. Importantly, the updated SC-GHG methodology does not yet reflect
interactions and feedback effects within, and across, Earth and human systems. For example, it
does not explicitly reflect potential interactions among damage categories, such as those
stemming from the interdependencies of energy, water, and land use. These, and other,
interactions and feedbacks were highlighted by the National Academies as an important area of
future research for longer-term enhancements in the SC-GHG estimation framework.
6.3 Health Effects Associated with Exposure to Criteria and Air Toxics Pollutants
Emissions sources impacted by this rule, including vehicles and power plants, emit pollutants
that contribute to ambient concentrations of ozone, PM, NO2, SO2, CO, and air toxics. This
chapter of the RIA discusses the health effects associated with exposure to these pollutants.
Additionally, because children have increased vulnerability and susceptibility for adverse
health effects related to air pollution exposures, EPA's findings regarding adverse effects for
children related to exposure to pollutants that are impacted by this rule are noted in this section.
The increased vulnerability and susceptibility of children to air pollution exposures may arise
because infants and children generally breathe more relative to their size than adults do, and
consequently may be exposed to relatively higher amounts of air pollution. (U.S. EPA 2009)
Children also tend to breathe through their mouths more than adults, and their nasal passages are
less effective at removing pollutants, which leads to greater lung deposition of some pollutants,
such as PM. (U.S. EPA 2019) (Foos, et al. 2008) Furthermore, air pollutants may pose health
risks specific to children because children's bodies are still developing (U.S. EPA 2021).209 For
example, during periods of rapid growth such as fetal development, infancy, and puberty, their
developing systems and organs may be more easily harmed. (U.S. EPA 2006, U.S. EPA 2005)
EPA produces the report titled "America's Children and the Environment," which presents
national trends on air pollution and other contaminants and environmental health of children.
(U.S. EPA 2022)
209 Children's environmental health includes conception, infancy, early childhood and through adolescence until 21
years of age as described in an EPA Memorandum (US EPA 2021).
6-15
-------
6.3.1 Particulate Matter
6.3.1.1 Background on Particulate Matter
Particulate matter (PM) is a complex mixture of solid particles and liquid droplets distributed
among numerous atmospheric gases which interact with solid and liquid phases. Particles in the
atmosphere range in size from less than 0.01 to more than 10 micrometers (|im) in diameter.
(U.S. EPA 2020) Atmospheric particles can be grouped into several classes according to their
aerodynamic diameter and physical sizes. Generally, the three broad classes of particles include
ultrafine particles (UFPs, generally considered as particles with a diameter less than or equal to
0.1 |im [typically based on physical size, thermal diffusivity, or electrical mobility]), "fine"
particles (PM2.5; particles with a nominal mean aerodynamic diameter less than or equal to 2.5
|im), and "thoracic" particles (PM10; particles with a nominal mean aerodynamic diameter less
than or equal to 10 |im). Particles that fall within the size range between PM2.5 and PM10, are
referred to as "thoracic coarse particles" (PM10-2.5, particles with a nominal mean aerodynamic
diameter greater than 2.5 |im and less than or equal to 10 |im). EPA currently has NAAQS for
PM2.5 and PM10 (40 CFR Part 50 2023, 40 CFR Part 53 2023, 40 CFR Part 58 2023).210
Most particles are found in the lower troposphere, where they can have residence times
ranging from a few hours to weeks. Particles are removed from the atmosphere by wet
deposition, such as when they are carried by rain or snow, or by dry deposition, when particles
settle out of suspension due to gravity. Atmospheric lifetimes are generally longest for PM2.5,
which often remains in the atmosphere for days to weeks before being removed by wet or dry
deposition. (U.S. EPA 2019) In contrast, atmospheric lifetimes for UFP and PM10-2.5 are shorter.
Within hours, UFP can undergo coagulation and condensation that lead to formation of larger
particles in the accumulation mode, or can be removed from the atmosphere by evaporation,
deposition, or reactions with other atmospheric components. PM10-2.5 are also generally removed
from the atmosphere within hours, through wet or dry deposition. (U.S. EPA 2019)
Particulate matter consists of both primary and secondary particles. Primary particles are
emitted directly from sources, such as combustion-related activities (e.g., industrial activities,
motor vehicle operation, biomass burning), while secondary particles are formed through
atmospheric chemical reactions of gaseous precursors (e.g., sulfur oxides (SOx), nitrogen oxides
(NOx), and volatile organic compounds (VOCs)). From 2000 to 2021, national annual average
ambient PM2.5 concentrations have declined by over 35 percent,211 largely reflecting reductions
in emissions of precursor gases.
6.3.1.2 Health Effects Associated with Exposure to Particulate Matter
Scientific evidence spanning animal toxicological, controlled human exposure, and
epidemiologic studies shows that exposure to ambient PM is associated with a broad range of
health effects. These health effects are discussed in detail in the Integrated Science Assessment
210 Regulatory definitions of PM size fractions, and information on reference and equivalent methods for measuring
PM in ambient air, are provided in 40 CFR parts 50, 53, and 58. With regard to national ambient air quality
standards (NAAQS) which provide protection against health and welfare effects, the 24-hour PM10 standard
provides protection against effects associated with short-term exposure to thoracic coarse particles (i.e., PM10-2.5).
211 See https://www.epa.gov/air-trends/particulate-matter-pm25-trends for more information.
6-16
-------
for Particulate Matter, which was finalized in December 2019 (2019 PM ISA) with a more
targeted evaluation of studies published since the literature cutoff date of the 2019 PM ISA in the
Supplement to the Integrated Science Assessment for PM (Supplement). (U.S. EPA 2019) (U.S.
EPA 2022) The PM ISA characterizes the causal nature of relationships between PM exposure
and broad health categories (e.g., cardiovascular effects, respiratory effects, etc.) using a weight-
of-evidence approach. (U.S. EPA 2019) Within this characterization, the PM ISA summarizes
the health effects evidence for short-term (i.e., hours up to one month) and long-term (i.e., one
month to years) exposures to PM2.5, PM10-2.5, and ultrafine particles, and concludes that
exposures to ambient PM2.5 are associated with a number of adverse health effects. The
following discussion highlights the PM ISA's conclusions and summarizes additional
information from the Supplement where appropriate, pertaining to the health effects evidence for
both short- and long-term PM exposures. Further discussion of PM-related health effects can also
be found in the 2022 Policy Assessment for the review of the PM NAAQS. (U.S. EPA 2022)
EPA has concluded that recent evidence in combination with evidence evaluated in the 2009
PM ISA supports a "causal relationship" between both long- and short-term exposures to PM2.5
and premature mortality and cardiovascular effects and a "likely to be causal relationship"
between long- and short-term PM2.5 exposures and respiratory effects. (U.S. EPA 2009)
Additionally, recent experimental and epidemiologic studies provide evidence supporting a
"likely to be causal relationship" between long-term PM2.5 exposure and nervous system effects,
and long-term PM2.5 exposure and cancer. Because of remaining uncertainties and limitations in
the evidence base, EPA determined the evidence is "suggestive of, but not sufficient to infer, a
causal relationship" for long-term PM2.5 exposure and reproductive and developmental effects
(i.e., male/female reproduction and fertility; pregnancy and birth outcomes), long- and short-term
exposures and metabolic effects, and short-term exposure and nervous system effects.
As discussed extensively in the 2019 PM ISA and the Supplement, recent studies continue to
support a "causal relationship" between short- and long-term PM2.5 exposures and mortality.
(U.S. EPA 2019) (U.S. EPA 2022) For short-term PM2.5 exposure, multi-city studies, in
combination with single- and multi-city studies evaluated in the 2009 PM ISA, provide evidence
of consistent, positive associations across studies conducted in different geographic locations,
populations with different demographic characteristics, and studies using different exposure
assignment techniques. Additionally, the consistent and coherent evidence across scientific
disciplines for cardiovascular morbidity, including exacerbations of chronic obstructive
pulmonary disease (COPD) and asthma, provide biological plausibility for cause-specific
mortality and total mortality. Recent epidemiologic studies evaluated in the Supplement,
including studies that employed alternative methods for confounder control, provide additional
support to the evidence base that contributed to the 2019 PM ISA conclusion for short-term
PM2.5 exposure and mortality (U.S. EPA 2022).
The 2019 PM ISA concluded a "causal relationship" between long-term PM2.5 exposure and
mortality. In addition to re-analyses and extensions of the American Cancer Society (ACS) and
Harvard Six Cities (HSC) cohorts, multiple new cohort studies conducted in the U.S. and
Canada, consisting of people employed in a specific job (e.g., teacher, nurse) and that apply
different exposure assignment techniques, provide evidence of positive associations between
long-term PM2.5 exposure and mortality. Biological plausibility for mortality due to long-term
PM2.5 exposure is provided by the coherence of effects across scientific disciplines for
cardiovascular morbidity, particularly for coronary heart disease, stroke, and atherosclerosis, and
6-17
-------
for respiratory morbidity, particularly for the development of COPD. Additionally, recent studies
provide evidence indicating that as long-term PM2.5 concentrations decrease there is an increase
in life expectancy. Recent cohort studies evaluated in the Supplement, as well as epidemiologic
studies that conducted accountability analyses or employed alternative methods for confounder
controls, support and extend the evidence base that contributed to the 2019 PM ISA conclusion
for long-term PM2.5 exposure and mortality.
A large body of studies examining both short- and long-term PM2.5 exposure and
cardiovascular effects supports and extends the evidence base evaluated in the 2009 PM ISA.
The strongest evidence for cardiovascular effects in response to short-term PM2.5 exposures is for
ischemic heart disease and heart failure. The evidence for short-term PM2.5 exposure and
cardiovascular effects is coherent across scientific disciplines and supports a continuum of
effects ranging from subtle changes in indicators of cardiovascular health to serious clinical
events, such as increased emergency department visits and hospital admissions due to
cardiovascular disease and cardiovascular mortality. For long-term PM2.5 exposure, there is
strong and consistent epidemiologic evidence of a relationship with cardiovascular mortality.
This evidence is supported by epidemiologic and animal toxicological studies demonstrating a
range of cardiovascular effects including coronary heart disease, stroke, impaired heart function,
and subclinical markers (e.g., coronary artery calcification, atherosclerotic plaque progression),
which collectively provide coherence and biological plausibility. Recent epidemiologic studies
evaluated in the Supplement, as well as studies that conducted accountability analyses or
employed alternative methods for confounder control, support and extend the evidence base that
contributed to the 2019 PM ISA conclusion for both short- and long-term PM2.5 exposure and
cardiovascular effects.
Studies evaluated in the 2019 PM ISA continue to provide evidence of a "likely to be causal
relationship" between both short- and long-term PM2.5 exposure and respiratory effects.
Epidemiologic studies provide consistent evidence of a relationship between short-term PM2.5
exposure and asthma exacerbation in children and COPD exacerbation in adults, as indicated by
increases in emergency department visits and hospital admissions, which is supported by animal
toxicological studies indicating worsening allergic airways disease and subclinical effects related
to COPD. Epidemiologic studies also provide evidence of a relationship between short-term
PM2.5 exposure and respiratory mortality. However, there is inconsistent evidence of respiratory
effects, specifically lung function declines and pulmonary inflammation, in controlled human
exposure studies. With respect to long-term PM2.5 exposure, epidemiologic studies conducted in
the U.S. and abroad provide evidence of a relationship with respiratory effects, including
consistent changes in lung function and lung function growth rate, increased asthma incidence,
asthma prevalence, and wheeze in children; acceleration of lung function decline in adults; and
respiratory mortality. The epidemiologic evidence is supported by animal toxicological studies,
which provide coherence and biological plausibility for a range of effects including impaired
lung development, decrements in lung function growth, and asthma development.
Since the 2009 PM ISA, a growing body of scientific evidence examined the relationship
between long-term PM2.5 exposure and nervous system effects, resulting for the first time in a
causality determination for this health effects category of a "likely to be causal relationship." The
strongest evidence for effects on the nervous system come from epidemiologic studies that
consistently report cognitive decrements and reductions in brain volume in adults. The effects
observed in epidemiologic studies in adults are supported by animal toxicological studies
6-18
-------
demonstrating effects on the brain of adult animals including inflammation, morphologic
changes, and neurodegeneration of specific regions of the brain. There is more limited evidence
for neurodevelopmental effects in children with some studies reporting positive associations with
autism spectrum disorder and others providing limited evidence of an association with cognitive
function. While there is some evidence from animal toxicological studies indicating effects on
the brain (i.e., inflammatory and morphological changes) to support a biologically plausible
pathway for neurodevelopmental effects, epidemiologic studies are limited due to their lack of
control for potential confounding by copollutants, the small number of studies conducted, and
uncertainty regarding critical exposure windows.
Building off the decades of research demonstrating mutagenicity, DNA damage, and other
endpoints related to genotoxicity due to whole PM exposures, recent experimental and
epidemiologic studies focusing specifically on PM2.5 provide evidence of a relationship between
long-term PM2.5 exposure and cancer. Epidemiologic studies examining long-term PM2.5
exposure and lung cancer incidence and mortality provide evidence of generally positive
associations in cohort studies spanning different populations, locations, and exposure assignment
techniques. Additionally, there is evidence of positive associations with lung cancer incidence
and mortality in analyses limited to never smokers. The epidemiologic evidence is supported by
both experimental and epidemiologic evidence of genotoxicity, epigenetic effects, carcinogenic
potential, and that PM2.5 exhibits several characteristics of carcinogens, which collectively
provide biological plausibility for cancer development and resulted in the conclusion of a "likely
to be causal relationship."
For the additional health effects categories evaluated for PM2.5 in the 2019 PM ISA,
experimental and epidemiologic studies provide limited and/or inconsistent evidence of a
relationship with PM2.5 exposure. As a result, the 2019 PM ISA concluded that the evidence is
"suggestive of, but not sufficient to infer a causal relationship" for short-term PM2.5 exposure
and metabolic effects and nervous system effects, and for long-term PM2.5 exposures and
metabolic effects as well as reproductive and developmental effects.
In addition to evaluating the health effects attributed to short- and long-term exposure to
PM2.5, the 2019 PM ISA also conducted an extensive evaluation as to whether specific
components or sources of PM2.5 are more strongly related with health effects than PM2.5 mass.
An evaluation of those studies resulted in the 2019 PM ISA concluding that "many PM2.5
components and sources are associated with many health effects, and the evidence does not
indicate that any one source or component is consistently more strongly related to health effects
than PM2.5 mass." (U.S. EPA 2019)
For both PM10-2.5 and ultrafine particles (UFPs), for all health effects categories evaluated, the
2019 PM ISA concluded that the evidence was "suggestive of, but not sufficient to infer, a causal
relationship" or "inadequate to determine the presence or absence of a causal relationship." For
PM10-2.5, although a Federal Reference Method (FRM) was instituted in 2011 to measure PM10-2.5
concentrations nationally, the causality determinations reflect that the same uncertainty identified
in the 2009 PM ISA with respect to the method used to estimate PM10-2.5 concentrations in
epidemiologic studies persists. Specifically, across epidemiologic studies, different approaches
are used to estimate PM10-2.5 concentrations (e.g., direct measurement of PM10-2.5, difference
between PM10 and PM2.5 concentrations), and it remains unclear how well correlated PM10-2.5
concentrations are both spatially and temporally across the different methods used.
6-19
-------
For UFPs, which have often been defined as particles less than 0.1 |im in diameter, the
uncertainty in the evidence for the health effect categories evaluated across experimental and
epidemiologic studies reflects the inconsistency in the exposure metric used (i.e., particle number
concentration, surface area concentration, mass concentration) as well as the size fractions
examined. In epidemiologic studies the size fraction examined can vary depending on the
monitor used and exposure metric, with some studies examining number count over the entire
particle size range, while experimental studies that use a particle concentrator often examine
particles up to 0.3 |im. Additionally, due to the lack of a monitoring network, there is limited
information on the spatial and temporal variability of UFPs within the United States as well as
population exposures to UFPs, which adds uncertainty to epidemiologic study results.
The 2019 PM ISA cites extensive evidence indicating that "both the general population as
well as specific populations and life stages are at risk for PM2.5-related health effects." (U.S.
EPA 2019) For example, in support of its "causal" and "likely to be causal" determinations, the
ISA cites substantial evidence for (1) PM-related mortality and cardiovascular effects in older
adults; (2) PM-related cardiovascular effects in people with pre-existing cardiovascular disease;
(3) PM-related respiratory effects in people with pre-existing respiratory disease, particularly
asthma exacerbations in children; and (4) PM-related impairments in lung function growth and
asthma development in children. The ISA additionally notes that stratified analyses (i.e., analyses
that directly compare PM-related health effects across groups) provide strong evidence for racial
and ethnic differences in PM2.5 exposures and in the risk of PM2.5-related health effects,
specifically within Hispanic and non-Hispanic Black populations with some evidence of
increased risk for populations of low socioeconomic status. Recent studies evaluated in the
Supplement support the conclusion of the 2019 PM ISA with respect to disparities in both PM2.5
exposure and health risk by race and ethnicity and provide additional support for disparities for
populations of lower socioeconomic status. (U.S. EPA 2022) Additionally, evidence spanning
epidemiologic studies that conducted stratified analyses, experimental studies focusing on animal
models of disease or individuals with pre-existing disease, dosimetry studies, as well as studies
focusing on differential exposure suggest that populations with pre-existing cardiovascular or
respiratory disease, populations that are overweight or obese, populations that have particular
genetic variants, and current/former smokers could be at increased risk for adverse PM2.5-related
health effects. The 2022 Policy Assessment for the review of the PM NAAQS also highlights
that factors that may contribute to increased risk of PM2.5-related health effects include life stage
(children and older adults), pre-existing diseases (cardiovascular disease and respiratory disease),
race/ethnicity, and socioeconomic status. (U.S. EPA 2022)
6.3.2 Ozone
6.3.2.1 Background on Ozone
Ground-level ozone pollution forms in areas with high concentrations of ambient nitrogen
oxides (NOx) and volatile organic compounds (VOCs) when solar radiation is strong. Major U.S.
sources of NOx are highway and nonroad motor vehicles, engines, power plants, and other
industrial sources; natural sources, such as soil, vegetation, and lightning, serve as smaller
sources. Vegetation is the dominant source of VOCs in the U.S. Volatile consumer and
commercial products, such as propellants and solvents, highway and nonroad vehicles, engines,
fires, and industrial sources also contribute to the atmospheric burden of VOCs at ground-level.
EPA currently has NAAQS for ozone (40 CFR Part 50 2023, 40 CFR Part 58 2023).
6-20
-------
The processes underlying ozone formation, transport, and accumulation are complex.
Ground-level ozone is produced and destroyed by an interwoven network of free radical
reactions involving the hydroxyl radical (OH), NO, NO2, and complex reaction intermediates
derived from VOCs. Many of these reactions are sensitive to temperature and available sunlight.
High ozone events most often occur when ambient temperatures and sunlight intensities remain
high for several days under stagnant conditions. Ozone and its precursors can also be transported
hundreds of miles downwind, which can lead to elevated ozone levels in areas with otherwise
low VOC or NOx emissions. As an air mass moves and is exposed to changing ambient
concentrations of NOx and VOCs, the ozone photochemical regime (relative sensitivity of ozone
formation to NOx and VOC emissions) can change.
When ambient VOC concentrations are high, comparatively small amounts of NOx catalyze
rapid ozone formation. Without available NOx, ground-level ozone production is severely
limited, and VOC reductions would have little impact on ozone concentrations. Photochemistry
under these conditions is said to be "NOx-limited." When NOx levels are sufficiently high, faster
NO2 oxidation consumes more radicals, dampening ozone production. Under these "VOC-
limited" conditions (also referred to as "NOx-saturated" conditions), VOC reductions are
effective in reducing ozone, and NOx can react directly with ozone, resulting in suppressed
ozone concentrations near NOx emission sources. Under these NOx-saturated conditions, NOx
reductions can actually increase local ozone under certain circumstances, but overall ozone
production (considering downwind formation) decreases, and even in VOC-limited areas, NOx
reductions are not expected to increase ozone levels if the NOx reductions are sufficiently large
enough to become NOx-limited.
6.3.2.2 Health Effects Associated with Exposure to Ozone
This section provides a summary of the health effects associated with exposure to ambient
concentrations of ozone.212 The information in this section is based on the information and
conclusions in the April 2020 Integrated Science Assessment for Ozone (Ozone ISA). (U.S. EPA
2020) The Ozone ISA concludes that human exposures to ambient concentrations of ozone are
associated with a number of adverse health effects and characterizes the weight of evidence for
these health effects.213 The following discussion highlights the Ozone ISA's conclusions
pertaining to health effects associated with both short-term and long-term periods of exposure to
ozone.
For short-term exposure to ozone, the Ozone ISA concludes that respiratory effects, including
lung function decrements, pulmonary inflammation, exacerbation of asthma, respiratory-related
hospital admissions, and mortality, are causally associated with ozone exposure. It also
concludes that metabolic effects, including metabolic syndrome (i.e., changes in insulin or
212 Human exposure to ozone varies over time due to changes in ambient ozone concentration and because people
move between locations which have notably different ozone concentrations. Also, the amount of ozone delivered to
the lung is influenced not only by the ambient concentrations but also by the breathing route and rate.
213 The ISA evaluates evidence and draws conclusions on the causal relationship between relevant pollutant
exposures and health effects, assigning one of five "weight of evidence" determinations: causal relationship, likely
to be a causal relationship, suggestive of a causal relationship, inadequate to infer a causal relationship, and not
likely to be a causal relationship. For more information on these levels of evidence, please refer to Table II in the
Preamble of the ISA.
6-21
-------
glucose levels, cholesterol levels, obesity, and blood pressure) and complications due to diabetes
are likely to be causally associated with short-term exposure to ozone and that evidence is
suggestive of a causal relationship between cardiovascular effects, central nervous system
effects, and total mortality and short-term exposure to ozone.
For long-term exposure to ozone, the Ozone ISA concludes that respiratory effects, including
new onset asthma, pulmonary inflammation, and injury, are likely to be causally related with
ozone exposure. The Ozone ISA characterizes the evidence as suggestive of a causal relationship
for associations between long-term ozone exposure and cardiovascular effects, metabolic effects,
reproductive and developmental effects, central nervous system effects, and total mortality. The
evidence is inadequate to infer a causal relationship between chronic ozone exposure and
increased risk of cancer.
Finally, interindividual variation in human responses to ozone exposure can result in some
groups being at increased risk for detrimental effects in response to exposure. In addition, some
groups are at increased risk of exposure due to their activities, such as outdoor workers and
children. The Ozone ISA identified several groups that are at increased risk for ozone-related
health effects. These groups are people with asthma, children and older adults, individuals with
reduced intake of certain nutrients (i.e., Vitamins C and E), outdoor workers, and individuals
having certain genetic variants related to oxidative metabolism or inflammation. Ozone exposure
during childhood can have lasting effects through adulthood. Such effects include altered
function of the respiratory and immune systems. Children absorb higher doses (normalized to
lung surface area) of ambient ozone, compared to adults, due to their increased time spent
outdoors, higher ventilation rates relative to body size, and a tendency to breathe a greater
fraction of air through the mouth. Children also have a higher asthma prevalence compared to
adults. Recent epidemiologic studies provide generally consistent evidence that long-term ozone
exposure is associated with the development of asthma in children. Studies comparing age
groups reported higher magnitude associations for short-term ozone exposure and respiratory
hospital admissions and emergency room visits among children than among adults. Panel studies
also provide support for experimental studies with consistent associations between short-term
ozone exposure and lung function and pulmonary inflammation in healthy children. Additional
children's vulnerability and susceptibility factors are listed in Section IX.G of the preamble.
6.3.3 Nitrogen Oxides
6.3.3.1 Background on Nitrogen Oxides
Oxides of nitrogen (NOx) refers to nitric oxide (NO) and nitrogen dioxide (NO2). Most NO2
is formed in the air through the oxidation of nitric oxide (NO) emitted when fuel is burned at a
high temperature. NOx is a criteria pollutant, regulated for its adverse effects on public health
and the environment, and highway vehicles are an important contributor to NOx emissions. NOx,
along with VOCs, are the two major precursors of ozone, and NOx is also a major contributor to
secondary PM2.5 formation.
6.3.3.2 Heath Effects Associated with Exposure to Nitrogen Oxides
The most recent review of the health effects of oxides of nitrogen completed by EPA can be
found in the 2016 Integrated Science Assessment for Oxides of Nitrogen - Health Criteria
(Oxides of Nitrogen ISA). (U.S. EPA 2016) The largest source of NO2 is motor vehicle
6-22
-------
emissions, and ambient NO2 concentrations tend to be highly correlated with other traffic-related
pollutants. Thus, a key issue in characterizing the causality of NCh-health effect relationships
was evaluating the extent to which studies supported an effect of NO2 that is independent of
other traffic-related pollutants. EPA concluded that the findings for asthma exacerbation
integrated from epidemiologic and controlled human exposure studies provided evidence that is
sufficient to infer a causal relationship between respiratory effects and short-term NO2 exposure.
The strongest evidence supporting an independent effect of NO2 exposure comes from controlled
human exposure studies demonstrating increased airway responsiveness in individuals with
asthma following ambient-relevant NO2 exposures. The coherence of this evidence with
epidemiologic findings for asthma hospital admissions and ED visits as well as lung function
decrements and increased pulmonary inflammation in children with asthma describe a plausible
pathway by which NO2 exposure can cause an asthma exacerbation. The 2016 ISA for Oxides of
Nitrogen also concluded that there is likely to be a causal relationship between long-term NO2
exposure and respiratory effects. This conclusion is based on new epidemiologic evidence for
associations of NO2 with asthma development in children combined with biological plausibility
from experimental studies.
In evaluating a broader range of health effects, the 2016 ISA for Oxides of Nitrogen
concluded that evidence is "suggestive of, but not sufficient to infer, a causal relationship"
between short-term NO2 exposure and cardiovascular effects and mortality and between long-
term NO2 exposure and cardiovascular effects and diabetes, birth outcomes, and cancer. In
addition, the scientific evidence is inadequate (insufficient consistency of epidemiologic and
toxicological evidence) to infer a causal relationship for long-term NO2 exposure with fertility,
reproduction, and pregnancy, as well as with postnatal development. A key uncertainty in
understanding the relationship between these non-respiratory health effects and short- or long-
term exposure to NO2 is co-pollutant confounding, particularly by other roadway pollutants. The
available evidence for non-respiratory health effects does not adequately address whether NO2
has an independent effect or whether it primarily represents effects related to other or a mixture
of traffic-related pollutants.
The 2016 ISA for Oxides of Nitrogen concluded that people with asthma, children, and older
adults are at increased risk for N02-related health effects. In these groups and lifestages, NO2 is
consistently related to larger effects on outcomes related to asthma exacerbation, for which there
is confidence in the relationship with NO2 exposure.
6.3.4 Sulfur Oxides
6.3.4.1 Background on Sulfur Oxides
Sulfur dioxide (SO2), a member of the sulfur oxide (SOx) family of gases, is formed from
burning fuels containing sulfur (e.g., coal or oil), extracting gasoline from oil, or extracting
metals from ore. SO2 and its gas phase oxidation products can dissolve in water droplets and
further oxidize to form sulfuric acid which reacts with ammonia to form sulfates, which are
important components of ambient PM.
6.3.4.2 Health Effects Associated with Exposure to Sulfur Oxides
This section provides an overview of the health effects associated with SO2. Additional
information on the health effects of SO2 can be found in the 2017 Integrated Science Assessment
6-23
-------
for Sulfur Oxides - Health Criteria (SOx ISA). (U.S. EPA 2017) Following an extensive
evaluation of health evidence from animal toxicological, controlled human exposure, and
epidemiologic studies, EPA has concluded that there is a causal relationship between respiratory
health effects and short-term exposure to SO2. The immediate effect of SO2 on the respiratory
system in humans is bronchoconstriction. People with asthma are more sensitive to the effects of
SO2, likely resulting from preexisting inflammation associated with this disease. In addition to
those with asthma (both children and adults), there is suggestive evidence that all children and
older adults may be at increased risk of S02-related health effects. In free-breathing laboratory
studies involving controlled human exposures to SO2, respiratory effects have consistently been
observed following 5-10 min exposures at SO2 concentrations > 400 ppb in people with asthma
engaged in moderate to heavy levels of exercise, with respiratory effects occurring at
concentrations as low as 200 ppb in some individuals with asthma. A clear concentration-
response relationship has been demonstrated in these studies following exposures to SO2 at
concentrations between 200 and 1000 ppb, both in terms of increasing severity of respiratory
symptoms and decrements in lung function, as well as the percentage of individuals with asthma
adversely affected. Epidemiologic studies have reported positive associations between short-term
ambient SO2 concentrations and hospital admissions and emergency department visits for asthma
and for all respiratory causes, particularly among children and older adults (> 65 years). The
studies provide supportive evidence for the causal relationship.
For long-term SO2 exposure and respiratory effects, EPA has concluded that the evidence is
suggestive of a causal relationship. This conclusion is based on new epidemiologic evidence for
positive associations between long-term SO2 exposure and increases in asthma incidence among
children, together with animal toxicological evidence that provides a pathophysiologic basis for
the development of asthma. However, uncertainty remains regarding the influence of other
pollutants on the observed associations with SO2 because these epidemiologic studies have not
examined the potential for co-pollutant confounding.
Consistent associations between short-term exposure to SO2 and mortality have been observed
in epidemiologic studies, with larger effect estimates reported for respiratory mortality than for
cardiovascular mortality. While this finding is consistent with the demonstrated effects of SO2 on
respiratory morbidity, uncertainty remains with respect to the interpretation of these observed
mortality associations due to potential confounding by various copollutants. Therefore, EPA has
concluded that the overall evidence is suggestive of a causal relationship between short-term
exposure to SO2 and mortality.
6.3.5 Carbon Monoxide
6.3.5.1 Background on Carbon Monoxide
Carbon Monoxide (CO) is a colorless, odorless gas formed by incomplete combustion of
carbon-containing fuels and by photochemical reactions in the atmosphere. Nationally,
particularly in urban areas, the majority of CO emissions to ambient air come from mobile
sources. (U.S. EPA 2010)
6.3.5.2 Health Effects Associated with Exposure to Carbon Monoxide
Information on the health effects of carbon monoxide (CO) can be found in the January 2010
Integrated Science Assessment for Carbon Monoxide (CO ISA). (U.S. EPA 2010) The CO ISA
6-24
-------
presents conclusions regarding the presence of causal relationships between CO exposure and
categories of adverse health effects.214 This section provides a summary of the health effects
associated with exposure to ambient concentrations of CO, along with the CO ISA
conclusions.215
Controlled human exposure studies of subjects with coronary artery disease show a decrease
in the time to onset of exercise-induced angina (chest pain) and electrocardiogram changes
following CO exposure. In addition, epidemiologic studies presented in the CO ISA observed
associations between short-term CO exposure and cardiovascular morbidity, particularly
increased emergency room visits and hospital admissions for coronary heart disease (including
ischemic heart disease, myocardial infarction, and angina). Some epidemiologic evidence is also
available for increased hospital admissions and emergency room visits for congestive heart
failure and cardiovascular disease as a whole. The CO ISA concludes that a causal relationship is
likely to exist between short-term exposures to CO and cardiovascular morbidity. It also
concludes that available data are inadequate to conclude that a causal relationship exists between
long-term exposures to CO and cardiovascular morbidity.
Animal studies show various neurological effects with in utero CO exposure. Controlled
human exposure studies report central nervous system and behavioral effects following low-level
CO exposures, although the findings have not been consistent across all studies. The CO ISA
concludes that the evidence is suggestive of a causal relationship with both short- and long-term
exposure to CO and central nervous system effects.
A number of studies cited in the CO ISA have evaluated the role of CO exposure in birth
outcomes such as preterm birth or cardiac birth defects. There is limited epidemiologic evidence
of a CO-induced effect on preterm births and birth defects, with weak evidence for a decrease in
birth weight. Animal toxicological studies have found perinatal CO exposure to affect birth
weight, as well as other developmental outcomes. The CO ISA concludes that the evidence is
suggestive of a causal relationship between long-term exposures to CO and developmental
effects and birth outcomes.
Epidemiologic studies provide evidence of associations between short-term CO
concentrations and respiratory morbidity such as changes in pulmonary function, respiratory
symptoms, and hospital admissions. A limited number of epidemiologic studies considered
copollutants such as ozone, SO2, and PM in two-pollutant models and found that CO risk
estimates were generally robust, although this limited evidence makes it difficult to disentangle
effects attributed to CO itself from those of the larger complex air pollution mixture. Controlled
human exposure studies have not extensively evaluated the effect of CO on respiratory
morbidity. Animal studies at levels of 50-100 ppm CO show preliminary evidence of altered
pulmonary vascular remodeling and oxidative injury. The CO ISA concludes that the evidence is
214 The ISA evaluates the health evidence associated with different health effects, assigning one of five "weight of
evidence" determinations: causal relationship, likely to be a causal relationship, suggestive of a causal relationship,
inadequate to infer a causal relationship, and not likely to be a causal relationship. For definitions of these levels of
evidence, please refer to Section 1.6 of the ISA.
215 Personal exposure includes contributions from many sources, and in many different environments. Total personal
exposure to CO includes both ambient and non-ambient components; and both components may contribute to
adverse health effects.
6-25
-------
suggestive of a causal relationship between short-term CO exposure and respiratory morbidity,
and inadequate to conclude that a causal relationship exists between long-term exposure and
respiratory morbidity.
Finally, the CO ISA concludes that the epidemiologic evidence is suggestive of a causal
relationship between short-term concentrations of CO and mortality. Epidemiologic evidence
suggests an association exists between short-term exposure to CO and mortality, but limited
evidence is available to evaluate cause-specific mortality outcomes associated with CO exposure.
In addition, the attenuation of CO risk estimates that was often observed in copollutant models
contributes to the uncertainty as to whether CO is acting alone or as an indicator for other
combustion-related pollutants. The CO ISA also concludes that there is not likely to be a causal
relationship between relevant long-term exposures to CO and mortality.
6.3.6 Diesel Exhaust
6.3.6.1 Background on Diesel Exhaust
Diesel exhaust is a complex mixture composed of particulate matter, carbon dioxide, oxygen,
nitrogen, water vapor, carbon monoxide, nitrogen compounds, sulfur compounds, and numerous
low-molecular-weight hydrocarbons. A number of these gaseous hydrocarbon components are
individually known to be toxic, including aldehydes, benzene, and 1,3-butadiene. The diesel
particulate matter present in diesel exhaust consists mostly of fine particles (< 2.5 |im), of which
a significant fraction is ultrafine particles (< 0.1 |im). These particles have a large surface area
which makes them an excellent medium for adsorbing organics and their small size makes them
highly respirable. Many of the organic compounds present in the gases and on the particles, such
as polycyclic organic matter, are individually known to have mutagenic and carcinogenic
properties.
Diesel exhaust varies significantly in chemical composition and particle sizes between
different engine types (heavy-duty, light-duty), engine operating conditions (idle, acceleration,
deceleration), and fuel formulations (high/low sulfur fuel). Also, there are emissions differences
between onroad and nonroad engines because the nonroad engines are generally of older
technology. After being emitted in the engine exhaust, diesel exhaust undergoes dilution as well
as chemical and physical changes in the atmosphere. The lifetimes of the components present in
diesel exhaust range from seconds to months.
6.3.6.2 Health Effects Associated with Exposure to Diesel Exhaust
In EPA's 2002 Diesel Health Assessment Document (Diesel HAD), exposure to diesel
exhaust was classified as likely to be carcinogenic to humans by inhalation from environmental
exposures, in accordance with the revised draft 1996/1999 EPA cancer guidelines. (U.S. EPA
1999, U.S. EPA 2002) A number of other agencies (National Institute for Occupational Safety
and Health, the International Agency for Research on Cancer, the World Health Organization,
California EPA, and the U.S. Department of Health and Human Services) made similar hazard
classifications prior to 2002. EPA also concluded in the 2002 Diesel HAD that it was not
possible to calculate a cancer unit risk for diesel exhaust due to limitations in the exposure data
for the occupational groups or the absence of a dose-response relationship.
In the absence of a cancer unit risk, the Diesel HAD sought to provide additional insight into
the significance of the diesel exhaust cancer hazard by estimating possible ranges of risk that
6-26
-------
might be present in the population. An exploratory analysis was used to characterize a range of
possible lung cancer risk. The outcome was that environmental risks of cancer from long-term
diesel exhaust exposures could plausibly range from as low as 10"5 to as high as 10"3. Because of
uncertainties, the analysis acknowledged that the risks could be lower than 10"5, and a zero risk
from diesel exhaust exposure could not be ruled out.
Noncancer health effects of acute and chronic exposure to diesel exhaust emissions are also of
concern to EPA. EPA derived a diesel exhaust reference concentration (RfC) from consideration
of four well-conducted chronic rat inhalation studies showing adverse pulmonary effects. The
RfC is 5 |ig/m3 for diesel exhaust measured as diesel particulate matter. This RfC does not
consider allergenic effects such as those associated with asthma or immunologic or the potential
for cardiac effects. There was emerging evidence in 2002, discussed in the Diesel HAD, that
exposure to diesel exhaust can exacerbate these effects, but the exposure-response data were
lacking at that time to derive an RfC based on these then-emerging considerations. The Diesel
HAD states, "With [diesel particulate matter] being a ubiquitous component of ambient PM,
there is an uncertainty about the adequacy of the existing [diesel exhaust] noncancer database to
identify all of the pertinent [diesel exhaust]-caused noncancer health hazards." The Diesel HAD
also noted "that acute exposure to [diesel exhaust] has been associated with irritation of the eye,
nose, and throat, respiratory symptoms (cough and phlegm), and neurophysiological symptoms
such as headache, lightheadedness, nausea, vomiting, and numbness or tingling of the
extremities." The Diesel HAD notes that the cancer and noncancer hazard conclusions applied to
the general use of diesel engines then on the market and as cleaner engines replace a substantial
number of existing ones, the applicability of the conclusions would need to be reevaluated.
It is important to note that the Diesel HAD also briefly summarizes health effects associated
with ambient PM and discusses EPA's then-annual PM2.5 NAAQS of 15 |ig/m3. In 2012, EPA
revised the level of the annual PM2.5 NAAQS to 12 |ig/m3 and in 2024 EPA revised the level of
the annual PM2.5 NAAQS to 9.0 |ig/m3. (U.S. EPA 2024) There is a large and extensive body of
human data showing a wide spectrum of adverse health effects associated with exposure to
ambient PM, of which diesel exhaust is an important component. The PM2.5 NAAQS provides
protection from the health effects attributed to exposure to PM2.5. The contribution of diesel PM
to total ambient PM varies in different regions of the country and, within a region, from one area
to another. The contribution can be high in near-roadway environments, for example, or in other
locations where diesel engine use is concentrated.
Since 2002, several new studies have been published which continue to report increased lung
cancer risk associated with occupational exposure to diesel exhaust from older engines. Of
particular note since 2011 are three new epidemiology studies that have examined lung cancer in
occupational populations, including truck drivers, underground nonmetal miners, and other
diesel motor-related occupations. These studies reported increased risk of lung cancer related to
exposure to diesel exhaust, with evidence of positive exposure-response relationships to varying
degrees. (Garshick 2012, Silverman 2012, Olsson 2011) These newer studies (along with others
that have appeared in the scientific literature) add to the evidence EPA evaluated in the 2002
Diesel HAD and further reinforce the concern that diesel exhaust exposure likely poses a lung
cancer hazard. The findings from these newer studies do not necessarily apply to newer
technology diesel engines (i.e., heavy-duty highway engines from 2007 and later model years),
since newer engines have large reductions in the emission constituents compared to older
technology diesel engines.
6-27
-------
In light of the growing body of scientific literature evaluating the health effects of exposure to
diesel exhaust, in June 2012 the World Health Organization's International Agency for Research
on Cancer (IARC), a recognized international authority on the carcinogenic potential of
chemicals and other agents, evaluated the full range of cancer-related health effects data for
diesel engine exhaust. IARC concluded that diesel exhaust should be regarded as "carcinogenic
to humans." (IARC, Diesel and gasoline engine exhausts and some nitroarenes 2013) This
designation was an update from its 1988 evaluation that considered the evidence to be indicative
of a "probable human carcinogen."
6.3.7 Air Toxics
Light- and medium-duty engine emissions contribute to ambient levels of air toxics that are
known or suspected human or animal carcinogens, or that have noncancer health effects. These
compounds include, but are not limited to, acetaldehyde, acrolein, benzene, 1,3-butadiene,
formaldehyde, naphthalene, and polycyclic organic matter. These compounds were all identified
as national or regional cancer risk drivers or contributors in the 2019 AirToxScreen Assessment.
(U.S. EPA 2022, U.S. EPA 2023)
6.3.7.1 Health Effects Associated with Exposure to Acetaldehyde
Acetaldehyde is classified in EPA's IRIS database as a probable human carcinogen, based on
nasal tumors in rats, and is considered toxic by the inhalation, oral, and intravenous routes. (U.S.
EPA 1991) The inhalation unit risk estimate (URE) in IRIS for acetaldehyde is 2.2 x 10"6 per
|ig/m3, (U.S. EPA 1991) Acetaldehyde is reasonably anticipated to be a human carcinogen by the
NTP in the 14th Report on Carcinogens and is classified as possibly carcinogenic to humans
(Group 2B) by the IARC. (NTP, Report on Carcinogens, Fourteenth Edition 2016) (IARC 1999)
The primary noncancer effects of exposure to acetaldehyde vapors include irritation of the
eyes, skin, and respiratory tract. (U.S. EPA 1991)In short-term (4 week) rat studies,
degeneration of olfactory epithelium was observed at various concentration levels of
acetaldehyde exposure. (Appleman, Woutersen and Feron 1982) Data from these studies were
used by EPA to develop an inhalation reference concentration of 9 |ig/m3. Some asthmatics have
been shown to be a sensitive subpopulation to decrements in functional expiratory volume
(FEV1 test) and bronchoconstriction upon acetaldehyde inhalation. (Myou, et al. 1993) Children,
especially those with diagnosed asthma, may be more likely to show impaired pulmonary
function and symptoms of asthma than are adults following exposure to acetaldehyde. (OEHHA
2014)
6.3.7.2 Health Effects Associated with Exposure to Benzene
EPA's Integrated Risk Information System (IRIS) database lists benzene as a known human
carcinogen (causing leukemia) by all routes of exposure, and concludes that exposure is
associated with additional health effects, including genetic changes in both humans and animals
and increased proliferation of bone marrow cells in mice. (U.S. EPA 2000, IARC 1982, Irons, et
al. 1992) EPA states in its IRIS database that data indicate a causal relationship between benzene
exposure and acute lymphocytic leukemia and suggest a relationship between benzene exposure
and chronic non-lymphocytic leukemia and chronic lymphocytic leukemia. EPA's IRIS
documentation for benzene also lists a range of 2.2 x 10-6 to 7.8 x 10-6 per |ig/m3 as the unit
6-28
-------
risk estimate (URE) for benzene.216 (U.S. EPA 2000) The International Agency for Research on
Cancer (IARC) has determined that benzene is a human carcinogen, and the U.S. Department of
Health and Human Services (DHHS) has characterized benzene as a known human carcinogen.
(IARC 2018, NTP, Report on Carcinogens, Fourteenth Edition 2016)
A number of adverse noncancer health effects, including blood disorders such as preleukemia
and aplastic anemia, have also been associated with long-term exposure to benzene. (Aksoy
1989, Goldstein 1988) The most sensitive noncancer effect observed in humans, based on current
data, is the depression of the absolute lymphocyte count in blood. (Rothman 1996, U.S. EPA
2002) EPA's inhalation reference concentration (RfC) for benzene is 30 |ig/m3. The RfC is
based on suppressed absolute lymphocyte counts seen in humans under occupational exposure
conditions. In addition, studies sponsored by the Health Effects Institute (HEI) provide evidence
that biochemical responses occur at lower levels of benzene exposure than previously known.
(O. Qu, et al. 2003, Q. Qu, et al. 2002, Lan, et al. 2004, Turtletaub and Mani 2003) EPA's IRIS
program has not yet evaluated these new data. EPA does not currently have an acute reference
concentration for benzene. The Agency for Toxic Substances and Disease Registry (ATSDR)
Minimal Risk Level (MRL) for acute exposure to benzene is 29 |ig/m3 for 1-14 days
exposure.217 (ATSDR, Toxicological profile for benzene 2007)
There is limited information from two studies regarding an increased risk of adverse effects to
children whose parents have been occupationally exposed to benzene. (Corti and Snyder 1996,
P. A., et al. 1991) Data from animal studies have shown benzene exposures result in damage to
the hematopoietic (blood cell formation) system during development. (Keller and Snyder 1986,
Keller and Snyder 1988, Corti and Snyder 1996) Also, key changes related to the development of
childhood leukemia occur in the developing fetus. (U.S. EPA 2002) Several studies have
reported that genetic changes related to eventual leukemia development occur before birth. For
example, there is one study of genetic changes in twins who developed T cell leukemia at nine
years of age. (Ford, et al. 1997)
6.3.7.3 Health Effects Associated with Exposure to 1,3-Butadiene
EPA has characterized 1,3-butadiene as carcinogenic to humans by inhalation. (U.S. EPA
2002)' (U.S. EPA 2002) The IARC has determined that 1,3-butadiene is a human carcinogen and
the U.S. DHHS has characterized 1,3-butadiene as a known human carcinogen. (IARC 1999)
(IARC 2008) (NTP 2016) (IARC 2012) There are numerous studies consistently demonstrating
that 1,3-butadiene is metabolized into genotoxic metabolites by experimental animals and
humans. The specific mechanisms of 1,3-butadiene-induced carcinogenesis are unknown;
however, the scientific evidence strongly suggests that the carcinogenic effects are mediated by
genotoxic metabolites. Animal data suggest that females may be more sensitive than males for
cancer effects associated with 1,3-butadiene exposure; there are insufficient data in humans from
which to draw conclusions about sensitive subpopulations. The URE for 1,3-butadiene is 3 x 10-
5 per |ig/m3, (U.S. EPA 2002) 1,3-butadiene also causes a variety of reproductive and
developmental effects in mice; no human data on these effects are available. The most sensitive
216 A unit risk estimate is defined as the increase in the lifetime risk of cancer of an individual who is exposed for a
lifetime to 1 |ig/m3 benzene in air.
217 A minimal risk level (MRL) is defined as an estimate of the daily human exposure to a hazardous substance that
is likely to be without appreciable risk of adverse noncancer health effects over a specified duration of exposure.
6-29
-------
effect was ovarian atrophy observed in a lifetime bioassay of female mice. (Bevan, Stadler and al
1996) Based on this critical effect and the benchmark concentration methodology, an RfC for
chronic health effects was calculated at 0.9 ppb (approximately 2 |ig/m3).
6.3.7.4 Health Effects Associated with Exposure to Formaldehyde
In 1991, EPA concluded that formaldehyde is a Class B1 probable human carcinogen based
on limited evidence in humans and sufficient evidence in animals. (U.S. EPA 1990) An
Inhalation URE for cancer and a Reference Dose for oral noncancer effects were developed by
EPA and posted on the IRIS database. Since that time, the NTP and IARC have concluded that
formaldehyde is a known human carcinogen. (NTP, Report on Carcinogens, Fourteenth Edition
2016, IARC 2006, IARC 2012)
The conclusions by IARC and NTP reflect the results of epidemiologic research published
since 1991 in combination with previous and more recent animal, human, and mechanistic
evidence. Research conducted by the National Cancer Institute reported an increased risk of
nasopharyngeal cancer and specific lymphohematopoietic malignancies among workers exposed
to formaldehyde. (Hauptmann, Lubin, et al. 2003, Hauptmann, Lubin, et al. 2004, Beane
Freeman, et al. 2009) A National Institute of Occupational Safety and Health study of garment
workers also reported increased risk of death due to leukemia among workers exposed to
formaldehyde. (Pinkerton 2004) Extended follow-up of a cohort of British chemical workers did
not report evidence of an increase in nasopharyngeal or lymphohematopoietic cancers, but a
continuing statistically significant excess in lung cancers was reported. (Coggon, et al. 2003)
Finally, a study of embalmers reported formaldehyde exposures to be associated with an
increased risk of myeloid leukemia but not brain cancer. (Hauptmann, et al. 2009)
Health effects of formaldehyde in addition to cancer were reviewed by the Agency for Toxic
Substances and Disease Registry in 1999, supplemented in 2010, and by the World Health
Organization. (ATSDR 1999, ATSDR 2010, IPCS 2002) These organizations reviewed the
scientific literature concerning health effects linked to formaldehyde exposure to evaluate
hazards and dose response relationships and defined exposure concentrations for minimal risk
levels (MRLs). The health endpoints reviewed included sensory irritation of eyes and respiratory
tract, reduced pulmonary function, nasal histopathology, and immune system effects. In addition,
research on reproductive and developmental effects and neurological effects were discussed
along with several studies that suggest that formaldehyde may increase the risk of asthma -
particularly in the young.
In June 2010, EPA released a draft Toxicological Review of Formaldehyde - Inhalation
Assessment through the IRIS program for peer review by the National Research Council (NRC)
and public comment. (U.S. EPA 2010) That draft assessment reviewed more recent research
from animal and human studies on cancer and other health effects. The NRC released their
review report in April 2011. (NRC, Review of the Environmental Protection Agency's Draft
IRIS Assessment of Formaldehyde 2011) EPA addressed the NRC (2011) recommendations and
applied systematic review methods to the evaluation of the available noncancer and cancer health
effects evidence and released a new draft IRIS Toxicological Review of Formaldehyde -
Inhalation in April 2022. (US EPA 2022) In this draft, updates to the 1991 IRIS finding include a
stronger determination of the carcinogenicity of formaldehyde inhalation to humans, as well as
characterization of its noncancer effects to propose an overall reference concentration for
inhalation exposure. The National Academies of Science, Engineering, and Medicine released
6-30
-------
their review of EPA's 2022 Draft Formaldehyde Assessment in August 2023, concluding that
EPA's "findings on formaldehyde hazard and quantitative risk are supported by the evidence
identified." (National Academies of Sciences. Engineering 2023) EPA is currently revising the
draft IRIS assessment in response to comments received.218
6.3.7.5 Health Effects Associated with Exposure to Naphthalene
Naphthalene is found in small quantities in gasoline and diesel fuels. Naphthalene emissions
have been measured in larger quantities in both gasoline and diesel exhaust compared with
evaporative emissions from mobile sources, indicating it is primarily a product of combustion.
Acute (short-term) exposure of humans to naphthalene by inhalation, ingestion, or dermal
contact is associated with hemolytic anemia and damage to the liver and the nervous system.
(U.S. EPA 1998) Chronic (long term) exposure of workers and rodents to naphthalene has been
reported to cause cataracts and retinal damage. (U.S. EPA 1998) Children, especially neonates,
appear to be more susceptible to acute naphthalene poisoning based on the number of reports of
lethal cases in children and infants (hypothesized to be due to immature naphthalene
detoxification pathways). (U.S. EPA 1998)EPA released an external review draft of a
reassessment of the inhalation carcinogenicity of naphthalene based on a number of recent
animal carcinogenicity studies. (U.S. EPA 1998) The draft reassessment completed external peer
review. (Oak Ridge Institute for Science and Education 2004) Based on external peer review
comments received, EPA is developing a revised draft assessment that considers inhalation and
oral routes of exposure, as well as cancer and noncancer effects (U.S. EPA 2023). The external
review draft does not represent official agency opinion and was released solely for the purposes
of external peer review and public comment. The NTP listed naphthalene as "reasonably
anticipated to be a human carcinogen" in 2004 on the basis of bioassays reporting clear evidence
of carcinogenicity in rats and some evidence of carcinogenicity in mice. (NTP, Report on
Carcinogens, Fourteenth Edition 2016) California EPA has released a risk assessment for
naphthalene, and the IARC has reevaluated naphthalene and re-classified it as Group 2B:
possibly carcinogenic to humans. (IARC 2002)
Naphthalene also causes a number of non-cancer effects in animals following chronic and
less-than-chronic exposure, including abnormal cell changes and growth in respiratory and nasal
tissues. (U.S. EPA 1998) The current EPA IRIS assessment includes noncancer data on
hyperplasia and metaplasia in nasal tissue that form the basis of the inhalation RfC of 3 |ig/m3.
(U.S. EPA 1998) The ATSDR MRL for acute and intermediate duration oral exposure to
naphthalene is 0.6 mg/kg/day based on maternal toxicity in a developmental toxicology study in
rats. (ATSDR 2005) ATSDR also derived an ad hoc reference value of 6 x 10-2 mg/m3 for acute
(<24-hour) inhalation exposure to naphthalene in a Letter Health Consultation dated March 24,
2014 to address a potential exposure concern in Illinois. (ATSDR 2014) The ATSDR acute
inhalation reference value was based on a qualitative identification of an exposure level
interpreted not to cause pulmonary lesions in mice. More recently, EPA developed acute RfCs
for 1-, 8-, and 24-hour exposure scenarios; the <24-hour reference value is 2 x 10-2 mg/m3. (U.S.
EPA 2022) EPA's acute RfCs are based on a systematic review of the literature, benchmark dose
218 For more information, see https://cfpub.epa.gov/ncea/iris_drafts/recordisplay.cfm?deid=248150#.
6-31
-------
modeling of naphthalene-induced nasal lesions in rats, and application of a PBPK
(physiologically based pharmacokinetic) model.
6.3.7.6 Health Effects Associated with Exposure to PAHs/POM
The term polycyclic organic matter (POM) defines a broad class of compounds that includes
the polycyclic aromatic hydrocarbon compounds (PAHs). One of these compounds, naphthalene,
is discussed separately in Section 6.2.7.5. POM compounds are formed primarily from
combustion and are present in the atmosphere in gas and particulate form as well as in some fried
and grilled foods. Epidemiologic studies have reported an increase in lung cancer in humans
exposed to diesel exhaust, coke oven emissions, roofing tar emissions, and cigarette smoke; all
of these mixtures contain POM compounds. (ATSDR 1995) (U.S. EPA 2002) In 1991, EPA
classified seven PAHs (benzo[a]pyrene, benz[a]anthracene, chrysene, benzo[b]fluoranthene,
benzo[k]fluoranthene, dibenz[a,h]anthracene, and indeno[l,2,3-cd]pyrene) as Group B2,
probable human carcinogens based on the 1986 EPA Guidelines for Carcinogen Risk
Assessment. (U.S. EPA 1991) Studies in multiple animal species demonstrate that
benzo[a]pyrene is carcinogenic at multiple tumor sites (alimentary tract, liver, kidney, respiratory
tract, pharynx, and skin) by all routes of exposure. An increasing number of occupational studies
demonstrate a positive exposure-response relationship with cumulative benzo[a]pyrene exposure
and lung cancer. The inhalation URE in IRIS for benzo[a]pyrene is 6 x 10"4 per |ig/m3 and the
oral slope factor for cancer is 1 per mg/kg-day. (U.S. EPA 2017)
Animal studies demonstrate that exposure to benzo[a]pyrene is also associated with
developmental (including developmental neurotoxicity), reproductive, and immunological
effects. In addition, epidemiology studies involving exposure to PAH mixtures have reported
associations between internal biomarkers of exposure to benzo[a]pyrene (benzo[a]pyrene diol
epoxide-DNA adducts) and adverse birth outcomes (including reduced birth weight, postnatal
body weight, and head circumference), neurobehavioral effects, and decreased fertility. The
inhalation RfC for benzo[a]pyrene is 2 x 10"6 mg/m3 and the RfD for oral exposure is 3 x 10"4
mg/kg-day. (U.S. EPA 2017)
6.3.8 Exposure and Health Effects Associated with Traffic
Locations in close proximity to major roadways generally have elevated concentrations of
many air pollutants emitted from motor vehicles. Hundreds of studies have been published in
peer-reviewed journals, concluding that concentrations of CO, CO2, NO, NO2, benzene,
aldehydes, PM, black carbon, and many other compounds are elevated in ambient air within
approximately 300-600 meters (about 1,000-2,000 feet) of major roadways. The highest
concentrations of most pollutants emitted directly by motor vehicles are found at locations within
50 meters (about 165 feet) of the edge of a roadway's traffic lanes.
A large-scale review of air quality measurements in the vicinity of major roadways between
1978 and 2008 concluded that the pollutants with the steepest concentration gradients in
vicinities of roadways were CO, UFPs, metals, elemental carbon (EC), NO, NOx, and several
VOCs. (Karner, Eisinger and Niemeier, Near-roadway air quality: synthesizing the findings from
real-world data 2010) These pollutants showed a large reduction in concentrations within 100
meters downwind of the roadway. Pollutants that showed more gradual reductions with distance
from roadways included benzene, NO2, PM2.5, and PM10. In reviewing the literature, Karner,
Eisinger, and Niemeier (Karner, Eisinger and Niemeier, Near-roadway air quality: synthesizing
6-32
-------
the findings from real-world data 2010) reported that results varied based on the method of
statistical analysis used to determine the gradient in pollutant concentration. More recent studies
continue to show significant concentration gradients of traffic-related air pollution around major
roads. (McDonald, et al. 2014, Kimbrough, Baldauf, et al. 2013, Kimbrough, Palma and Baldauf
2014, Kimbrough, Owen, et al. 2017, Hilker, et al. 2019, Grivas, et al. 2019, Apte, et al. 2017,
Dabek-Zlotorzynska, et al. 2019, Gu, et al. 2018) There is evidence thatEPA's regulations for
vehicles have lowered the near-road concentrations and gradients. (Sarnat, et al. 2018) Starting
in 2010, EPA required through the NAAQS process that air quality monitors be placed near
high-traffic roadways for determining concentrations of CO, NO2, and PM2.5 (in addition to those
existing monitors located in neighborhoods and other locations farther away from pollution
sources). The monitoring data for NO2 indicate that in urban areas, monitors near roadways often
report the highest concentrations of NO2. (Gantt, Owen and Watkins 2021, Lai, Ramaswami and
Russell 2020)
For pollutants with relatively high background concentrations relative to near-road
concentrations, detecting concentration gradients can be difficult. For example, many carbonyls
have high background concentrations because of photochemical breakdown of precursors from
many different organic compounds. However, several studies have measured carbonyls in
multiple weather conditions and found higher concentrations of many carbonyls downwind of
roadways. (Liu, et al. 2006, Cahill, Charles and Seaman 2010) These findings suggest a
substantial roadway source of these carbonyls.
In the past 30 years, many studies have been published with results reporting that populations
who live, work, or go to school near high-traffic roadways experience higher rates of numerous
adverse health effects, compared to populations far away from major roads.219 In addition,
numerous studies have found adverse health effects associated with spending time in traffic, such
as commuting or walking along high-traffic roadways, including studies among children. (Laden,
et al. 2007, Peters, et al. 2004, Zanobetti, et al. 2009, Adar, et al. 2007)
Numerous reviews of this body of health literature have been published. In a 2022 final
report, an expert panel of the Health Effects Institute (HEI) employed a systematic review
focusing on selected health endpoints related to exposure to traffic-related air pollution. (HEI
2022)220 The HEI panel concluded that there was a high level of confidence in evidence between
long-term exposure to traffic-related air pollution and health effects in adults, including all-
cause, circulatory, and ischemic heart disease mortality. (Boogaard, et al. 2022) The panel also
found that there is a moderate-to-high level of confidence in evidence of associations with
asthma onset and acute respiratory infections in children and lung cancer and asthma onset in
adults. This report follows on an earlier expert review published by HEI in 2010, where it found
strongest evidence for asthma-related traffic impacts. Other literature reviews have been
published with conclusions generally similar to the HEI panels' conclusions. (Boothe and
Shendell 2008, Salam, Islam and Gilliland 2008, Sun, Zhang and Ma 2014, Raaschou-Nielsen
219 In the widely used PubMed database of health publications, between January 1, 1990 and December 31, 2021,
1,979 publications contained the keywords "traffic, pollution, epidemiology," with approximately half the studies
published after 2015.
220 This more recent review focused on health outcomes related to birth effects, respiratory effects, cardiometabolic
effects, and mortality.
6-33
-------
and Reynolds 2006) Additionally, in 2014, researchers from the U.S. Centers for Disease
Control and Prevention (CDC) published a systematic review and meta-analysis of studies
evaluating the risk of childhood leukemia associated with traffic exposure and reported positive
associations between "postnatal" proximity to traffic and leukemia risks, but no such association
for "prenatal" exposures. (Boothe, et al. 2014) The U.S. Department of Health and Human
Services' National Toxicology Program (NTP) published a monograph including a systematic
review of traffic-related air pollution and its impacts on hypertensive disorders of pregnancy.
The NTP concluded that exposure to traffic-related air pollution is "presumed to be a hazard to
pregnant women" for developing hypertensive disorders of pregnancy. (NTP 2019)
For several other health outcomes there are publications to suggest the possibility of an
association with traffic-related air pollution, but insufficient evidence to draw definitive
conclusions. Among these outcomes are neurological and cognitive impacts (e.g., autism and
reduced cognitive function, academic performance, and executive function) and reproductive
outcomes (e.g., preterm birth, low birth weight). (Volk, et al. 2011, Franco-Suglia, et al. 2007,
Power, et al. 2011, Wu, et al. 2011, Stenson, et al. 2021, Gartland, et al. 2022)
Numerous studies have also investigated potential mechanisms by which traffic-related air
pollution affects health, particularly for cardiopulmonary outcomes. For example, numerous
studies indicate that near-roadway exposures may increase systemic inflammation, affecting
organ systems, including blood vessels and lungs. (Riediker 2007, Alexeef, et al. 2011, Eckel, et
al. 2011, Zhang, et al. 2009) Additionally, long-term exposures in near-road environments have
been associated with inflammation-associated conditions, such as atherosclerosis and asthma.
(Adar, Klein, et al. 2010, Kan, et al. 2008, McConnell, et al. 2010)
The risks associated with residence, workplace, or schools near major roads are of potentially
high public health significance due to the large population in such locations. The 2013 U.S.
Census Bureau's American Housing Survey (AHS) was the last AHS that included whether
housing units were within 300 feet of an "airport, railroad, or highway with four or more
lanes."221 The 2013 survey reports that 17.3 million housing units, or 13 percent of all housing
units in the U.S., were in such areas. Assuming that populations and housing units are in the
same locations, this corresponds to a population of more than 41 million U.S. residents within
300 feet of high-traffic roadways or other transportation sources. According to the Central
Intelligence Agency's World Factbook, based on data collected between 2012-2014, the United
States had 6,586,610 km of roadways, 293,564 km of railways, and 13,513 airports. As such,
highways represent the overwhelming majority of transportation facilities described by this
factor in the AHS.
Scientific literature suggests that some sociodemographic factors may increase susceptibility
to the effects of traffic-associated air pollution. For example, several studies have found stronger
adverse health associations in children experiencing chronic social stress, such as living in
violent neighborhoods or in homes with low incomes or high family stress. (Islam, et al. 2011,
Clougherty, et al. 2007, Chen, et al. 2008, Long, Lewis and Langpap 2021) HEI's 2022 critical
review of traffic and health studies mentions additional potential mediators or effect modifiers of
the relationship between traffic-related air pollution and health, including preexisting morbidities
(e.g., obesity, hypertension), the built environment (i.e., green space, walkability), and
221 The variable was known as "ETRANS" in the questions about the neighborhood.
6-34
-------
socioeconomic characteristics, but notes that additional research is needed to better understand
such interactions. (HEI 2022)
EPA conducted a study to estimate the number of people living near truck freight routes in the
United States, which includes many large highways and other routes where light- and medium-
duty vehicles operate. (U.S. EPA 2021) Based on a population analysis using the U.S.
Department of Transportation's (USDOT) Freight Analysis Framework 4 (FAF4) and population
data from the 2010 decennial census, an estimated 72 million people live within 200 meters of
these FAF4 roads, which are used by all types of vehicles (U.S. DOT 2023).222,223 anaiySjs
includes the population living within twice the distance of major roads compared with the
analysis of housing units near major roads described above in this section. The larger distance
and other methodological differences explain the difference in the two estimates for populations
living near major roads.
In examining schools near major roadways, we used the Common Core of Data from the U.S.
Department of Education, which includes information on all public elementary and secondary
schools and school districts nationwide.224'225 To determine school proximities to major
roadways, we used a geographic information system to map each school and roadways based on
the U.S. Census's TIGER roadway file. Ten million students attend public schools within 200
meters of major roads, about 20 percent of the total number of public school students in the U.S.,
and about 800,000 students attend public schools within 200 meters of primary roads.226'227 We
found that students of color were overrepresented at schools within 200 meters of primary
roadways, and schools within 200 meters of primary roadways had a disproportionately greater
population of students eligible for free or reduced-price lunches. Black students represent 22
percent of students at schools located within 200 meters of a primary road, compared to 17
percent of students in all U.S. schools. Hispanic students represent 30 percent of students at
schools located within 200 meters of a primary road, compared to 22 percent of students in all
U.S. schools. (Pedde and Bailey 2011)
222 FAF4 is a model from the USDOT's Bureau of Transportation Statistics (BTS) and Federal Highway
Administration (FHWA), which provides data associated with freight movement in the U.S. It includes data from the
2012 Commodity Flow Survey (CFS), the Census Bureau on international trade, as well as data associated with
construction, agriculture, utilities, warehouses, and other industries. FAF4 estimates the modal choices for moving
goods by trucks, trains, boats, and other types of freight modes. It includes traffic assignments, including truck flows
on a network of truck routes
223 The same analysis estimated the population living within 100 meters of a FAF4 truck route is 41 million.
224 This information is available at: http://nces.ed.gov/ccd/.
225 TIGER/Line shapefiles for the year 2010. [Online at https://www.census.gov/geographies/mapping-files/time-
series/geo/tiger-line-file.2010.html]
226 Here, "major roads" refer to those TIGER classifies as either "Primary" or "Secondary." The Census Bureau
describes primary roads as "generally divided limited-access highways within the Federal interstate system or under
state management." Secondary roads are "main arteries, usually in the U.S. highway, state highway, or county
highway system."
227 For this analysis we analyzed a 200-meter distance based on the understanding that roadways generally influence
air quality within a few hundred meters from the vicinity of heavily traveled roadways or along corridors with
significant trucking traffic. See U.S. EPA, 2014. Near Roadway Air Pollution and Health: Frequently Asked
Questions. EPA-420-F-14-044.
6-35
-------
While near-roadway studies focus on residents near roads or others spending considerable
time near major roads, the duration of commuting results in another important contributor to
overall exposure to traffic-related air pollution. Studies of health that address time spent in transit
have found evidence of elevated risk of cardiac impacts. (Riediker, Cascio, et al. 2004, Peters, et
al. 2004, Adar, Gold and Coull 2007) Studies have also found that school bus emissions can
increase student exposures to diesel-related air pollutants, and that programs that reduce school
bus emissions may improve health and reduce school absenteeism. (Sabin, et al. 2005, Li, N and
Ryan 2009, Austin, Heutel and Kreisman 2019, Adar, D.Souza and Sheppard 2015) Lastly,
EPA's Exposure Factor Handbook indicates that, on average, Americans spend more than an
hour traveling each day, bringing nearly all residents into a high-exposure microenvironment for
part of the day. (U.S. EPA 2016) 228
6.4 Welfare Effects Associated with Exposure to Criteria and Air Toxics Pollutants
This section discusses the environmental effects associated with non-GHG pollutants affected
by this rule, specifically PM, ozone, NOx, SOx, and air toxics.
6.4.1 Visibility Degradation
Visibility can be defined as the degree to which the atmosphere is transparent to visible light.
(NRC 1993) Visibility impairment is caused by light scattering and absorption by suspended
particles and gases. It is dominated by contributions from suspended particles except under
pristine conditions. Fine particles with significant light-extinction efficiencies include sulfates,
nitrates, organic carbon, elemental carbon, sea salt, and soil. (J. e. Hand 2011, Sisler 1996)
Visibility is important because it has direct significance to people's enjoyment of daily activities
in all parts of the country. Individuals value good visibility for the well-being it provides them
directly, where they live and work and in places where they enjoy recreational opportunities.
Visibility is also highly valued in significant natural areas, such as national parks and wilderness
areas, and special emphasis is given to protecting visibility in these areas. For more information
on visibility, see the final 2019 PM ISA. (U.S. EPA 2019)
The extent to which any amount of light extinction affects a person's ability to view a scene
depends on both scene and light characteristics. For example, the appearance of a nearby object
(e.g., a building) is generally less sensitive to a change in light extinction than the appearance of
a similar object at a greater distance. See Figure 6-1 for an illustration of the important factors
affecting visibility. (Malm 2016)
228 It is not yet possible to estimate the long-term impact of growth in telework associated with the COVID-19
pandemic on travel behavior. There were notable changes during the pandemic. For example, according to the 2021
American Time Use Survey, a greater fraction of workers did at least part of their work at home (38%) as compared
with the 2019 survey (24%). [Online at https://www.bls.gov/news.release/atus.nrO.htm],
6-36
-------
Characteristics of Observer
Optical Characteristics of Illumination
itlcal Characteristics of
Optical Characteristics of
Intervene
• Color
• Contrast Detail (Texture)
• Form
• Brightness
• Detection Thresholds
• Psychological Response to
Incoming Light
• Value Judgements
• Light Added to Sight Path by
Particles and Cases
• Image-Forming Light Subtracted
from Sight Path by Scattering
and Absorption
Sunlight ^ jf'
scattered ^
from ground
scattered into
sight path
Image-forming
light absorbed
Image-forming
light scattered
out of sight path
Light from clouds
scattered into
sight path ^
Figure 6-1: Important Factors Involved in Seeing a Scenic Vista (Malm, 2016).
EPA is working to address visibility impairment. Reductions in air pollution from
implementation of various programs called for in the Clean Air Act Amendments of 1990
(CAAA) have resulted in substantial improvements in visibility and will continue to do so in the
future. Nationally, because trends in haze are closely associated with trends in particulate sulfate
and nitrate emissions due to the relationship between their concentration and light extinction,
visibility trends have improved as emissions of SO2 and NOx have decreased over time due to air
pollution regulations such as the Acid Rain Program. (U.S. EPA 2019)However, in the western
part of the country, changes in total light extinction were smaller, and the contribution of
particulate organic matter to atmospheric light extinction was increasing due to increasing
wildfire emissions. (Hand, et al. 2020)
In the Clean Air Act Amendments of 1977, Congress recognized visibility's value to society
by establishing a national goal to protect national parks and wilderness areas from visibility
impairment caused by manmade pollution (42 USC §7491 (a) 2013). In 1999, EPA finalized the
regional haze program (64 FR 35714) to protect the visibility in Mandatory Class I Federal areas.
There are 156 national parks, forests, and wilderness areas categorized as Mandatory Class I
Federal areas (62 FR 38680-38681, July 18, 1997). These areas are defined in CAA Section 162
as those national parks exceeding 6,000 acres, wilderness areas and memorial parks exceeding
5,000 acres, and all international parks that were in existence on August 7, 1977. Figure 6-2
shows the location of the 156 Mandatory Class I Federal areas.
6-37
-------
• NPS Units
• FWS Units
• FS Units
Figure 6-2: Mandatory Class I Federal Areas in the U.S.
Vltpn Mard> HP
* Rainbow Lake, W1 and Bradwell Bay, FL are Class 1 Areas
where visibility is not an important air quality related value
Produced by NPS Air Resources Division
EPA has also concluded that PM2.5 causes adverse effects on visibility in other areas that are
not targeted by the Regional Haze Rule, such as urban areas, depending on PM2.5 concentrations
and other factors such as dry chemical composition and relative humidity (i.e., an indicator of the
water composition of the particles). The secondary (welfare-based) PM NAAQS provide
protection against visibility effects. In recent PM NAAQS reviews, EPA evaluated a target level
of protection for visibility impairment that is expected to be met through attainment of the
existing secondary PM standards.
6.4.1.1 Visibility Monitoring
In conjunction with the U.S. National Park Service, the U.S. Forest Service, other Federal
land managers, and State organizations in the U.S., EPA has supported visibility monitoring in
national parks and wilderness areas since 1988. The monitoring network was originally
established at 20 sites, but it has now been expanded to 152 sites that represent all but one of the
156 Mandatory Federal Class I areas across the country (see Figure 6-2). This long-term
visibility monitoring network is known as IMPROVE (Interagency Monitoring of Protected
Visual Environments).
IMPROVE provides direct measurement of particles that contribute to visibility impairment.
The IMPROVE network employs aerosol measurements at all sites, and optical and scene
measurements at some of the sites. Aerosol measurements are taken for PM10 and PM2.5 mass,
and for key constituents of PM2.5, such as sulfate, nitrate, organic and elemental carbon (OC and
EC), and other elements that can be used to estimate soil dust and sea salt contributions.
Measurements for specific aerosol constituents are used to calculate "reconstructed" aerosol light
extinction by multiplying the mass for each constituent by its empirically-derived scattering
and/or absorption efficiency, with adjustment for the relative humidity. The IMPROVE program
utilizes both an "original" and a "revised" reconstruction formula for this purpose, with the latter
6-38
-------
explicitly accounting for sea salt concentrations. Knowledge of the main constituents of a site's
light extinction "budget" is critical for source apportionment and control strategy development.
In addition to this indirect method of assessing light extinction, there are optical measurements
which directly measure light extinction or its components. Such measurements are made
principally with a nephelometer to measure light scattering; some sites also include an
aethalometer for light absorption; and a few sites use a transmissometer, which measures total
light extinction. Scene characteristics are typically recorded using digital or video photography
and are used to determine the quality of visibility conditions (such as effects on color and
contrast) associated with specific levels of light extinction as measured under both direct and
aerosol-related methods. Directly measured light extinction is used under the IMPROVE
protocol to cross check that total light extinction calculated from the IMPROVE reconstruction
formula are consistent with directly measured extinction. Aerosol-derived light extinction from
the IMPROVE equation is used to document spatial and temporal trends and to determine how
changes in atmospheric constituents would affect future visibility conditions.
Annual average visibility conditions (reflecting light extinction due to both anthropogenic and
non-anthropogenic sources) vary regionally across the U.S. Figures 13-1 through 13-14 in the
PM ISA detail the percent contributions to particulate light extinction for ammonium nitrate and
sulfate, EC and OC, and coarse mass and fine soil, by month. (U.S. EPA 2019)
6.4.2 Plant and Ecosystem Effects of Ozone
The welfare effects of ozone include effects on ecosystems, which can be observed across a
variety of scales, i.e., subcellular, cellular, leaf, whole plant, population, and ecosystem. When
ozone effects that begin at small spatial scales, such as the leaf of an individual plant, occur at
sufficient magnitudes (or to a sufficient degree), they can result in effects being propagated
higher and higher levels of biological organization. For example, effects at the individual plant
level, such as altered rates of leaf gas exchange, growth and reproduction, can, when widespread,
result in broad changes in ecosystems, such as productivity, carbon storage, water cycling,
nutrient cycling, and community composition.
Ozone can produce both acute and chronic injury in sensitive plant species depending on the
concentration level and the duration of the exposure. (U.S. EPA 2020) In those sensitive
species,229 effects from repeated exposure to ozone throughout the growing season of the plant
can tend to accumulate, so that even relatively low concentrations experienced for a longer
duration have the potential to create chronic stress on vegetation. (U.S. EPA 2020)230 Ozone
damage to sensitive plant species includes impaired photosynthesis and visible injury to leaves.
The impairment of photosynthesis, the process by which the plant makes carbohydrates (its
source of energy and food), can lead to reduced crop yields, timber production, and plant
productivity and growth. Impaired photosynthesis can also lead to a reduction in root growth and
carbohydrate storage below ground, resulting in other, more subtle plant and ecosystems impacts
(U.S. EPA 2020). These latter impacts include increased susceptibility of plants to insect attack,
229 Only a small percentage of all the plant species growing within the U.S. (over 43,000 species have been
catalogued in the USDA PLANTS database) have been studied with respect to ozone sensitivity.
230 The concentration at which ozone levels overwhelm a plant's ability to detoxify or compensate for oxidant
exposure varies. Thus, whether a plant is classified as sensitive or tolerant depends in part on the exposure levels
being considered.
6-39
-------
disease, harsh weather, interspecies competition, and overall decreased plant vigor. The adverse
effects of ozone on areas with sensitive species could potentially lead to species shifts and loss
from the affected ecosystems,231 resulting in a loss or reduction in associated ecosystem goods
and services. (73 FR 16493-16494 2008) Additionally, visible ozone injury to leaves can result in
a loss of aesthetic value in areas of special scenic significance like national parks and wilderness
areas and reduced use of sensitive ornamentals in landscaping. (U.S. EPA 2020) In addition to
ozone effects on vegetation, newer evidence suggests that ozone affects interactions between
plants and insects by altering chemical signals (e.g., floral scents) that plants use to communicate
to other community members, such as attraction of pollinators.
The Ozone ISA presents more detailed information on how ozone affects vegetation and
ecosystems. (U.S. EPA 2020) The Ozone ISA reports causal and likely causal relationships
between ozone exposure and a number of welfare effects and characterizes the weight of
evidence for different effects associated with ozone.232 The Ozone ISA concludes that visible
foliar injury effects on vegetation, reduced vegetation growth, reduced plant reproduction,
reduced productivity in terrestrial ecosystems, reduced yield and quality of agricultural crops,
alteration of below-ground biogeochemical cycles, and altered terrestrial community
composition are causally associated with exposure to ozone. It also concludes that increased tree
mortality, altered herbivore growth and reproduction, altered plant-insect signaling, reduced
carbon sequestration in terrestrial ecosystems, and alteration of terrestrial ecosystem water
cycling are likely to be causally associated with exposure to ozone.
6.4.3 Deposition
Deposited airborne pollutants contribute to adverse effects on ecosystems, and to soiling and
materials damage. These welfare effects result mainly from exposure to excess amounts of
specific chemical species, regardless of their source or predominant form (particle, gas, or
liquid). Nitrogen and sulfur tend to comprise a large portion of PM in many locations; however,
gas-phase forms of oxidized nitrogen and sulfur also cause adverse ecological effects. The
following characterizations of the nature of these environmental effects are based on information
contained in the 2019 PM ISA, and the 2020 Integrated Science Assessment for Oxides of
Nitrogen, Oxides of Sulfur, and Particulate Matter - Ecological Criteria. (U.S. EPA 2020, U.S.
EPA 2019)
6.4.3.1 Deposition of Nitrogen and Sulfur
Nitrogen and sulfur interactions in the environment are highly complex, as shown in Figure
6-3. (U.S. EPA 2020)Both nitrogen and sulfur are essential, and sometimes limiting, nutrients
needed for growth and productivity of ecosystem components (e.g., algae, plants). In terrestrial
and aquatic ecosystems, excesses of nitrogen or sulfur can lead to acidification and nutrient
231 Per footnote above, ozone impacts could be occurring in areas where plant species sensitive to ozone have not yet
been studied or identified.
232 The Ozone ISA evaluates the evidence associated with different ozone related health and welfare effects,
assigning one of five "weight of evidence" determinations: causal relationship, likely to be a causal relationship,
suggestive of a causal relationship, inadequate to infer a causal relationship, and not likely to be a causal
relationship. For more information on these levels of evidence, please refer to Table II of the ISA.
6-40
-------
enrichment. (U.S. EPA 2020) In addition, in aquatic ecosystems, sulfur deposition can increase
mercury methylation.
Sunlight
VOC
| Foliar and
teffocts
Oxidation Dissolution
S02 *¦ H2S04 *¦ 2H* +SOA*
NO, >¦ HNOj » H**NOj"
NO,
Wet Deposition
H*.NH4*. NOj , SO42
Ambient Air
Concentration
Dry deposition s02
NO,, NH„ SO, NO,
NHj | Deposition
A
-ft' ilUSM
Acidification of water ~ Eutrophication
Ecological
Effect
Figure 6-3: Nitrogen and Sulfur Cycling, and Interactions in the Environment.
6.4.3.1.1 Ecological Effects of Acidification
Deposition of nitrogen and sulfur can cause acidification, which alters biogeochemistry and
affects animal and plant life in terrestrial and aquatic ecosystems across the U.S. Soil
acidification is a natural process, but is often accelerated by acidifying deposition, which can
decrease concentrations of exchangeable base cations in soils. (U.S. EPA 2020) Biological
effects of acidification in terrestrial ecosystems are generally linked to aluminum toxicity and
decreased ability of plant roots to take up base cations. (U.S. EPA 2020) Decreases in the acid
neutralizing capacity and increases in inorganic aluminum concentration contribute to declines in
zooplankton, macro invertebrates, and fish species richness in aquatic ecosystems. (U.S. EPA
2020)
Geology (particularly surficial geology) is the principal factor governing the sensitivity of
terrestrial and aquatic ecosystems to acidification from nitrogen and sulfur deposition. (U.S. EPA
2020) Geologic formations having low base cation supply generally underlie the watersheds of
acid-sensitive lakes and streams. Other factors contribute to the sensitivity of soils and surface
waters to acidifying deposition, including topography, soil chemistry, land use, and hydrologic
flow path. (U.S. EPA 2020)
6-41
-------
6.4.3.1.1.1 Aquatic Acidification
Aquatic effects of acidification have been well studied in the U.S. and elsewhere at various
trophic levels. These studies indicate that aquatic biota have been affected by acidification at
virtually all levels of the food web in acid sensitive aquatic ecosystems. Effects have been most
clearly documented for fish, aquatic insects, other invertebrates, and algae. Biological effects are
primarily attributable to a combination of low pH and high inorganic aluminum concentrations.
Such conditions occur more frequently during rainfall and snowmelt that cause high flows of
water, and less commonly during low-flow conditions, except where chronic acidity conditions
are severe. Biological effects of episodes include reduced fish condition factor, changes in
species composition, and declines in aquatic species richness across multiple taxa, ecosystems,
and regions.
Because acidification primarily affects the diversity and abundance of aquatic biota, it also
affects the ecosystem services, e.g., recreational and subsistence fishing, that are derived from
the fish and other aquatic life found in these surface waters. For example, in the northeastern
United States, the surface waters affected by acidification are a source of food for some
recreational and subsistence fishermen and for other consumers with particularly high rates of
self-caught fish consumption, such as the Hmong and Chippewa ethnic groups. (Hutchison 1994,
Peterson, etal. 1994)
6.4.3.1.1.2 Terrestrial Acidification
Acidifying deposition has altered major biogeochemical processes in the U.S. by increasing
the nitrogen and sulfur content of soils, accelerating nitrate and sulfate leaching from soil to
drainage waters, depleting base cations (especially calcium and magnesium) from soils, and
increasing the mobility of aluminum. Inorganic aluminum is toxic to some tree roots. Plants
affected by high levels of aluminum from the soil often have reduced root growth, which restricts
the ability of the plant to take up water and nutrients, especially calcium. (U.S. EPA 2020) These
direct effects can, in turn, influence the response of these plants to climatic stresses such as
droughts and cold temperatures. They can also influence the sensitivity of plants to other
stresses, including insect pests and disease leading to increased mortality of canopy trees. (Joslin
1992) In the U.S., terrestrial effects of acidification are best described for forested ecosystems
(especially red spruce and sugar maple ecosystems) with additional information on other plant
communities, including shrubs and lichen. (U.S. EPA 2020)
Both coniferous and deciduous forests throughout the eastern U.S. have experienced gradual
losses of base cation nutrients from the soil historically due to accelerated leaching from
acidifying deposition. This change in cation availability may reduce the quality of forest nutrition
over the long term. Evidence suggests that red spruce and sugar maple in some areas in the
eastern U.S. have experienced declining health because of this deposition. For red spruce (Picea
rubens), dieback or decline has been observed across high elevation landscapes of the
northeastern U.S. and, to a lesser extent, the southeastern U.S., and acidifying deposition has
been implicated as a causal factor. (DeHayes, et al. 1999)
6-42
-------
6.4.3.1.2 Ecological Effects from Nitrogen Enrichment
6.4.3.1.2.1 Aquatic Enrichment
Eutrophication in estuaries is associated with a range of adverse ecological effects including
low dissolved oxygen (DO), harmful algal blooms (HABs), loss of submerged aquatic vegetation
(SAV), and low water clarity. Low DO disrupts aquatic habitats, causing stress to fish and
shellfish, which, in the short-term, can lead to episodic fish kills and, in the long-term, can
damage overall growth in fish and shellfish populations. In addition to often being toxic to fish
and shellfish and leading to fish kills and aesthetic impairments of estuaries, HABs can, in some
instances, also be harmful to human health. SAV provides critical habitat for many aquatic
species in estuaries and, in some instances, can also protect shorelines by reducing wave
strength; therefore, declines in SAV due to nutrient enrichment are an important source of
concern. Low water clarity is in part the result of accumulations of both algae and sediments in
estuarine waters. In addition to contributing to declines in SAV, high levels of turbidity also
degrade the aesthetic qualities of the estuarine environment.
In estuaries, nitrogen from the atmosphere and other sources contributes to increased primary
productivity leading to eutrophication. An assessment of estuaries nationwide by the National
Oceanic and Atmospheric Administration (NOAA) concluded that 64 estuaries (out of 99 with
available data) suffered from moderate or high levels of eutrophication due to excessive inputs of
both nitrogen (N) and phosphorus. (Bricker, et al. 2007) Source apportionment data in the 2008
NOxSOx ISA and the 2020 NOxSOxPM ISA indicate that atmospheric contributions to estuarine
nitrogen are heterogeneous across the U.S., ranging from <10% to approximately 70% of total
estuary nitrogen inputs. (U.S. EPA 2020) Estuaries are an important source of food production,
in particular fish and shellfish. These complex systems are capable of supporting large stocks of
resident commercial species, and they serve as the breeding grounds and interim habitat for
several migratory species. Eutrophication in estuaries may also affect the demand for seafood
(after well-publicized toxic blooms), water-based recreation, and erosion protection provided by
SAV.
6.4.3.1.2.2 Terrestrial Enrichment
Terrestrial enrichment occurs when terrestrial ecosystems receive nitrogen loadings in excess
of natural background levels, through either atmospheric deposition or direct application.
Atmospheric nitrogen deposition is associated with changes in the types and number of species
and biodiversity in terrestrial systems. This occurs because increased nitrogen affects
competition between plant species, with certain species responding in growth more vigorously
than others, leading to overall declines in species richness. Nitrogen enrichment occurs over a
long time period; as a result, it may take as many as 50 years or more to see changes in
ecosystem conditions and indicators. One of the main provisioning services potentially affected
by nitrogen deposition is grazing opportunities offered by grasslands for livestock production in
the Central U.S. Although nitrogen deposition on these grasslands can offer supplementary
nutritive value and promote overall grass production, there are concerns that fertilization may
favor invasive grasses and shift the species composition away from native grasses. This process
may ultimately reduce the productivity of grasslands for livestock production.
Terrestrial enrichment also affects habitats, for example the Coastal Sage Scrub (CSS) and
Mixed Conifer Forest (MCF) habitats which are an integral part of the California landscape.
6-43
-------
Together the ranges of these habitats include the densely populated and valuable coastline and
the mountain areas. Numerous threatened and endangered species at both the state and federal
levels reside in CSS and MCF. Nutrient enrichment can also increase fire risk by encouraging the
growth of more flammable, non-native grasses, thereby increasing the fuel load and increasing
fire frequency.
6.4.3.1.3 Vegetation Effects Associated with Gaseous Sulfur Dioxide, Nitric
Oxide, Nitrogen Dioxide, Peroxyacetyl Nitrate, and Nitric Acid
Uptake of gaseous pollutants in a plant canopy is a complex process involving adsorption to
surfaces (leaves, stems, and soil) and absorption into leaves. These pollutants penetrate into
leaves through the stomata, although there is evidence for limited pathways via the cuticle. (U.S.
EPA 2020) Pollutants must be transported from the bulk air to the leaf boundary layer in order to
reach the stomata. When the stomata are closed, as occurs under dark or drought conditions,
resistance to gas uptake is very high and the plant has a very low degree of susceptibility to
injury. In contrast, mosses and lichens do not have a protective cuticle barrier to gaseous
pollutants or stomates and are generally more sensitive to gaseous sulfur and nitrogen than
vascular plants. (U.S. EPA 2020)
Acute foliar injury from SO2 usually happens within hours of exposure, involves a rapid
absorption of a toxic dose, and involves collapse or necrosis of plant tissues. Another type of
visible injury is termed chronic injury and is usually a result of variable SO2 exposures over the
growing season. Besides foliar injury, chronic exposure to low SO2 concentrations can result in
reduced photosynthesis, growth, and yield of plants. (U.S. EPA 2022) These effects are
cumulative over the season and are often not associated with visible foliar injury. As with foliar
injury, these effects vary among species and growing environment. SO2 is also considered the
primary factor causing the death of lichens in many urban and industrial areas. (Hutchinson,
Maynard and Geiser 1996)
Similarly, in sufficient concentrations, nitric oxide (NO), nitrogen dioxide (NO2),
peroxyacetyl nitrate (PAN), and nitric acid (HNO3) can have phytotoxic effects on plants such as
decreasing photosynthesis and inducing visible foliar injury. It is also known that these gases can
alter the nitrogen cycle in some ecosystems, especially in the western U.S., and contribute to
nitrogen saturation. Further, there are several lines of evidence that past and current HNO3
concentrations may be contributing to the decline in lichen species in the Los Angeles basin.
(Riddell, Nash and Padgett 2008)
6.4.3.1.4 Mercury Methylation
Mercury is a persistent, bioaccumulative toxic metal that is emitted in three forms: gaseous
elemental Hg (HgO), oxidized Hg compounds (Hg+2), and particle-bound Hg (HgP).
Methylmercury (MeHg) is formed by microbial action in the top layers of sediment and soils
after Hg has precipitated from the air and deposited into waterbodies or land. Once formed,
MeHg is taken up by aquatic organisms and bioconcentrates up the aquatic food web. Larger
predatory fish may have MeHg concentrations many times higher, typically on the order of one
million times, than the concentrations in the freshwater body in which they live. The NOx SOx
ISA—Ecological Criteria concluded that evidence is sufficient to infer a causal relationship
between sulfur deposition and increased mercury methylation in wetlands and aquatic
environments. (U.S. EPA 2020) Specifically, there appears to be a relationship between SO42"
6-44
-------
deposition and mercury methylation; however, the rate of mercury methylation varies according
to several spatial and biogeochemical factors whose influence has not been fully quantified.
Therefore, the correlation between SO42" deposition and MeHg cannot yet be quantified for the
purpose of interpolating the association across waterbodies or regions. Nevertheless, because
changes in MeHg in ecosystems represent changes in significant human and ecological health
risks, the association between sulfur and mercury cannot be neglected. (U.S. EPA 2020)
6.4.3.2 Deposition of Metallic and Organic Constituents of PM
Several significant ecological effects are associated with the deposition of chemical
constituents of ambient PM such as metals and organics. (U.S. EPA 2020, U.S. EPA 2009) The
trace metal constituents of PM can include cadmium, copper, chromium, mercury, nickel, zinc,
and lead. Organic pollutants that may be associated with PM encompass several chemical classes
including persistent organic pollutants (POPs), polyaromatic hydrocarbons (PAHs), and
polybrominated diphenyl ethers (PBDEs). Direct effects of exposures to PM may occur via
deposition (e.g., wet, dry or occult) to vegetation surfaces, while indirect exposure may occur via
deposition to soils or surface waters where the deposited constituents of PM then interact with
biota residing in these ecosystems. While both fine and coarse-mode particles have the potential
to affect plants and other organisms, more often the chemical constituents drive the ecosystem
response to PM. (Grantz, Garner and Johnson 2003) Ecological effects of PM include direct
effects on plant foliage and indirect effects such as contribution to total metal loading resulting in
alteration of soil biogeochemistry and microbial communities, reduced growth and reproduction
in plants and animals, and contribution to total loading of organics which bioaccumulate and
biomagnify in terrestrial and aquatic biota.
Particulate matter can adversely impact plants and ecosystem services provided by plants by
deposition to vegetative surfaces. (U.S. EPA 2020, U.S. EPA 2009) Particulates deposited on the
surfaces of leaves and needles can alter plant metabolism and photosynthesis by the blocking of
sunlight. PM deposition near sources of heavy deposition can obstruct stomata (limiting gas
exchange), and damage leaf surfaces. (U.S. EPA 2020, U.S. EPA 2009) Plants growing on
roadsides exhibit impact damage from near-road PM deposition, having higher levels of organics
and heavy metals, and accumulating salt from road de-icing during winter months. (U.S. EPA
2020) In addition, atmospheric PM can scatter direct solar radiation to diffuse radiation. (U.S.
EPA 2020) Decreases in crop yields (a provisioning ecosystem service) due to reductions in solar
radiation have been attributed to regional scale air pollution in counties with especially severe
regional haze. (Chameides, et al. 1999)
In addition to damage to plant surfaces, deposited PM can be taken up by plants from soil or
foliage. (U.S. EPA 2020, U.S. EPA 2009) Copper, zinc, and nickel have been shown to be
directly toxic to vegetation under field conditions. (U.S. EPA 2020) The ability of vegetation to
take up heavy metals is dependent upon the amount, solubility, and chemical composition of the
deposited PM. Uptake of PM by plants from soils and vegetative surfaces can disrupt
photosynthesis, alter pigments and mineral content, reduce plant vigor, decrease frost hardiness,
and impair root development.
Particulate matter can also contain organic air toxic pollutants, including PAHs, which are a
class of polycyclic organic matter (POM). PAHs can accumulate in soils, sediments, flora, and
fauna. The bioavailability of PM-associated organics is dependent upon the physical, chemical,
and biological conditions under which an organism is exposed at a particular geographic
6-45
-------
location. (U.S. EPA 2020) Different species can have different uptake rates of PAHs.
Biomagnification of organics has been extensively documented in aquatic and terrestrial
ecosystems for several decades, and these compounds are detected in biota at remote locations
due to long-range atmospheric transport processes. (U.S. EPA 2020)
Contamination of plant leaves by heavy metals can lead to elevated concentrations in the soil.
Trace metals absorbed into the plant, frequently by binding to the leaf tissue, and then are shed
when the leaf drops. As the fallen leaves decompose, the heavy metals are transferred into the
soil. (Cotrufo, et al. 1995, Niklinska, Laskowski and Maryanski 1998) Many of the major
indirect plant responses to PM deposition are chiefly soil-mediated and depend on the chemical
composition of individual components of deposited PM. Upon entering the soil environment, PM
pollutants can alter ecological processes of energy flow and nutrient cycling, inhibit nutrient
uptake to plants, change microbial community structure, and affect biodiversity. Accumulation
of heavy metals in soils depends on factors such as local soil characteristics, geologic origin of
parent soils, and metal bioavailability. Heavy metals such as zinc, copper, and cadmium, and
some pesticides can interfere with microorganisms that are responsible for decomposition of soil
litter, an important regulating ecosystem service that serves as a source of soil nutrients. (U.S.
EPA 2020) Surface litter decomposition is reduced in soils having high metal concentrations.
Soil communities have associated bacteria, fungi, and invertebrates that are essential to soil
nutrient cycling processes. Changes to the relative species abundance and community
composition are associated with deposited PM to soil biota. (U.S. EPA 2020)
Atmospheric deposition can be the primary source of some organics and metals to watersheds.
Deposition of PM to surfaces in urban settings increases the metal and organic component of
storm water runoff. (U.S. EPA 2020) This atmospherically-associated pollutant burden can then
be toxic to aquatic biota. The contribution of atmospherically deposited PAHs to aquatic food
webs was demonstrated in high elevation mountain lakes with no other anthropogenic
contaminant sources. (U.S. EPA 2020)Metals associated with PM deposition limit
phytoplankton growth, affecting aquatic trophic structure. The Western Airborne Contaminants
Assessment Project (WACAP) collected data on contaminant transport and PM depositional
effects on sensitive ecosystems in the Western U.S. (Landers, et al. 2008) In this project, the
transport, fate, and ecological impacts of anthropogenic contaminants from atmospheric sources
were assessed from 2002 to 2007 in seven ecosystem components (air, snow, water, sediment,
lichen, conifer needles, and fish) in eight core national parks. The study concluded that
bioaccumulation of semi-volatile organic compounds occurred throughout park ecosystems, an
elevational gradient in PM deposition exists with greater accumulation in higher altitude areas,
and contaminants accumulate in proximity to individual agriculture and industry sources, which
is counter to the original working hypothesis that most of the contaminants would originate from
Eastern Europe and Asia.
6.4.3.3 Materials Damage and Soiling
Building materials including metals, stones, cements, and paints undergo natural weathering
processes from exposure to environmental elements (e.g., wind, moisture, temperature
fluctuations, sunlight, etc.). Pollution can worsen and accelerate these effects. Deposition of PM
is associated with both physical damage (materials damage effects) and impaired aesthetic
qualities (soiling effects). Wet and dry deposition of PM can physically affect materials, adding
to the effects of natural weathering processes, by potentially promoting or accelerating the
6-46
-------
corrosion of metals, degrading paints and deteriorating building materials such as stone, concrete
and marble. (U.S. EPA 2020, U.S. EPA 2022) The effects of PM are exacerbated by the
presence of acidic gases and can be additive or synergistic depending on the complex mixture of
pollutants in the air and surface characteristics of the material. Acidic deposition has been shown
to have an effect on materials including zinc/galvanized steel and other metal, carbonate stone
(such as monuments and building facings), and surface coatings (paints). (Irving 1991) The
effects on historic buildings and outdoor works of art are of particular concern because of the
uniqueness and irreplaceability of many of these objects. In addition to aesthetic and functional
effects on metals, stone, and glass, altered energy efficiency of photovoltaic panels by PM
deposition is also becoming an important consideration for impacts of air pollutants on materials.
6.4.4 Welfare Effects of Air Toxics
Emissions from producing, transporting, and combusting fuel contribute to ambient levels of
pollutants that contribute to adverse effects on vegetation. VOCs, some of which are considered
air toxics, have long been suspected to play a role in vegetation damage. (U.S. EPA 1991) In
laboratory experiments, a wide range of tolerance to VOCs has been observed. (Cape, et al.
2003) Decreases in harvested seed pod weight have been reported for the more sensitive plants,
and some studies have reported effects on seed germination, flowering, and fruit ripening.
Effects of individual VOCs or their role in conjunction with other stressors (e.g., acidification,
drought, temperature extremes) have not been well studied. In a recent study of a mixture of
VOCs including ethanol and toluene on herbaceous plants, significant effects on seed production,
leaf water content, and photosynthetic efficiency were reported for some plant species. (Cape, et
al. 2003)
Research suggests an adverse impact of vehicle exhaust on plants, which has in some cases
been attributed to aromatic compounds and in other cases to NOx. (Viskari 2000, Ugrekhelidze,
Korte and Kvesitadze 1997, Kammerbauer, et al. 1987) The impacts of VOCs on plant
reproduction may have long-term implications for biodiversity and survival of native species
near major roadways. Most of the studies of the impacts of VOCs on vegetation have focused on
short-term exposure and few studies have focused on long-term effects of VOCs on vegetation
and the potential for metabolites of these compounds to affect herbivores or insects.
6.5 Criteria Pollutant Human Health Benefits
The light-duty passenger cars and light trucks and medium-duty vehicles subject to the final
standards are significant sources of mobile source air pollution, including directly-emitted PM2.5
as well as NOx and VOC emissions (both precursors to ozone formation and secondarily-formed
PM2.5). The final program will reduce exhaust emissions of these pollutants from the regulated
vehicles, which will in turn reduce ambient concentrations of ozone and PM2.5. Emissions from
upstream sources will likely increase in some cases (e.g., power plants) and decrease in others
(e.g., refineries). We project that in total, the final standards will result in substantial net
reductions of emissions of pollutants like PM2.5, NOx, and VOCs and a net increase in emissions
of SO2. Emissions changes attributable to the final standards are presented in Chapter 8 of the
RIA. Exposures to ambient pollutants such as PM2.5 and ozone are linked to adverse
environmental and human health impacts, such as premature deaths and non-fatal illnesses (as
explained in Chapters 6.2 and 6.3). Reducing human exposure to these pollutants results in
significant and measurable health benefits.
6-47
-------
Changes in ambient concentrations of ozone, PM2.5, and air toxics that will result from the
standards are expected to improve human health by reducing premature deaths and other serious
human health effects, and they are also expected to result in other important improvements in
public health and welfare (see Chapters 6.2 and 6.3). Children, especially, benefit from reduced
exposures to criteria and toxic pollutants because they tend to be more sensitive to the effects of
these respiratory pollutants. Ozone and particulate matter have been associated with increased
incidence of asthma and other respiratory effects in children, and particulate matter has been
associated with a decrease in lung maturation.
This section discusses the economic benefits from reductions in adverse health and
environmental impacts resulting from criteria pollutant emission reductions that can be expected
to occur as a result of the final emission standards. When feasible, EPA conducts full-scale
photochemical air quality modeling to demonstrate how its national mobile source regulatory
actions affect ambient concentrations of regional pollutants throughout the United States. The
estimation of the human health impacts of a regulatory action requires national-scale
photochemical air quality modeling to conduct a full-scale assessment of PM2.5 and ozone-
related health benefits.
EPA conducted an air quality modeling analysis of a regulatory scenario involving light- and
medium-duty vehicle emission reductions and corresponding changes in "upstream" emission
sources like EGU (electric generating unit) emissions and refinery emissions (see RIA Chapter
8). Decisions about the emissions and other elements used in the air quality modeling were made
early in the analytical process for this rulemaking. Accordingly, the air quality analysis reflects
the impacts of a policy scenario that is slightly different than the final standards, however, we
view the results of the modeling analysis as the best representation of the final rulemaking's
impacts on PM2.5 and ozone in 2055. For a complete description of the modeled air quality
scenario and the results of that analysis, including a full analysis of PM2.5- and ozone-related
health benefits in 2055, see Chapter 7. Because the air quality analysis was only conducted for
one future year (2055), a year when the regulatory scenario will be fully implemented and when
most of the regulated fleet will have turned over, we used the OMEGA-based emissions analysis
(see RIA Chapter 8) and benefit-per-ton (BPT) values to estimate the criteria pollutant (PM2.5)
health benefits of the final and alternative standards.
The BPT approach estimates the monetized economic value of PM2.5-related emission
reductions or increases (such as direct PM, NOx, and SO2) due to implementation of the final
program. Similar to the SC-GHG approach for monetizing reductions in GHGs, the BPT
approach monetizes health benefits of avoiding one ton of PIVh.s-related emissions from a
particular onroad mobile or upstream source. The value of health benefits from reductions (or
increases) in PM2.5 emissions associated with the standards were estimated by multiplying PM2.5-
related BPT values by the corresponding annual reduction (or increase) in tons of directly-
emitted PM2.5 and PM2.5 precursor emissions (NOx and SO2).
The BPT approach monetizes avoided premature deaths and illnesses that are expected to
occur as a result of reductions in directly-emitted PM2.5 and PM2.5 precursors. A chief limitation
to using PM2.5-related BPT values is that they do not reflect benefits associated with reducing
ambient concentrations of ozone, direct exposure to NO2, or exposure to mobile source air
toxics, nor do they account for improved ecosystem effects or visibility. The estimated benefits
6-48
-------
of the standards would be larger if we were able to monetize these unquantified benefits at this
time.
Using the BPT approach, we estimate the annualized value of PIVh.s-related benefits for the
final program between 2027 and 2055 (discounted back to 2027) is $5.3 to $11 billion assuming
a 3-percent discount rate and $3.7 to $7.2 billion assuming a 7-percent discount rate. Benefits are
reported in year 2022 dollars and reflect the PIVh.s-related benefits associated with reductions in
NOx, SO2, and direct PM2.5 emissions. Because premature mortality typically constitutes the vast
majority of monetized benefits in a PM2.5 benefits assessment, we present a range of PM benefits
based on risk estimates reported from two different long-term exposure studies using different
cohorts to account for uncertainty in the benefits associated with avoiding PM-related premature
deaths. Tables of the monetized PM2.5-related benefits of the standards can be found in RIA
Chapter 9.
6.5.1 Approach to Estimating Human Health Benefits
This section summarizes EPA's approach to estimating the economic value of the PM--
related benefits for this rulemaking. We use a BPT approach that is conceptually consistent with
EPA's use of BPT estimates in its regulatory analyses (U.S. EPA 2018) (U.S. EPA 2023). In this
approach, the PM2.5-related BPT values are the total monetized human health benefits (the sum
of the economic value of the reduced risk of premature death and illness) that are expected from
reducing one ton of NOx, SO2, or directly-emitted PM2.5.
The mobile sector BPT estimates used in this analysis were published in 2019 but have been
updated to be consistent with the health benefits Technical Support Document (Benefits TSD)
that accompanied the 2023 PM NAAQS Reconsideration Proposal. (Wolfe, et al. 2019) (U.S.
EPA 2022) (U.S. EPA 2023). The Benefits TSD details the approach used to estimate the PM2.5-
related benefits reflected in these BPTs. The upstream Refinery and EGU BPT estimates used in
this analysis were also recently updated to be consistent with the Benefits TSD (U.S. EPA 2023).
We multiply these BPT values by national reductions in annual emissions in tons to estimate the
total monetized human health benefits associated with the standards.
Our procedure for calculating BPT values follows three steps:
1. Using source apportionment photochemical modeling, predict annual average ambient
concentrations of NOx, SO2, and primary PM2.5 that are attributable to each source sector
(Onroad Heavy-Duty Diesel, Onroad Heavy Duty Gas, Refineries, and Electricity Generating
Units), for the Continental U.S. (48 states). This yields the estimated ambient pollutant
concentrations to which the U.S. population is exposed.
2. For each sector, estimate the health impacts, and economic value of those impacts,
associated with the attributable ambient concentrations of NOx, SO2, and primary PM2.5 using
the environmental Benefits Mapping and Analysis Program-Community Edition (BenMAP-CE)
6-49
-------
(U.S. EPA 2023).233 This yields the estimated total monetized value of health effects associated
with exposure to the relevant pollutants by sector.
3. For each sector, divide the monetary value of health impacts by the inventory of associated
precursor emissions. That is, primary PM2.5 benefits for a given sector are divided by direct
PM2.5 emissions from that same sector, sulfate benefits are divided by SO2 emissions, and nitrate
benefits are divided by NOx emissions. This yields the estimated monetary value of one ton of
sector-specific direct PM2.5, SO2, or NOx emissions.
The quantified and monetized PM2.5 health categories that are included in the BPT values are
summarized in Table 6-2.
Chapter 7.4.6 lists the ozone, PM2.5, SO2, and NOx health and welfare categories that are not
quantified and monetized by the BPT approach and are therefore not included in the estimated
benefits analysis for this rulemaking.
233 BenMAP-CE is an open-source computer program developed by the EPA that calculates the number and
economic value of air pollution-related deaths and illnesses. The software incorporates a database that includes
many of the concentration-response relationships, population files, and health and economic data needed to quantify
these impacts. Information on BenMAP is found at: https://www.epa.gov/benmap/benmap-community-edition, and
the source code is available at: https://github.eom/BenMAPCE/BenMAP-CE.
6-50
-------
Table 6-2 Human Health Effects of PM2.5
Effect (age)
Effect
Effect
More
Quantified
Monetized
Information
Adult premature mortality based on cohort study
/
/
PM ISA
estimates (>17 or >64)
Infant mortality (<1)
/
/
PM ISA
Non-fatal heart attacks (>18)
/
/
PM ISA
Hospital admissions - cardiovascular (all)
sf
sf
PM ISA
Hospital admissions - respiratory (<19 and >64)
/
/
PM ISA
Hospital admissions - Alzheimer's disease (>64)
sf
sf
PM ISA
Hospital admissions - Parkinson's disease (>64)
/
/
PM ISA
Emergency department visits - cardiovascular (all)
sf
sf
PM ISA
Emergency department visits - respiratory (all)
/
/
PM ISA
Emergency hospital admissions (>65)
sf
sf
PM ISA
Non-fatal lung cancer (>29)
/
/
PM ISA
Stroke incidence (50-79)
sf
sf
PM ISA
New onset asthma (< 12)
/
y
PM ISA
Exacerbated asthma - albuterol inhaler use
sf
sf
PM ISA
(asthmatics. 6-13)
Lost work days (18-64)
PM ISA
Other cardiovascular effects (e.g.. doctor's visits.
—
—
PM ISA1
prescription medication)
Other respiratory effects (e.g.. pulmonary function.
—
—
PM ISA1
other ages)
Other cancer effects (e.g.. mutagenicity.
—
—
PM ISA1
genotoxicity)
Other nervous system effects (e.g.. dementia)
—
—
PM ISA1
Metabolic effects (e.g.. diabetes, metabolic
—
—
PM ISA1
syndrome)
Reproductive and developmental effects (e.g.. low
—
—
PM ISA1
birth weight, pre-term births)
^ We assess these benefits qualitatively due to epidemiological or economic data limitations.
Of the PM-related health endpoints listed in Table 6-2, EPA estimates the incidence of air
pollution effects for only those classified as either "causal" or "likely-to-be-causal" in the 2019
PM Integrated Science Assessment (ISA) and the 2022 PM ISA update (U.S. EPA 2019) (U.S.
EPA 2022).234 The full complement of human health effects associated with PM remains
unquantified because of current limitations in methods or available data. Thus, our quantified
PM-related benefits omit a number of known or suspected health effects linked with PM, either
because appropriate health impact functions are not available or because outcomes are not easily
interpretable (e.g., changes in heart rate variability).
234 The ISA synthesizes the toxicological, clinical, and epidemiological evidence to determine whether each
pollutant is causally related to an array of adverse human health outcomes associated with either acute (i.e., hours-
or days-long) or chronic (i.e. years-long) exposure. For each outcome, the ISA reports this relationship to be causal,
likely to be causal, suggestive of a causal relationship, inadequate to infer a causal relationship, or not likely to be a
causal relationship.
6-51
-------
We anticipate the standards will also yield benefits from reduced exposure to ambient
concentrations of ozone. However, the complex, non-linear photochemical processes that govern
ozone formation prevent us from developing reduced-form ozone BPT values for mobile
sources. The BPT approach also omits health effects associated with ambient concentrations of
NO2 as well as criteria pollutant-related welfare effects such as improvements in visibility,
reductions in materials damage, ecological effects from reduced PM deposition, ecological
effects from reduced nitrogen emissions, and vegetation effects from reduced ozone exposure.
We also do not provide estimated monetized benefits due to reductions in mobile source air
toxics. This is primarily because currently available tools and methods to assess air toxics risk
from mobile sources at the national scale are not adequate for extrapolation to incidence
estimation or benefits assessment.
6.5.2 Estimating PM2.5-attributable Adult Premature Death
Of the PM2.5-related health endpoints listed in Table 6-2, adult premature deaths typically
account for the majority of total monetized PM benefits and are thus the primary component of
the PM2.5-related BPT values. In this section, we provide more detail on PM mortality effect
coefficients and the concentration-response functions that underlie the BPT values.
A substantial body of published scientific literature documents the association between PM2.5
concentrations and the risk of premature death (U.S. EPA 2019) (U.S. EPA 2022). This body of
literature reflects thousands of epidemiology, toxicology, and clinical studies. The PM ISA,
completed as part of the review of the recently finalized PM NAAQS Reconsideration and
reviewed by the Clean Air Scientific Advisory Committee (CASAC) (Sheppard 2022),
concluded that there is a causal relationship between mortality and both long-term and short-term
exposure to PM2.5 based on the full body of scientific evidence. The size of the mortality effect
estimates from epidemiologic studies, the serious nature of the effect itself, and the high
monetary value ascribed to prolonging life make mortality risk reduction the most significant
health endpoint quantified in this analysis. EPA selects Hazard Ratios235 from cohort studies to
estimate counts of PM-related premature death, following a systematic approach detailed in the
Benefits TSD.
For adult PM-related mortality, the BPT values are based on the risk estimates from two
alternative long-term exposure mortality studies: the National Health Interview Survey (NHIS)
cohort study (Pope III et al. 2019) and an extended analysis of the Medicare cohort (Wu et al.
2020). In past analyses, EPA has used two alternate estimates of mortality: one from the
American Cancer Society cohort and one from the Medicare cohort (Turner 2016) (Di 2017)
respectively. We use a risk estimate from the Pope III et al., 2019 study in place of the risk
estimate from the Turner et al., 2016 analysis, as it: (1) includes a longer follow-up period that
includes more recent (and lower) PM2.5 concentrations; (2) the NHIS cohort is more
representative of the U.S. population than is the ACS cohort with respect to the distribution of
individuals by race, ethnicity, income, and education.
235 A Hazard Ratio is a measure of how often a particular event happens in one group compared to how often it
happens in another group, such as mortality associated with exposure to PM2.5.
6-52
-------
Based on the 2022 Supplement to the PM ISA, EPA substituted a risk estimate from Wu et
al., 2020 in place of a risk estimate from Di et al., 2017. These two epidemiologic studies share
many attributes, including the cohort and model used to characterize population exposure to
PM2.5. As compared to Di et al., 2017, Wu et al., 2020 includes a longer follow-up period and
reflects more recent PM2.5 concentrations.
The PM ISA also concluded that the scientific literature supports the use of a no-threshold
log-linear model to portray the PM-mortality concentration-response relationship while
recognizing potential uncertainty about the exact shape of the concentration-response
relationship. The 2019 PM ISA, which informed the final 2024 PM NAAQS Reconsideration,
reviewed available studies that examined the potential for a population-level threshold to exist in
the concentration-response relationship. Based on such studies, the ISA concluded that "evidence
from recent studies reduce uncertainties related to potential co-pollutant confounding and
continues to provide strong support for a linear, no-threshold concentration-response
relationship." Consistent with this evidence, the Agency historically has estimated health impacts
above and below the prevailing NAAQS.
6.5.3 Economic Value of Health Benefits
The BPT values used in this analysis are a reduced-form approach for relating emission
reductions to reductions in ambient concentrations of PM2.5 and associated improvements in
human health. Reductions in ambient concentrations of air pollution generally decrease the risk
of future adverse health effects by a small amount for a large population. To monetize these
benefits, the appropriate economic measure is willingness to pay (WTP) for changes in risk of a
health effect. For some health effects, such as hospital admissions, WTP estimates are generally
not available, so we use the cost of treating or mitigating the effect. These cost-of-illness (COI)
estimates generally (although not necessarily in every case) understate the true value of
reductions in risk of a health effect. They tend to reflect the direct expenditures related to
treatment, but not the value of avoided pain and suffering from the health effect. The WTP and
COI unit values for each endpoint are provided in the Benefits TSD. These unit values were used
to monetize the underlying health effects included in the PM2.5 BPT values.
Avoided premature deaths typically account for the majority of monetized PM2.5-related
benefits. The economics literature concerning the appropriate methodology for valuing
reductions in premature mortality risk is still developing and is the subject of continuing
discussion within the economics and public policy analysis community. Following the advice of
the SAB's Environmental Economics Advisory Committee (SAB-EEAC), EPA currently uses
the value of statistical life (VSL) approach in calculating estimates of mortality benefits. This
calculation provides the most reasonable single estimate of an individual's WTP for reductions
in mortality risk (U.S. EPA-SAB 2000). The VSL approach is a summary measure for the value
of small changes in mortality risk experienced by a large number of people.
EPA consulted several times with the SAB-EEAC on valuing mortality risk reductions and
continues work to update the Agency's guidance on the issue. Until updated guidance is
available, EPA determined that a single, peer-reviewed estimate applied consistently best reflects
the SAB-EEAC advice we have received. Therefore, EPA applies the VSL that was vetted and
endorsed by the SAB in the Agency's Guidelines for Preparing Economic Analyses (U.S. EPA
2016). This VSL value is the mean of the values reported in 26 labor market and contingent
valuation studies published between 1974 and 1991. The mean VSL across these studies is $4.8
6-53
-------
million (1990$). We then adjust this VSL to account for the currency year and to account for
income growth from 1990 to the analysis year. Specifically, the VSL applied in this analysis in
2022 dollars after adjusting for income growth to 2022 is $12.6 million.
EPA is committed to using scientifically sound, appropriately reviewed evidence in valuing
changes in the risk of premature death and continues to engage with the SAB to identify
scientifically sound approaches to update its mortality risk valuation estimates. Most recently,
the Agency proposed new meta-analytic approaches for updating its estimates, which were
subsequently reviewed by the SAB-EEAC (U.S. EPA 2017). EPA is taking the SAB's formal
recommendations under advisement.
6.5.4 Dollar Value per Ton of Directly-Emitted PM2.5 and PM2.5 Precursors
The value of health benefits from reductions in PM2.5 emissions associated with the standards
were estimated by multiplying PM2.5-related BPT values by the corresponding annual reduction
in tons of directly-emitted PM2.5 and PM2.5 precursor emissions (NOx and SO2). As explained
above, the PM2.5 BPT values represent the monetized value of human health benefits, including
reductions in both premature mortality and nonfatal illnesses. Table 6-3 presents the PM2.5 BPT
values estimated from two different PM-related premature mortality cohort studies, Wu et al.,
2020 (the Medicare cohort study) and Pope III et al., 2019 (the NHIS cohort study). The table
reports different values by source and pollutant because different pollutant emissions do not
equally contribute to ambient PM2.5 formation and different emissions sources do not equally
contribute to population exposure and associated health impacts. BPT values are also estimated
using either a 3 percent or 7 percent discount rate to account for a "cessation" lag between the
change in PM exposures and the total realization of changes in mortality effects. The source
sectors include: onroad light-duty gasoline cars, onroad light-duty gasoline trucks, onroad light-
duty diesel cars/trucks, electricity generating units, and refineries. We note that reductions in
medium-duty vehicle emissions are monetized using light-duty BPT values.
Detailed tables of the monetized PM2.5-related benefits of the standards can be found in RIA
Chapter 9.
6-54
-------
Table 6-3: PM2.5-related Benefit Per Ton values (2022$) associated with the reduction of
NOx, SO2 and directly emitted PM2.5 emissions for (A) Onroad light-duty gasoline cars, (B)
Onroad light-duty gasoline trucks, (C) Onroad light-duty diesel cars/trucks, (D) Electricity
Generating Units, and (E) Refineries.
A. Oiii'oihI Light-Duty Giisolinc Oils
NOX S02
Direct PM
: 3% Discount Rate
7% Discount Rate
3% Discount Rate
7% Discount Rate
3% Discount Rate
7% Discount Rate
Wu
Pope
Wu
: Pope
Wu
Pope
Wu
Pope
Wu
Pope
Wu
Pope
2025
$7,230
i $15,400
$6,490
i$13,800
$128,000
$274,000
$115,000 !
$246,000
$709,000
$1,520,000
$637,000 :
$1,360,000
2030
: $8,160
: $16,800
$7,330
: $15,100
; $147,000
$303,000
: $132,000 :
$273,000
; $814,000 :
$1,680,000
$731,000
$1,510,000
2035
i $9,200
: $18,500
$8,260
= $16,600
: $169,000 ;
$341,000
: $152,000
$307,000
: $939,000 i
$1,890,000
$843,000 !
$1,700,000
2040
: $10,100
s $19,900
$9,050
: $17,900
: $191,000
$378,000
$172,000 :
$340,000
$1,060,000
$2,100,000
$953,000 i
$1,890,000
2045
i $10,700
$21,000
$9,640
> $18,900
$211,000 ;
$413,000
$190,000 ;
$371,000
$1,170,000
$2,290,000
$1,050,000 i
$2,060,000
2050
$11.200
; $21,600
$10,000
$19,500
$229,000
$443,000
$206,000 ;
$398,000
: $1,270,000
$2,450,000
$1,140,000 ;
$2,200,000
2055
; $11,700
: $22,500
$10,500
; $20,300
$249,000
$477,000
$224,000
$429,000
: $1,370,000
$2,630,000
$1,240,000 :
$2,360,000
NOX
IS. Oiii'oihI Light-Duty Giisolinc Trucks
S02
Direct PM
i 3% Discount Rate
7% Discount Rate
3% Discount Rate
7% Discount Rate
3% Discount Rate
7% Discount Rate
Wu
Pope
Wu
: Pope
Wu
Pope
Wu
Pope
Wu
Pope
Wu
Pope
2025
i $6,550
$13,900
$5,880
: $12,500
$102,000
$219,000
$91,700
$197,000
$597,000
$1,280,000 ;
$536,000 :
$1,150,000
2030
: $7,400
$15,200
$6,640
: $13,700
; $117,000 :
$243,000
: $105,000 ;
$218,000
$685,000 :
$1,420,000
$615,000 :
$1,270,000
2035
$8,360
; $16,800
$7,510
$15,100
; $135,000
$272,000
; $121,000 ;
$245,000
$789,000
$1,590,000
$708,000
$1,430,000
2040
: $9,190
: $18,200
$8,250
: $16,400
: $152,000
$302,000
$137,000 :
$271,000
: $889,000 :
$1,760,000 :
$798,000
$1,580,000
2045
: $9,820
$19,200
$8,820
; $17,300
$168,000 =
$329,000
$151,000 ;
$296,000
: $979,000 ;
$1,910,000 :
$880,000
$1,720,000
2050
: $10,300
$19,900
$9,220
; $17,900
; $182,000 :
$352,000
$163,000 ;
$316,000
$1,060,000:
$2,040,000 ;
$950,000
$1,840,000
2055
i $10,800
$20,800
$9,700
: $18,700
$197,000
$378,000
$177,000 :
$340,000
: $1,140,000
$2,190,000 :
$1,030,000 :
$1,970,000
NOX
Oiii'oihI Light-Duty Diesel Curs/'Trucks
S02
Direct PM
: 3% Discount Rate
7% Discount Rate
3% Discount Rate
7% Discount Rate
3% Discount Rate
7% Discount Rate
: Wu
Pope
Wu
: Pope
Wu
Pope
Wu
Pope
Wu
Pope
Wu
Pope
2025
$5,790
$12,300
$5,200
: $11,100
$305,000 :
$655,000
$274,000
$589,000 ;
$489,000
: $1,050,000 :
$439,000 :
$942,000
2030
: $6,550
$13,500
$5,880
; $12,100
$349,000
$725,000
$314,000
$652,000
$560,000
; $1,160,000 :
$503,000 :
$1,040,000
2035
: $7,400
$14,900
$6,640
; $13,400
$402,000
$813,000
$361,000
$731,000 ;
$646,000
, $1,300,000
$580,000 :
$1,170,000
2040
5 $8,130
$16,100
$7,310
: $14,500
$453,000
$900,000
$407,000
$810,000 :
$728,000
. $1,440,000
$654,000
$1,300,000
2045
: $8,700
$17,000
$7,820
; $15,300
$500,000 :
$980,000
$449,000
$882,000 i
$803,000
: $1,570,000
$721,000
$1,410,000
2050
: $9,100
$17,700
$8,180
> $15,900
: $541,000 =
$1,050,000
$486,000
$944,000 :
$868,000
: $1,680,000
$780,000 :
$1,510,000
2055
: $9,570
$18,400
$8,600
: $16,600
: $587,000
$1,130,000
: $528,000
$1,010,000:
$939,000
: $1,800,000
$844,000
$1,620,000
NOX
: 3% Discount Rate 7% Discount Rate :
I). ICIcctricitv Gciicriiting I nits
S02
Direct PM
3% Discount Rate
7% Discount Rate
3% Discount Rate
7% Discount Rate
2025
2030
2035
2040
Wu
$7,470
: $8,370
$9,370
: $10,200
Pope
$15,800 i
$17,100 :
$18,700
$20,000
Wu
$6,710
$7,530
$8,420
$9,130
Pope
$14,200
$15,400 ;
$16,900 ;
$18,000 =
Wu
$55,200
$62,300
$69,900
$76,400
Pope
$118,000
$129,000
$141,000
$152,000
Wu
$49,700
$56,000
$62,900
$68,700
Pope
$106,000
$116,000
$127,000
$136,000
Wu
$110,000
$125,000
$142,000
$158,000
Pope
$235,000
$258,000
$287,000
$314,000
Wu
$98,400
$112,000
$128,000
$142,000
Pope
$211,000
$232,000
$258,000
$283,000
I'.. Refineries
NOX
: 3% Discount Rate 7% Discount Rate
S02
Direct PM
3% Discount Rate
7% Discount Rate
3% Discount Rate
7% Discount Rate
Wu : Pope Wu : Pope Wu Pope Wu Pope Wu Pope Wu Pope
2025 : $22,500 ! $48,300 ; $20,200 ! $43,400 $49,600 ; $107,000 : $44,500 ; $96,400 $358,000 ; $776,000 : $322,000 ; $698,000
2030 ; $24,800 $51,500 $22,300 : $46,300 $54,800 : $114,000 $49,200 : $103,000 $395,000 : $826,000 : $355,000 : $743,000
2035 : $28,500 $57,500 $25,600 : $51,800 = $62,700 : $127,000 $56,400 > $115,000 $453,000 $923,000 * $407,000 $831,000
2040 $31,900 ; $63,300 $28,700 : $56,900 $70,100 s $140,000 $63,000 ; $126,000 : $509,000 - $1,020,000 $458,000 $915,000
Notes: All estimates are rounded to three significant figures. The benefit-per-ton estimates presented in this table assume either a 3 percent or 7 percent
discount rate in the valuation of premature mortality to account for a twenty-year segmented cessation lag. Benefit-per-ton values were estimated for
years 2025, 2030, 2035, 2040, 2045, 2050 and 2055 for mobile sources, and for years 2025, 2030, 2035 and 2040 for EGUs and refineries. We hold
values constant for intervening years (e.g., the 2025 values are assumed to apply to years 2021 -2024, and so on). We hold 2040 values constant out to
2055 for EGUs and Refineries.
6-55
-------
6.5.5 Characterizing Uncertainty in the Estimated Benefits
There are likely to be sources of uncertainty in any complex analysis using estimated
parameters and inputs from numerous models, including this analysis. The Benefits TSD details
our approach to characterizing uncertainty in both quantitative and qualitative terms. That TSD
describes the sources of uncertainty associated with key input parameters including emissions
inventories, air quality data from models (with their associated parameters and inputs),
population data, population estimates, health effect estimates from epidemiology studies,
economic data for monetizing benefits, and assumptions regarding the future state of the country
(i.e., regulations, technology, and human behavior). Each of these inputs is uncertain and affects
the size and distribution of the estimated benefits. When the uncertainties from each stage of the
analysis are compounded, even small uncertainties can have large effects on the total quantified
benefits.
The BPT approach is a simplified approach that relies on additional assumptions and has its
own limitations, some of which are described in Chapter 6.5.6. Additional uncertainties related
to key assumptions underlying the estimates for PM2.5-related premature mortality described in
Section 6.3.1.2 of this chapter include the following:
• We assume that all fine particles, regardless of their chemical composition, are equally
potent in causing premature mortality. This is an important assumption because PM2.5
varies considerably in composition across sources, but the scientific evidence is not
yet sufficient to allow differentiation of effect estimates by particle type. The PM ISA,
which was reviewed by CASAC, concluded that "across exposure durations and health
effects categories ... the evidence does not indicate that any one source or component
is consistently more strongly related with health effects than PM2.5 mass." (U.S. EPA
2019)
• We assume that the health impact function for fine particles is log-linear down to the
lowest air quality levels modeled in this analysis. Thus, the estimates include health
benefits from reducing fine particles in areas with varied concentrations of PM2.5,
including both regions that are in attainment with the fine particle standard and those
that do not meet the standard down to the lowest modeled concentrations. The PM ISA
concluded that "the majority of evidence continues to indicate a linear, no-threshold
concentration-response relationship for long-term exposure to PM2.5 and total
(nonaccidental) mortality." (U.S. EPA 2019)
• We assume that there is a "cessation" lag between the change in PM exposures and the
total realization of changes in mortality effects. Specifically, we assume that some of
the incidences of premature mortality related to PM2.5 exposures occur in a distributed
fashion over the 20 years following exposure based on the advice of the SAB-HES,
which affects the valuation of mortality benefits at different discount rates. The above
assumptions are subject to uncertainty (U.S. EPA-SAB 2005). Similarly, we assume
there is a cessation lag between the change in PM exposures and both the development
and diagnosis of lung cancer.
6-56
-------
6.5.6 Benefit-per-Ton Estimate Limitations
All BPT estimates have inherent limitations. One limitation of using the PM2.5-related BPT
approach is an inability to provide estimates of the health and welfare benefits associated with
exposure to ozone, welfare benefits, and some unquantified health benefits associated with
PM2.5, as well as health and welfare benefits associated with ambient NO2 and SO2.
Table 6-4 presents a selection of unquantified criteria pollutant health and welfare benefits
categories. Another limitation is that the mobile sector-specific air quality modeling that
underlies the PM2.5 BPT values did not provide estimates of the PM2.5-related benefits associated
with reducing VOC emissions, but these unquantified benefits are generally small compared to
benefits associated with other PM2.5 precursors.
6-57
-------
Table 6-4: Unquantified Health and Welfare Benefits Categories
( iitijiorv I IH|U;|ntitled I tTict
Nonfatal morbidity from exposure to ozone
Reduced incidence of morbidity from
exposure to N02
Reduced visibility impairment
Reduced effects on materials
Reduced effects from PM deposition (metals
and organics)
Reduced vegetation and ecosystem effects
from exposure to ozone
Improved Ilummi Ileiilth
Premature respiratory mortality from short-
term exposure (0-99)
Premature respiratory mortality from long-
term exposure (age 30-99)
Hospital admissions—respiratory (ages 65-
99)
Emergency department visits—respiratory
(ages 0-99)
Asthma onset (0-17)
Asthma symptoms/exacerbation (asthmatics
age 5-17)
Allergic rhinitis (hav fever) svmptoms (ages
3-17)
Minor restricted-activity days (age 18-65)
School absence days (age 5-17)
Decreased outdoor worker productivity (age
18-65)
Metabolic effects (e.g., diabetes)
Other respiratory effects (e.g., premature
aging of lungs)
Cardiovascular and nervous system effects
Reproductive and developmental effects
Asthma hospital admissions
Chronic lung disease hospital admissions
Respiratory emergency department visits
Asthma exacerbation
Acute respiratory symptoms
Premature mortality
Other respiratory effects (e.g., airway
hyperresponsiveness and inflammation, lung
function, other ages and populations)
Improved l.nvironnunt
Visibility in Class 1 areas
Visibility in residential areas
Household soiling
Materials damage (e.g., corrosion, increased
wear)
Effects on individual organisms and
ecosystems
Visible foliar injury on vegetation
Reduced vegetation growth and reproduction
Yield and quality of commercial forest
products and crops
Damage to urban ornamental plants
Carbon sequestration in terrestrial
ecosystems
Recreational demand associated with forest
aesthetics
Other non-use effects
Ecosystem functions (e.g., water cycling,
biogeochemical cycles, net primary
productivity, leaf-gas exchange, community
composition)
More
Information
Ozone ISA3
Ozone ISA3
Ozone ISA3
Ozone ISA3
Ozone ISA3
Ozone ISA3
Ozone ISA3
Ozone ISA3
Ozone ISA3
Ozone lSAb
Ozone lSAb
Ozone lSAb
Ozone lSAb
Ozone ISa'3
N02 lSAa
N02 lSAa
N02 lSAa
N02 lSAa
N02 lSAa
. a,b,c
N02 ISA
N02 ISA1
b,c
PM ISA
PM lSAa
PM lSAa'b
PM lSAb
PM lSAb
Ozone ISA3
Ozone ISA3
Ozone ISA3
Ozone lSAb
Ozone ISA3
Ozone lSAb
Ozone ISA
Ozone ISA
6-58
-------
C iltlJJOIT
liiH|uiiiitifk'(l littcct
More
Infoi'miition
Reduced effects from acid deposition
Recreational fishing
NOX SOX ISAa
Tree mortality and decline
NOX SOX ISAb
Commercial fishing and forestry effects
NOX SOX ISAb
Recreational demand in terrestrial and
NOX SOX ISAb
aquatic ecosystems
Other non-use effects
NOX SOX ISAb
Ecosystem functions (e.g., biogeochemical
NOX SOX ISAb
cycles)
Reduced effects from nutrient enrichment
Species composition and biodiversity in
NOX SOX ISAb
terrestrial and estuarine ecosystems
Coastal eutrophication
NOX SOX ISAb
Recreational demand in terrestrial and
NOX SOX ISAb
estuarine ecosystems
Other non-use effects
NOX SOX ISAb
Ecosystem functions (e.g., biogeochemical
NOX SOX ISAb
cycles, fire regulation)
Reduced vegetation effects from ambient
Injury to vegetation from S02 exposure
NOX SOX ISAb
exposure to S02 and NOx
Injury to vegetation from NOx exposure
NOX SOX ISAb
a We assess these benefits qualitatively due to data and resource limitations for this RIA.
b We assess these benefits qualitatively because we do not have sufficient confidence in available data or methods.
: ° We assess these benefits qualitatively because current evidence is only suggestive of causality or there are other significant concerns over
: the strength of the association.
There are also benefits associated with reductions in air toxic pollutant emissions that would
result from the program (see RIA Chapter 6.3.7) but that the PIVh.s-related BPT approach does
not capture. While EPA continues to work to improve its benefits estimation tools, there remain
critical limitations for estimating incidence and assessing benefits of reducing air toxics.
National-average BPT values reflect the geographic distribution of the underlying modeled
emissions used in their calculation, which may not exactly match the geographic distribution of
the emission reductions that would occur due to a specific rulemaking. Similarly, BPT estimates
may not reflect local variability in population density, meteorology, exposure, baseline health
incidence rates, or other local factors for any specific location. For instance, even though we
assume that all fine particles have equivalent health effects, the BPT estimates vary across
precursors depending on the location and magnitude of their impact on PM2.5 levels, which
drives population exposure. The photochemically-modeled emissions of the onroad mobile- and
upstream sector-attributable PM2.5 concentrations used to derive the BPT values may not match
the change in air quality resulting from the control strategies associated with the final standards.
For this reason, the PM-related health benefits reported here may be larger, or smaller, than those
that would be realized through this rulemaking.
Given the uncertainty that surrounds BPT analysis, EPA systematically compared benefits
estimated using its BPT approach (and other reduced-form approaches) to benefits derived from
full-form photochemical model representation. This work is referred to as the "Reduced Form
Tool Evaluation Project" (Project), which began in 2017, and the initial results were available at
the end of 2018. The Agency's goal was to better understand the suitability of alternative
reduced-form air quality modeling techniques for estimating the health impacts of criteria
pollutant emissions changes in EPA's benefit-cost analysis. The Project analyzed air quality
6-59
-------
policies that varied in the magnitude and composition of their emissions changes and in the
emissions source affected (e.g., on-road mobile, industrial point, or electricity generating units).
The policies also differed in terms of the spatial distribution of emissions and concentration
changes, and in their impacts on directly-emitted PM2.5 and secondary PM2.5 precursor emissions
(NOx and SO2).
For scenarios where the spatial distribution of emissions was similar to the inventories used to
derive the BPT, the Project found that total PM2.5 BPT-derived benefits were within
approximately 10 percent to 30 percent of the health benefits calculated from full-form air
quality modeling, though the discrepancies varied by regulated scenario and PM2.5 species. The
scenario-specific emission inputs developed for the Project, and a final project report, are
available online (U.S. EPA 2019). We note that the BPT values used to monetize the benefits of
the final standards were not part of the Project, though we believe they are our best estimate of
the stream of benefits associated with the rulemaking absent year-over-year air quality modeling
and we have confidence in the BPT approach and the appropriateness of relying on BPT health
estimates for this rulemaking. EPA continues to research and develop reduced-form approaches
for estimating PM2.5 benefits.
6-60
-------
Chapter 6 References
40 CFRPart 50. 2023. https://www.ecfr.gov/current/title-40/chapter-I/subchapter-C/part-50.
40 CFRPart 53. 2023. https://www.ecfr.gov/current/title-40/chapter-I/subchapter-C/part-53.
40 CFRPart 58. 2023. https://www.ecfr.gov/current/title-40/chapter-I/subchapter-C/part-58.
42 USC §7491 (a). 2013. https://www.govinfo.gov/content/pkg/USCODE-2013-
title42/html/USCODE-2013-ti tle42-chap85-subchapI-partC-subpartii-sec7491.htm.
73 FR 16493-16494 . 2008.
81 FR 54422. 2016. "Finding that Greenhouse Gas Emissions From Aircraft Cause or Contribute
to Air Pollution That May Reasonably Be Anticipated To Endanger Public Health and Welfare."
August 15.
Adar, S., G. Adamkiewicz, D.R. Gold, J. Schwartz, B.A. Coull, and Ft. Suh. 2007. "Ambient and
microenvironmental particles and exhaled nitric oxide before and after a group bus trip." Environ
Health Perspect 115: 507-512.
Adar, S.D., D.R. Gold, and B.A. Coull. 2007. "Focused exposure to airborne traffic particles and
heart rate variability in the elderly." Epidemiology 18: 95-103.
Adar, S.D., J. D.Souza, and L. Sheppard. 2015. "Adopting clean fuels and technologies on
school buses. Pollution and health impacts in children." Am J Respir Crit Care Med 191 (12):
1413-1421.
Adar, S.D., R. Klein, E.K. Klein, and et al. 2010. "Air pollution and the microvasculature: a
cross-sectional assessment of in vivo retinal images in the population-based Multi-Ethnic Study
of Atherosclerosis." PLoS Med 7 (11): E1000372.
Aksoy, M. 1989. "Hematotoxicity and carcinogenicity of benzene." Environ. Health Perspect 82:
193-197.
Alexeef, S.E., B.A. Coull, A. Gryparis, and et al. 2011. "Medium-term exposure to traffic-related
air pollution and markers of inflammation and endothelial function." Environ Health Perspect
119: 481-486.
Anthoff, D, and R. S. J. Tol. 2013b. "The uncertainty about the social cost of carbon: A
decomposition analysis using FUND." Climate Change 117(3), 515-530.
doi:https://doi.org/10.1007/sl 0584-013-0706-7.
Anthoff, D, and R.S.J Tol. 2013. "Erratum to: The uncertainty about the social cost of carbon: A
decomposition analysis using FUND." Climatic Change 121(2), 413.
Appleman, L.M., R.A. Woutersen, and V.J. Feron. 1982. "Inhalation toxicity of acetaldehyde in
rats. I. Acute and subacute studies." Toxicology 23: 293-297.
Apte, J.S., K.P. Messier, S. Gani, M. Brauer, T.W. Kirchstetter, M.M. Lunden, J.D. Marshall,
C.J. Portier, R.C.H. Vermeulen, and S.P. Hamburg. 2017. "High-Resolution Air Pollution
6-61
-------
Mapping with Google Street View Cars: Exploiting Big Data." Environ Sci Technol 51: 6999-
7008.
Arrow, K, M Cropper, C Gollier, B Groom, G Heal, R Newell, W Nordhaus, R Pindyck, W
Pizer, and P Portney. 2013. "Determining benefits and costs for future generations." Science
341(6144) : 349-350.
ATSDR. 2010. Addendum to the Toxicological Profile for Formaldehyde. Washington, DC: U.S.
Department of Health and Human Services (HHS).
ATSDR. 2014. Letter Health Consultation, Radiac Abrasives, Inc. Chicago, Illinois: CDC.
https://www.atsdr.cdc.gov/HAC/pha/RadiacAbrasives/Radiac%20Abrasives,%20Inc.%20_%20L
HC%20(Final)%20_%2003 -24-2014%20(2)_5 08. pdf.
ATSDR. 2007. Toxicological profile for benzene. Atlanta: U.S. Agency for Toxic Substances
and Disease Registry (ATSDR), U.S. Department of Health and Human Services, Public Health
Service. http://www.atsdr.cdc.gov/ToxProfiles/tp3.pdf.
ATSDR. 1999. Toxicological Profile for Formaldehyde. Washington, DC: U.S. Department of
Health and Human Services.
ATSDR. 2005. Toxicological Profile for Naphthalene, 1-Methylnaphthalene, and 2-
Methylnaphthalene. CDC. https://www.atsdr.cdc.gov/ToxProfiles/tp67-p.pdf.
ATSDR. 1995. Toxicological profile for Polycyclic Aromatic Hydrocarbons (PAHs). Atlanta,
GA: Agency for Toxic Substances and Disease Registry, U.S. Department of Health and Human
Services, Public Health Service, http://www.atsdr.cdc.gov/ToxProfiles/TP.asp?id=122&tid=25.
Auffhammer, M. 2018. "Quantifying economic damages from climate change." Journal of
Economic Perspectives 32(4): 33-52.
Austin, W., G. Heutel, and D. Kreisman. 2019. "School bus emissions, student health and
academic performance." Econ Edu Rev 70: 108-112.
Bauer, MD, and GD Rudebusch. 2020. "Interest rates under falling stars." American Economic
Review 110(5): 1316-54.
Bauer, MD, and GD Rudebusch. 2023. "The rising cost of climate change: evidence from the
bond market." The Review of Economics and Statistics 105(5): 1255-1270.
Beane Freeman, L. E., A. Blair, J. H. Lubin, P. A. Stewart, R. B. Hayes, R. N. Hoover, and M.
Hauptmann. 2009. "Mortality from lymphohematopoietic malignancies among workers in
formaldehyde industries: The National Cancer Institute cohort." J. National Cancer Inst. 101:
751-761.
Bevan, C, J.C. Stadler, and et al. 1996. "Subchronic toxicity of 4-vinylcyclohexene in rats and
mice by inhalation." Fundamentals of Applied Toxicology 1-10.
Blunden, J, and T Boyer. 2022. "State of the Climate in 2021." Bulletin of the American
Meteorological Society 103 (8). https://doi.org/10.1175/2022BAMSStateoftheClimate.
6-62
-------
Boogaard, H., Patton. A.P., R.W. Atkinson, J.R. Brook, H.H. Chang, D.L. Crouse, J.C. Fussell,
et al. 2022. "Long-term exposure to traffic-related air pollution and selected health outcomes: a
systematic review and meta-analysis." Environ Intl 164: 107262.
Boothe, V.L., and D.G. Shendell. 2008. "Potential health effects associated with residential
proximity to freeways and primary roads: review of scientific literature, 1999-2006." J Environ
Health 70: 33-41.
Boothe, VL., T.K. Boehmer, A.M. Wendel, andF.Y. Yip. 2014. "Residential traffic exposure
and childhood leukemia: a systematic review and meta-analysis." Am J Prev Med 46: 413-422.
Bricker, S., B. Longstaff, W. Dennison, A. Jones, K. Boicourt, C. Wicks, and J. Woerner. 2007.
Effects of Nutrient Enrichment In the Nation's Estuaries: A Decade of Change. Silver Spring,
MD: NOAA Coastal Ocean Program Decision Analysis Series No. 26. National Centers for
Coastal Ocean Science.
Bureau of Economic Analysis (BEA). 2021. "Table 1.1.9." National Income and Product
Accounts. Accessed 2023.
Cahill, T.M., M.J. Charles, and V.Y. Seaman. 2010. Development and application of a sensitive
method to determine concentrations of acrolein and other carbonyls in ambient air. Health
Effects Institute Research Report 149. https://www.healtheffects.org/system/files/Cahilll49.pdf.
Cape, JN, ID Leith, J Binnie, J Content, M Donkin, M Skewes, DN Price, AR Brown, and AD
Sharpe. 2003. "Effects of VOCs on herbaceous plants in an open-top chamber experiment."
Environ. Pollut 124: 341-343.
Carleton, T., Jina, A., Delgado, M., Greenstone, M., Houser, T., Hsiang, S., Hultgren, A., Kopp,
R.E., McCusker, K.E., Nath, I., Rising, J., Ashwin, A., Seo, H., Viaene, A., Yaun, J., and Zhang,
A.,. 2022. "Valuing the Global mortality Consequences of Climate Change Accounting for
Adaptation Costs and Benefits." The Quarterly Journal of Economics 137(4): 2037-2105.
Chameides, W.L., H. Yu, S.C. Liu, M. Bergin, X. Zhou, L. Mearns, G. Wang, et al. 1999. "Case
study of the effects of atmospheric aerosols and regional haze on agriculture: an opportunity to
enhance crop yields in China through emission controls?" PNAS 96: 13626-13633.
Chen, E., H.M. Schrier, R.C. Strunk, and et al. 2008. "Chronic traffic-related air pollution and
stress interact to predict biologic and clinical outcomes in asthma." Environ Health Perspect 116:
970-975.
CIL, Climate Impact Lab. 2023. Documentation for Data-driven Spatial Climate Impact Model
(DSCIM).
https://impactlab.org/wpcontent/uploads/2023/10/DSCIM_UserManual_Version092023-
EPA.pdf.
Clougherty, J.E., J.I. Levy, L.D. Kubzansky, and et al. 2007. "Synergistic effects of traffic-
related air pollution and exposure to violence on urban asthma etiology." Environ Health
Perspect 115: 1140-1146.
Coggon, D, EC Harris, J Poole, and KT Palmer. 2003. "Extended follow-up of a cohort of British
chemical workers exposed to formaldehyde." J National Cancer Inst. 95: 1608-1615.
6-63
-------
Corti, M., and C.A. Snyder. 1996. "Influences of gender, development, pregnancy and ethanol
consumption on the hematotoxicity of inhaled 10 ppm benzene." Arch Toxicol 70: 209-217.
Cotrufo, M.F., De Santo A.V., Alfani A., Bartoli G., and De Cristofaro A. 1995. "Effects of
urban heavy metal pollution on organic matter decomposition in Quercus ilex L. Woods."
Environmental Pollution 89 (1): 81-87.
Cropper, ML, MC Freeman, B Groom, and WA Pizer. 2014. "Declining discount rates."
American Economic Review 104(5): 538-43.
Dabek-Zlotorzynska, E., V. Celo, L. Ding, D. Herod, C-H. Jeong, G. Evans, and N. Hilker. 2019.
"Characteristics and sources of PM2.5 and reactive gases near roadways in two metropolitan areas
in Canada." Atmos Environ 218: 116980.
DeHayes, D.H., P.G. Schaberg, G.J. Hawley, and G.R. Strimbeck. 1999. "Acid rain impacts on
calcium nutrition and forest health." Bioscience 49 (10): 789-800.
Di, Q, Wang, Y, Zanobetti, A, Wang, Y, Koutrakis, P, Choirat, C, Dominici, F and Schwartz,
JD. 2017. "Air pollution and mortality in the Medicare population." New England Journal of
Medicine 376 (26): 2513-2522.
Eckel, S.P., K. Berhane, M.T. Salam, and et al. 2011. "Residential Traffic-related pollution
exposure and exhaled nitric oxide in the Children's Health Study." Environ Health Perspect.
doi:10.1289/ehp,1103516.
EPA. 2023f. "Supplementary Material for the Regulatory Impact Analysis for the Final
Rulemaking: Standards of Performance for New, Reconstructed, and Modified Sources and
Emissions Guidelines for Existing Sources: Oil and Natural Gas Sector Climate Review." EPA
Report on the Social Cost of Greenhouse Gases: Estimates Incorporating Recent Scientific
Advances, Washington, DC. doi:Docket ID No. EPA-HQ-OAR-2021-0317.
Foos, B., M. Marty, J. Schwartz, W. Bennet, J. Moya, A.M. Jarabek, and A.G. Salmon. 2008.
"Focusing on children's inhalation dosimetry and health effects for risk assessment: An
introduction." J Toxicol Environ Health (71A): 149-165.
Ford, AM, MS Pombo-de-Oliveira, KP McCarthy, JM MacLean, KC Carrico, RF Vincent, and
M. Greaves. 1997. "Monoclonal origin of concordant T-cell malignancy in identical twins."
Blood 89: 281-285.
Franco-Suglia, S., A. Gryparis, R.O. Wright, and et al. 2007. "Association of black carbon with
cognition among children in a prospective birth cohort study." Am J Epidemiol,
doi: 10.1093/aj e/kwm3 08.
Gantt, B, R.C. Owen, and N. Watkins. 2021. "Characterizing nitrogen oxides and fine particulate
matter near major highways in the United States using the National Near-road Monitoring
Network." Environ Sci Technol 55: 2831-2838.
Garshick, Eric, Francine Laden, Jaime E. Hart, Mary E. Davis, Ellen A. Eisen, and Thomas J.
Smith. 2012. "Lung cancer and elemental carbon exposure in trucking industry workers."
Environmental Health Perspectives 120 (9): 1301-1306.
6-64
-------
Gartland, N, H.E. Aljofi, K. Dienes, L.A. Munford, A.L. Theakston, and M. van Tongeren. 2022.
"The effects of traffic air pollution in and around schools on executive function and academic
performance in children: a rapid review." Int J Environ Res Public Health 10: 749.
Goldstein, B.D. 1988. "Benzene toxicity." Occupational Medicine - State of the Art Reviews 3:
541-554.
Grantz, DA, JHB Garner, and DW Johnson. 2003. "Ecological effects of particulate matter."
Environ Int 29: 213-239.
Grivas, G., I. Stavroulas, E. Liakakou, D.G. Kaskaoutis, A. Bougiatioti, D. Paraskevopoulou, E.
Gerasopoulos, and N. Mihalopoulos. 2019. "Measuring the spatial variability of black carbon in
Athens during wintertime." Air Quality, Atmosphere & Health 12: 1405-1417.
Gu, P., H.Z. Li, Q. Ye, and et al. 2018. "Intercity variability of particulate matter is driven by
carbonaceous sources and correlated with land-use variables." Environ Sci Technol 52: 11545-
11554.
Hand, J.L. et al. 2011. Spatial and Temporal Variability of Haze and its Constituents in the
United States. IMPROVE Report 0737-5352 (HERO #3121721).
http://vista.cira.colostate.edu/improve/wp-
content/uploads/2016/08/IMPRO VE_V_FullReport.pdf.
Hand, JL, AJ Prenni, S Copeland, BA Schichtel, and WC Malm. 2020. "Thirty years of the Clean
Air Act Amendments: Impacts on haze in remote regions of the United States (1990-2018)."
Atmos Environ 243: 117865.
Hartin, C. 2023. "Advancing the estimation of future climate impacts within the United States."
Earth System Dynamics 14: 1015-1037. https://dio.org/10.5194/esd-14-1015-2023.
Hauptmann, M„ Stewart P. A., Lubin J. H., L. E. Beane Freeman, R. W. Hornung, R. F. Herrick,
R. N. Hoover, J. F. Fraumeni, and R. B. Hayes. 2009. "Mortality from lymphohematopoietic
malignancies and brain cancer among embalmers exposed to formaldehyde." Journal of the
National Cancer Institute 101: 1696-1708.
Hauptmann, M., J. H. Lubin, P. A. Stewart, R. B. Hayes, and A. Blair. 2003. "Mortality from
lymphohematopoetic malignancies among workers in formaldehyde industries." Journal of the
National Cancer Institute 95: 1615-1623.
Hauptmann, M., J. H. Lubin, P. A. Stewart, R. B. Hayes, and A. Blair. 2004. "Mortality from
solid cancers among workers in formaldehyde industries." American Journal of Epidemiology
159: 1117-1130.
HEI. 2022. HEI Panel on the Health Effects of Long-Term Exposure to Traffic-Related Air
Pollution (2022) Systematic review and meta-analysis of selected health effects of long-term
exposure to traffic-related air pollution. Health Effects Institute Special Report 23.
https://www.healtheffects.org/system/files/hei-special-report-23_l.pdf.
Hilker, N., J.W. Wang, C-H. Jong, R.M. Healy, U. Sofowote, J. Debosz, Y. Su, et al. 2019.
"Traffic-related air pollution near roadways: discerning local impacts from background." Atmos.
Meas. Tech 12: 5247-5261.
6-65
-------
Hope, C. 2013. "Critical issues for the calculation of the social cost of CO2: why the estimates
from PAGE09 are higher than those from PAGE2002." Climate Change 117(3), 531-543.
doi:https://doi.org/10.1007/sl 0584-012-0633-z.
Howard, PH, and T Sterner. 2017. "Few and not so far between: a meta-analysis of climate
damage estimates." Environmental Resource Economics 68(1): 197-225.
Hutchinson, J, D Maynard, and L Geiser. 1996. Air quality and lichens - a literature review
emphasizing the Pacific Northwest, USA. Pacific Northwest Region Air Resource Management
Program; U.S. Forest Service; U.S. Department of Agriculture (USDA).
Hutchison, R., and C.E. Kraft. 1994. "Hmong Fishing Activity and Fish Consumption." Journal
of Great Lakes Research 20 (2): 471-487.
IARC. 2013. "Diesel and gasoline engine exhausts and some nitroarenes." IARC Monographs
105. http://monographs.iarc.fr/ENG/Monographs/voll05/index.php.
IARC. 1982. IARC monographs on the evaluation of carcinogenic risk of chemicals to humans.
Lyons, France: International Agency for Research on Cancer, World Health Organization.
IARC. 2008. "Monographs on the evaluation of carcinogenic risk of chemicals to humans, 1,3-
butadiene, ethylene oxide, and vinyl halides, Vol 97."
IARC. 2018. Monographs on the evaluation of carcinogenic risks to humans. Lyons, France:
International Agency for Research on Cancer, World Health Organization.
http://publications.iarc.fr/Book-And-Report-Series/Iarc-Monographs-On-The-Identification-Of-
Carcinogenic-Hazards-To-Humans/Benzene-2018.
IARC. 2012. Monographs on the Evaluation of Carcinogenic Risks to Humans, Volume 100F:
Formaldehyde. International Agency for Research on Cancer.
IARC. 2006. Monographs on the Evaluation of Carcinogenic Risks to Humans, Volume 88:
Formaldehyde, 2-Butoxyethanol and l-tert-Butoxypropan-2-ol. International Agency for
Research on Cancer.
IARC. 2002. Monographs on the Evaluation of the Carcinogenic Risk of Chemicals for Humans.
Lyons, France: International Agency for Research on Cancer .
IARC. 1999. Re-evaluation of some organic chemicals, hydrazine, and hydrogen peroxide. Lyon,
France: IARC Monographs on the Evaluation of Carcinogenic Risk of Chemical to Humans Vol.
71.
IPCC. 2021. Climate Change 2021 - The Physical Science Basis. Online: Cambridge University
Press. https://www.ipcc.ch/report/ar6/wgl/downloads/report/IPCC_AR6_WGI_Chapter07.pdf.
—. 2021a. Climate Change 2021: The Physical Science Basis. Vol. Contribution of Working
Group I to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change.
Online: Cambridge University Press. doi:https://www.ipcc.ch/report/ar6/wgl.
—. 2019. "Climate Change and Land: an IPCC special report on climate change, desertification,
land degradation, sustainable land management, food security, and greenhouse gas fluxes in
terrestrial ecosystems."
6-66
-------
—. 2018. "Global Warming of 1.5 °C. An IPCC Special Report on the impacts of global
warming of 1.5 °C above pre-industrial levels ..."
IPCC. 2018. "Global Warming of 1.5°C: An IPCC Special Report on the impacts of global
warming of 1.5°C above pre-industrial levels and related global greenhouse gas emission
pathways."
—. 2019. "Special Report on the Ocean and Cryosphere in a Changing Climate."
—. 2021. "Summary for Policymakers. In: Climate Change 2021: The Physical Science Basis.
Contribution of Working Group I to the Sixth Assessment Report of the Intergovernmental Panel
on Climate Change." Cambridge, United Kingdom and New York, NY, USA: Cambridge
University Press. 3-32. doi:10.1017/9781009157896.001.
—. 2013. "Summary for Policymakers. In: Climate Change 2023: Synthesis Report. Contribution
of Working Groups I, II and III to the Sixth Assessment Report of the Intergovernmental Panel
on Climate Change." Geneva. doi:10.59327/IPCC/AR6-9789291691647.001.
IPCS. 2002. Concise International Chemical Assessment Document 40. Formaldehyde. World
Health Organization.
Irons, R.D., W.S. Stillman, D.B. Colagiovanni, and V.A. Henry. 1992. "Synergistic action of the
benzene metabolite hydroquinone on myelopoietic stimulating activity of
granulocyte/macrophage colony-stimulating factor in vitro." Proc. Natl. Acad. Sci. 89: 3691-
3695.
Irving, P.M. 1991. Acidic deposition: State of science and technology. Vols I-IV. Washington,
DC: The U.S. National Acid Precipitation Assessment Program, 24-76.
Islam, T., R. Urban, W.J. Gauderman, and et al. 2011. "Parental stress increases the detrimental
effect of traffic exposure on children's lung function." Am J Respir Crit Care Med.
IWG. 2016a. Addendum to Technical Support Document on Social Cost of Carbon for
Regulatory Impact Analysis under Executive Order 12866: Application of the Methodology to
Estimate the Social Cost of Methane and the Social Cost of Nitrous Oxide.
https://www.epa.gov/sites/default/files/2016-12/documents/addendum_to_sc-
ghg_tsd_august_2016. pdf.
IWG. 2021. Social Cost of Carbon, Methane, and Nitrous Oxide: Interim Estimates under
Executive Order 13990. Technical Support Government, Interagency Working Group on Social
Cost of Carbon, United States Government.
—. 2010. Technical Support Document: Social Cost of Carbon for Regulatory Impact Analysis
under Executive Order 12866. Accessed 2023. https://www.epa.gov/sites/default/files/2016-
12/documents/ scc_tsd_2010.pdf.
—. 2021. Technical Support Document: Social Cost of Carbon, Methane, and Nitrous Oxide
Interim Estimates under Executive Order 13990. Accessed 2023.
https://www.whitehouse.gov/wp-
content/uploads/2021/02/TechnicalSupportDocument_SocialCostofCarbonMethaneNitrouSOxid
e.pdf.
6-67
-------
—. 2013. Technical Support Document: Technical Update of the Social Cost of Carbon for
Regulatory Impact Analysis Under Executive Order 12866.
https://www.ourenergypolicy.org/wp-
content/uploads/2013/06/social_cost_of_carbon_for_ria_2013_update.pdf.
—. 2016b. Technical Support Document: Technical Update of the Social Cost of Carbon for
Regulatory Impact Analysis Under Executive Order 12866. Accessed 2023.
https://www.epa.gov/sites/default/files/2016- 12/documents/sc_CO2_tsd_august_2016.pdf.
Joslin, J.D., Kelly, J.M., and van Miegroet, H. 1992. "Soil chemistry and nutrition of North
American spruce-fir stands: evidence for recent change." Journal of Environmental Quality 21:
12-30.
Kammerbauer, H., H. Selinger, R. Rommelt, A. Ziegler-Jons, D. Knoppik, and B. Hock. 1987.
"Toxic components of motor vehicle emissions for the spruce Picea abies." Environ. Pollut. 48:
235-243.
Kan, H., G. Heiss, K.M. Rose, and et al. 2008. "Prospective analysis of traffic exposure as a risk
factor for incident coronary heart disease: The Atherosclerosis Risk in Communities (ARIC)
study." Environ Health Perspect 116: 1463-1468.
Karner, A.A., D.S. Eisinger, andD.A. Niemeier. 2010. "Near-roadway air quality: synthesizing
the findings from real-world data." Environ Sci Technol 44: 5334-5344.
Keller, KA, and CA. Snyder. 1988. "Mice exposed in utero to 20 ppm benzene exhibit altered
numbers of recognizable hematopoietic cells up to seven weeks after exposure." Fundam Appl
Toxicol 10: 224-232.
Keller, KA, and CA. Snyder. 1986. "Mice exposed in utero to low concentrations of benzene
exhibit enduring changes in their colony forming hematopoietic cells." Toxicology 42: 171-181.
Kimbrough, S., R.C. Owen, M. Snyder, and J. Richmond-Bryant. 2017. "NO to N02 Conversion
Rate Analysis and Implications for Dispersion Model Chemistry Methods using Las Vegas,
Nevada Near-Road Field Measurements." Atmos Environ 165: 23-24.
Kimbrough, S., R.W. Baldauf, G.S.W. Hagler, R.C. Shores, W. Mitchell, D.A. Whitaker, C.W.
Croghan, and D.A. Vallero. 2013. "Long-term continuous measurement of near-road air
pollution in Las Vegas: seasonal variability in traffic emissions impact on air quality." Air Qual
Atmos Health 6: 295-305.
Kimbrough, S., T. Palma, and R.W. Baldauf. 2014. "Analysis of mobile source air toxics
(MSATs)—Near-road VOC and carbonyl concentrations." Journal of the Air &Waste
Management Association 64 (3): 349-359. doi:10.1080/10962247.2013.863814.
Laden, F., J.E. Hart, T.J. Smith, M.E. Davis, and E. Garshick. 2007. "Cause-specific mortality in
the unionized U.S. trucking industry." Environmental Health Perspect 115: 1192-1196.
Lai, Raj M, Anu Ramaswami, and Armistead G Russell. 2020. "Assessment of the Near-Road
(monitoring) Network including comparison with nearby monitors within U.S. cities."
Environmental Research Letters (15): 114026. doi:10.1088/1748-9326/ab8156.
6-68
-------
Lan, Qing, L. Zhang, G. Li, and R., et al. Vermeulen. 2004. "Hematotoxically in Workers
Exposed to Low Levels of Benzene." Science 306: 1774-1776.
Landers, DH, SL Simonich, DA Jaffe, LH Geiser, DH Campbell, AR Schwindt, CB Schreck, et
al. 2008. The Fate, Transport and Ecological Impacts of Airborne Contaminants in Western
National Parks (USA). EPA/600/R-07/138. Corvallis, Oregon: U.S. Environmental Protection
Agency, Office of Research and Development, NHEERL, Western Ecology Division.
Li, C., Q. N, and P.H. Ryan. 2009. "School bus pollution and changes in the air quality at
schools: a case study." J Environ Monit 11: 1037-1042.
Liu, W., J. Zhang, J.l Kwon, and et al. 2006. "Concentrations and source characteristics of
airborne carbonyl compounds measured outside urban residences." J Air Waste Manage Assoc
56: 1196-1204.
Long, D., D. Lewis, and C. Langpap. 2021. "Negative traffic externalities and infant health: the
role of income heterogeneity and residential sorting." Environ and Resource Econ 80: 637-674.
Malm, William. 2016. Visibility - The Seeing of Near and Distant Landscape Features. Elsevier.
McConnell, R., T. Islam, K. Shankardass, and et al. 2010. "Childhood incident asthma and
traffic-related air pollution at home and school." Environ Health Perspect 1021-1026.
McDonald, B.C., Z.C. McBride, E.W. Martin, and R.A. Harley. 2014. "High-resolution mapping
of motor vehicle carbon dioxide emissions." J. Geophys. Res.Atmos 119: 5283-5298.
doi: 10.1002/2013JD021219.
McDuffie, EE, MC Sarofim, W Raich, M Jackson, H Roman, K Seltzer, BH Henderson, et al.
2023. "The social cost of ozone-related mortality impacts from methane emissions." Earth's
Future 11(9). doi:https://doi.org/10.1029/2023EF003853.
Millar, RJ, ZRNicholls, P Friedlingstein, and MR Allen. 2017. "A modified impulse-response
representation of the global near-surface air temperature and atmospheric concentration response
to carbon dioxide emissions." Atmospheric Chemistry and Physics 17(11): 7213-7228.
Myou, S., M. Fujimura, Nishi K., T. Ohka, and T. and Matsuda. 1993. "Aerosolized
acetaldehyde induces histamine-mediated bronchoconstriction in asthmatics." Am. Rev. Respir.
Dis 148 (4 Pt 1): 940-943.
NASEM. 2016. "Attribution of Extreme Weather Events in the Context of Climate Change."
Washington, D.C.: National Academies of Sciences, Engineering, and Medicine, The National
Academies Press, https://doi.org/10.17226/21852.
—. 2019. "Climate Change and Ecosystems." Washington, D.C.: National Academies of
Sciences, Engineering, and Medicine, The National Academies Press.
doi:https://doi. org/10.17226/25504.
—. 2017. "Valuing Climate Damages: Updating Estimation of the Social Cost of Carbon
Dioxide." Washinton, D.C.: National Academies of Sciences, Engineering, and Medicine, The
National Academies Press, https://doi.org/10.17226/24651.
6-69
-------
National Academies of Sciences. Engineering, and Medicine. 2023. Review of EPA's 2022 Draft
Formaldehyde Assessment. The National Academies Press, doi: 10.17226/27153.
National Academies. 2017. Valuing Climate Damages: Updating Estimation of the Social Cost
of Carbon Dioxide. Washington, D.C.: The National Academies Press.
Newell, RG, WA Pizer, and BC Prest. 2022. "A discounting rule for the social cost of carbon."
Journal of the Association of Environmental and Resource Economists 9(5): 1017-1046.
NIFC. 2021. "Total wildland fires and acres (1983-2020)." National Interagency Fire Center.
Niklinska, M, R Laskowski, and M. Maryanski. 1998. "Effect of heavy metals and storage time
on two types of forest litter: basal respiration rate and exchangeable metals." Ecotoxicological
Environmental Safety 41: 41.
NOAA. 2022. "State of the Climate 2021." https://www.ncei.noaa.gov/bams-state-of-climate.
Nordhaus, W. 2014. "Esitmates of the social cost of carbon: concepts and results from the DICE
2013R model and alternative approaches." Journal of the Association of Environmental
Economists 1(1/2): 273-312.
Nordhaus, W.D. 2010. "Economic aspects of global warming in a post-Copenhagen
environment. ." Proceedings of the National Academy of Sciences of the United States of
America 107(26), 11721-11726.
NRC. 1993. Protecting Visibility in National Parks and Wilderness Areas. Washington, DC:
National Research Council, National Academy of Sciences Committee on Haze in National
Parks and Wilderness Areas, https://nap.nationalacademies.org/initiative/committee-on-haze-in-
national-parks-and-wilderness-areas.
NRC. 2011. Review of the Environmental Protection Agency's Draft IRIS Assessment of
Formaldehyde. Washington, DC: National Research Council, National Academies Press.
http://books.nap.edu/openbook.php?record_id=13142.
NTP. 2019. NTP Monograph on the Systematic Review of Traffic-related Air Pollution and
Hypertensive Disorders of Pregnancy. National Toxicology Program Monograph 7.
https://ntp.niehs.nih.gov/ntp/ohat/trap/mgraph/trap_final_508.pdf.
NTP. 2016. Report on Carcinogens, Fourteenth Edition. Washington, DC: National Toxicology
Program, U.S. Department of Health and Human Services, Public Health Service.
https://ntp.niehs.nih.gov/go/rocl4.
Oak Ridge Institute for Science and Education. 2004. "External Peer Review for the IRIS
Reassessment of the Inhalation Carcinogenicity of Naphthalene."
http://cfpub. epa.gov/ncea/cfm/recordisplay. cfm?deid=84403.
OEHHA, California. 2014. "TSD for Noncancer RELs: Appendix D. Individual, Acute, 8-Hour,
and Chronic Reference Exposure Level Summaries."
https://oehha.ca.gov/media/downloads/crnr/appendixdlfinal.pdf.
6-70
-------
Olsson, Ann C., et al. 2011. "Exposure to diesel motor exhaust and lung cancer risk in a pooled
analysis from case-control studies in Europe and Canada." American journal of respiratory and
critical care medicine 183 (7): 941-948.
OMB. 2022. "Climate Risk Exposure: An Assessment of the Federal Government's Financial
Risks to Climate Change, White Paper." Office of Management and Budget, April.
OMB, Office of Management and Budget. 2003. Circular A-4, Regulatory Analysis.
Washington, DC: Office of Management and Budget.
https://obamawhitehouse.archives.gov/omb/circulars_a004_a-4.
P.A., McKinney, F.E. Alexander, R.A. Cartwright, and et al. 1991. "Parental occupations of
children with leukemia in west Cumbria, north Humberside, and Gateshead." Br. Med. J 302:
681-686.
Pedde, M., and C. Bailey. 2011. Identification of Schools within 200 Meters of U.S. Primary and
Secondary Roads. Memorandum to the docket, U.S. Environmental Protection Agency.
Peters, A., S. von Klot, M. Heier, I. Trentinaglia, A. Hormann, H.E. Wichmann, and H. Lowel.
2004. "Exposure to traffic and the onset of myocardial infarction." New England J Med 351:
1721-1730.
Peterson, D.E., M.S. Kanarek, M.A. Kuykendall, J.M. Diedrich, H.A. Anderson, P.L.
Remington, and T.B. Sheffy. 1994. "Fish Consumption Patterns and Blood Mercury Levels in
Wisconsin Chippewa Indians." Archives of Environmental Health 49 (1): 53-58.
Pindyck, RS. 2017. "Comments on Proposed Rule and Regulatory Impact Analysis on the Delay
and Suspension of Certain Requirements for Waster Prevention and Resource Conservation."
Accessed Comment submitted on Nov. 6, 2017. https://downloads.regulations.gov/EPA-HQ-
OAR-2018-0283- 6184/attachment_6.pdf.
Pinkerton, L. E. 2004. "Mortality among a cohort of garment workers exposed to formaldehyde:
an update." Occup. Environ. Med. 61: 193-200.
Pope III et al., CA, Lefler, JS, Ezzati, M, Higbee, JD, Marshall, JD, Kim, S-Y, Bechle, M,
Gilliat, KS, Vernon, SE and Robinson, AL. 2019. "Mortality risk and fine particulate air
pollution in a large, representative cohort of US adults." Environmental Health Perspectives 127
(7): 077007.
Power, M.C., M.G. Weisskopf, S.E. Alexeef, and et al. 2011. "Traffic-related air pollution and
cognitive function in a cohort of older men." Environ Health Perspect 682-687.
Qu, O., R. Shore, G. Li, X. Jin, C.L. Chen, B. Cohen, A. Melikian, et al. 2003. Validation &
Evaluation of Biomarkers in Workers Exposed to Benzene in China HEI Report 115. Health
Effects Institute .
Qu, Q., R. Shore, G. Li, X. Jin, L.C. Chen, and et al. B. Cohen. 2002. "Hematological changes
among Chinese workers with a broad range of benzene exposures." American Journal of
Industrial Medicine 42: 275-285.
Raaschou-Nielsen, O., and P Reynolds. 2006. "Air pollution and childhood cancer: a review of
the epidemiological literature." Int J Cancer 118: 2920-2929.
6-71
-------
Ramsey, FP. 1928. "A mathematical theory of saving." The Economic Journal 38(152): 543-559.
Rennert, K, BC Prest, WA Pizer, RG Newell, D Anthoff, C Kingdon, L Rennels, et al. 2022a.
"The social cost of carbon: Advances in long-term probabilistic projections of population, GDP,
emissions, and discount rates." Brookings Papers on Economic Activity Fall 2021 223-305.
Rennert, K, F Errickson, BC Prest, L Rennels, R Newell, W Pizer, C Kingdon, J Wingenroth,
and R Cooke. 2022. "Comprehensive evidence implies a higher social cost of CO2." Nature
610(7933): 687-692.
Rennert, K., Prest, B.C., Pizer, W.A., Newell, R.G., Anthoff, D., Kingdon, C., Rennels, L.,
Cooke, R., Raftery, A.E., Sevcikova, H. and Errickson, F. 2021. "The social cost of carbon:
Advances in long-term probablisitic projections of poulation, GDP, emissions, and discount
rates." Brookings Papers on Economic Activity 223-305.
Riddell, J., T.H. Nash, and P. Padgett. 2008. "The effect of HN03 gas on the lichen Ramalina
menziesii." Flora - Morphology, Distribution, Functional Ecology of Plants 203 (1): 47-54.
Riediker, M. 2007. "Cardiovascular effects of fine particulate matter components in highway
patrol officers." Inhal Toxicol 19: 99-105.
Riediker, M., W.E. Cascio, T.R. Griggs, and et al. 2004. "Particulate matter exposure in cars is
associated with cardiovascular effects in healthy young men." Am J Respir Crit Care Med 169.
Rode, A, T Carleton, M Delgado, M Greenstone, T Houser, S Hsiang, A Hultgren, et al. 2021.
"Estimating a social cost of carbon for global energy consumption." Nature 598(7880): 308-314.
Rose, S, D Turner, G Blanford, J Bistline, F de la Chesnaye, and T Wilson. 2014.
"Understanding the Social Cost of Carbon: A Technical Assessment." EPRI Technical Update
Report, Palo Alto, CA.
Rothman, N., G.L. Li, M. Dosemeci, W.E. Bechtold, G.E. Marti, Y.Z. Wang, M. Linet, L.Q. Xi,
W. Lu, M.T. Smith, N. Titenko-Holland, L.P. Zhang, W. Blot, S.N. Yin, and R.B. Hayes. 1996.
"Hematotoxicity among Chinese workers heavily exposed to benzene." American Journal of
Industrial Medicine 29: 236-246.
Sabin, L., E. Behrentz, A.M. Winer, and et al. 2005. "Characterizing the range of children's air
pollutant exposure during school bus commutes." J Expo Anal Environ Epidemiol 15: 377-387.
Salam, M.T., T. Islam, and F.D. Gilliland. 2008. "Recent evidence for adverse effects of
residential proximity to traffic sources on asthma." Curr Opin Pulm Med 14: 3-8.
Sarnat, J.A., A. Russell, D. Liang, J.L Moutinho, R. Golan, R. Weber, D. Gao, et al. 2018.
Developing Multipollutant Exposure Indicators of Traffic Pollution: The Dorm Room Inhalation
to Vehicle Emissions (DRIVE) Study. Health Effects Institute Research Report Number 196.
https://www.healtheffects.org/publication/developing-multipollutant-exposure-indicators-traffic-
pollution-dorm-room-inhalation.
Sheppard, EL (2022). 2022. "Letter from Elizabeth A. (Lianne) Sheppard, Chair, Clean Air
Scientific Advisory Committee, to Administrator Michale Regan. ." Re:CASAC Review of the
EPA's Supplement to the 2019 Integrated Science Assessment for Particulate Matter (External
Review Draft - October 2021). EPA-CASAC-22-001, March 18.
6-72
-------
Silverman, D. T., Samanic, C. M., Lubin, J. H., Blair, A. E., Stewart, P. A., Vermeulen, R., &
Attfield, M. D. 2012. "The diesel exhaust in miners study: a nested case-control study of lung
cancer and diesel exhaust." Journal of the National Cancer Institute.
Sisler, J.F. 1996. Spatial and seasonal patterns and long-term variability of the composition of
the haze in the United States: an analysis of data from the IMPROVE network. CIRA Report,
ISSN 0737-5352-32, Colorado State University.
Smith, CJ, PM Forster, M Allen, N Leach, RJ Millar, GA Passerello, and LA Regayre. 2018.
"FAIR vl.3: a simple emissions-based impulse response and carbon cycle model." Geosci.
Model Dev. 11(6): 2273-2297. doi:https://doi.org/10.5194/gmd-l 1-2273-2018.
Stenson, C., A.J. Wheeler, A. Carver, and et al. 2021. "The impact of traffic-related air pollution
on child and adolescent academic performance: a systematic review." Environ Intl 155: 106696.
Sun, X., S. Zhang, and X. Ma. 2014. "No association between traffic density and risk of
childhood leukemia: a meta-analysis." Asia Pac J Cancer Prev 15: 5229-5232.
Tol, R. 2009. An analysis of mitigation as a response to climate change. Copenhagen Consensus
on Climate, Copenhagen Consensus Center.
Turner, MC, Jerrett, M, Pope, A, III, Krewski, D, Gapstur, SM, Diver, WR, Beckerman, BS,
Marshall, JD, Su, J, Crouse, DL and Burnett, RT (2016). 2016. "Long-term ozone exposure and
mortality in a large prospective study." American Journal of Respiratory and Critical Care
Medicine 193 (10): 1134-1142.
Turtletaub, K.W., and C. Mani. 2003. Benzene metabolism in rodents at doses relevant to human
exposure from Urban Air. HEI Report 113. Health Effects Institutue.
U.S. Dept of State and U.S. Executive Office of the President. 2021. "The Long-Term Strategy
of the United States: Pathways to Net-Zero Greenhouse Gas Emissions by 2050." Washington,
D.C., November.
U.S. DOT. 2023. Freight Analysis Framework.
https://ops.fhwa.dot.gov/freight/freight_analysis/faf/.
U.S. EPA. 2002. 1,3-Butadiene IRIS page.
https://cfpub. epa.gov/ncea/iris2/chemicalLanding. cfm?substance_nmbr=139.
U.S. EPA. 2023. 2019 AirToxScreen Risk Drivers. Washington, DC: U.S. Environmental
Protection Agency. https://www.epa.gov/AirToxScreen/airtoxscreen-risk-drivers.
—. 2021. "2021 Policy on Children's Health." October 5.
https ://www.epa. gov/system/files/documents/2021-10/2021 -policy-on-childrens-health.pdf.
U.S. EPA. 2022. "2023 PMNAAQS Reconsideration Proposal RIA." EPA-HQ-OAR-2019-
0587.
U.S. EPA. 2006. "A Framework for Assessing Health Risks of Environmental Exposures to
Children. EPA/600/R-05/093F." Washinton, DC.
—. 2022. America's Children and the Environment. May 16. Accessed March 7, 2023.
https://www.epa.gov/americaschildrenenvironment.
6-73
-------
—. 2023. BenMAP-CE. Accessed March 10, 2023. https://github.com/BenMAPCE/BenMAP-
CE.
U.S. EPA. 2022. Derivation of an acute reference concentration for inhalation exposure to
naphthalene. EPA/600/R-21/292. Washington, D.C.: U.S. Environmental Protection Agency.
https://cfpub. epa.gov/ncea/risk/recordisplay. cfm?deid=355035.
U.S. EPA. 1991. Drinking Water Criteria Document for Polycyclic Aromatic Hydrocarbons
(PAHS). ECAO-CIN-OOIO. EPA Research and Development.
U.S. EPA. 1991. Effects of organic chemicals in the atmosphere on terrestrial plants. EPA/600/3-
91/001. Washington, DC: U.S. Environmental Protection Agency.
U.S. EPA. 2023. "Estimating PM2.5- and Ozone-Attributable Health Benefits. Technical Support
Document (TSD) for the 2023 PM NAAQS Reconsideration Proposal." EPA-HQ-OAR-2019-
0587.
U.S. EPA. 2018. "Estimating the Benefit per Ton of reducing PM2.5 Precursors from 17 Sectors."
https://www.epa.gov/sites/default/files/2018-
02/documents/sourceapportionmentbpttsd_2018.pdf.
U.S. EPA. 2021. Estimation of Population Size and Demographic Characteristics among People
Living Near Truck Routes in the Conterminous United States. Memorandum to the Docket, U.S.
Environmental Protection Agency.
U.S. EPA. 2016. Exposure Factors Handbook: 2011 Edition. Chapter 16. U.S. Environmental
Protection Agency, https://www.epa.gov/sites/production/files/2015-09/documents/efh-
Chapterl6.pdf.
U.S. EPA. 1990. Formaldehyde (CASRN 50-00-0) Chemical Assessment Summary.
Washington, DC: U.S. Environmental Protection Agency, Integrated Risk Information Systems
(IRIS). https://cfpub.epa.gov/ncea/iris2/chemicalLanding.cfm?substance_nmbr=419.
U.S. EPA. 1999. Guidelines for Carcinogen Risk Assessment (Review Draft) NCEA-F-0644.
Washinton, DC: U.S. Environmental Protection Agency. Accessed January 11, 2023.
https://cfpub. epa.gov/ncea/risk/recordisplay. cfm?deid=54932.
U.S. EPA. 2016. "Guidelines for Preparing Economic Analyses." Accessed March 10, 2023.
https://www.epa.gov/sites/default/files/2017-08/documents/ee-0568-50.pdf.
U.S. EPA. 2002. Health Assessment Document for Diesel Engine Exhaust. EPA/600/8-90/057F.
Washington. DC: U.S. Environmental Protection Agency.
U.S. EPA. 2002. "Health Assessment of 1,3-Butadiene." EPA 600-P-98-001F.
U.S. EPA. 2000. Integrated Risk Information System File for Benzene. Washington, DC: U.S.
Environmental Protection Agency.
https://cfpub. epa.gov/ncea/iris2/chemicalLanding. cfm?substance_nmbr=276.
U.S. EPA. 1991. Integrated Risk Information System File of Acetaldehyde. Washington, D.C.:
Research and Development, National Center for Environmental Assessment, U.S. Environmental
Protection Agency. https://cfpub.epa.gov/ncea/iris2/chemicalLanding.cfm?substance_nmbr=290.
6-74
-------
U.S. EPA. 2010. Integrated Science Assessment (ISA) for Carbon Monoxide (Final Report)
EPA/600/R-09/019F. Washington, DC: U.S. Environmental Protection Agency.
U.S. EPA. 2020. Integrated Science Assessment (ISA) for Oxides of Nitrogen, Oxides of Sulfur
and Particulate Matter Ecological Criteria (Final Report) EPA/600/R-20/278. Washington, D.C.:
U.S. Environmental Protection Agency.
U.S. EPA. 2020. Integrated Science Assessment (ISA) for Ozone and Related Photochemical
Oxidants (Final Report) EPA/600/R-20/012. Washington, DC: US Environmental Protection
Agency.
U.S. EPA. 2019. "Integrated Science Assessment (ISA) for Particulate Matter (Final Report)."
Office of Research and Development, Center for Public Health and Environmental Assessment
EPA-HQ-OAR-2020-0272.
U.S. EPA. 2017. Integrated Science Assessment (ISA) for Sulfur Oxides - Health Criteria (Final
Report) EPA/600/R-17/451. Washington, DC: U.S. Environmental Protection Agency.
—. 2016. "Integrated Science Assessment for Oxides of Nitrogen - Health Criteria (Final
Report)." National Center for Environmental Assessment, Research Triangle Park, NC. July.
https://cfpub. epa.gov/ncea/isa/recordisplay. cfm?deid=310879.
U.S. EPA. 2019. "Integrated Science Assessment for Particulate Matter (Final Report)
EPA/600/R-06/129F." Washington, DC, Ch 4 "Overall Conclusions" p. 4-1.
U.S. EPA. 2009. Integrated Science Assessment for Particulate Matter (Final Report)
EPA/600/R-08/139F. Washinton, D.C.: U.S. Environmental Protection Agency.
—. 2023. IRIS - Naphthalene CASRN 91-20-3.
https://cfpub. epa.gov/ncea/iris_drafts/recordisplay. cfm?deid=340791.
U.S. EPA. 2009. Metabolically-derived ventilation rates: A revised approach based upon oxygen
consumption rates EPA/600/R-06/129F. Washington, DC: Office of Research and Development.
http://cfpub. epa.gov/ncea/cfm/recordisplay. cfm?deid=202543.
—. 2024. National Ambient Air Quality Standards (NAAQS) for PM. February 15.
https://www.epa.gov/pm-pollution/national-ambient-air-quality-standards-naaqs-pm.
U.S. EPA. 2022. Policy Assessment (PA) for the Reconsideration of the National Ambient Air
Quality Standards for Particulate Matter (Final Report) EPA/452/R-22-004. Washington, D.C.:
U.S. Environmental Protection Agency.
U.S. EPA. 2020. Policy Assessment (PA) for the Review of the National Ambient Air Quality
Standards for Particulate Matter (Final Report) EPA/452/R-20/002. Washington DC: US EPA.
—. 2019. "Reduced Form Evaluation Project Report." https://www.epa.gov/benmap/reduced-
form-evaluation-proj ect-report.
—. 2017. "SAB Review of EPA's Proposed Methodology for Updating Mortality Risk Valuation
Estimates for Policy Analysis."
https://nepis. epa.gov/Exe/ZyPDF.cgi/P100ROQR.PDF?Dockey=P 100ROQR.PDF.
6-75
-------
U.S. EPA. 2022. "Standards of Performance for New, Reconstructed, and Modified Sources and
Emissions Guidelines for Existing Sources: Oil and Natural Gas Sector Climate Review EPA-
HQ-0 AR-2021 -0317."
U.S. EPA. 2022. "Supplement to the 2019 Integrated Science Assessment for Particulate Matter
(Final Report)." Office of Research and Development, Center for Public Health and
Environmental Assessment EPA-600-R-19-188.
U.S. EPA. 2005. Supplemental guidance for assessing susceptibility from early-life exposure to
carcinogens. EPA/630/R-03/003F. Washington, DC: Risk Assessment Forum.
https://www3.epa.gov/airtoxics/childrens_supplement_final.pdf.
—. 2021. "Technical Documentation on the Framework for Evaluating Damages and Impacts
(FrEDI) EPA 43O-R-21-004." https://www.epa.gov/cira/fredi.
U.S. EPA. 2022. Technical Support Document (TSD) EPA Air Toxics Screening Assessment.
2018 AirToxScreen TSD. Washington, D.C.: U.S. Environmental Protection Agency.
https://www.epa.gov/system/files/documents/2023-02/AirToxScreen_2018%20TSD.pdf.
U.S. EPA. 2023. "Technical Support Document: Estimating the Benefit per Ton of Reducing
Directly-Emitted PM2.5, PM2.5 Precursors and Ozone Precursors from 21 Sectors."
U.S. EPA. 2002. Toxicological Review of Benzene (Noncancer Effects) EPA/63 5/R-02/00 IF.
Washington, DC: U.S. Environmental Protection Agency, Integrated Risk Information System
(IRIS), National Center for Environmental Assessment.
U.S. EPA. 2017. Toxicological Review of Benzo[a]pyrene. Washington, DC: U.S.
Environmental Protection Agency.
https://cfpub.epa.gov/ncea/iris/iris_documents/documents/toxreviews/0136tr.pdf.
U.S. EPA. 2010. Toxicological Review of Formaldehyde (CAS No. 50-00-0) - Inhalation
Assessment: In Support of Summary Information on the Integrated Risk Information System
(IRIS). External Review Draft. EPA/635/R-10/002A. Washington, DC: U.S. Environmental
Protection Agency. http://cfpub.epa.gov/ncea/iris_drafts/recordisplay.cfm?deid=223614.
U.S. EPA. 1998. Toxicological Review of Naphthalene (Reassessment of the Inhalation Cancer
Risk). Washington, D.C.: Environmental Protection Agency, Integrated Risk Information
System, Research and Development, National Center for Environmental Assessment.
https://cfpub. epa.gov/ncea/iris_drafts/recordisplay. cfm?deid=56434.
U.S. EPA-SAB. 2005. "Advisory Council on Clean Air Compliance Analysis Response to
Agency Request on Cessation Lag." EPA-COUNCIL-LTR-05-001. December.
https://council.epa.gov/ords/sab/f?p=104:12:968651521971.
—. 2000. "An SAB Report on EPA's White Paper Valuing the Benefits of Fatal Cancer Risk
Reduction." https://www.epa.gov/system/files/documents/2022-03/86189901_0.pdf.
Ugrekhelidze, D., F, Korte, and G. Kvesitadze. 1997. "Uptake and transformation of benzene and
toluene by plant leaves." Ecotox. Environ. Safety 37: 24-29.
US EPA. 2021. "Climate Change and Social Vulnerability in the United States: A Focus on Six
Impacts." EPA 430-R-21-003.
6-76
-------
US EPA. 2022. "IRIS Toxicological Review of Formaldehyde-Inhalation (External Review
Draft)." EPA/635/R-22/039.
USGCRP. 2023. "Ch. 1. Overview: Understanding risks, impacts, and responses. In: Fifth
National Climate Assessment." Washington, D.C.: U.S. Global Change Research Program,
doi :https://doi.org/10.7930/NCA5.2023. CHI.
—. 2017. "Climate Science Special Report: Fourth National Climate Assessment, Volume I."
Washington, D.C.: U.S. Global Change Research Program. doi:10.7930/J0J964J6.
—. 2018. "Impacts, Risks, and Adaptation in the United States: Fourth National Climate
Assessment, Volume II." Washington, D.C.: U.S. Global Change Research Program.
doi:10.7930/NCA4.2018.
—. 2018. "Impacts, Risks, and Adaptation in the United States: Fourth National Climate
Assessment, Volume II." Washington, D.C.: U.S. Global Change Research Program.
doi:10.7930/NCA4.2018.
—. 2018. "Impacts, Risks, and Adaptation in the United States: Fourth National Climate
Assessment, Volume II." Washington, D.C.: U.S. Global Change Research Program.
https://www.nifc.gov/fireInfo/fireInfo_stats_totalFires.html.
—. 2016. "The Impacts of Climate Change on Human Health in the United States: A Scientific
Assessment."
Viskari, E.L. 2000. "Epicuticular wax of Norway spruce needles as indicator of traffic pollutant
deposition." Water, Air, and Soil Pollut. 121: 327-337.
Volk, H.E., I. Hertz-Picciotto, L. Delwiche, and et al. 2011. "Residential proximity to freeways
and autism in the CHARGE study." Environ Health Perspect 119: 873-877.
WMO. 2018. "Scientific Assessment of Ozone Depletion: 2018 Global Ozone Research and
Monitoring Project - Report No. 58." Geneva, Switzerland: World Meteorological Association.
Wolfe, P., K. Davidson, C. Fulcher, N. Fann, M. Zawacki, and K. R. Baker. 2019. "Monetized
Health Benefits Attributable to Mobile Source Emission Reductions across the United States in
2025." Science of the Total Environment (650): 2490-2498.
doi :https://doi.org/10.1016/J.SCITOTENV.2018.09.273.
Wu et al., X, Braun, D, Schwartz, J, Kioumourtzoglou, M and Dominici, F. 2020. "Evaluating
the impact of long-term exposure to fine particulate matter on mortality among the elderly."
Science Advances 6 (29): eaba5692.
Wu, J., M. Wilhelm, J. Chung, and et al. 2011. "Comparing exposure assessment methods for
traffic-related air pollution in and adverse pregnancy outcome study." Environ Res 111: 6685-
6692.
Zanobetti, A., P.H. Stone, F.E. Spelzer, J.D. Schwartz, B.A. Coull, H.H. Suh, B.D. Nearling,
M.A. Mittleman, R.L. Verrier, and D.R. Gold. 2009. "T-wave alternans, air pollution and traffic
in high-risk subjects." Am J Cardiol 104: 665-670.
6-77
-------
Zhang, J., J.E. McCreanor, P. Cullinan, and et al. 2009. "Health effects of real-world exposure
diesel exhaust in persons with asthma." Res Rep Health Effects Inst 138.
Ziska, L, A Crimmins, A Auclair, S DeGrasse, J F Garofalo, A S Khan, I Loladze, et al. 2016.
"Ch. 7: Food Safety, Nutrition, and Distribution. The Impacts of Climate Change on Human
Health in the United States: A Scientific Assessment." Washington, D.C.: U.S. Global Change
Research Program. 189-216.
6-78
-------
Chapter 7: Analysis of Air Quality Impacts of Light- and Medium-Duty
Vehicles Regulatory Scenario
For this final rule, EPA conducted an air quality modeling (AQM) analysis of the proposed
standards involving light- and medium-duty "onroad" vehicle emission reductions and
corresponding changes in "upstream" emission sources like EGUs (electric generating units) and
refineries. The analysis provides insight into the air quality impacts associated with emissions
increases and decreases from these multiple sectors.
This chapter presents a discussion of current air quality in Chapter 7.1, information about the
inventory used in the AQM analysis in Chapter 7.2, details related to the methodology used for
the AQM analysis in Chapter 7.3, results of the AQM analysis in Chapter 7.4, and quantified and
monetized benefits of the analysis in Chapter 7.5. Chapter 7.6 presents results of a demographic
analysis based on the AQM.
7.1 Current Air Quality
In this section, we present information related to current air pollutant concentrations and
deposition amounts. This provides context for the modeled projections presented in Chapter 7.4.
7.1.1 PM2.5 Concentrations
As described in Chapter 6 of this RIA, PM causes adverse health effects, and EPA has set
NAAQS to protect against those health effects. There are two primary NAAQS for PM2.5: an
annual standard (9.0 micrograms per cubic meter ([j,g/m3)) and a 24-hour standard (35 (j,g/m3),
and there are two secondary NAAQS for PM2.5: an annual standard (15.0 jag/ m3) and a 24-hour
standard (35 (J,g/m3). The initial PM2.5 standards were set in 1997 and revisions to the standards
were finalized in 2006, in 2012, retained in 2020, and finalized in 2024. On February 7, 2024,
EPA finalized a rule to revise the primary annual PM2.5 standard to 9.0 [j,g/m3. (U.S. EPA 2024)
There are areas of the country that are currently in nonattainment for the annual and 24-hour
primary PM2.5 NAAQS. As of November 30, 2023, more than 19 million people lived in the 3
areas that are designated as nonattainment for the 1997 annual PM2.5 NAAQS. (U.S. EPA 2023)
Also, as of November 30, 2023, more than 31 million people lived in the 11 areas that are
designated as nonattainment for the 2006 24-hour PM2.5 NAAQS, and more than 20 million
people lived in the 5 areas designated as nonattainment for the 2012 annual PM2.5 NAAQS. (U.S.
EPA 2023) (U.S. EPA 2023) In total, there are currently 12 PM2.5 nonattainment areas with a
population of more than 32 million people. (U.S. EPA 2023)236 Nonattainment areas for the
PM2.5 NAAQS, as of November 30, 2023, are pictured in Figure 7-1.
236 The population total is calculated by summing, without double counting, the 1997, 2006 and 2012 PM2.5
nonattainment populations contained in the Criteria Pollutant Nonattainment Summary report
(https://www.epa.gov/green-book/green-book-data-download).
7-1
-------
Counties Designated Nonattainment
for PM-2.5 (1997. 2006. and/or 2012 Standards)
Nonattainment areas are indicated by colof
\Mien only a portion or a county is shown in color
it indicates that onty that part ot the county is within
a nonattainment aroa boundary
Designated Nonattainment
HI All three PM-2.5 Stanoards
E3 Both 2006 and 2012 PM-2 5
\ I Both 1997 and 2006 PM-2 5
¦¦ 2012 PM 2 5 onty
I I 2006 PM-2 5 onty
rn~l 1997 PM-2 5 only
Figure 7-1: Counties designated nonattainment for PM2.5 (1997, 2006, and/or 2012
standards).
The final standards will take effect in 2027 and may assist areas with attaining the NAAQS
and may relieve areas with already stringent local regulations from some of the burden
associated with adopting additional local controls. The rule may also provide assistance to
counties with ambient concentrations near the level of the NAAQS who are working to ensure
long-term attainment or maintenance of the PM2.5 NAAQS.
7.1.2 Ozone Concentrations
As described in Chapter 6 of this RIA, ozone causes adverse health effects, and EPA has set
national ambient air quality standards (NAAQS) to protect against those health effects. The
primary NAAQS for ozone, established in 2015 and retained in 2020, is an 8-hour standard with
a level of 0.07 ppm. (U.S. EPA 2020) EPA is also implementing the previous 8-hour ozone
primary standard, set in 2008 at a level of 0.075 ppm. As of November 30, 2023, there were 34
ozone nonattainment areas for the 2008 primary ozone NAAQS, composed of 133 full or partial
counties, with a population of more than 90 million (see Figure 7-2); there were 46 ozone
nonattainment areas for the 2015 primary ozone N AAQS, composed of 191 full or partial
counties, with a population of more than 115 million (see Figure 7-3). (U.S. EPA 2023) (U.S.
EPA 2023) In total, there are currently (as of November 30, 2023) 54 ozone nonattainment areas
with a population of more than 119 million people. (U.S. EPA 2023).237
237 The total population is calculated by summing, without double counting, the 2008 and 2015 ozone nonattainment
populations contained in the Criteria Pollutant Nonattainment Summary report (https://www.epa.gov/green-
book/green-book-data-download).
7-2
-------
8-Hour Ozone Nonattainment Areas {2008 Standard)
Nonattainment areas are indicated by color
When only a portion of a county is s
it indicates that only that part of the
a nonattainment area boundary
I I Severe 15
| Serious
I | Moderate
I | Marginal
Figure 7-2: 8-Hour ozone nonattainment areas (2008 Standard).
8-Hour Ozone Nonattainment Areas (2015 Standard)
Nonattainment areas are indicated by color
\Mien only a portion of a county is shown in color,
it indicates that only that part of the county is within
a nonattainment area boundary
For the Ozone-8Hr (2015) Louisville, KY-IN nonattainment area the Ohio portion was redesignated on July 5 2022 The Kentucky portion has not been redesignated
The Kentucky portion of the Louisville area was reclassified from Marginal to Moderate on November 7 2022
The entire area is not considered in maintenance until all states in a multi-state area are redesignated
Figure 7-3: 8-Hour ozone nonattainment areas (2015 Standard).
States with ozone nonattainment areas are required to take action to bring those areas into
attainment. The attainment date assigned to an ozone nonattainment area is based on the area's
classification. The attainment dates for areas designated nonattainment for the 2008 8-hour
ozone NAAQS are in the 2015 to 2032 timeframe, depending on the severity of the problem in
each area. Attainment dates for areas designated nonattainment for the 2015 ozone NAAQS are
7-3
-------
in the 2021 to 2038 timeframe, again depending on the severity of the problem in each area.
The standards will take effect starting in MY 2027 and may assist areas with attaining the
NAAQS and may relieve areas with already stringent local regulations from some of the burden
associated with adopting additional local controls. The rule may also provide assistance to
counties with ambient concentrations near the level of the NAAQS who are working to ensure
long-term attainment or maintenance of the NAAQS.
7.1.3 NO2 Concentrations
There are two primary NAAQS for NO2: an annual standard (53 ppb) and a 1-hour standard
(100 ppb).238 In 2010, EPA established requirements for monitoring NO2 near roadways
expected to have the highest concentrations of NO2 within large cities. Monitoring within this
near-roadway network began in 2014, with additional sites deployed in the following years. At
present, there are no nonattainment areas for NO2.
7.1.4 SO2 Concentrations
EPA most recently completed a review of the primary SO2 NAAQS in February 2019 and
decided to retain the existing 2010 SO2 NAAQS. (US EPA 2023) The current primary NAAQS
for SO2 is a 1-hour standard of 75 ppb.239 As of November 30, 2023, there are 40 counties that
make up 30 SO2 nonattainment areas, with a population of over 2 million people. (U.S. EPA
2023).
238 The statistical form of the 1-hour NAAQS for N02 is the 3-year average of the yearly distribution of 1-hour daily
maximum concentrations.
239 The statistical form of the 1-hour NAAQS for SO2 is the 3-year average of the 99th percentile of 1-hour daily
maximum concentrations.
7-4
-------
S02 Nonattainment Areas (2010 Standard)
S02 Nonattainment Areas
tP
Nonattainment areas are indicated by color
When only a portion of a county is shown in color,
it indicates that only that part of the county is within
a nonattainment area boundary
Figure 7-4: Counties designated nonattainment for SO2 (2010 standard).
7.1.5 CO Concentrations
There are two primary NAAQS for CO: an 8-hour standard (9 ppm) and a 1 -hour standard (35
ppm). There are currently no CO nonattainment areas; as of September 27, 2010, all CO
nonattainment areas had been redesignated to attainment.
7.1.6 Air Toxics Concentrations
The most recent available data indicate that millions of Americans live in areas where air
toxics pose potential health concerns. (U.S. EPA 2022) The levels of air toxics to which people
are exposed vary depending on where people live and work and the kinds of activities in which
they engage, as discussed in detail in EPA's 2007 Mobile Source Air Toxics Rule. (U.S. EPA
2007) According to EPA's 2017 National Emissions Inventory (NEI), mobile sources were
responsible for 39 percent of outdoor anthropogenic toxic emissions. Further, mobile sources
were the largest contributor to national average risk of cancer and immunological and respiratory
health effects from directly emitted pollutants, according to EPA's Air Toxics Screening
Assessment (AirToxScreen) for 2019. (U.S. EPA 2022)240 Mobile sources are also significant
contributors to precursor emissions which react to form air toxics. (Cook, et al. 2020)
Formaldehyde is the largest contributor to cancer risk of all 72 pollutants quantitatively assessed
in the 2019 AirToxScreen. Mobile sources were responsible for 26 percent of primary
anthropogenic emissions of this pollutant in the 2017 NEI and are significant contributors to
2411 AirToxScreen also includes estimates of risk attributable to background concentrations, w hich includes
contributions from long-range transport, persistent air toxics, and natural sources; as well as secondary
concentrations, where toxics are formed via secondary formation. Mobile sources substantially contribute to long-
range transport and secondarily formed air toxics.
7-5
-------
formaldehyde precursor emissions. Benzene is also a large contributor to cancer risk, and mobile
sources account for about 60 percent of average exposure to ambient concentrations.
7.1.7 Deposition
Over the past two decades, the EPA has undertaken numerous efforts to reduce nitrogen and
sulfur deposition across the U.S. Analyses of monitoring data for the U.S. show that deposition
of nitrogen and sulfur compounds has decreased over the last 25 years. At 34 long-term
monitoring sites in the eastern U.S., where data are most abundant, average total nitrogen
deposition decreased by 43 percent between 1989-1991 and 2014-2016. (U.S. EPA 2022)
Although total nitrogen deposition has decreased over time, many areas continue to be negatively
impacted by deposition.
7.1.8 Visibility
As of November 30, 2023, over 32 million people live in areas that are designated
nonattainment for the PM2.5 NAAQS. Overall, the evidence is sufficient to conclude that a causal
relationship exists between PM and visibility impairment. (U.S. EPA 2019) Thus, the
populations who live in nonattainment areas and travel to these areas will likely be experiencing
visibility impairment. Additionally, while visibility trends have improved in Mandatory Class I
Federal areas, these areas continue to suffer from visibility impairment. (US EPA 2023) (US
EPA 2018) (US EPA 2020)241 In summary, visibility impairment is experienced throughout the
U.S., in multi-state regions, urban areas, and remote Mandatory Class I Federal areas.
7.2 Emissions Modeling for Air Quality Analysis
Air pollution emission inventories are an important input to AQM. This section describes the
modeled changes to onroad emissions from light- and medium-duty vehicles, as well as modeled
emission changes from "upstream" sectors like electricity generating units (EGUs) and refineries.
Emission inventories for unchanging sectors are referenced in the AQM Memo to the Docket.
(U.S. EPA 2024b)
For this analysis, AQM was performed for a 2016 base case, a 2055 reference scenario, and a
2055 light- and medium-duty vehicle (LMDV) policy scenario. The "reference" scenario
represents projected 2055 emissions and air quality without any additional LMDV controls. The
LMDV "policy" scenario is based on the proposed standards. In this scenario, we estimated that
battery electric vehicle (BEV) penetration would reach 71 percent for passenger cars and 66
percent for light-duty trucks in model year 2050. The policy scenario also assumes a phase-in of
gasoline particulate filters (GPF) for gasoline vehicles beginning for model year 2027 and later.
Air quality modeling was done for the future year 2055 when the LMDV policy scenario will be
fully implemented and when most of the regulated fleet will have turned over. The emissions
used for the policy scenario were the same as those in the reference scenario for all emissions
sectors except for onroad mobile source emissions, EGU emissions, and petroleum sector
emissions (specifically refineries, crude oil production well sites, and natural gas production well
sites). The net changes in emissions for these sectors is summarized in Table 7-11 below.
241 Mandatory Class I Federal areas are the 156 national parks and wilderness areas where state and federal agencies
work to improve visibility.
7-6
-------
The model used for the air quality analysis is the Community Multiscale Air Quality (CMAQ)
model which requires hourly emissions of specific gas and particle species for the horizontal and
vertical grid cells contained within the modeled region (i.e., modeling domain). Additional
information on projecting air quality model-ready emissions is included in the AQM Memo to
the Docket.
7.2.1 Onroad Vehicle Emission Estimates with MOVES
7.2.1.1 Overview
EPA's MOtor Vehicle Emission Simulator (MOVES) is a state-of-the-science emissions
modeling system that estimates air pollution emissions for criteria air pollutants, greenhouse
gases, and air toxics. MOVES covers light-, medium-, and heavy-duty onroad vehicles such as
cars, trucks and buses, and other mobile sources. MOVES accounts for the phase-in of federal
emissions standards, vehicle and equipment activity, fuels, temperatures, humidity, and emission
control activities such as inspection and maintenance (I/M) programs. (U.S. EPA 2023) Unlike
the OMEGA model described elsewhere in the RIA, MOVES can be used to estimate emissions
for specific counties as done here to capture geographical and temporal variation in onroad
vehicle emissions.
Table 7-1 summarizes the change in total onroad emissions between the reference scenario
and the policy scenario in calendar year 2055 as modeled for air quality analysis. Substantial
reductions are seen for all pollutants.
Table 7-1: Total onroad emissions impact in AQM policy scenario in 2055
Reference Policy Change in Percent
Pollutant Scenario Scenario Emissions
(tons/yr) (tons/vr) (tons/vr) ' c,cncc
PM2.5 34.667 26.342 -8'.325 -24%
NOx 403.861 319.169 -84.692 -21%
SO2 f>-458 4.124 -2.334 -36%
VOC 502.643 337.484 -165.159 -33%
7.2.1.2 MOVES versions used for air quality modeling
To generate the onroad emission inventories used for this AQM analysis, we developed
internal regulatory versions of MOVES4. These versions incorporated all the substantive features
of the public MOVES4.0 released August 30, 2023, but lacked some user-support tools and
documentation, and used slightly different input databases. Table 7-2 lists the code and database
versions associated with each of the three scenarios.
Table 7-2: MOVES versions for AQM scenarios
Scenario MOVES Code MOVES Database
2016 Base Year MOVES4.RC2 movcsdb20230515
2055 Reference Scenario MOVES4.R1 movcsdb20230713
2055 Policv Scenario MOVES4.R2 movcsdb20230817
The code and database differences between these MOVES versions and the public
MOVES4.0 are detailed in a docket memo. (Mo 2024) MOVES4.0 updates were peer reviewed
7-7
-------
under EPA's peer review policy. (U.S. EPA 2015)(ERG 2023) (U.S. EPA 2023) Developing
onroad inventories for the LMDV policy scenario required additional revised inputs as described
in Section 7.2.1.4.
For both the reference and policy scenario, county-specific age distributions and fuel mix
inputs were derived to preserve current differences between counties, such that counties with
newer-than-average vehicle fleets and more light-duty electric vehicles than average in calendar
year 2020 also have newer fleets and more electric vehicles in calendar year 2055. Additional
detail is provided in the AQM Memo to the Docket.
7.2.1.3 Modeling the Reference scenario with MOVES
The 2055 reference scenario was modelled with a slightly revised version of MOVES4.0. In
particular, the electric vehicle fleet-wide CO2 values for light-duty vehicles in the reference
scenario were updated to be consistent with the OMEGA No Action case in the NPRM analysis.
Similarly, LD electric vehicle sales were estimated using values from the NPRM OMEGA No
Action case, reaching 33 percent and 36 percent in model year 2050 for light-duty trucks and
passenger cars, respectively.
Electric vehicle fractions for heavy-duty vehicles were updated to reflect EPA's waiver of
preemption for California's Advanced Clean Trucks (ACT) regulation. The national impact of
the ACT was estimated based on the implementation of the rule in California and seven other
states which had adopted the ACT at the time of analysis. In tandem with that work, heavy-duty
electric vehicle adoption rates were updated for states that had not adopted the ACT. Finally, the
energy efficiency of HD BEVs and Fuel Cell Electric Vehicles (FCEVs) was updated for
consistency with EPA's analysis with the EPA TRUCS model. (Sui 2023)
Electric vehicle fractions for medium-duty vehicles were based on the updated modeling of
ACT similarly to heavy-duty vehicles, except that adoption rates in non-ACT states were
modeled using OMEGA.
7.2.1.4 Modeling the Policy scenario with MOVES
The policy scenario was modeled with a light- and medium-duty fleet that phased-in new
vehicle BEV sales based on NPRM OMEGA modeling of the proposed rule and an analysis of
the expected national impact of California's ACT regulation. We assumed the required
improvement in average CO2 emissions for light-duty vehicles. For HC and NOx emissions, we
modeled reduced fleet-wide emissions consistent with the proposed new bin structure. For PM,
we modeled reduced LD gasoline vehicle organic carbon and elemental carbon rates consistent
with predicted impact of GPFs based on OTAQ literature review and testing as described in
7.2.1.4.4.1. Vehicle age distributions were the same as in MOVES4.0. More details on each of
these changes are provided below.
7.2.1.4.1 EV sales and stock
The policy scenario EV penetrations (fraction of new sales) for light-duty passenger cars and
light-duty trucks were modeled in MOVES based on OMEGA EV outputs. These penetrations
are fleet-wide BEV penetration by model year, with separate values for light-duty passenger cars
and light-duty trucks. We assume BEV penetration will reach 71 percent for passenger cars and
66 percent for light-duty trucks in model year 2050. For medium-duty class 2b and 3, the policy
7-8
-------
scenario was modeled based on an analysis of OMEGA NPRM outputs for the proposed
standards and the expected national impact of California's ACT regulation. The distribution of
EV sales among counties was similar to the reference scenario and is discussed in detail in the
AQM Memo to the Docket.
Vehicle age distributions were the same as in the reference scenario.
7.2.1.4.2 Internal Combustion Engine Vehicle Energy Consumption
For the policy scenario modeling, the internal combustion engine vehicle (ICEV) energy rates
(MY2027-MY2060) were adjusted to match rates from OMEGA modeling of a scenario in
which EV sales were estimated as described above, and ICEV rates were limited by light-duty
fleet-wide average standards that assume zero tailpipe CO2 g/mi for BEVs and allow averaging
between ICEVs and electric vehicles.
Energy consumption changes for medium-duty class 2b and 3 were driven by the EV fraction
update in the policy scenario (described above). No additional adjustments were applied to the
class 2b and 3 ICEV energy consumption rates to meet the LMDV policy scenario standards.
7.2.1.4.3 ICEV HC and NOx
After accounting for projected increases in the sales of electric vehicles following the onset
of the phase-in period for the rule, we concluded that ICE vehicles meeting Tier-3 standards for
NMOG+NOx as modelled in MOVES would comply with fleet-average requirements under the
current rulemaking as well. Accordingly, the base emission rates for NOx and total hydrocarbons
(THC) were not modified for the analyses supporting this RIA. This conclusion was reached by
simulating FTP composite emissions for NMOG+NOx in MOVES using rates for the start- and
running-exhaust emissions processes and applies to both light-duty and medium-duty vehicles.
However, the analyses also account for the possibility that the increased sales of electric
vehicles could allow manufacturers to certify ICE vehicles to NMOG+NOx levels higher than
they might absent the rule. This possibility was accounted for by assuming that manufacturers
would certify ICE vehicles to the highest level achievable given assumed levels of EV sales in a
given model year and within standard levels allowed by the rule. For light-duty vehicles, Tier-3
Bins above Bin70 were excluded to reflect the proposed bin changes under the LMDV. For
medium-duty vehicles, fleet-average requirements are stringent enough that standard levels
above Binl60 are effectively excluded. Adjustments to emissions levels to reflect these
assumptions were developed and applied at the MOVES source type level through the
emissionRateAdjustment table. For light-duty vehicles, adjustments range from 1.08-1.54. For
medium-duty vehicles, adjustments ranged from 0.90-1.40 and 0.33-1.14 for vehicles in classes
2b and 3, respectively.
Additionally, the rule's provisions to control NMOG+NOx start emissions for "intermediate"
soak periods were represented by reducing start emissions for soak periods between 40 minutes
and 12 hours, covering MOVES operating modes 104 to 107. After verifying that Tier-3 rates at
45 minutes and 12 hours (operating modes 103 and 108) would meet requirements specific for
those periods, we projected reduced emissions levels for the intervening periods by interpolating
linearly between the levels at modes 103 and 108. These modifications were applied directly to
the base rates for THC, CO, and NOx in the targeted operating modes in the MOVES
emissionRateByAge table. Reductions are largest for NOx, THC, and CO, in that order.
7-9
-------
7.2.1.4.4 ICEV PM rates
PM emissions reductions were modeled for light-duty gasoline vehicles for model years 2027
and later. The modeled reductions were based on present-day GPFs as the best current PM
reduction technology. GPFs filter PM in the exhaust, thus directly reducing PM emissions. The
filter effectiveness differs for elemental carbon (EC) PM (MOVES Pollutant ID 112) and for
non-EC PM (MOVES pollutant ID 118). To model the addition of GPFs, we apply a
proportional reduction to the relevant start and running exhaust PM emission rates. In this case,
the reductions are applied to start and running emissions for light-duty cars and trucks for
gasoline, diesel and E85 fuels (MOVES fuelTypelD in 1,2,5).242 For class 2b and 3 trucks, the
reductions were applied for gasoline trucks only. Note, for MY 2010 and later, the MOVES
emission rates for class 2b and 3 diesel trucks already included reductions representing control
from diesel particulate filters (DPFs).
While we modeled substantial reductions in exhaust PM due to the rule, we model no change
in brake and tire wear emissions, modeling BEVs with the same brake and tire wear as the
ICEVs they replace.243
7.2.1.4.4.1 PM emission reduction fractions
The reduction fractions applied to both EC and non-EC PM are derived from laboratory
testing of a lightly loaded underfloor catalyzed GPF. (Bohac and Ludlum 2023) For that study,
EC and organic carbon (OC) measurements were made using the NIOSH 870 method. Here, we
use the observed reduction in EC to determine the reduction fraction for the MOVES EC
pollutant. We use the observed OC reduction as the reduction fraction to apply to the MOVES
non-EC PM pollutant. OC is not identical to non-EC PM because OC measurements do not
include information about other elemental components of the particulate matter such as
hydrogen, nitrogen, oxygen, calcium, and metallic ash components. For modeling purposes, we
assume that the other components of non-EC PM are filtered by the GPF in the same proportion
as the OC part.
The reduction factors for the start operating modes come from the study's 25°C FTP cycle
tests. For running emissions excluding MOVES operating modes 30 and 40, the reduction factors
come from averaging the results of the 60mph and highway fuel economy (HWFET) tests. The
reduction fractions for operating modes 30 and 40 are from the US06 test. Finally, to avoid
computational issues that arise from setting emission factors to zero, reductions originally
reported as 100% were adjusted to 99.9%. The final PM reductions by operating mode are
summarized in Table 7-3 below.
242 While GPFs are relevant only for gasoline and E85 vehicles, in MOVES, the emission rates for light-duty
gasoline vehicles were also applied to light-duty diesel. This has a negligible impact on calendar year 2055
emissions since we model the diesel fraction of the light-duty sales as less than 0.002% for all model years after
2018.
243 Road dust, including road wear, is not modelled by MOVES, but is included in the air quality modeling as an
area source. We modelled no difference in road dust from EVs as compared to ICEVs.
7-10
-------
Table 7-3: PM reduction by MOVES operating mode
Operating Modes EC Reduetion (%) Non-EC PM Reduetion (%)
0 - 29
99.9
75
30
98.5
80
33 - 39
99.9
75
40
98.5
80
101 - 108
99.9
91
7.2.1.4.4.2 PM reduction phase-in
To model the air quality modeling policy scenario, we applied the PM reduction phase-in
fractions shown in Table 7-4.
Table 7-4: PM control fraction by MOVES reg class and model year
Model Year Reg Class 20 Reg Class 30 Reg Class 41
2026 0 0 0
2027 0.4 0.2 0
2028 0.8 0.4 0
2029 1 0.5 0
2030+ 1 1 1
The phase-in was combined with the reduction factors for each operating mode to create
weighted reduction factors for each model year. Finally, the weighted reduction factors were
applied to the original MOVES base emission rates to create a set of new, lower PM emission
rates.
7.2.1.4.4.3 PM update for LEV rates
The phase-in described above overlaps with the California 1 mg/mile PM standard that is
relevant for California and for other states that have adopted California requirements under
Clean Air Act Section 177. Prior to phasing in GPF-equivalent PM rates, the rates in the
MOVES emissionRateByAgeLEV table were lower than the rates in the MOVES default
emissionRateByAge table. For passenger cars, the policy scenario default values are lower than
the LEV table values starting in model year 2027, and for light-duty trucks, the policy scenario
default values are lower starting in 2030. Therefore, for the policy scenario modeling, the
emissionRateByAgeLEV table was updated by dropping the rates for those years where the new
default emission rates are lower than the LEV rates.
7.2.2 Upstream Emission Estimates for AO Modeling
This section describes emission impacts estimated for the following "upstream" emission
sources: EGU emissions (Chapter 7.2.2.1), refinery emissions (Chapter 0), emissions from crude
oil production well sites and pipeline pumps (Chapter 7.2.2.3), and emissions from natural gas
production well sites and pipeline pumps (Chapter 7.2.2.4).
EPA estimates that total upstream emissions in the policy scenario will decrease compared to
the reference scenario.
Table 7-5 presents the net impact of the upstream sources by pollutant in 2055. The impacts
include a projected increase in emissions from EGUs, as well as increased emissions projected
7-11
-------
from natural gas production well sites and pipeline pumps, due to a projected increase in natural
gas fueled EGUs. The emission impacts also include a projected decrease in emissions from
refineries and crude production wells and pipeline pumps due to assumed decreases in activity at
refineries related to a decrease in demand for liquid fuels for light- and medium-duty vehicles.
Table 7-5: Total upstream emissions impact in AQM policy scenario in 2055
Pollutant
Reference
Scenario
(tons/vr)
Policy
Scenario
(tons/vr)
Change in
Emissions
(tons/vr)
Percent
Difference
PM2.5
64.115
62,722
-1.393
-2%
NOx
814.881
805.238
-9.643
-1%
SO2
142.170
139.241
-2.929
-2%
VOC
2.852.174
2.823.145
-29.029
-1%
There is uncertainty about the impact of reduced demand for petroleum fuels on refinery
activity and emissions. In response to comments, we have updated our estimates of the impacts
of reduced domestic fuel demand on U.S. refining. In the NPRM illustrative AQ analysis, we
projected that the LMDV regulatory scenario would result in lower demand for onroad fuels and
therefore reduce emissions from fuel refineries. The NPRM assumed that most of the refined
product demand caused by the proposed rulemaking would result in a similar reduction in U.S.
refinery operations (93 percent), and a sensitivity analysis was performed where U.S. refineries
continued to operate at their current capacities. As noted by commenters, there are good
economic reasons why U.S. refineries might continue to operate despite reduced domestic
demand, leading to increased exports. Therefore, for this final rule analysis, we assumed that
more of the drop in domestic demand would be offset by increased exports than in the NPRM
analysis (see discussion in Chapter 0).
7.2.2.1 Electricity Generating Units (EGUs)
The EGU emissions inventories used in the air quality analysis were developed from 2050
outputs of the 2022 post-IRA Reference Case run of the Integrated Planning Model (IPM). This
version of IPM included EGU fleet information and rules and regulations that were final at the
time the IPM version was finalized, and included supply-side impacts (production and
investment tax credits) associated with the Inflation Reduction Act (IRA). (U.S. EPA 2023)
More detail on the rules and regulations included in this version of IPM, as well as additional
information on the IPM version, can be found in the AQM Memo to the Docket.
Emissions of select pollutants from EGUs in 2050 (representing 2055 levels) are shown in
Table 7-6. The policy scenario caused an increase in emissions of all pollutants, which is
expected as the policy case includes an increase in electric vehicles.
7-12
-------
Table 7-6: EGU emissions impact in AQM inventories in 2055244
Pollutant
Reference
Scenario
(tons/vr)
Policy Scenario
(tons/vr)
Change in
Emissions
(tons/vr)
Percent
Difference
PM2.5
26.420
27.459
1.039
4%
NOx
95.934
97.539
1.605
2%
SO2
17.117
19.063
1.946
11%
VOC
17.023
17.490
467
3%
7.2.2.2 Refineries
The reference scenario refinery emission inventories used in the air quality analysis were a
subset of the refinery emissions in the 2016v3 emissions modeling platform that were projected
to 2050 using the reference case modeled by EIA in its 2023 Annual Energy Outlook (AEO)
(U.S. EIA 2023) (U.S. EPA 2023). Pollutant-specific adjustment factors were developed and
then applied to the reference scenario inventory to generate the policy scenario inventory. These
adjustment factors are presented in
Table 7-7 and account for reduced domestic fuel demand in response to the policy scenario
(proposed standards).
As mentioned above, in the NPRM air quality analysis, we assumed that 7% of the reduced
domestic demand for refined fuels would be made up by an increase in exports, based on a
comparison of the reference case and low economic growth case in AEO 2021. We received
comments from several organizations that refineries would increase exports more than we
assumed. After taking into consideration stakeholder comments, the more desirable economic
conditions for refiners in the U.S., and the recent closures and conversions of some U.S.
refineries, we have updated our projection of how refineries will be impacted by this rulemaking.
For this final rule AQM analysis, we estimated policy case refinery emissions by assuming that
U.S. refineries would increase exports to offset half of the projected reductions in domestic
demand for liquid fuels. Thus, the total decrease in refinery activity, measured in gallons of
gasoline and diesel refined, is half of the estimated drop in domestic fuel demand. Additional
detail on how the adjustment factors were calculated is available in the AQM Memo to the
Docket.
Table 7-7: Adjustment factors to apply to 2050 refinery inventory
Pollutant Policy Scenario
PM2.5 °.86
NOx 0.86
SO2 0.87
VOC 0.87
244IPM output for a set of years with the furthest out year being 2055. The 2050 output was used in the air quality
analysis and was assumed to represent 2055 to avoid any "end of timeframe" issues with using the furthest out year
output from the model.
7-13
-------
Emissions decreases of select pollutants from refineries in 2055 are shown in
Table 7-8.
Table 7-8: Refinery emissions impact in AQM inventories in 2055
Pollutant
Reference
Scenario
(tons/vr)
Policy Scenario
(tons/vr)
Change in
Emissions
(tons/vr)
Percent
Difference
PM2.5
18.867
16.399
-2.467
-13%
NOx
80.188
69.720
-10.468
-13%
SO2
25.846
22,779
-3,067
-12%
VOC
62.842
55,637
-7,205
-12%
7.2.2.3 Crude Production Well Sites and Pipeline Pumps
The reference case emission inventories for crude production well sites and associated
pipeline pumps used in the air quality analysis were developed from emissions in the 2016v3
emissions modeling platform that were projected to 2050 using the reference case modeled by
EIA in its 2023 AEO (U.S. EIA 2023) (U.S. EPA 2023). Policy case emissions were decreased
(through application of an adjustment factor) to account for lower activity due to lower domestic
demand for liquid fuels. This adjustment factor is 0.98 and additional detail on how the
adjustment factor was calculated is available in the AQM Memo to the Docket. Decreases in
emissions of select pollutants from crude production well sites and pipeline pumps in 2055 are
shown in Table 7-9.
Table 7-9: Crude production well site and pipeline pump impact in AQM inventories in
2055
Pollutant
Reference
Scenario
(tons/vr)
Policy Scenario
(tons/vr)
Change in
Emissions
(tons/vr)
Percent
Difference
PM2.5
5.102
5.000
-102
-2%
NOx
238.895
234.117
-4.778
-2%
SO2
93.330
91.464
-1.867
-2%
VOC
1.667.134
1.633.791
-33.343
-2%
7.2.2.4 Natural Gas Production Well Sites and Pipeline Pumps
The reference case emission inventories for natural gas production well sites and associated
pipeline pumps used in the air quality analysis were developed from emissions in the 2016v3
emissions modeling platform that were projected to 2050 using the reference case modeled by
EIA in its 2023 AEO (U.S. EIA 2023) (U.S. EPA 2023). Policy case emissions were increased
(through application of an adjustment factor) to account for increased activity at natural gas
production well sites and pipeline pumps consistent with increased demand for natural gas fueled
EGUs. This adjustment factor is 1.01 and additional detail on how the adjustment factor was
calculated is available in the AQM Memo to the Docket. Increases in emissions of select
pollutants from natural gas production well sites and pipeline pumps in 2055 are shown in
7-14
-------
Table 7-10.
Table 7-10: Natural gas production well and pipeline pump impact in AQM inventories in
2055
Pollutant
Reference
Scenario
(tons/yr)
Policy Scenario
(tons/yr)
Change in
Emissions
(tons/yr)
Percent
Difference
PM2.5
13.726
13.863
137
1%
NOx
399.863
403.862
3.999
1%
SO2
5.876
5.935
59
1%
VOC
1.105.175
1.116.227
11.052
1%
7.2.2.5 Limitations of the Upstream Inventory
There is uncertainty associated with the upstream inventory. The air quality analysis assumes
that there is no change in mandated renewable fuel volumes and percentages, that refineries will
decrease some of their activity rather than export additional fuels, that the decreased production
occurs at the same rate at all refineries and at all crude production wells and pipeline pumps, and
that increased production occurs at the same rate at all natural gas production wells and pipeline
pumps. In addition, projections out to 2055 inherently are less certain than projections that do not
go out as far into the future. Lastly, the upstream emissions inventory does not account for all
upstream sources related to vehicles, fuels, and electricity generation, such as charging
infrastructure, storage of petroleum fuels, battery manufacture, etc.
7.2.3 Combined Onroad and Upstream Emission Impacts
Total onroad, upstream, and net emissions of select pollutants in 2055 are shown in
Table 7-11. The policy case has lower combined onroad and upstream emissions than the
reference case, driven by reductions in the onroad sector.
Table 7-11: Net impacts" on criteria pollutant emissions from the LMDV regulatory
scenario.
2055 AQM Policy Scenario
(tons/yr)
2055 AQM Reference Scenario
(tons/vr)
Pollutant Onroad Upstream
PM2.5
34.667
64.115
NOx
403.861
814.881
SO2
6.458
142.170
VOC
502.643
2.852.174
Total
Onroad
and
Upstream
98.782
1.218.742
148.628
3.354.817
Onroad Upstream
26.342
319.169
4.124
337.484
62,722
805.238
139.241
2.823.145
Net
Percent
Total
Emissions
Change
Onroad
Impact
Emissions
and
(tons/vr)
Impact
Upstream
89.063
-9.719
-10%
1.124.407
-94.335
-8%
143.365
-5.263
-4%
3.160.629
-194.188
-6%
1 Emissions reductions arc presented as negative numbers and emissions increases as positive numbers.
7-15
-------
7.3 Air Quality Modeling Methodology
In this section we present information related to the methods used in the air quality analysis
for this final rule. Additional information is available in the AQM Memo to the Docket. (U.S.
EPA 2024b)
7.3.1 Air Quality Model
CMAQ is a non-proprietary computer model that simulates the formation and fate of
photochemical oxidants, primary and secondary PM concentrations, acid deposition, and air
toxics, over regional and urban spatial scales for given inputs of meteorological conditions and
emissions. CMAQ includes numerous science modules that simulate the emission, production,
decay, deposition, and transport of organic and inorganic gas-phase and particle pollutants in the
atmosphere. The CMAQ model is a well-known and well-respected tool and has been used in
numerous national and international applications. (US EPA 2023) The AQM completed for this
rule used the 2016v3 platform with the most recent multi-pollutant CMAQ code available at the
time of AQM (CMAQ version 5.4). (UNC Institute for the Environment 2023) The 2016 CMAQ
runs utilized the CB6r5 chemical mechanism (Carbon Bond with linearized halogen chemistry)
for gas-phase chemistry, and AER07 (aerosol model with non-volatile primary organic aerosol)
for aerosols. The CMAQ model is regularly peer-reviewed; CMAQ versions 5.2 and 5.3 beta
were most recently peer-reviewed in 2019 for the U.S. EPA. (Versar, Inc 2019)
7.3.2 Model Domain and Configuration
The CMAQ modeling analyses used a domain covering the continental United States, as shown
in Figure 7-5. This single domain covers the entire continental U.S. (CONUS) and large portions
of Canada and Mexico using 12 km x 12 km horizontal grid spacing. The 2016 simulation used a
Lambert Conformal map projection centered at (-97, 40) with true latitudes at 33 and 45 degrees
north. The model extends vertically from the surface to 50 millibars (approximately 17,600
meters) using a sigma-pressure coordinate system with 35 vertical layers. Table 7-12 provides
some basic geographic information regarding the CMAQ domains and Table 7-13 provides the
vertical layer structure for the CMAQ domain.
Table 7-12: Geographic elements of domains used in air quality modeling.
CMAQ Modeling Configuration
Grid Resolution
Map Projection
Coordinate Center
True Latitudes
12 km National Grid
Lambert Conformal Projection
97 deg W. 40 deg N
33 deg N and 45 deg N
396 x 246 x 35
35 Layers: Surface to 50 millibar level
(see
Dimensions
Vertical extent
Table 7-13)
7-16
-------
Table 7-13: Vertical layer structure for CMAQ domain.
Vertical
Sigma P
Pressure
Approximate
Layers
(mb)
Height (m)
35
0.0000
50.00
17.556
34
0.0500
97.50
14.780
33
0.1000
145.00
12.822
32
0.1500
192.50
11.282
31
0.2000
240.00
10.002
30
0.2500
287.50
8.901
29
0.3000
335.00
7.932
28
0.3500
382.50
7.064
27
0.4000
430.00
6,275
26
0.4500
477.50
5,553
25
0.5000
525.00
4.885
24
0.5500
572.50
4.264
23
0.6000
620.00
3.683
22
0.6500
667.50
3.136
21
0.7000
715.00
2.619
20
0.7400
753.00
2.226
19
0.7700
781.50
1.941
18
0.8000
810.00
1.665
17
0.8200
829.00
1.485
16
0.8400
848.00
1.308
15
0.8600
867.00
1.134
14
0.8800
886.00
964
13
0.9000
905.00
797
12
0.9100
914.50
714
11
0.9200
924.00
632
10
0.9300
933.50
551
9
0.9400
943.00
470
8
0.9500
952.50
390
7 |
0.9600
962.00
3 11
6
0.9700
971.50
232
5
0.9800
981.00
154
4
0.9850
985.75
115
3
0.9900
990.50
77
2
0.9950
995.25
38
1
0.9975
997.63
19
o
1.0000
1000.00
0 ""
7-17
-------
7.3.3 Model Inputs
The key inputs to the CMAQ model include emissions from anthropogenic and biogenic
sources, meteorological data, and initial and boundary conditions. The emissions inputs are
summarized above in Chapter 7.2.
The CMAQ meteorological input files were derived from simulations of the Weather
Research and Forecasting Model (WRF) version 3.8 for the entire 2016 year. (Skamarock 2008)
(U.S. EPA 2019) The WRF Model is a state-of-the-science mesoscale numerical weather
prediction system developed for both operational forecasting and atmospheric research
applications. (National Center for Atmospheric Research 2022) The meteorological outputs from
WRF were processed to create 12 km model-ready inputs for CMAQ using the Meteorology-
Chemistry Interface Processor (MCIP) version 4.3. These inputs included hourly varying
horizontal wind components (i.e., speed and direction), temperature, moisture, vertical diffusion
rates, and rainfall rates for each grid cell in each vertical layer. (Byun, Ching and EPA 1999)
The boundary and initial species concentrations were provided by a northern hemispheric
CMAQ modeling platform for the year 2016. (Henderson 2018) (Mathur 2017) The hemispheric-
scale platform uses a polar stereographic projection at 108 km resolution to completely and
7-18
-------
continuously cover the northern hemisphere for 2016. Meteorology is provided by WRF v3.8.
Details on the emissions used for hemispheric CMAQ can be found in the 2016 hemispheric
emissions modeling platform TSD. (U.S. EPA 2019) The atmospheric processing
(transformation and fate) was simulated by CMAQ (v5.2.1) using the CB6r3 and the aerosol
model with non-volatile primary organic carbon (AE6nvPOA). The CMAQ model also included
the on-line windblown dust emission sources (excluding agricultural land), which are not always
included in the regional platform but are important for large-scale transport of dust.
7.3.4 Model Evaluation
The CMAQ predictions for ozone, fine particulate matter, sulfate, nitrate, ammonium, organic
carbon, elemental carbon, nitrogen and sulfur deposition, and specific air toxics (acetaldehyde,
benzene, and formaldehyde) from the 2016 base scenario were compared to measured
concentrations in order to evaluate the ability of the modeling platform to replicate observed
concentrations. This evaluation was comprised of statistical and graphical comparisons of paired
modeled and observed data. Details on the model performance evaluation, including a
description of the methodology, the model performance statistics, and results, are provided in the
AQM Memo to the Docket. (U.S. EPA 2024b)
7.3.5 Model Simulation Scenarios
As part of our analysis for this rulemaking, the CMAQ modeling system was used to calculate
annual PM2.5 concentrations and projected design values, 8-hour maximum average ozone season
(April - Sept) concentrations and design values, annual NO2, SO2, and CO concentrations, annual
and seasonal (summer and winter) air toxics concentrations, and annual nitrogen and sulfur
deposition for each of the following scenarios:
• 2016 base year;
• 2055 reference;
• 2055 light- and medium-duty policy scenario based on proposed standards.
Decisions about the emissions and other elements used in the air quality modeling were made
early in the analytical process for the final rulemaking and the decision was made to model the
proposed standards. Accordingly, the air quality analysis does not fully represent the final
regulatory scenario; however, we consider the modeling results to be a fair reflection of the
impact the standards will have on air quality in 2055.
When possible, we use the predictions from the CMAQ model in a relative sense by
combining the 2016 base-year predictions with predictions from the future-year scenario and
applying these modeled ratios to ambient air quality observations to estimate future-year
concentrations. The PM2.5 and ozone concentrations are modeled using this relative method. The
ambient air quality observations are average conditions, on a site-by-site basis, for a period
centered around the model base year (i.e., 2014-2018).
The projected PM2.5 concentrations and design values, and 8-hour ozone concentrations and
design values, were calculated using the approach identified in EPA's guidance on air quality
modeling attainment demonstrations. (US EPA 2018)
7-19
-------
Additionally, we conducted an analysis to compare the absolute differences between the
future year reference and policy scenario for annual and seasonal acetaldehyde, benzene,
formaldehyde, and naphthalene, as well as annual NO2, SO2, CO, and nitrate/sulfate deposition.
These data were not compared in a relative sense due to the limited observational data available.
7.4 Results of Air Quality Analysis
For this final rule, EPA conducted an air quality modeling analysis of the proposed standards
involving light- and medium-duty "onroad" vehicle emission reductions and corresponding
changes in "upstream" emission sources like EGU (electric generating unit) emissions and
refinery emissions. We also modeled a sensitivity case that examined only the air quality impacts
of the onroad emissions changes from the proposed standards. This "onroad-only" sensitivity
case assumed no change in emissions from upstream sources and is based on the onroad
emission inventories described in Chapter 7.2.1.
In this section, we summarize the results of our AQM based on the projected emission
impacts of the policy scenario, as well as the onroad-only sensitivity case. Air quality modeling
was done for the future year 2055, which is when the program will be fully implemented and
when most of the regulated fleet will have turned over. The "reference" scenario represents
projected 2055 air quality without the policy scenario, and the "policy" scenario represents
projected 2055 emissions with the proposed standards. As described in Chapter 7.2, in this
scenario we estimated that battery electric vehicle (BEV) penetration would reach 71 percent for
passenger cars and 66 percent for light-duty trucks in model year 2050. The policy case also
assumes a phase-in of gasoline particulate filters (GPF) for gasoline vehicles beginning in model
year 2027.
7.4.1 PM2.5
7.4.1.1 Overall Projected PM2.5 Impacts
This section summarizes projected changes in PM2.5 concentrations in 2055 from the rule.
Figure 7-6 presents the absolute changes in annual average PM2.5 concentrations in 2055
between the reference and the policy scenario and indicates that there will be widespread
decreases due to the projected reductions in primary PM, NOx, SO2, and VOC emissions, see
Table 7-11 (blue areas). In a few isolated areas, this rule is expected to result in increases in
annual average PM2.5, likely due to increases in EGU emissions. We expect the power sector to
become cleaner over time as a result of the IRA and future policies, which will reduce the air
quality impacts of EGUs. Less than 0.1% of CMAQ grid cells are projected to have increases in
annual average PM2.5 concentration of >0.025 [j,g/m3 (yellow and red areas), and only 0.1% of
the population of CONUS lives in those areas.
7-20
-------
'A
Max: 0.2013 Min 0.3602
*. V
r
>0.100
/ \ r . ?
^ V '
1
0.075
4 jj^
0.050
0.025
0.000
f/ P'
y
-0.025
-0.050
Y
\
.V>
I
-0.075
<-0.100
*1 1 -
aJ . n •
Figure 7-6: Projected changes in annual average P1VI2.5 concentrations in 2055 due to the
rule.
The rule will decrease annual average PM2.5 concentrations by an average of 0.02 ug/nr' in
2055, with a maximum decrease of 0.36 (ig/m3 and a maximum increase of 0.20 (jg/m3. The
population-weighted average change in annual average PM2.5 concentrations will be a decrease
of 0.04 jig/m5 in 2055. We also expect that this rule's reductions in directly emitted PM2.5 will
contribute to reductions in PM2.5 concentrations near roadways, although our air quality
modeling is not of sufficient resolution to capture that impact.
7.4.1.2 Onroad-Only Projected PM2.5Impact
We also modeled an "onroad-only" sensitivity case. Figure 7-7 presents the absolute changes
in annual average PM2.5 concentrations in 2055 between the reference and onroad-only
sensitivity scenarios.
7-21
-------
b
9 \
H \ *¦*
Max: le-04 Min: -0.129.7
ML
w.
,' %-s
> • .-¦" '+MJ!
' ' — 1- i-
-t -
.- r
,f»v',* /
l f
i
x i
- ¦ _
Figure 7-7: Projected changes in annual average P1VI2.5 concentrations in 2055 from
"onroad-only" emissions changes.
When only the onroad emissions impacts of the rule are considered, annual average PM2.5
concentrations will decrease by an average of 0.02 (ig/m3 in 2055, with a maximum decrease of
0.13 ug/m \ The population-weighted average change in annual average PM2.5 concentrations
attributable to the onroad emissions reductions will be a decrease of 0.04 pg/nr' in 2055.
7.4.1.3 Projected Annual PM2.5 Design Value Impacts in 2055
This section summarizes the impacts of the final rule on projected annual PM2.5 design value
concentrations in 2055, based on our CMAQ modeling. Figure 7-8 presents the changes in
annual PM2.5 design value concentrations in 2055. Not all counties have monitor data that meet
the requirements to calculate a design value concentration; counties without a calculated design
value are left white on the map.
7-22
-------
Annual PM2.5 County-High Design Value (DV) Differences : 2055gh_ctl minus 2055gh_ref
Hi > -0.010 to <- 0.010 Counfces 39
> OOIO to <= 0 020 : Counties 1
t > -0.040 to <= -0.020 ; counties 199
» > 0 020 to <= 0,040 : Counties 0
> -0.020 to <= -0O10 ; Counties 50
Figure 7-8: Projected change in annual P1VI2.5 Design Values in 2055 due to the rule.
As shown in Figure 7-8, the majority of the design value decreases in 2055 are less than 0.1
ug/nv\ A total of 561 counties were modeled to estimate projected design value changes; the
mean impact of the rule on annual PM2.5 design values in these counties is a decrease of 0.04
ug/m\ The maximum projected decrease in an annual PM2.5 design value in 2055 is 0.14 jag/m3
in Los Angeles County, California.
These modeling results project that the rule will not have considerable impact on the number
of counties that are projected to be above and below the level of the 2012 annual primary PM2.5
NAAQS. Forty-four counties are projected to have concentrations above the level of the 2024
standard in the 2055 reference scenario. While the number of counties with projected design
values above the level of the NAAQS is less certain than the average projected changes in design
values, we estimate that the rule will reduce annual PM2.5 design values in two counties (Contra
Costa County, California and Butler County, Ohio) from above the level of the 2024 standard, to
below the standard. The projected population in these counties in 2055 is over 2 million people.
While the air quality modeling results suggest that annual PM2.5 design values will decrease as a
result of emissions changes from the rule in the vast majority of counties, there is one county in
2055 that is projected to have an increase in modeled annual PM2.5 design value concentration
(Washington County, Utah).
7.4.2 Ozone
7.4.2.1 Overall Projected Ozone Impacts
This section summarizes projected changes in ozone concentrations in 2055 from the rule
Figure 7-8 presents the absolute changes in 8-hour ozone maximum average concentrations over
the ozone season (April - September) in 2055 between the reference and the policy scenario and
indicates that there will be widespread decreases (blue areas). In a few isolated areas, this rule is
7-23
-------
expected to result in increases in 8-hour maximum average ozone, likely due mainly to increases
in EGU emissions. We expect the power sector to become cleaner over time as a result of the
IRA and future policies, which will reduce the air quality impacts of EGUs. Less than 0.1% of
CMAQ grid cells are projected to have increases in 8-hour maximum ozone concentration of
>0.1 ppb (yellow and red areas), and only 0.1% of the population of CONUS lives in those areas.
I
>0.40
0.30
0.20
0.10
>
0.00 -g.
Q.
-0.10
-0.20
-0.30
<0.40
Figure 7-8: Projected changes in 8-hour maximum average ozone concentrations in 2055
ozone season due to the rule.
The rule will decrease 8-hour maximum average ozone concentrations by an average of 0.09
ppb in 2055, with a maximum decrease of 0.71 ppb and a maximum increase of 0.36 ppb. The
population-weighted average change in 8-hour maximum average ozone concentrations will be a
decrease of 0.16 ppb in 2055.
7.4.2.2 Onroad-Only Projected Ozone Impacts
We also modeled an "onroad-only" sensitivity case. Figure 7-9 presents the absolute changes
in 8-hour maximum average ozone concentrations in 2055 between the reference and onroad-
only sensitivity scenarios.
7-24
-------
>0.40
0,30
0.20
0.10
>
0.00 £
Q.
-0.10
-0.20
-0.30
<-0.40
Figure 7-9: Projected changes in 8-hour maximum average ozone concentrations in 2055
ozone season from "onroad-only" emissions changes.
When only the onroad emissions impacts of the rule are considered, 8-hour maximum average
ozone concentrations will decrease by an average of 0.09 ppb in 2055, with a maximum decrease
of 0.70 ppb. The population-weighted average change in 8-hour maximum average ozone
concentrations attributable to the onroad emissions reductions will be a decrease of 0.16 ppb in
2055.
7.4.2.3 Projected Ozone Design Value Impacts in 2055
This section summarizes the impacts of the final rule on projected ozone design value
concentrations in 2055, based on our CMAQ modeling. Figure 7-10 presents the annual
maximum 8-hour ozone design value concentrations in 2055. Not all counties have monitor data
that meet the requirements to calculate a design value concentration; counties without a
calculated design value are left white on the map.
7-25
-------
MDA8 Ozone County (4th High) Design Value Difference: 2055gh_ctl minus 2055gh_ref
Legend I ppb )
I <* 0.800 : Counties 16
I > 0 800 to <* -0.600 : Counties 69
I > 0 600 to <- 0.400 : Counties 769
: > 0 400 to <- 0.200 Counties 143
> -0-200 to «¦» 0 000 Counties 41
> 0.000 to <- 0 200 ; Counties 0
I > 0.200 to <= 0 400 : Counties 0
I > 0.400 to <= 0 600 : Counties 0
I > 0 600 to <= 0 800 : Counties 0
I > 0.800 : Counties 0
Figure 7-10 Projected change in 8-hour ozone design values in 2055 due to the Rule.
As shown in Figure 7-10, the majority of the design value decreases in 2055 are less than 1
ppb. A total of 538 counties were modeled to estimate projected design value changes; the mean
impact of the rule on 8-hour ozone design values in these counties is a decrease of 0.4 ppb. The
maximum projected decrease in 8-hour ozone design value in 2055 is 1.2 ppb in Los Angeles,
Riverside, and San Bernardino Counties, California.
The number of counties with projected design values above the level of the NAAQS is less
certain than the average projected changes in design values. That said, these modeling results
project that the rule will not have an impact on the number of counties that are projected to be
above and below the level of the 2015 ozone NAAQS.
7.4.3 Nth
7.4.3.1 Overall Projected NOi Impacts
This section summarizes projected changes in NO2 concentrations in 2055 from the rule.
Figure 7-11 presents the absolute changes in annual average NO2 concentrations in 2055 and
indicates that there will be decreases in many urban areas (blue areas). In a few isolated areas,
this rule is expected to result in increases in annual average NO2, likely due to increases in EGU
emissions. We expect the power sector to become cleaner over time as a result of the IRA and
future policies, which will reduce the air quality impacts of EGUs. Only 0.1% of CMAQ grid
cells are projected to have increases in annual average N02 concentration of >0.025 ppb (yellow
and red areas), and only 0.1% of the population of CONUS lives in those areas.
Max 0.000 Mrn 1.200
7-26
-------
I .
I >0.050
B 0.037
i
~
i
0.025
0.013
>
0.000 £
-0.013
-0.025
J $-
-0.037
<-0.050
Max: 0-1104 Min:-0 M21
Figure 7-11: Projected changes in annual average NO2 concentrations in 2055 due to the
rule.
The rule will decrease annual average NO2 concentrations by an average of 0.01 ppb in 2055,
with a maximum decrease of 0.34 ppb and a maximum increase of 0.11 ppb. The population-
weighted average change in annual average NO2 concentrations will be a decrease of 0.08 ppb in
7.4.3.2 Onroad-Only Projected NO2 Impacts
We also modeled an "onroad-only" sensitivity case. Figure 7-12 presents the absolute changes
in annual average NO2 concentrations in 2055 between the reference and onroad-only sensitivity
scenarios.
2055.
7-27
-------
i
>v
A
*
4ax: 0.0002 Min
: -0.2786^ \
Figure 7-12: Projected changes in annual average NCte concentrations in 2055 from
"onroad-only" emissions changes.
When only the onroad emissions impacts of the rule are considered, annual average NO2
concentrations will decrease by an average of 0.01 ppb in 2055, with a maximum decrease of
0.28 ppb. The population-weighted average change in annual average NO2 concentrations
attributable to the onroad emissions reductions will be a decrease of 0.07 ppb in 2055.
7.4.4 SO2
7.4.4.1 Overall Projected SO2 Impacts
This section summarizes projected changes in SO2 concentrations in 2055 from the rule.
Figure 7-13 Figure 7-13 presents the absolute changes in annual average SO2 concentrations in
2055. In some areas there will be decreases (blue areas), and in some areas there will be
increases, likely due to increases in EGU emissions. We expect the power sector to become
cleaner over time as a result of the IRA and future policies, which will reduce the air quality
impacts of EGUs. Only 0.2% of CMAQ grid cells are projected to have increases in annual
average SO2 concentration of >0.005 ppb (yellow and red areas), and less than 0.1% of the
population of CONUS lives in those areas.
7-28
-------
t
I
>0.020
0.015
0.010
0.005
0.000
-0.005
-0.010
-0.015
<-0.020
a.
cl
I \ ' /
Max: 0.318 Min: -0.2576 %
Figure 7-13: Projected changes in annual average SO2 concentrations in 2055 due to the
rule.
The rule will decrease annual average SO2 concentrations by an average of 0.001 ppb in 2055,
with a maximum decrease of 0.26 ppb and a maximum increase of 0.32 ppb. The population-
weighted average change in annual average SO2 concentrations will be a decrease of 0.003 ppb
in 2055.
7.4.4.2 Onroad-Only Projected SO2 Impacts
We also modeled an "onroad-only" sensitivity case. Figure 7-14 presents the absolute changes
in annual average SO2 concentrations in 2055 between the reference and onroad-only sensitivity
scenarios.
7-29
-------
Figure 7-14: Projected changes in annual average SO2 concentrations in 2055 from
"onroad-onlv" emissions changes.
When only the onroad emissions impacts of the rule are considered, annual average SO2
concentrations will decrease by an average of 0.0002 ppb in 2055, with a maximum decrease of
0.01 ppb. The population-weighted average change in annual average SO2 concentrations
attributable to the onroad emissions reductions will be a decrease of 0.001 ppb in 2055.
7.4.5 Carbon Monoxide
7.4.5.1 Overall Projected CO Impacts
This section summarizes projected changes in CO concentrations in 2055 from the rule.
Figure 7-15 presents the absolute changes in annual average CO concentrations in 2055 between
the reference and the policy scenario and indicates that there will be decreases in the vast
majority of the country (blue areas). No CMAQ grid cells are projected to have increases in
annual average CO concentration of >0.4 ppb (yellow and red areas).
7-30
-------
Figure 7-15: Projected changes in annual average CO concentrations in 2055 due to the
rule.
The rule will decrease annual average CO concentrations by an average of 0.85 ppb in 2055,
with a maximum decrease of 16.37 ppb and a maximum increase of 0.07 ppb. The population-
weighted average change in annual average CO concentrations will be a decrease of 3.74 ppb in
2055.
7.4.5.2 Onroad-only projected CO impacts of rulemaking
We also modeled an "onroad-only" sensitivity case. Figure 7-16 presents the absolute changes
in annual average CO concentrations in 2055 between the reference and onroad-only sensitivity
scenarios.
7-31
-------
diff CO 2016 Annual avg 2055gh_onronly - 2055gh_ref
L '¦ '
Max: 0.0067 Min: -16:3703 \
I
>1.60
1.20
0.80
0.40
>
0.00
Q.
-0.40
-0.80
-1.20
<1.60
Figure 7-16: Projected changes in annual average CO concentrations in 2055 from
"onroad-only" emissions changes.
When only the onroad emissions impacts of the rule are considered, annual average CO
concentrations will decrease by an average of 0.85 ppb in 2055, with a maximum decrease of
16.37 ppb. The population-weighted average change in annual average CO concentrations
attributable to the onroad emissions reductions will be a decrease of 3.75 ppb in 2055.
7.4.6 Air Toxics
7.4.6.1 Overall Projected Air Toxics Impacts
This section summarizes projected changes in concentrations of select air toxics in 2055 from
the rule. Our modeling indicates that the rule will have relatively little impact on national
average ambient concentrations of the modeled air toxics in 2055. Figure 7-17 to Figure 7-21
present the absolute changes in annual average acetaldehyde, benzene, 1,3-butadiene,
formaldehyde, and naphthalene concentrations in 2055 between the reference and the policy
scenario. National average annual concentrations in 2055 will decrease for all pollutants,
although in some localized areas, this rule is expected to result in increases in annual average
concentration of acetaldehyde, benzene, 1,3-butadiene or formaldehyde, likely due to increases
in EGU emissions. We expect the power sector to become cleaner over time as a result of the
IRA and future policies, which will reduce the air quality impacts of EGUs.
The projected impact of the standards on average air toxics concentrations in 2055 are
presented in Table 7-14.
7-32
-------
Table 7-14: Projected changes in annual average air toxics concentrations in 2055 due to
the rule.
Pollutant
Unit
Average change
Acclaldchvde
jig/m.l
-0.0021
Benzene
PPb
-0.0007
1.3-Buladiene
jig/m.l
-0.0001
Formaldehyde
PPb
-0.0023
Naphthalene
jig/m.l
-0.00004
7-33
-------
*
I 1
£4 * ; "P | * , jj
Jf h
ft-v '» !-4 I
Wax: 0.0094 Min: -0.02.59 N !
K
'¦&-, ¦¦ ^ „.
I
>6.00e-03
4.50e-03
3.00e-03
1.50e-03
0.00e+00
-1.50e-03
-3.00e-03
-4.50e-03
<-6.00e-03
%
<-50
H -50 to-25
Hi -25 to-10.0
^ -10.0 to -5.0
-5.0 to -2.5
-2.5 to-1.00
-1.00 to 1.00
1.00 to 2.5
2.5 to 5.0
¦¦ 5.0 to 10.0
^ 10.0 to 25
¦¦ 25 to 50
Figure 7-17: Projected a) absolute changes and b) percent changes in annual average
acetaldehyde concentrations in 2055 due to the rule.
Max: 0.518 Min: -3.7905
7-34
-------
0.0099 Min: -0.0396
'
>2.00e-03
1.50e-03
1.00e-03
5.00e-04
>
0.00e+00 -g
-5.00e-04
-1.00e-03
-1.50e-03
<-2.00e-03
%
¦¦ <-50
¦¦ -50 to-25
H -25 to-10.0
^ -10.0 to -5.0
-5.0 to-2.5
-2.5 to-1.00
-1.00 to 1.00
1.00 to 2.5
2.5 to 5.0
^ 5.0 to 10.0
Hi 10.0 to 25
^ 25 to 50
H > 50
Figure 7-18: Projected a) absolute changes and b) percent changes in annual average
benzene concentrations in 2055 due to the rule.
Max: 0.8727 Min: -10.6606
7-35
-------
|>2.00e-04
1.50e-04
1.00e-04
5.00e-05
0.00e+00
-5.00e-05
— l.OOe 04
I -1.50e-04
<-2.00e-04
< -50
-50 to -25
-25 to -10.0
-10.0 to -5.0
-5.0 to -2.5
-2.5 to-1.00
-1.00 to 1.00
1.00 to 2.5
2.5 to 5.0
5.0 to 10.0
10.0 to 25
25 to 50
> 50
/
Max: le-04 Min: -0.0102
Figure 7-19: Projected a) absolute changes and b) percent changes in annual average 1,3-
butadiene concentrations in 2055 due to the rule.
7-36
-------
I
W
f
x
\.
\ t
Max: 0.0852 Min: -0.1839
I
>8.00e-03
6.00e-03
4.00e-03
2.00e-03
0.00e+00
o.
-2.00e-03
-4.00e-03
-6.00e-03
<-8.00e-03
< -50
-50 to -25
-25 to-10.0
-10.0 to -5.0
-5.0 to -2.5
-2.5 to-1.00
-1.00 to 1.00
1.00 to 2.5
2.5 to 5.0
5.0 to 10.0
10.0 to 25
25 to 50
> 50
Figure 7-20: Projected a) absolute changes and b) percent changes in annual average
formaldehyde concentrations in 2055 due to the rule.
7-37
-------
Max: 0.0 Mm: -0.0013
I
>2.00e-04
1.50e-04
1.00e-04
5.00e-05
0.00e+00
-5.00e-05
-1.00e-04
-1.50e-04
<-2.00e-04
%
< -50
a -50 to-25
¦¦ -25 to-10.0
¦ -10.0 to-5.0
-5.0 to -2.5
-2.5 to-1.00
-1.00 to 1.00
1.00 to 2.5
2.5 to 5.0
^ 5.0 to 10.0
Hi 10.0 to 25
^ 25 to 50
¦¦ >50
Figure 7-21: Projected a) absolute changes and b) percent changes in annual average
naphthalene concentrations in 2055 due to the rule.
Max; 0.1315 Min: -5.52
7-38
-------
7.4.6.2 Onroad-Only Projected Air Toxics Impacts
We also modeled an "onroad-only" sensitivity case. Figure 7-22 through Figure 7-26 present
the absolute changes in annual average air toxic concentrations in 2055 between the reference
and onroad-only sensitivity scenarios.
Summary statistics for the projected "onroad-only" impact of the standards on air toxics
concentrations in 2055 are presented in Table 7-15.
Table 7-15: Projected changes in annual average air toxics concentrations in 2055 due to
onroad-only emissions changes.
Pollutant
Unit
Average
change
Acclaldchvde
jig/m.l
-0.0019
Benzene
PPb
-0.0006
1.3-Biiladicne
jig/m.l
-0.0001
Formaldehyde
PPb
-0.0022
Naphthalene
jig/m.l
-0.00004
7-39
-------
I
>6.00e-03
4.50e-03
3.00e-03
1.50e-03
0.00e+00
-1.50e-03
-3.00e-03
-4.50e-03
<-6.00e-03
%
< -50
-50 to -25
-25 to-10.0
-10.0 to -5.0
-5.0 to -2.5
-2.5 to-1.00
-1.00 to 1.00
1.00 to 2.5
2.5 to 5.0
5.0 to 10.0
10.0 to 25
25 to 50
> 50
Figure 7-22: Projected a) absolute changes and b) percent changes in annual average
acetaldehyde concentrations in 2055 from "onroad-onlv" emissions changes.
7-40
-------
j?.
/ V
• >.,2.00e-03
1.50e-03
1.00e-03
5.00e-04
>
0.00e+00 -g
a
-5.00e-04
-1.00e-03
-1.50e-03
<-2.00e-03
< -50
-50 to -25
-25 to -10.0
-10.0 to -5.0
-5.0 to-2.5
-2.5 to-1.00
-1.00 to 1.00
1.00 to 2.5
2.5 to 5.0
5.0 to 10.0
10.0 to 25
25 to 50
> 50
Figure 7-23: Projected a) absolute changes and b) percent changes in annual average
benzene concentrations in 2055 from ffonroad-onlyff emissions changes.
7-41
-------
I
• f
¦s.
AT
vj,
I
V.
r
4*
*
Max: 0.0 Min:-0.0036"
I
>2.00e-04
1.50e-04
1.00e-04
5.00e-05
0.00e+00
-5.00e-05
-1.00e-04
-1.50e-04
<-2.00e-04
< -50
-50 to -25
-25 to -10.0
-10.0 to -5.0
-5.0 to -2.5
-2.5 to-1.00
-1.00 to 1.00
1.00 to 2.5
2.5 to 5.0
5.0 to 10.0
10.0 to 25
25 to 50
> 50
Max: 1.6571 Min: -29.4655
Figure 7-24: Projected a) absolute changes and b) percent changes in annual average 1,3-
butadiene concentrations in 2055 from "onroad-only" emissions changes.
7-42
-------
a)
Max:-0.0 Min:-0.0167*,
{ '
\ \
I
>8.00e-03
6.00e-03
4.00e-03
2.00e-03
0.00e+00 |
Q.
-2.00e-03
-4.00e-03
-6.00e-03
<-8.00e-03
%
¦¦ <-50
¦¦ -50 to-25
H -25 to-10.0
¦¦ -10.0 to-5.0
-5.0 to -2.5
-2.5 to -1.00
-1.00 to 1.00
1.00 to 2.5
2.5 to 5.0
¦¦i 5.0 to 10.0
¦¦ 10.0 to 25
¦¦ 25 to 50
> 50
Figure 7-25: Projected a) absolute changes and b) percent changes in annual average
formaldehyde concentrations in 2055 from ,fonroad-onlyff emissions changes.
Wax; -0.0082 Men: -1J3
7-43
-------
Max: 0.0 Min: -0.0013 J
I
>2.00e-04
1.50e-04
1.00e-04
5.00e-05
0.00e+00
-5.00e-05
-1.00e-04
-1.50e-04
<-2.00e-04
%
H <-50
^ -50 to-25
^ -25 to-10.0
¦ -10.0 to-5.0
-5.0 to -2.5
-2.5 to-1.00
-1.00 to 1.00
1.00 to 2.5
2.5 to 5.0
5.0 to 10.0
¦¦ 10.0 to 25
M 25 to 50
H >50
Max: 0.1014 Min: -4.147
Figure 7-26: Projected a) absolute changes and b) percent changes in annual average
naphthalene concentrations in 2055 from "onroad-only" emissions changes.
7-44
-------
7.4.7 Deposition
7.4.7.1 Overall Projected Nitrogen and Sulfur Deposition Impacts
This section summarizes projected changes in nitrogen (N) and sulfur (S) deposition in 2055
from the rule. Figure 7-27 and Figure 7-28 present the absolute changes in annual N and S
deposition in 2055, respectively. Less than 0.1% of CMAQ grid cells are projected to have
increases in annual average N deposition of >0.01 ppb (yellow and red areas), and less than 0.1%
of the population of CON US lives in those areas. Only 1.0% of CMAQ grid cells are projected to
have increases in annual average S deposition of >0.01 ppb (yellow and red areas), and only
0.1% of the population of COMJS lives in those areas.
I
>0.100
0.075
0.050
0.025
0.000 ¦
-0.025
-0.050
-0.075
<-0.100
Figure 7-27: Projected changes in annual nitrogen deposition in 2055 due to the rule.
The rule will decrease annual average N deposition concentrations by an average of 0.04
kgN/ha in 2055, with a maximum decrease of 0.76 kgN/ha and a maximum increase of 0.14
kgN/ha. The population-weighted average change in annual average N deposition concentrations
will be a decrease of 0.17 kgN/ha in 2055.
: 0.1436 Min
: -0.76]
.12
7-45
-------
Figure 7-28: Projected changes in annual sulfur deposition in 2055 due to the rule.
The rule will decrease annual average S deposition concentrations by an average of 0.002
kgS/ha in 2055, with a maximum decrease of 0.29 kgS/ha and a maximum increase of 0.70
kgS/ha. The population-weighted average change in annual average S deposition concentrations
will be a decrease of 0.01 kgS/ha in 2055.
7.4.7.2 Onroad-Only Projected Nitrogen and Sulfur Deposition Impacts
We also modeled an "onroad-only" sensitivity case (i.e., without including any changes to
emissions from the upstream sources included in the policy scenario). Figure 7-29 presents the
absolute changes in annual N deposition in 2055 between the reference and onroad-only
sensitivity scenarios and Figure 7-30 presents the absolute changes in annual S deposition.
7-46
-------
-0.0002 Min
: -O-Tisll
Figure 7-29: Projected changes in annual nitrogen deposition in 2055 from "onroad-only"
emissions changes.
When only the onroad emissions impacts of the policy scenario are considered, annual
average N deposition concentrations will decrease by an average of 0.04 kgN/ha in 2055, with a
maximum decrease of 0.77 kgN/ha. The population-weighted average change in annual average
N deposition concentrations attributable to the onroad emissions reductions will be a decrease of
0.17 kgN/ha in 2055.
"A
b
Jt
I * is ' . Vl"
"J ~ 4
i
J
r
I'
kliii W
* £k>»- v-S-
p im ¦
~
W
"7 V
. *- r
«
* ;• * -
' -
# . ¦ r. —}. 1 - "T
\ '
Max: le-04 Min: -0.02(^7
.. ' \
* w p\
AX '1
I
>5.00e-03
3.75e-03
2.50e-03
1.25e-03
0.006+00
~>
-1.25e-03
-2.50e-03
-3.75e-03
<-5.00e-03
Figure 7-30: Projected changes in annual sulfur deposition in 2055 from "onroad-only"
emissions changes.
7-47
-------
When only the onroad emissions impacts of the policy scenario are considered, annual
average S deposition concentrations will decrease by an average of 0.0008 kgS/ha in 2055, with
a maximum decrease of 0.02 kgS/ha. The population-weighted average change in annual average
S deposition concentrations attributable to the onroad emissions reductions will be a decrease of
0.004 kgS/ha in 2055.
7.5 Ozone and Particulate Matter Health Benefits
As described in this Chapter, EPA conducted an air quality modeling analysis of a regulatory
scenario in 2055 involving light- and medium-duty vehicle emission reductions and
corresponding changes in "upstream" emission sources like EGU (electric generating unit)
emissions and refinery emissions. Year 2055 was selected as a year that best represents the fleet
turning over to nearly full implementation of the final standards. Decisions about the emissions
and other elements used in the air quality modeling were made early in the analytical process for
the final rulemaking. Accordingly, the air quality analysis does not fully represent the final
regulatory scenario; however, we consider the modeling results to be a fair reflection of the
impact the standards will have on PM2.5 and ozone air quality, as well as associated health
impacts, in the snapshot year of 2055. In contrast, the OMEGA-based emissions analysis (see
RIA Chapter 8) does represent the final form of the standards. As a result, we used the OMEGA-
based emissions analysis and benefit-per-ton (BPT) values to estimate the criteria pollutant
(PM2.5) health benefits of these standards. RIA Chapter 6.4 describes the benefit-per-ton
valuation methodology and RIA Chapter 9 presents the PM2.5-related health benefits.
The AQM analysis supports the conclusion that in 2055, the standards will result in
widespread decreases in ozone and PM2.5 that will lead to substantial improvements in public
health and welfare. Using the AQM results, we have quantified and monetized health impacts in
2055, representing a LMDV regulatory scenario described in RIA Chapter 7.2. The approach we
used to estimate health benefits is consistent with the approach described in the technical support
document (TSD) that was published for the 2023 PM NAAQS Reconsideration Proposal. (U.S.
EPA 2023)
Table 7-16 reports the PM2.5- and ozone-attributable effects we quantified and those we did
not quantify in this benefits analysis. The list of benefit categories not quantified is not
exhaustive. The table below omits welfare effects such as acidification and nutrient enrichment.
7-48
-------
Category
Premature
mortality
from
exposure
to PM2.5
Nonfatal
morbidity
from
exposure
to PM2.5
Table 7-16: Health effects of ambient ozone and PM2.5.
Effect Effect Effect More Information
Mortality
from
exposure
to ozone
Nonfatal
morbidity
from
exposure
to ozone
Adult premature mortality from long-term exposure (age >17
or >64)
Infant mortality (age <1)
Effect
Quantified
/
/
Effect
Monetized
/
/
Non-fatal heart attacks (>18) y
Hospital admissions - cardiovascular (all) y y
Hospital admissions - respiratory (<19 and >64) y y
Hospital admissions - Alzheimer's disease (>64) y •/
Hospital admissions - Parkinson's disease (>64) y y
Emergency department visits - cardiovascular (all) y ya
Emergency department visits - respiratory (all) y ya
Emergency hospital admissions (>65) y y
Non-fatal lung cancer (>29)2 y y
Out-of-hospital cardiac arrest (all) y y
Stroke incidence (50-79) y y
New onset asthma (<12) y y
Exacerbated asthma-albuterol inhaler use (asthmatics, 6-13) y y
Lost work days (18-64) y y
Minor restricted-activity days (18-64) y y
Other cardiovascular effects (e.g., other ages) — —
Other respiratory effects (e.g., pulmonary function, non-asthma ; — —
ER visits, non-bronchitis chronic diseases, other ages and
populations)
Other nervous system effects (e.g., autism, cognitive decline, — —
dementia)
Metabolic effects (e.g., diabetes) — —
Reproductive and developmental effects (e.g., low birth weight, ^ — —
pre-term births, etc.)
Cancer, mutagenicity, and genotoxicity effects — —
Premature respiratory mortality from short-term exposure (0- y y
99) j ;
Premature respiratory mortality from long-term exposure (age y y :
30-99) j
Hospital admissions—respiratory (ages 65-99) y y
Emergency department visits—respiratory (ages 0-99) y y
Asthma onset (0-17) y y
Asthma symptoms/exacerbation (asthmatics age 5-17) y y
Allergic rhinitis (hay fever) symptoms (ages 3-17) y y
Minor restricted-activity days (age 18-65) y y
School absence days (age 5-17) y y
Decreased outdoor worker productivity (age 18-65) — —
Metabolic effects (e.g., diabetes) — —
Other respiratory effects (e.g., premature aging of lungs) — —
Cardiovascular and nervous system effects — —
Reproductive and developmental effects — —
1 Valuation estimate excludes initial hospital and/or emergency department visits.
' Not quantified due to data availability limitations and/or because current evidence is only suggestive of causality.
PM ISA
PM ISA
PM ISA
PM ISA
PM ISA
PM ISA
PM ISA
PM ISA
PM ISA
PM ISA
PM ISA
PM ISA
PM ISA
PM ISA
PM ISA
PM ISA
PM ISA
PM ISAb
PM ISAb
PM ISAb
PM ISAb
PM ISAb
PM ISAb
Ozone ISA
Ozone ISA
Ozone ISA
Ozone ISA
Ozone ISA
Ozone ISA
Ozone ISA
Ozone ISA
Ozone ISA
Ozone ISAb
Ozone ISAb
Ozone ISAb
Ozone ISAb
Ozone ISAb
7-49
-------
Below we report the estimated number and economic value of reduced premature deaths and
illnesses in 2055 attributable to the modeled regulatory scenario along with the 95 percent
confidence interval. Table 7-17 reports the number of reduced deaths and illnesses associated
with reductions in PM2.5, along with their monetized economic value.
Table 7-18 reports the number of reduced ozone-related deaths and illnesses, along with their
monetized economic value. Table 7-19 reports total benefits associated with the regulatory
scenario in 2055, reflecting alternative combinations of the economic value of PM2.5- and ozone-
related premature deaths summed with the economic value of illnesses for each discount rate.
7-50
-------
Table 7-17: Quantified and monetized avoided PM2.5-related premature mortalities and
illnesses of the regulatory
scenario in 2055 (95% confidence interval)".
Avoided I'M
Point ICstiinute
Yiiluiition (Millions, 2II22S)
Outcomes
All-Cause Mortality
(Wu et al. 2020) (65-99)
1.000
< 2°-ob •
$14,000 ($1,300 to $37,000)
(900 to 1.100)
; 3% :
$14,000 ($1,300 to $36,000)
; 7% i
$12,000 ($1,100 to $32,000)
(Pope III et al. 2019) (18-99)
2.000
: 2% i
$27,000 ($2,500 to $74,000)
(1,400 to 2,500)
; 3% f
$27,000 ($2,400 to $72,000)
i 7
$24,000 ($2,200 to $65,000)
| (Woodruff 2008) (0-0)
1.6
$23 '
(-0.98 to 4.0)
($-13 to $92)
ER visits, respiratory
ER visits, All Cardiac Outcomes
310
$0.47
(-120 to 720)
($-0.18 to $1.1)
ER visits, respiratory
560
$0.64
(110 to 1,200)
($0.13 to $1.3)
Hospital Admissions
HA, Alzheimers Disease
530
$8.3
(390 to 660)
($6.2 to $10)
HA, Cardio-, Cerebro- and
150
$3.0
Peripheral Vascular Disease
(110 to 190)
($2.2 to $3.9)
HA, Parkinsons Disease
61
$1.0
(31 to 90)
($0.52 to $1.5)
HA, Respiratory-2 HA, All
91
$2.0
Respiratory
(31 to 150)
($0.42 to $3.4)
Respiratory Incidence
Incidence, Asthma
2,100
1 2% :
$120
(2,000 to 2,200)
($110 to $130)
f"3% V
$120
($110 to $130)
T 7% V
$76
($71 to $80)
Incidence, Hay Fever/Rhinitis
14,000
$11
(3,300 to 24,000)
($2.6 to $19)
Incidence, Lung Cancer
74
i 2%
$3.1
(22 to 120)
($0.94 to $5.2)
3% ;
$2.6
($0.78 to $4.3)
| 7% '1
$1.9
($0.59 to $3.2)
Incidence, Out of Hospital
15
; 2% ;
$0.70
Cardiac Arrest
(-6.1 to 34)
($-0.29 to $1.6)
; 3% ;
$0.70
($-0.29 to $1.6)
; 7% ;
$0.70
($-0.28 to $1.6)
; Asthma Symptoms, Albuterol use
400,000
$0.18
i (-200,000 to 980,000)
($-0,089 to $0.45)
Additional Morbidity
Acute Myocardial Infarction,
33
] "2% I
$2.2
Effects
Nonfatal
(19 to 46)
($1.3d to $3.0)
1' 3%
$2.1
($1.2 to $3.0)
; 7% :
$2.1
($1.2 to $2.9)
Incidence, Stroke
58
$2.6
(15 to 100)
($0.67 to $4.4)
Minor Restricted Activity Days
680,000
$65
(550,000 to 800,000)
($34 to $99)
Work Loss Days
110,000
$25
(96,000 to 130,000)
($21 to $29)
a Values rounded to two significant figures.
We discount the value of those avoided health outcomes that are expected to accrue over more than a single year. Note that for Asthma Incidence
and Out of Hospital Cardiac Arrests, we do not yet have discounted value streams using a 2 percent discount rate. We repeat the 3 percent values in
these instances.
7-51
-------
Table 7-18: Quantified and monetized avoided ozone-related premature mortalities and
illnesses of the regulatory scenario in 2055 (95% confidence interval)".
Avoided Ozone Outeomcs
Avoided Outeomcs
Valuation
(Millions, 2022S)
Avoided
Long-term
(Turner 2016)
550
2%b
$7,600
Premature
Exposure
(380 to 710)
($680 to $21,000)
Respiratory
10/
Mortalities
J 70 :
$7,400
($660 to $20,000)
| 7% ;
$6,600
($600 to $18,000)
Short-Term
(Katsouvanni 2009) and
25
$370
Exposure
(Zanobetti 2008). pooled
(10 to 39)
($30 to $1,100)
Morbidity
Long-term
Asthma Onset
3,700
2%
$210
Effects
Exposure
(3.200 to 4.200)
($180 to $250)
10/
J /o
$210
7%
($180 to $250)
I/O
$130
($110 to $150)
Allergic Rhinitis
22.000
$17
Symptoms
(12.000 to 32.000)
($9.1 to $25)
Short-Term
Hospital Admissions -
73
$3.5
Exposure
Respiratory
(-19 to 160)
($-0.91 to $7.7)
ER Visits - Respiratory
1.300
$1.5
(350 to 2.700)
($0.41 to $3.1)
Asthma Symptoms
690,000
$210
(-86.000 to 1.400.000)
($-25 to $430)
MRADs
350,000
$34
(140.000 to 560.000)
($12 to $63)
School Absences
250.000
$34
(-36.000 to 530.000)
($-4.8 to $72)
a Values rounded to two significant figures.
k We discount the value of those avoided health outcomes that are expected to accrue over n tl a single year. Note that for Asthma
Onset we do not yet have discounted value streams using a 2 percent discount rate. We repeat the J percent values in this instance.
7-52
-------
Table 7-19: Total PM2.5 and ozone benefits of the regulatory scenario in 2055 (95%
confidence interval, billions of 2022 dollars)a,b.
PIVI2.5 Ozone Total
Benefits using PM2.5-rclalcd mortality estimate from Pope III el al.. 2019 (Pope III el al. 2019) and o/.onc-relaled
mortality estimate from Turner el al.. 2016 (Turner 2016)
2% Discount Rate $28 $8.7 $36
($2.7 - $74) ($0.65 - $23) ($3.3 - $98)
3% Discount Rate $27 $8.5 $35
($2.6-$72) ($0.63-$23) ($3.2 - $95)
7% Discount Rate $24 $7.6 $32
($2.3-$65) ($0.50-$21) ($2.8 - $86)
Benefits using PM2.5-rclalcd mortality estimate from W11 el al.. 2020 (W11 el al. 2020) and a pooled o/.onc-relaled
mortality estimate from Katsouvanni cl al.. 2009 (Katsouvanni 2009) and Zanobctti cl al.. 2008 (Zanobctti 2008)
2% Discount Rate $ 14 $1.5 $ 16
($1.5-$38) ($-0.0010-$3.8) ($1.5-$41)
3% Discount Rate $14 $1.5 $15
($1.4-$36) ($-0.0010-$3.8) ($1.4-$40)
7% Discount Rate $12 $1.4 $14
($1.3-$33) ($-0.071-$3.7) ($1.2 - $37)
a Values rounded to two significant figures.
k The benefits associated with the standards presented here do not include the full complement of health and environmental benefits that, if
quantified and monetized, would increase the total monetized benefits.
In any complex analysis using estimated parameters and inputs from numerous models, there
are likely to be many sources of uncertainty. The health benefits TSD that was published for the
2023 PM NAAQS Reconsideration Proposal details our approach to characterizing uncertainty in
both quantitative and qualitative terms. That TSD describes the sources of uncertainty associated
with key input parameters including emissions inventories, air quality data from models (with
their associated parameters and inputs), population data, population estimates, health effect
estimates from epidemiology studies, economic data for monetizing benefits, and assumptions
regarding the future state of the country (i.e., regulations, technology, and human behavior).
Each of these inputs is uncertain and affects the size and distribution of the estimated benefits.
When the uncertainties from each stage of the analysis are compounded, even small uncertainties
can have large effects on the total quantified benefits.
7-53
-------
7.6 Demographic Analysis
7.6.1 Overview
When feasible, EPA conducts full-scale photochemical air quality modeling to demonstrate
how its national mobile source regulatory actions affect ambient concentrations of regional
pollutants throughout the United States. As described in this Chapter, we conducted air quality
modeling of future (2055) projections of PM2.5 and ozone concentrations in a "baseline" scenario
absent the rule and in a "control" scenario that assumes the rule is in place. These baseline and
control scenarios are also used as inputs to the health benefits analysis (see RIA Chapter 7.5).
The ozone and PM2.5 health benefits that are projected to result from the rule will be substantial.
This air quality modeling data can also be used to conduct an analysis of how human
exposure to future air quality varies with population characteristics relevant to potential
environmental justice concerns in scenarios with and without the rule in place. Although the
spatial resolution of the air quality modeling is not sufficient to capture very local heterogeneity
of human exposures, particularly the pollution concentration gradients near roads, the air quality
modeling data can be used to observe demographic trends at a national scale.
We conducted an analysis using the air quality modeling data to demonstrate how this rule
will affect different population groups with potential EJ concerns throughout the U.S. This rule
applies nationally and will be implemented consistently throughout the nation. Specifically,
because this final rule affects both onroad and upstream emissions, and because PM emission
precursors and ozone can undergo long-range transport, it is appropriate to conduct a national-
scale EJ assessment of the contiguous U.S.245 As depicted in the maps presented in RIA Chapter
7.4, these reductions will be geographically widespread. Taking these factors into consideration,
this demographic analysis evaluates both national average exposures and the distribution of
exposure outcomes that will result from the final rule.
7.6.2 Air Quality, Population and Demographic Data
We began with projected 2055 baseline and control scenarios of modeled PM2.5 and ozone
concentration data (described in RIA Chapter 7.4). Ambient air quality concentration data
(annual average |ig/m3 for PM2.5 and April-September daily maximum 8-hour average ppb for
ozone) was estimated at a standard grid resolution of 12km x 12km across the contiguous United
States (CONUS).
The analysis also used population projections based on proprietary economic forecasting
models developed by Woods and Poole in 2015 (Woods & Poole 2015). The Woods and Poole
database contains county-level projections of population by age, sex, and race out to 2060,
relative to a baseline using the 2010 Census data. Population projections for each county are
determined simultaneously with every other county in the U.S to consider patterns of economic
growth and migration. The sum of growth in county-level populations is constrained to equal a
previously determined national population growth, based on Bureau of Census estimates
245 As we note in Chapter 7.3.2, the CMAQ modeling output we use as an input to the demographic analysis projects
ambient concentrations of air pollution over a domain that is limited to the continental United States and portions of
Canada and Mexico. We therefore are unable to conduct the same demographic analysis for Alaska, Hawaii, Puerto
Rico, and the Pacific Islands.
7-54
-------
(Hollmann, F. et al. 2000). According to Woods and Poole, linking county-level growth
projections together and constraining to a national-level total growth avoids potential errors
introduced by forecasting each county independently. Total projected population in 2055, and
2055 population stratified by race/ethnicity, was extracted from the Environmental Benefits
Mapping and Analysis Program - Community Edition (BenMAP-CE) at the same 12km x 12km
grid resolution as the air quality data to allow for the estimation of human exposure to PM2.5 and
ozone in scenarios both without and with the rule in place.246'247
The population variables considered in this EJ assessment are described in Table 7-20. These
variables are relevant to potential EJ concerns and are consistent with those first presented in the
recently finalized "Regulatory Impact Analysis of the Standards of Performance for New,
Reconstructed, and Modified Sources and Emissions Guidelines for Existing Sources: Oil and
Natural Gas Sector Climate Review" (U.S. EPA 2023). The variables include race and ethnicity
(Hispanic, Non-Hispanic White, Non-Hispanic Black, Non-Hispanic Asian, Non-Hispanic
Native American), linguistic isolation (those not fluent in English), educational attainment (those
25 and older with and without high school education), poverty status (those below the federal
poverty line and those below 200% of the federal poverty line), Tribal lands, urban status
(metropolitan area or not), historically redlined areas (HOLC Grades A-C and HOLC Grade D),
life expectancy (those at or below the 25th percentile of life expectancy), health insurance status
(insured or uninsured), and employment status (employed or unemployed). We have also added
disability status (disabled or not disabled) to this analysis.
We note that for all variables except race/ethnicity, we applied recent measures of population
characteristics to the projected population in 2055 (see Table 7-20). The projected populations
for each of the population variables in 2055 were then extracted from BenMAP-CE at the same
12km x 12km grid resolution as the air quality data to allow for the estimation of human
exposure to PM2.5 and ozone in scenarios both without and with the rule in place. The use of this
data to project conditions in the future is inherently uncertain since measures of recent
population characteristics are not necessarily predictors of the status of future populations.
246 Information about the BenMAP-CE tool can be found here: https://www.epa.gov/benmap. Additional
information regarding the population projections used in this analysis can be found in Appendix J of the BenMAP-
CE User's Manual (https://www.epa.gov/benmap/benmap-ce-manual-and-appendices).
247 In 2055, we estimate that there are 446 million people projected to be living in the contiguous United States; 209
million are projected to be Non-Hispanic-White (NH-White) and 236 million are projected to be people of color. To
put these projections into perspective, 2010 populations for the contiguous United States were 201 million for NH-
White and 106 million for people of color.
7-55
-------
Table 7-20: Demographic Population Variables Included in
the EJ Analysis.
Population
Groups
Age
Range
Spatial Scale
Data Source
Race/Ethnicity
Hispanic, Non-Hispanic (NH)
White, NH Native American, NH
Asian, NH Black
0-99
Census Tract
Woods & Poole0
Poverty Status
Above Federal Poverty Line (FPL),
Below FPL, Above Twice FPL,
Below Twice FPL
0-99
Census Tract
American Community
Survey (2015-2019)d
Urban Status
Metropolitan (Metro) Area,
Non-Metro Area
0-99
County
USD A Rural Urban
Continuum Code6
Historical
HOLC Grades A-C, HOLC Grade
0-99
Census Tract
Home Owners'
Redlining51
D ("Redlined")
Loan Corporation
Linguistic
Isolation
Linguistically Isolated (English <
"well"), Fluent in English
0-99
Census Tract
American Community
Survey (2015-2019)d
Educational
Attainment
No High School Degree, High
School Degree or More
25-99
Census Tract
American Community
Survey (2015-2019)d
Disability Status
Not Disabled, Disabled
0-99
Census Tract
American Community
Survey (2015-2019)d
Employment
Status
Employed, Unemployed
0-99
County
U.S. Census Bureau, 2017
to 202 lg
Health
Uninsured, Insured
0-64
County
U.S. Census Bureau, 2015
Insurance Status
to2019h
Tribal Land
Tribal Land, Not Tribal Land
0-99
Census Tract
Bureau of Indian Affairs1
Life
Bottom 25%ile; Top 75%ile
0-99
Census Tract
CDC USALEEP, 2010 to
Expectancy13
2015>
a The variable "redlined areas" is used to assess exposure in communities with a legacy of discriminatory land use designations and siting
decisions (i.e., historically redlined areas).
k The life expectancy variable is one way to assess cumulative exposures and impacts. The variable differentiates between populations with
differing baseline health levels and measures the average life expectancy within a census tract. For average life expectancy, low values indicate a
higher overall burden or cumulative risk, while higher values indicate a lower overall burden or cumulative risk.
° Population data is projected out to the future year 2055 based on economic forecasting models developed by Woods and Poole, Inc. (Woods &
Poole, 2015). The Woods and Poole database contains county-level projections of population by race/ethnicity out to 2060, relative to a baseline
using 2010 Census data.
d The American Community Survey (ACS) data represent 5-year average estimates from 2015 to 2019.
6 The US Department of Agriculture's 2013 Rural Urban Continuum Codes (RUCC) classify metropolitan counties by the population size of their
metropolitan area, and nonmetropolitan counties by the degree of urbanization and adjacency to a metro area,
f
Graded census tracts developed by Noelke et al. (2022) from digitized Home Owners' Loan Corporation (HOLC) residential security maps
overlaid onto 2010 Census tracts. Each census tract is classified as being covered by "Mainly A," "Mainly B," "Mainly C," and "Mainly D"
grading, corresponding to coverage of different hazard ratings from original HOLC maps. Census tracts labeled "HOLC Grade D" are
categorized as redlined areas and census tracts that were mainly "HOLC Grades A-C" are categorized as not redlined.
Q
County-level unemployment rates are from the Bureau of Labor Statistics from 2017 to 2021.
^ County-level data from the Small Area Health Insurance Estimates (SAHIE) collected by the U.S. Census Bureau from 2015 to 2019.
1 Tribal lands are defined by the Bureau of Indian Affairs (bia.gov).
J Census tract-level life expectancy estimates for the period 2010-2015 from CDC's U.S. Small-area Life Expectancy Estimates Project
(USALEEP).
7-56
-------
7.6.3 National Population-Weighted Average Concentration Analysis
In Table 7-21 and Table 7-22, we present the national population-weighted average PM2.5 and
ozone concentrations in 2055, respectively, for each specific population group in scenarios
without (baseline) and with (control) the rule in place. We also present the reduction in PM2.5
and ozone (from baseline to control) for each population group along with the relative reduction
from baseline expressed as a percentage. To highlight the changes in each category, results are
color-coded by air quality (lighter colors represent lower average concentrations and darker
coloring represents higher average concentrations). We note that on average, all population
groups will benefit from reductions in exposure to ambient PM2.5 and ozone due to the final rule.
7-57
-------
Table 7-21: Population-weighted averages for the reference, control, absolute difference,
and relative difference (in percentage terms) for each population group for PM2.5
reductions in 2055 associated with the final rule.
7-58
-------
Table 7-22: Population-weighted averages for the reference, control, absolute difference,
and relative difference (in percentage terms) for each population group for ozone
reductions in 2055 associated with the final rule.
Ozone Concentrations (ppb)
Demographic
Population
Absolute
Percent
Variable
(million)
Baseline
Control
Difference
Difference
Total U.S. (CONUS)
446
38.93
38.76
0.164
0.42%
Hispanic
127
40.98
40.81
0.168
0.41%
NH White
209
37.99
37.83
0.157
0.41%
NH Asian
43
40.27
40.11
0.168
0.42%
NH Black
63
36.88
36.70
0.176
0.48%
NH Native American
3.4
40.94
40.81
0.125
0.31%
Fluent in English
424
38.82
38.66
0.164
0.42%
Linguistically Isolated
22
40.94
40.78
0.161
0.39%
With HS Education
268
38.74
38.58
0.163
0.42%
Without HS Education
46
39.63
39.47
0.163
0.41%
Disabled
56
38.42
38.26
0.160
0.42%
Not Disabled
390
39.00
38.83
0.164
0.42%
Below Poverty
70
38.92
38.76
0.162
0.41%
Above Poverty
376
38.93
38.76
0.164
0.42%
Below Twice Poverty
146
38.91
38.75
0.162
0.42%
Above Twice Poverty
299
38.93
38.77
0.165
0.42%
Tribal Lands
4.6
39.99
39.85
0.138
0.35%
Not Tribal Land
441
38.92
38.75
0.164
0.42%
Metro
391
39.30
39.13
0.170
0.43%
Non-Metro
55
36.28
36.16
0.121
0.33%
HOLC Grades: A-C
50
40.51
40.37
0.143
0.35%
HOLC Grade: D - Redlined Areas
18
39.74
39.61
0.131
0.33%
Top 75%ile Life Expectancy
329
39.35
39.18
0.162
0.41%
Bottom 25%ile Life Expectancy
88
37.35
37.19
0.168
0.45%
Has Health Insurance
317
39.06
38.89
0.164
0.42%
No Health Insurance
39
38.37
38.20
0.167
0.43%
Employed
209
38.94
38.77
0.163
0.42%
Unemployed
11
39.50
39.34
0.164
0.42%
7-59
-------
As shown in Table 7-21, the 2055 population-weighted baseline concentration of PM2.5 across
the total CONUS population is 6.89 |ig/m3. Certain population groups have population-weighted
average baseline concentrations that are higher than the national average (the shaded
concentrations in Table 7-21 and Table 7-22), indicating disproportionate exposures in the future
baseline. For example, both Hispanics and those who are linguistically isolated are among the
population groups that have higher population-weighted average baseline PM2.5 concentrations
compared to the national average, at 7.54 and 7.84 |ig/m3, respectively. Large urban areas that
were designated by HOLC as Grades A-C and Grade D are also among the population groups
that have the highest baseline PM2.5 concentrations compared to the national average, at 7.65 and
7.89 |ig/m3, respectively. In general, those who are most exposed to elevated concentrations of
PM2.5 in the 2055 baseline will also experience some of the greatest absolute and relative
reductions in PM2.5 exposure. However, most population groups are projected to experience an
absolute reduction in PM2.5 that is similar to the national average of 0.044 |ig/m3, and only minor
differences in relative terms.
As shown in Table 7-22, the national population-weighted baseline ozone concentration is
38.93 ppb. Hispanics, Non-Hispanic Native Americans, and those who are linguistically isolated
have higher baseline population-weighted averages, at 40.98, 40.94 and 40.94 ppb, respectively.
Ozone is a more regional pollutant that is formed in the atmosphere and can undergo long-range
transport, and the reduction in ozone from this final rule is relatively consistent across the
population groups. Most population groups are projected to experience an absolute difference in
the ozone concentration that is similar to the national average of 0.16 ppb, and only very minor
differences in relative terms. The Non-Hispanic Black population will experience the greatest
absolute and percent reduction in ozone concentration.
In summary, projected disparities in 2055 are not likely mitigated or exacerbated by the rule
for most of the population groups evaluated, due to the relatively similar pollution concentration
reductions across demographic groups, especially for ozone. However, for some population
groups, nationally-averaged exposure disparity is mitigated to a small degree in both absolute
and relative terms.
7-60
-------
7.6.4 National Distributional Analysis
While the national average results described above can provide some insight when comparing
within and across population groups, they do not provide information on the full distribution of
concentration impacts. This is because both population groups and ambient concentrations can
be unevenly distributed across the spectrum of exposures, meaning that average exposures may
mask important regional disparities. We therefore conducted a distributional analysis using
cumulative distribution plots of pollution exposure reductions for each EJ population variable for
this final rule. However, given the spatial scale of our air quality modeling data (12km grid cell
resolution), this distributional analysis does not reflect near-roadway impacts.
To evaluate how the distribution of exposure reductions varies within and across population
groups, we plot the full array of exposure reductions projected to be experienced by the entirety
of each population group for PM2.5 and ozone (Figure 7-31 and Figure 7-32, respectively). The
distribution plots present the running sum of each group's total population on the y-axes
expressed as a percentage (i.e., cumulative percent of population). By constructing the
cumulative percent metric, we are able to directly compare exposure reductions across
population groups with different population sizes. The x-axes show reductions in PM2.5 and
ozone concentrations from low to high concentration reductions. Similar to the national average
analysis described above, pollution concentrations are at a 12km grid cell resolution and
population demographics have been area-weighted at a 12km grid cell resolution for consistency.
In other words, plots compare the running sum of each population group against PM2.5 and ozone
concentration reductions such that populations whose trendlines are further right on the plot have
a higher proportion of their population experiencing larger reductions in pollution concentrations
as a result of the final rule.
7-61
-------
Employed
Unemployed
000 0.02 0 04 0.06 0.08 0.10 0.12 0 14 016 0 IS 0 20 0 22 0 24
Above US Education
Below HS Education
Insured
Uninsured
Hispanic
NH White
NH Asian
NH Black
NH Native Am
Tribal Land
Not Tribal Land
Non-Metro Areas
Metro Areas
HOLC Grades A-C
HOLC Grade D
Not Lingistically Isolated
Linguistically Isolated
0.14 0.16 0.18 0.20 0.22
Above Poverty
Below Povery
Above Twice Poverty
Below Twice Poverty
0 16 0 18 0 20 0 22
Life Expectancy: Top 75%
Life Expectancy Bottom 25%
Life Expectancy Not Available
0.06 0.08 0.10 0.12 0.14 0 16 0 18 0 20 0 22
2055 PM:? Concentration Reductions (jigm3)
0 02 0 04 0.06 0.08 0.10 0.12 0 14 0 16 0 18 0 20 0 22
2055 PM;. Concentration Reductions (fig/m*)
Figure 7-31: Distribution of PM2.5 concentration reductions (fig/m3) for each population
group in 2055 from this final rule.
7-62
-------
0.3 0.4
Tribal Land
Not Tribal Land
0.6 0.7
Non-Metro Areas
Metro Areas
HOLC Grades A-C
HOLC Grade D
Not Lingjstically Isolated
Linguistically Isolated
too
E
100
90
90
80
80
s
70
70
¦-C!
re
3
60
60
a«
o
50
50
0-
Vh
40
40
°
Above Poverty
30
Below Povery
30
20
Above Twice Poverty
20
10
Below Twice Poverty
10
0
0
-Insured
" Uninsured
-Not Disabled
- Disabled
0.2 0.3 0.4 0.5 0.6
2055 Ozone Concentration Reductions (ppb)
-Above HS Education
- Below HS Education
Employed
Unemployed
Life Expectancy: Top 75%
Life Expectancy: Bottom 25%
Lite Expectancy Not Available
2055 Ozone Concentration Reductions (ppb)
Figure 7-32: Distribution of ozone concentration reductions (ppb) for each population
group in 2055 from this final rule.
7-63
-------
For most of the population groups, the small differences in the distributional plots suggest that
the rule is not likely to exacerbate nor mitigate PM2.5 or ozone exposure concerns. However,
differences in the distribution of impacts between some groups do exist.
Nearly the entire distribution of PM2.5 reductions for Hispanic, NH Black, and NH Asian
populations are to the right of the distribution of reductions for NH White and NH Native
American populations, meaning those population groups experience larger reductions in PM2.5
pollution concentrations as a result of the final rule (Figure 7-31, Panel A). However, the
differences in reductions are comparatively small between population groups, and all
race/ethnicity groups are projected to benefit from the final rule. A similar, though less
pronounced, trend in race/ethnicity distributions can also be observed for reductions in ozone
exposures (Figure 7-32, Panel A).
Most notably, the distribution plots show that populations who live in urban centers
(metropolitan and HOLC-graded areas), and those who are linguistically isolated, are more likely
to experience larger reductions in PM2.5 concentrations than their comparison groups (Figure 7-
31, Panels B and D). Because these population groups also have higher average baseline
exposures (see Table 7-21), the likelihood that a greater percentage of these population groups
experience larger reductions may somewhat mitigate existing disparities in PM2.5 exposure.
The distribution plot of ozone reductions in metropolitan areas (Figure 7-32, Panel B) is also
to the right of the distribution of reductions in non-metropolitan areas, likely reflecting the
regional nature of ozone formation and the breadth of metropolitan area definitions.
We also note that tribal areas experience lower PM2.5 and ozone concentration reductions
compared to reductions experienced in non-Tribal areas (Figure 7-31, Panel C and Figure 7-32,
Panel C).
7.6.5 Uncertainty in the Demographic Analysis
The results of this demographic analysis are dependent on the available input data and its
associated uncertainty. As we note in both the air quality modeling and health benefits chapters,
uncertainties exist along the entire pathway from emissions to air quality to population
projections and exposure. This analysis is subject to these same sources of uncertainty.
A limitation of this analysis is the 12km x 12km horizontal grid spacing of the air quality
modeling domain. Such resolution is unable to capture the heterogeneity of human exposures to
pollutants within that area, especially pollutant concentration gradients that exist near roads. EPA
is considering how to better estimate the near-roadway air quality impacts of its regulatory
actions and how those impacts are distributed across populations.
Another key source of uncertainty is the accuracy of the projected concentrations of PM2.5 and
ozone in 2055. Assumptions that influence projections of future air quality include emissions in
the future and the meteorology used to model air quality (2016 conditions). For example, in a
few isolated areas, increased electricity generation is projected to increase ambient SO2, PM2.5,
ozone, or some air toxics. However, we expect those projected impacts will decrease over time
as the electric power sector becomes cleaner as a result of the IRA and future policies. We
therefore urge caution when interpreting with precision the magnitude and location of emissions
and pollution concentrations in the future. We also note that decisions about the emissions and
other elements used in the air quality modeling were made early in the analytical process for the
7-64
-------
final rulemaking. Accordingly, the air quality analysis does not fully represent the final
regulatory scenario; however, we consider the modeling results to be a fair reflection of the
impact the standards will have on air quality in 2055.
There is also inherent uncertainty in the Woods & Poole-based populations projected out to
2055. As mentioned above, the population projections are based on proprietary economic
forecasting models developed by Woods and Poole in 2015 and are relative to a baseline using
the 2010 Census data. Underlying the population projections are forecasted variables such as
income, employment, and population. Each of these forecasts require many assumptions:
economy-wide modeling to project income and employment, net migration rates based on
employment opportunities and taking into account fertility and mortality, and the estimation of
age/sex/race distributions at the county-level based on historical rates of mortality, fertility, and
migration. To the extent these patterns and assumptions have changed since the population
projections were estimated, and to the extent that these patterns and assumptions may change in
the future, we would expect the projections of future population would be different than those
used in this analysis. EPA continues to investigate how best to incorporate population projections
into our analyses.
Many of the population variables used in this analysis (see Table 7-20) are based on data from
the American Community Survey representing five-year average data collected between 2015-
2019, or other recent data sources, that are not projected into the future. The use of this data to
project conditions in the future is inherently uncertain since measures of recent population
characteristics are not necessarily predictors of the status of future populations. We intend to
continue to refine demographic analyses in future rulemakings.
7-65
-------
Chapter 7 References
Bohac, Stanislav V., and Scott Ludlum. 2023. "Characterization of a Lightly Loaded Underfloor
Catalyzed Gasoline Particulate Filter in a Turbocharged Light Duty Truck." Journal of
Engineering fro Gas Turbines and Power 021024-1 - 021024-7.
Byun, D.W., J.K.S. Ching, and US EPA. 1999. "Science Algorithms of EPA Models-3
Community Multiscale Air Quality (CMAQ) Modeling System."
Cook, Rich, Sharon Phillips, Madeleine Strum, Alison Eyth, and James Thurman. 2020.
"Contribution of mobile sources to secondary formation of carbonyl compounds." Journal of the
Air and Waste Management Association.
ERG. 2023. "EXTERNAL PEER REVIEW OF MOVES3.R1 REPORTS: UPDATES TO THE
MODELING OF ELECTRIC VEHICLES AND UPDATES TO REFUELING AND NH3
CRITERIA EMISSIONS, Final Peer Review Report."
https://cfpub. epa.gov/si/si_public_file_download. cfm?p_download_id=546439&Lab=OTAQ.
Henderson, B. et al. 2018. "Hemispheric-CMAQ Application and Evaluation for 2016." 2019
CMAS Conference .
Hollmann, F. et al. 2000. Methodology and assumptionsfor the population projections of the
United States: 199 to 2100 (Population Division Working Paper No. 38). Washington, DC: US
Bureau of the Census.
Katsouyanni, K, Samet, JM, Anderson, HR, Atkinson, R, Le Tertre, A, Medina, S, Samoli, E,
Touloumi, G, Burnett, RT, Krewski, D, Ramsay, T, Dominici, F, Peng, RD, Schwartz, J,
Zanobetti, A and Committee, HEIHR (2009). 2009. "Air pollution and health: a European and
North American approach (APHENA)." Res Rep Health Eff Inst 142: 5-90.
Mathur, R. et al. 2017. "Extending the Community Multiscale Air Quality (CMAQ) Modeling
System to Hemispheric Scales: Overview of Process Considerations and Initial Applications."
Atmospheric Chemistry Physics 12449-12474.
Mo, Tiffany. 2024. "MOVES Versions used in FRM AQM for the Mult-Pollutant Emissions
Standards for Model Years 2027 and Later Light-Duty and Medium-Duty Vehicles."
Memorandum to Docket EPA-HQ-OAR-2022-0829. February 10.
National Center for Atmospheric Research. 2022. Weather Research & Forecasting Model
(WRF), v4.42. Accessed 2023. https://www.mmm.ucar.edu/models/wrf.
Pope III et al., CA, Lefler, JS, Ezzati, M, Higbee, JD, Marshall, JD, Kim, S-Y, Bechle, M,
Gilliat, KS, Vernon, SE and Robinson, AL. 2019. "Mortality risk and fine particulate air
pollution in a large, representative cohort of US adults." Environmental Health Perspectives 127
(7): 077007.
Skamarock, W.C. et al. 2008. "A Description of the Advanced Research WRF Version 3."
Sui, Lang. 2023. "Memorandum to Docket EPA-HQ-OAR-2022-0985. "Heavy-Duty
Technology Resource Use Case Scenario Tool (HD TRUCS)"." April.
7-66
-------
Turner, MC, Jerrett, M, Pope, A, III, Krewski, D, Gapstur, SM, Diver, WR, Beckerman, BS,
Marshall, JD, Su, J, Crouse, DL and Burnett, RT (2016). 2016. "Long-term ozone exposure and
mortality in a large prospective study." American Journal of Respiratory and Critical Care
Medicine 193 (10): 1134-1142.
U.S. EIA. 2023. Annual Energy Outlook 2023. Accessed 2023.
https://www.eia.gov/outlooks/aeo/tables_ref.php.
U.S. EPA. 2019. " PM ISA."
—. 2022. 2019 AirToxScreen: Assessment Results. https://www.epa.gov/AirToxScreen/2019-
airtoxscreen-assessment-results.
—. 2023. 8-Hour Ozone (2008) Nonattainment Area Summary. November 30. Accessed
December 2023. https://www3.epa.gov/airquality/greenbook/hnsum.html.
—. 2023. 8-Hour Ozone (2015) Nonattainment Area Summary. November 30. Accessed
December 2023. https://www3.epa.gov/airquality/greenbook/jnsum.html.
U.S. EPA. 2007. "Control of Hazardous Air Pollutants from Mobile Sources Regulatory Impact
Analysis." EPA 420-R-07-002. https://nepis.epa.gov/Exe/ZyPdf.cgi?Dockey=P1004LNN.PDF.
—. 2023. "EPA Responses to MOVES3.R1 Peer Review Comments to Updates on the Modeling
of Electric Vehicles and Updates to Refueling and NH3 Criteria Emissions." EPA Science
Inventory. Accessed March 5, 2024.
https://cfpub. epa.gov/si/si_public_file_download. cfm?p_download_id=546440&Lab=OTAQ.
U.S. EPA. 2023. "Estimating PM2.5- and Ozone-Attributable Health Benefits. Technical Support
Document (TSD) for the 2023 PM NAAQS Reconsideration Proposal." EPA-HQ-OAR-2019-
0587,.
—. 2023. Green Book Data Download. November 30. Accessed December 2023.
https://www.epa.gov/green-book/green-book-data-download.
U.S. EPA. 2024b. "Memo to the Docket: Air Quality Modeling Analysis for the Light- and
Medium- Duty 2027 Multipollutant Final Rule." EPA-HQ-OAR-2022-0829.
U.S. EPA. 2019. "Meteorological Model Performance for Annual 2016 Simulation WRF v3.8."
—. 2024. National Ambient Air Quality Standards (NAAQS) for PM. Accessed February 2024.
https://www.epa.gov/pm-pollution/national-ambient-air-quality-standards-naaqs-pm.
U.S. EPA. 2023. Overview of EPA's MOtor Vehicle Emission Simulator (MOVES4). Ann
Arbor: Office of Transportation and Air Quality. US Environmental Protection Agency.
—. 2020. Ozone National Ambient Air Quality Standards (NAAQS). December 23. Accessed
January 20, 2023. https://www.epa.gov/ground-level-ozone-pollution/ozone-national-ambient-
air-quality-standards-naaqs.
—. 2023. PM-2.5 (1997) Nonattainment Area Summary. November 30. Accessed December
2023. https://www3.epa.gov/airquality/greenbook/qnsum.html.
7-67
-------
—. 2023. PM-2.5 (2006) Nonattainment Area Summary. November 30. Accessed December
2023. https://www3.epa.gov/airquality/greenbook/rnsum.html.
—. 2023. PM-2.5 (2012) Nonattainment Area Summary. November 30. Accessed December
2023. https://www3.epa.gov/airquality/greenbook/knsum.html.
—. 2023. Power Sector Modeling. Accessed 2023. https://www.epa.gov/power-sector-
modeling/post-ira-2022-reference-case.
U.S. EPA. 2023. "Regulatory Impact Analysis of the Standards of Performance for New,
Reconstructed, and Modified Sources and Emissions Guidelines for Existing Sources: Oil and
Natural Gas Sector Climate Review." Office of Air Quality Planning and Standards, Health and
Environmental Impacts Division, Research Triangle Park, NC. Accessed January 11, 2024.
https://www.epa.gov/system/files/documents/2023-12/eol2866_oil-and-gas-nsps-eg-climate-
review-2060-avl6-ria-20231130.pdf.
—. 2022. Report on the Environment, technical documentation for acid deposition,
https ://cfpub. epa.gov/roe/indicator. cfm?i=l.
—. 2023. Sulfur Dioxide (2010) Nonattainment Area Summary. November 30. Accessed
December 2023. https://www3 .epa.gov/airquality/greenbook/tnsum.html.
U.S. EPA. 2023. "Technical Support Document (TSD): Preparation of Emissions Inventories for
the 2016v3 North American Emissions Modeling Platform, EPA-454/B-23-002."
https://www.epa.gov/system/files/documents/2023-
03/2016v3 EmisMod TSD_January2023_ 1 .pdf.
U.S. EPA. 2022. "Technical Support Document EPA Air Toxics Screening Assessment. 2017
AirToxScreen TSD." https://www.epa.gov/system/files/documents/2022-
03/airtoxscreen_2017tsd.pdf.
U.S. EPA. 2023. "Technical Support Document: Preparation of Emissions Inventories for the
2016v3 North American Emissions Modeling Platform EPA-454/B-23-002."
https://www.epa.gov/air-emissions-modeling/2016v3-platform.
U.S. EPA. 2019. "Technical Support Document: Preparation of Emissions Inventories for the
Version 7.1 2016 hemispheric Emissions Modeling Platform." https://www.epa.gov/air-
emissions-modeling/2016-version-71 -technical-support-document.
—. 2015. "U.S. Environmental Protection Agency Peer Review Handbook." EPA-100-B15-001.
https://www.epa.gov/sites/default/files/2020-
08/documents/epa_peer_review_handbook_4th_edition.pdf.
UNC Institute for the Environment. 2023. CMAS: CMAQ. https://www.cmascenter.org/cmaq/.
US EPA. 2023. CMAQ: The Community Multiscale Air Quality Modeling System.
https://www.epa.gov/cmaq.
US EPA. 2018. "Modeling Guidance for Demonstrating Air Quality Goals for Ozone, PM2.5 and
Regional Haze." https://www.epa.gov/sites/default/files/2020-10/documents/o3-pm-rh-
modeling_guidance-2018. pdf.
7-68
-------
—. 2023. Primary National Ambient Air Quality Standard (NAAQS) for Sulfur Dioxide.
https://www.epa.gov/SO2-pollution/primary-national-ambient-air-quality-standard-naaqs-sulfur-
di oxide.
—. 2023. Regional Haze Program, https://www.epa.gov/visibility/regional-haze-program.
—. 2020. Regional Haze Storymap.
https://epa.maps.arcgis.com/apps/Cascade/index.html?appid=e4dbe2263elf49fb849aflc73a04e2
f2.
—. 2018. Report on the Environment: Regional Haze,
https ://cfpub. epa.gov/roe/indicator. cfm?i=21.
Versar, Inc. 2019. "The Sixth External Peer Review of the Community Multiscale Air Quality
(CMAQ) Modeling System." https://www.epa.gov/sites/production/files/2019-
08/documents/ sixth_cmaq_peer_review_comment_report_6.19.19.pdf.
Woodruff, T.J., J. Grillo, and K.C. Schoendorf. 2008. "Air pollution and postneonatal infant
mortality in the United States, 1999-2002." Environmental Health Perspectives 116(1): 110-
115.
Woods & Poole. 2015. "Complete Demographic Database." https://www.woodsandpoole.com/.
Wu et al., X, Braun, D, Schwartz, J, Kioumourtzoglou, M and Dominici, F. 2020. "Evaluating
the impact of long-term exposure to fine particulate matter on mortality among the elderly."
Science Advances 6 (29): eaba5692.
Zanobetti, A and Schwartz, J. 2008. "Mortality displacement in the association of ozone with
mortality: an analysis of 48 cities in the United States ." Am J Respir Crit Care Med 177 (2):
184-189.
7-69
-------
Chapter 8: OMEGA Physical Effects of the Final Standards and Alternatives
This chapter describes the methods and approaches used within the OMEGA model to
estimate physical effects of the final standards and alternatives. Physical effects refer to emission
inventories, fuel consumption, oil imports, vehicle miles traveled including effects associated
with the rebound effect, and safety effects. The cost and benefits of the final standards are tied
directly to these physical effects and are discussed in Chapter 9 of this RIA.
We have made several changes to the calculation of physical effects since the NPRM. Those
are:
1) We have updated to Annual Energy Outlook ("AEO") AEO 2023 from AEO 2021. This
impacts fleet mix and liquid fuel prices, the latter of which impacts fuel costs per mile
estimates and rebound VMT for any liquid fueled vehicles including PHEVs.
2) We no longer use AEO for electricity prices; instead we use EPA estimates based on IPM
modeling as discussed in Chapter 5 of this RIA. This impacts fuel costs per mile estimates
and rebound VMT for PHEVs and BEVs where applicable.
3) We have updated our safety effects to be consistent with the August 2023 CAFE proposed
rule.
4) We have corrected the electricity consumption estimates to include charging losses between
the charge point and the vehicle battery. These losses were inadvertently excluded from the
NPRM analysis.
5) We have updated our EGU inventory estimates via use of updated EGU inventory modeling
as discussed in Chapter 5 of this RIA.
6) We have updated our refinery inventory estimates via use of updated refinery emission
modeling as discussed in Chapter 7 of this RIA along with employing an updated and more
comprehensive methodology to estimating refinery emissions.
7) We have updated our estimate of the share of domestic liquid fuel demand reduction leading
to reduced domestic refining.
8) We have updated our estimate of the share of domestic liquid fuel demand reduction leading
to reduced oil imports and its impact on energy security.
8.1 The OMEGA "Context"
OMEGA makes use of projections of fleet size, market shares, fuel prices, vehicle miles
travelled (VMT), etc., from the Annual Energy Outlook ("AEO"). Any AEO can be used
provided the input files are made available to OMEGA. For this final analysis, EPA has used
AEO 2023. (U.S. EIA 2023) AEO 2023 was done assuming that the future fleet would comply
with the 2022 CAFE FRM (NHTSA 2022) and would include impacts associated with the
Inflation Reduction Act (IRA). However, the way that the IRA was reflected in AEO 2023 was
not as impactful as we believe it should have been. Hence, when running OMEGA, the first
scenario run, the context run, reflects EPA's 2021 LD GHG FRM which was similar in
stringency to the 2022 CAFE FRM, and does not reflect IRA tax credits. This context OMEGA
8-1
-------
run, or session, is then used as a reference session from which future fleet VMT and rebound
VMT can be calculated, as described below.
8.2 The Analysis Fleet and the Legacy Fleet
OMEGA uses as a "base year fleet," a comprehensive list of vehicles sold in a recent model
year. This base year fleet includes all models of vehicles, their sales, and a long list of attributes
such as their curb weights, their footprints, and the primary GHG technologies on those vehicles.
For this analysis, EPA is using the MY 2022 light-duty fleet as the base year fleet, updated from
the MY 2019 fleet used in the NPRM, and the same MY 2020 medium-duty base year fleet that
we used in the NPRM. When OMEGA runs, it begins with the 2023 calendar year as the first
year of the analysis and uses the fleet of vehicles contained in the base year fleet as the starting
point for the analysis. These MY 2023 and later vehicles are referred to as the "analysis fleet."
Vehicles that exist in the fleet prior to the first year of the analysis (i.e., MY 2023) are
referred to as the "legacy fleet." These vehicles include vehicles of all ages that exist in the fleet
as of calendar year 2022. Those vehicles are "aged out" of the fleet over the course of running
the analysis. The legacy fleet vehicles are not changed in any way within OMEGA other than
being scrapped (aging out) and driving fewer miles per year as they age.
Figure 8-1 shows ICE vehicle stock—liquid-fueled vehicles including HEVs—and Figure 8-2
shows BEV and PHEV stock. The ICE vehicle stock can be seen to be aging out of the fleet as
the BEV and PHEV stock grows. Figure 8-3 shows the total vehicle stock with growth
representing economic and population growth going forward. Figure 8-3 also highlights that
OMEGA scales fleet stock for consistency with AEO projections given that each alternative,
despite differences in ICE, HEV, PHEV, and BEV numbers have the same total stock.
8-2
-------
300,000,000
100,000,000
50,000,000
0
2027 2031 2035 2039 2043 2047 2051 2055
Figure 8-1: ICE vehicle stock in OMEGA effects calculations
300,000,000
250,000,000
200,000,000
2027 2031 2035 2039 2043 2047 2051 2055
8-3
-------
Figure 8-2: BEV & PHEV stock in OMEGA effects calculations.
350,000,000
300,000,000
250,000,000
200,000,000
150,000,000
100,000,000
50,000,000
0
2027 2031 2035 2039 2043 2047 2051 2055
Figure 8-3: Light- and medium-duty stock in OMEGA effects calculations.
Figure 8-4 through Figure 8-7 show the share of ICE (including HEV), PHEV, and BEV
vehicles in the total light- and medium-duty stock for the calendar years 2027 through 2055 in
the No Action, Final, Alternative A, and Alternative B scenarios, respectively.
• ••••• No Action
AltB
Final
— • -AltA
8-4
-------
Figure 8-4: Share of ICE (including HEV), PHEV and BEV in the total light- and medium-
duty stock in the No Action scenario.
Figure 8-5: Share of ICE (including HEV), PHEV and BEV in the total light- and medium-
duty stock under the Final standards.
8-5
-------
Figure 8-6: Share of ICE (including HEV), PHEV and BEV in the total light- and medium-
duty stock under Alternative A.
Figure 8-7: Share of ICE (including HEV), PHEV and BEV in the total light- and medium-
duty stock under Alternative B.
8.3 Estimating Vehicle, Fleet, and Rebound VMT
OMEGA uses a static set of mileage accumulation rates based on body style. OMEGA uses
three self-explanatory body styles: sedan wagon; cuv suv van; and pickup. All vehicles in both
8-6
-------
the analysis and legacy fleets are characterized as being of one of these body styles. The rates at
which each body style is aged-out of the fleet, or re-registered, and the miles driven by age are
shown in Table 8-1 for light-duty and medium-duty. The same values are used in both the
analysis and the legacy fleets based on vehicle age.
Table 8-1: Mileage accumulation and re-registration rates used for light-duty.
Mileage Accumulation ; Re-Registration Rate
Age :
Sedan Wagon
¦ CUV SUV Van
Pickup =
Sedan Wagon
CUV SUV Van ;
Pickup
0 :
15,922
16,234
: 18,964 :
100.0%
100.0%
100.0%
" 1 |
15,379
15,805
: 17,986 !
98.8%
98.8%
97.8%
2 ;
14,864
; 15,383
" 17,076 ;
97.7%
97.7%
96.3%
"3 1
14,378
14,966
f 16,231
96.1%
96.1%
94.3%
4 i
13,917
14,557
15.119
94.5%
94.5% ;
93.1%
5 i
13,481
14,153
f 14,726 :
93.0%
93.0%
91.5%
6
13,068
13,756
; 14,060 ;
91.1%
91.1% !
89.3%
7 1
12,677
13,366
; 13,448 1
89.1%
89.1%
87.0%
8
' 12,305
12,982
i 12,886 "i
86.9%
86.9%
84.1%
9 !
11,952
12,605
r 12,372 ]
84.0%
84.0% !
"79.6%
in I
11,615
12,234
11,903 {
80.0%
80.0%
74.2%
11 '
11,294
11,870
: 11,476 I
75.6%
75.6%
69.2%
12
10,986
11,512
r 11,088;
70.6%
70.6% j
64.1%
13
10,690
11,161
,' 10,737 ':
65.3%
65.3%
58.3%
14
10,405
10,816
; 10,418 ';
59.5%
59.5%
" 53.5%
15
10,129
10,477
10,131 ;
53.1%
53.1%
48.6%
16
9,860
10,146
f 9,871
45.8%
45.8%
44.2%
17 1
9,597
9,820
i 9,635 's
38.3%
38.3%
39.8%
18 "]
9,338
9,501
! 9,421
30.8%
30.8%
35.2%
19
9,081
9,189
! 9,226
24.1%
24.1%
30.9%
20 i
8,826
8,883
9,047 "
18.3%
18.3%
26.7%
'21 ''
8,570
8,583
; 8,882
13.9%
13.9%
22.8%
"22
8,313
8,290
i 8,726 " :
10.7%
10.7%
20.2%
23 !
8,051
8,004
! 8,577 !
8.2%
8.2%
17.5%
24
7,785
7,724
; 8,433
6.3%
6.3% 1
15.8%
25
7,511
7,450
; 8,290 ;
5.1%
5.1%
14.5%
26
7,229
7,183
8,146 "1
4.2%
4.2%
13.9%
27
6,938
6,923
; 7,998 '
3.4%
3.4% ;
12.5%
28
6,635
6,669
;' 7,842
2.8%
2.8%
11.1%
29 I
6,319
6,421
i 7,676 1
2.4%
2.4% !
10.3%
3d
5,988
6,180
f 7,497 1
0.0%
0.0%
9.3%
31 1
5,641
5,946
: 7,302 :
0.0%
0.0%
8.3%
32
5,277
5,718
r 7,089
0.0%
0.0% 1
7.3%
"33 "'i
4,893
5,496
:' 6,853 i
0.0%
0.0%
6.2%
34
4,488
5,281
i" 6,593""!
0.0%
0.0%
5.0%
35
4,061
5,072
I 6,305
0.0%
0.0%
CO
36 "1
3,610
4,870
! 5,987 ;
0.0%
0.0% i
2.7%
37 " |
3,133
4,674
^ 5,635 I
0.0%
0.0%
0.0%
8-7
-------
8.3.1 OMEGA "Context" VMT
When running OMEGA, the mileage accumulation rates and re-registration rates shown in
Table 8-1 are used for all vehicles in both the analysis and legacy fleets at the indicated ages. To
ensure that the "context" VMT (i.e., the total VMT of the analysis and legacy fleets) travels the
number of miles projected by EIA's Annual Energy Outlook, OMEGA adjusts the VMT of every
vehicle such that the total fleet VMT in any calendar year will equal that projected in AEO. This
is done by determining the ratio of the AEO projection for a given calendar year to that given
calendar year's total VMT within OMEGA estimated using the static mileage accumulation rates
shown in Table 8-1. That ratio is then applied to every vehicle's "static" VMT to arrive at a
"context" VMT. This way, the fleet context VMT within OMEGA will be equivalent to the fleet
VMT projected by AEO. Importantly, this context VMT does not yet include any rebound VMT,
which is discussed below.
FleetVMTAE0
VehicleVMTcontext = VehicleVMTstatic x
FleetVMT0MEGA
Where,
VehicleVMTconte* = miles driven in OMEGA scenario 0 (the NTR without IRA impacts (light-
duty) and the heavy-duty phase 2 GHG FRM plus impacts of the Advanced Clean Trucks
program (medium-duty))
VehicleVM'/'suuc = miles driven using values shown in Table 8-1
I'leelVMIstatic = the projected annual VMT in the AEO report being used
FleetVMTouEGA = the calculated annual VMT within OMEGA using VehicleVMTstatic
values
8.3.2 Context Fuel Costs Per Mile
The VMT rebound effect is discussed in detail in Chapter 4.2. Estimates of "rebound" miles
driven depends, traditionally, on changing fuel prices and their effect on the number of miles
people drive—as fuel prices rise and the cost per mile of driving increases, people drive less. In
OMEGA, we estimate the rebound effect not based on changing fuel prices, but rather on the
changing cost per mile of driving for vehicles of different fuel consumption. In other words,
someone that has purchased a new vehicle that consumes less fuel per mile might drive that
vehicle more than if they would have continued to drive their prior vehicle that consumed more
fuel per mile. As such, OMEGA's estimate of rebound VMT does not include any rebound VMT
in the legacy fleet since the fuel consumption characteristics of the legacy fleet are not changing.
For the analysis fleet, OMEGA first determines the fuel cost per mile for each base year fleet
vehicle in every calendar year included in the analysis. This way, the base year fleet vehicle's
fuel consumption characteristics are not changing through the years but its fuel costs per mile are
due to changing fuel prices. These fuel costs per mile are then sales weighted by context size
-------
class248 and by non-BEV vs. BEV powertrains. These sales-weighted fuel costs per mile for
every non-BEV or BEV context size class vehicle are then used as the context fleet, or context
vehicle, fuel costs per mile.
In subsequent OMEGA scenarios, which include unique GHG standards that can result in
unique fuel consumption characteristics for all vehicles, the fuel costs per mile for those
individual vehicles are determined and compared to their non-BEV or BEV context size class
fuel cost per mile in each year of its life. This way, the fuel cost per mile of each vehicle in
OMEGA can be compared to the context fuel cost per mile of similarly categorized vehicles to
determine how its miles would be estimated to change due to the policy.
8.3.3 Rebound VMT
As discussed in Chapter 4.2, rebound VMT depends on the elasticity of demand for more
driving. The input values used in the analysis were -0.1 for ICE vehicles and PHEVs and zero for
BEVs. We have used a value of zero for BEV vehicles as explained in Chapter 4 of this RIA.
The cost per mile of operation and the subsequent rebound effect miles can then be calculated as:
\Tnile )onroad $
C02 gallon
gallon
Where,
CPM= cost per mile
charge efficiency = 0.9 to capture losses between the charge point and the vehicle battery
CO2per gallon is for the applicable liquid fuel (8887 for gasoline, 10180 for diesel)
$ values are retail values for electricity and the applicable liquid fuel
Note that the kWh per mile and CO2 per mile values in the cost per mile equation are
weighted values that account for the share of operation on electricity versus liquid fuel. For
BEVs the liquid fuel portion of the equation would be zero while the electricity portion would be
zero for pure ICE vehicles and HEVs. PHEVs would make use of both the electricity and liquid
fuel portions of the equation.
( M-nnlicV CPCOT] tPXt /
VehicleVMTrebound = VehicleVMTC0ntext x Elasticity x ^ J-
LrM context
Where,
VehicIeVMTxcUmnA = the rebound miles driven
VehicleVMTcoa*sxt = the context VMT discussed above
CPM =
rkWh\
\mile ) onroad
x
$
charge efficiency kWh
+
248 OMEGA has 14 context size classes: Minicompact, Subcompact, Compact, Midsize, Large, Two Seater, Small
Crossover, Large Crossover, Small Pickup, Large Pickup, Small Van, Large Van, Small Utility, Large Utility.
8-9
-------
Elasticity = elasticity of demand
CPMpoiicy = the cost per mile in the policy scenario
CPMContext = the cost per mile in the context scenario for similarly categorized vehicles based
on context size class and non-BEV vs. BEV.
And to calculate vehicle miles traveled in the policy scenario:
VehicleVMTpoiicy VehicleVMTC0ntext + V6hicl6VMTrei)0unci
Where,
VehicleVMTpoiicy = the policy VMT
VehicleVMTCOntext= the context VMT discussed above
VehicIeVMTxcUmnA = the rebound miles driven
8.3.4 Summary of VMT in the Analysis
The analysis fleet VMT will vary depending on the rebound elasticities used and the level of
GHG standards, the latter of which impact the fuel consumption characteristics of the future
fleet. The OMEGA No Action VMT and the projected fleet VMT under the final standards and
alternative standards are shown in Table 8-2. Table 8-3 shows the rebound VMT.
8-10
-------
Table 8-2: VMT summary, light-duty and medium-duty (billion miles).
Calendar
; OMEGA
¦ OMEGA
OMEGA
OMEGA
Year
; No Action
Final
Alternative A
Alternative B
2027
3,151
; 3,151
3,151
3,151
2028
3,177
1 3,177
3,178
3,177
2029
3,197
; 3,198
3,199
3,198
2030
3,215
f 3,216
3,218
3,217
2031
r 3,229
! 3,231
" 3,233
3,231
2032
3,243
i 3,244
3,247
3,244
2033
3,260
; 3,264
3,267
3,263
2034
: 3,281
! 3,288
3,290
3,286
2035
3,303
; 3,311
3,314
3,310
2036
3,320
l 3,331
3,334
3,329
2037
: 3,340
s 3,353
3,355
3,351
2038
3,362
| 3,377
3,379
3,375
2039
3,384
: 3,402
3,404
3,399
2040
3,410
f 3,429
3,431
3,426
2041
3,435
! 3,457 '
3,459
3,454
2042
3,461
; 3,484
3,486
3,481
2043
3,487
3,512
3,513
3,509
2044
3,515
; 3,542
3,543
3,539
2045
3,543
1 3,571
3,573
3,568
' 2046
3,577
3,607
3,608
3,603
2047
3,612
; 3,643
3,643
3,639
2048
3,648
; 3,681
3,681
3,677
2049
3,684
i 3,718
3,718
3,714
2050
3,725
1' 3,759
3,759
3,755
2051
3,765
: 3,800
3,800
3,797
2052
; 3,806
; 3,842
3,842
3,838
2053
f' 3,848
! 3,884
3,884
3,880
2054
3,890
f 3,926
3,926
3,923
2055
3,932
1 3,969
3,969
3,965
8-11
-------
Table 8-3: Rebound VMT relative to no action, light-duty and medium-duty (billion miles).
Calendar
OMEGA
OMEGA
OMEGA
Year
Final
Alternative A
Alternative B
2027
0.0136
0.0288
0.0175
2028
0.374
0.933
0.357
2029
0.711
1.8
0.679
2030
1.04
2.69
1.15
2031
1.35
3.47
1.48
2032
1.56
4
1.62
2033
4.02
6.4
3.13
2034
6.56
8.72'
5.05
2035
8.79
10.9
7.09
2036
11.1
13.2
9.12
2037
13.3
15.4
11.1
2038
15.5
17.5
13.1
2039
17.6
19.5
15
2040
19.6
21.3
16.9
2041
21.5
23.1
18.7
2042
23.3
24.8
20.4
2043
24.9
26.3
22
2044
26.6
27.8
23.5
2045
28.1
29.2
24.9
' 2046
29.8
30.8
26.5
2047
31.2
31.9
27.7
2048
32.4
33
28.9
2049
33.5
33.9
29.9
2050
34.5
34.6
30.8
2051
35.1
35.2
31.5
2052
35.5
35.6
32
2053
36
35.9
32.5
2054
36.4
36.2
32.9
2055
36.6
36.7
33.1
i Values show 3 significant digits.
8.4 Estimating Safety Effects
OMEGA estimates safety effects consistent with methods used in past light-duty GHG
analyses and consistent with the methods developed by NHTSA for use in the CAFE
Compliance and Effects Modeling System (CCEMS). In fact, the inputs used in OMEGA are
identical to inputs used by NHTSA in CCEMS in support of their August 2023 CAFE NPRM
(NHTSA 2023). NHTSA is the government entity tasked with regulating vehicle safety and, as
such, NHTSA has the foremost experts in the field. EPA has worked closely with NTHSA
through the years of joint GHG/CAFE regulatory development and has weighed in extensively
on the statistical analyses used in estimating vehicle safety effects. That said, EPA has always
used modeling parameters in OMEGA that are identical to those used by NHTSA in the
CCEMS.
As noted, OMEGA uses vehicle travel fatality rates and safety values associated with mass
reduction that have been generated by NHTSA. These fatality rates and safety values and how
they are generated are described at length in the regulatory documents supporting NHTSA's
2023 CAFE NPRM. (NHTSA 2023) The discussion here does not attempt to provide that same
level of detail and is meant only to summarize the NHTSA analysis to help in understanding the
input values used in OMEGA.
The safety analysis is meant to capture effects associated with three factors:
• Changes in vehicle mass or weight;
8-12
-------
• Changes associated with fleet composition including car, CUV, SUV, pickup shares,
and fleet turnover; and,
• The potential for additional safety impacts associated with additional driving (i.e., the
"rebound effect" as mentioned in Chapter 8.3) that might arise from lower fuel costs
resulting from more stringent GHG standards.
In the following, we first cover the base fatality rates of vehicles in the legacy fleet. We then
cover the changes to those fatality rates associated with changes in vehicle mass and changes in
the analysis fleet composition. We then summarize the calculation approach to estimating
fatalities within OMEGA and present results.
8.4.1 Fatality Rates used in OMEGA
To estimate the impact of the standards on safety, NHTSA uses statistical models that
explicitly incorporate variation in the safety performance of individual vehicle model years.
They use a model for fatalities that tracks vehicles from when they are produced and sold, enter
the fleet, gradually age, and are ultimately retired from service. NHTSA also considers how
newer technologies are likely to affect the safety of both individual vehicles and the combined
fleet. The overall safety of the light-duty vehicle fleet during any future calendar year is
determined by the safety performance of the individual model year cohorts comprising it at the
ages they will have reached during that year, the representation of each model year cohort in that
(calendar) year's fleet, and a host of external factors that fluctuate over time, such as driver
demographics and behavior, economic conditions, traffic levels, and emergency response and
medical care. Combining forecasts of future crash rates for individual model year cohorts at
different ages with the composition of the vehicle fleet produces baseline forecasts of fatalities.
Regulatory alternatives that establish new standards for future model years can change these
forecasts by altering the representation of different model year cohorts making up the future
light-duty fleet. (U.S. NHTSA 2022) NHTSA's work produces estimates of fatality rates for each
model year making up the fleet during each future calendar year, and the process is continued
until calendar year 2050. Multiplying these rates by the estimated number of miles driven by
vehicles of each model year in use during a future calendar year produces baseline estimates of
total fatalities.
As an example, Figure 8-8 illustrates the recent history and baseline forecast of the overall
fatality rate for occupants of cars and light trucks. According to NHTSA, the sharp rise in the
fatality rate for 2020 coincided with the steep drop in car and light truck VMT during that year
due to the COVID-19 pandemic and accompanying restrictions on activity, combined with an
increased number of fatalities in 2020. These rates are also used as the basis for estimating future
fatalities and for estimating changes in safety resulting from reductions in the mass of new
vehicles, additional rebound-effect driving, and changes in the numbers of cars and light trucks
from different model years making up each calendar year's fleet. The underlying causes and
methods for estimating each of those three sources of changes in safety are discussed in detail in
various sub-sections of Chapter 7 of the Technical Support Document (TSD) accompanying
NHTSA's 2023 CAFE NPRM. (NHTSA 2023)
8-13
-------
16 »
14
1 1 £. \
s \
> \
J 10 \
Historical Data \ i
a \ /
8 /
I Forecast Values
&
Q.
.Si 6
**
**
A
Li- 4
2
A
1990 1995 2000 2005 2010 2015 2020 2025 2030 2035 2040 2045 2050
Figure 8-8 Recent and Projected Future Fatality Rates for Cars and Light Trucks (NHTSA
2023)
8.4.2 Calculating Safety Effects tied to Vehicle Weight Changes
To calculate the safety effects associated with changes to vehicle weight, OMEGA makes use
of fatality rate changes per billion miles of vehicle travel associated with vehicles of different
body styles—as with base fatality rates, these are developed by NHTSA—and weight changes
determined within OMEGA as vehicles change to meet future GHG standards. The first of these
factors are, as noted, developed by NHTSA through an analytical process that is detailed in their
2023 CAFE NRPM. (NHTSA 2023) OMEGA makes use of the input parameters used by
NHTSA in CCEMS model runs supporting their 2023 NPRM. Those values are shown in Table
8-4 and are used for both light-duty and medium-duty vehicles with medium-duty vans using the
cuv_suv entries. Note that these values differ slightly from those used in the NPRM analysis with
the changes resulting from the inclusion of all-wheel drive passenger cars in the regressions.
Table 8-4: Safety values used in OMEGA (U.S. NHTSA 2022).
Body style
NHTSA Safety Class
Threshold
(lbs)
Change per 100 lbs below threshold
Change per 100 lbs at or above threshold
sedan
PC
3201
0.0112
0.0089
pickup
LT/SUV
5014
0.0031
-0.0061
CUV SUV
CUV/Minivan
3872
-0.0025
-0.0025
For example, the base fatality rate for a pickup would change by -0.0061 for every 100
pounds of weight reduced over 5,014 pounds. However, if that vehicle had a starting weight of
5,064 pounds and its weight was reduced by 100 pounds, then the first 50 of those pounds would
reduce the base fatality rate by -0.00305 (-0.0061 per 100 pounds but for only 50 pounds) and
the next 50 pounds would increase the base fatality rate by 0.00155 (0.0031 per 100 pounds but
for only 50 pounds). In other words, reducing pickup weight above 5,014 pounds reduces
8-14
-------
fatalities while reducing pickup weight below 5,104 increases fatalities. In contrast, increasing
pickup weight above 5,014 pounds increases fatalities while increasing pickup weight below
5,014 pounds reduces fatalities.
Therefore, OMEGA first determines the weight change of the given vehicle. This is calculated
as the curb weight of the vehicle in the policy scenario (i.e., the final weight) relative to the curb
weight of the vehicle in the base year fleet.
DeltaWeight = FinalWeight — B as eYearWeight
Where,
DeltaWeight = the change in weight, where a weight reduction will be a negative value
FinalWeight = the weight of the vehicle in the policy scenario
BaseYearWeight = the weight of the vehicle in the base year fleet
Knowing the delta weight, OMEGA then determines the weight change above and below the
threshold for the body style of the given vehicle. Importantly, because OMEGA sometimes
increases the curb weight of vehicles (e.g., due to conversion to BEV), whether the weight
change is positive (increased weight) or negative (decreased weight) is important given the
safety values and their signs as shown in Table 8-4.
To determine the pounds changed below the threshold, and whether they involve increased or
decreased weight, OMEGA uses the logic shown below:
If: Threshold < Base Weight and Threshold < FinalWeight'.
Then: DeltaPoundshcUm = 0
Else if: BaseWeight < Threshold and FinalWeight < Threshold.
Then: DeltaPoundshcUm = I'inalWeight - BaseWeight
Else if: BaseWeight < Threshold< FinalWeight.
Then: DeltaPoundshcUm = Threshold - BaseWeight
Else if: FinalWeight < Threshold< BaseWeight.
Then: DeltaPoundshcUm = I'inalWeight - Threshold
To determine the pounds changed above the threshold, and whether they involve increased or
decreased weight, OMEGA uses the logic shown below:
If: BaseWeight < Threshold and FinalWeight < Threshold
Then: DeltaPoundsabo\e = 0
Else if: Threshold <= BaseWeight and Threshold <= FinalWeight
Then: DeltaPoundsabo\e = FinalWeight - BaseWeight
Else if: BaseWeight <= Threshold<= FinalWeight.
8-15
-------
Then: DeltaPoundsabove = FinalWeight - Threshold
Else if: FinalWeight <= Threshold <= BaseWeight:
Then: DeltaPoundsabo\e = Threshold - BaseWeight
The sales weighted curb weights and changes to those curb weights in the No Action and
under the Final standards and for select years are shown in Table 8-5.
Table 8-5: Light- and medium-duty fleet-weighted attributes in the OMEGA safety
analysis for the No Action and Final Standards (pounds)*.
Body
Calendar
No Action
Final Curb
Change in
Change in Curb
Change in Curb
Style
Year
Curb
Weight
Curb
Weight Below
Weight Above
Weight
Weight
Threshold
Threshold
sedan
2027
3,527
3,527
0
0
0
2032
3,523
3,554
31
3
28
2040
3,482
3,598
113
8
105
2050
3,443
3,651
204
9
195
2055
3,437
3,666
226
8
218
pickup
2027
5,193
5,194
1
0
1
2032
5,326
5,376
50
8
41
2040
5,374
5,473
100
30
70
2050
5,343
5,414
76
28
49
2055
5,370
5,423
57
20
37
cuvsuv
2027
4,328
4,328
0
0
0
2032
4,301
4,335
34
6
28
2040
4,236
4,326
88
10
78
2050
4,209
4,330
118
8
111
2055
4,218
4,340
120
6
114
* The threshold weights are shown in Table 8-4. Fleet weighting here reflects the entire light- and medium-duty
stock of vehicles, not just new sales.
With the weight change above and below the threshold, OMEGA calculates the fatality rate
changes as shown below:
RateChangebeiow = ChangePerlOOPoundsbeiow x (—DeltaPoundsbeiow)
RateChangeabove = Change Per 100 Poundsabove x (—DeltaPoundsabove)
Where,
RateChange = the change in fatality rate below/above the weight threshold for the given body
style as shown in Table 8-4; the base fatality rate that is changed by this rate change is discussed
in the next section.
ChangePerlOOPounds = the applicable value for the given body style as shown in Table 8-4
DeltaPounds = the applicable value according to the logic described above.
8-16
-------
8.4.3 Calculating Fatalities
OMEGA first calculates the fatality rate of a given vehicle in the given policy scenario. This
is done using the equation below.
Fatality Ratepolicy = Fatality Rate base x (1 + RateChangebelow) x (1 + RateChangeabove)
Where,
FatalityRatepo\icy = the fatality rate per billion miles traveled in the policy scenario
FatalityRatebase = the fatality rate per billion miles traveled in the base case (Chapter 8.4.1)
RateChange = the appliable result for the calculations described above (Chapter 8.4.2)
The number of fatalities in the given policy scenario are then calculated as:
FatalitiesvoiiCy = Fatality Rat6p0nCy x VrAfT'pOjjCy/10^
Where,
FatalitieSpoiicy = the number of fatalities in the policy scenario
FatalityRatepo\icy = the fatality rate in the policy, as described above
VMTPo\icy = the vehicle miles traveled in the policy, as described in Chapter 8.3
8.4.4 Summary of Safety Effects in the Analysis
Table 8-6 shows the number of fatalities estimated in the No Action case (i.e., the EPA 2021
FRM remains in place) and the Final standards and Alternatives. Table 8-7 shows fatality rate
impacts per billion miles of vehicle travel.
8-17
-------
Table 8-6: Fatalities per year, light-duty and medium-duty.
Calendar
; No Action
: Final
Alternative A
Alternative B
Final
Alternative A
Alternative ]
Year
% Change
% Change
% Change
2027
15,829
f 15,830 ;
15,835
15,830
0.01%
0.04%
0.01%
2028
15,597
r15,6011
15,611
15,600
0.02%
0.09%
0.02%
2029
15,410
! 15,417 :
15,432
15,416
0.04%
0.14%
0.04%
2030
15,275
: 15,286 1
15,304
15,284
0.07%
0.20%
0.06%
2031
15,172
15,188 "
15,209
15,186
0.11%
0.24%
0.09%
2032
15,107
: 15,127 ;
15,148
15,125
0.13%
0.27%
0.12%
2033
15,089
r 15,128 :
15,147
15,120
0.26%
0.39%
0.20%
2034
15,109
i 15,164 :
15,183
15,151
0.36%
0.49%
0.28%
2035 "
1 15,147
i 15,2151
15,232
15,200
0.45%
0.56%
0.35%
2036
15,182
f 15,260 ;
15,276
15,244
0.52%
0.62%
0.41%
2037
15,236
: 15,323 1
15,337
15,306
0.57%
0.67%
0.46%
2038
15,315
15,408 :
15,422
15,391
0.61%
0.70%
0.50%
2039
15,400
! 15,499 i
15,514
15,483
0.64%
0.74%
0.54%
2040
15,504
; 15,607
15,621
15,592
0.66%
0.76%
0.57%
2041
l 15,613
j 15,720 '•
15,734
15,706
0.68%
0.78%
0.59%
2042
15,724
' 15,834 :
15,849
15,822
0.70%
0.79%
0.62%
2043
15,838
f 15,951 |
15,966
15,941
0.71%
0.81%
0.64%
2044
15,962
16,079:
16,093
16,068
0.73%
0.82%
0.66%
2045
j 16,087
P 16,206 :
16,220
16,196
0.74%
0.82%
0.68%
' 2046
16,234
1 16,358
16,369
16,346
0.77%
0.83%
0.69%
2047
; 16,388
16,515 1
16,524
16,502
0.78%
0.83%
0.70%
2048
16,552
; 16,683 ]
16,690
16,669
0.79%
0.83%
' 0.71%
2049
16,716
f 16,851 i
16,854
16,835
0.80%
0.83%
0.71%
2050
16,897
; 17,033
17,036
17,018
0.81%
0.82%
0.72%
2051
17,078
i 17,216 1
17,218
17,200
0.81%
0.82%
0.72%
2052
17,257
! 17,397 "j
17,398
17,381
0.81%
0.82%
0.72%
2053
17,440
r 17,581 1
17,581
17,565
0.81%
0.81%
0.71%
2054
17,624
' 17,766 =
17,765
17,749
0.81%
0.80%
0.71%
2055
17,810
f 17,952 i
17,952
17,934
0.80%
0.80%
0.70%
8-18
-------
Table 8-7: Fatality rate impacts, light-duty and medium-duty (fatalities per billion miles).
Calendar
; No Action
: Final
Alternative A
Alternative B
Final
; Alternative A
Alternative ]
Year
% Change
% Change
% Change
2027
5.02
5.02
5.03
5.02
0.01%
0.04%
0.01%
2028
4.91
: 4.91
4.91
4.91
0.00%
! 0.09%
0.02%
2029
4.82
1 4.82
4.83
4.82
0.01%
0.14%
0.04%
2030
4.75
i ' 4.75
4.76
4.75
0.02%
0.19%
0.06%
2031
4.70
r 4.70
4.71
4.70
0.04%
0.24%
0.09%
2032
4.66
i 4.66
4.67
4.66
0.06%
0.27%
0.12%
2033
4.62
; 4.63
4.64
4.63
0.19%
0.41%
0.23%
2034
4.60
: 4.61
4.62
4.61
0.30%
0.53%
0.32%
2035
4.57
i 4.59
4.60
4.59
0.38%
0.61%
0.40%
2036
4.56
4.58
4.59
4.58
0.45%
0.68%
0.47%
2037
4.54
I 4.57
4.58
4.57
0.51%
0.73%
0.53%
2038
4.53
f 4.56
4.57
4.56
0.55%
0.78%
0.57%
2039
f 4.53
; 4.55 ""
4.56
4.55
0.59%
0.81%
0.62%
2040
4.52
| '4.55'"
4.56
4.55
0.61%
0.84%
0.65%
2041
"" 4.52
1 5 5 ""
4.56
4.55
0.63%
0.86%
0.68%
2042
4.51
! 1.5 1
4.55
4.54
" 0.65%
0.88%
0.70%
2043
r 4.5i
; 4.54
4.55
4.54
0.67%
0.89%
0.73%
2044
4.51
: 4.54
4.55
4.54
0.69%
0.90%
0.75%
2045
4.50
4.54
4.55
4.54
0.71%
0.91%
0.77%
' 2046
4.50
; 4.53
4.54
4.54
0.74%
0.93%
' 0.79%
2047
4.50
f 4.53""
4.54
4.53
0.76%
0.93%
0.79%
2048
4.50
r 4.53
4.54
' 4.53
0.78%
1 0.93%
0.80%
2049
4.50
r 4.53
4.54
4.53
0.79%
0.92%
0.81%
2050
4.49
j 4.53
4.54
4.53
0.80%
0.92%
0.81%
2051
4.49
: 4.53
4.53
4.53
0.81%
! 0.92%
0.81%
2052
; 4.49
r 4.53""
4.53
4.53
0.81%
0.91%
0.81%
2053
4.49
153
4.53
4.53
0.81%
i 0.90%
0.80%
2054
4.49
f 4.53"'
4.53
4.52
0.81%
0.89%
0.80%
2055
4.49
r 4.52
4.53
4.52
0.80%
0.89%
0.79%
8.5 Estimating Fuel Consumption in OMEGA
8.5.1 Drive Cycles for Onroad Fuel Consumption
To develop a best mix of regulatory cycles representing typical onroad vehicle operation,
EPA used two sources: the MOVES light-duty drive cycles and associated weights, and
aggregate vehicle behavior gleaned from California Real Emissions Assessment Logging
(REAL) data.
The MOVES model uses 18 representative cycles. For each cycle, the time, distance, and
energy expenditure at each speed was calculated, then binned in 0.5 mph increments.
Additionally, the average speed and positive kinetic energy ("PKE;" a measure of driver
aggressiveness) was calculated. The energy expenditure was calculated using the equivalent test
weight (ETW) and road load of a nominal vehicle. Nominal vehicle characteristics were
determined using the MOVES average passenger car and light truck parameters, weighted by
vehicle miles traveled (VMT). The statistics for all cycles were combined and weighted based on
the VMT associated with each cycle. The end result indicated an average speed of 36.6 mph and
a PKE of 3700 km/hr2
From the California REAL data, the average vehicle speed and positive kinetic energy was
determined across a range of vehicles. These data indicated an average speed higher than that
from the MOVES model data (41.1 mph), but a similar PKE (3900 km/hr2).
8-19
-------
To represent onroad behavior, EPA began with the energy expenditure distribution from the
MOVES data, as shown in Figure 8-9. The MOVES energy expenditure distribution is shown
compared to the energy expenditure distribution of the "city" (FTP) and highway (HW) cycles,
weighted 55%/45%. As can be seen, the 55/45 FTP/HW cycle has peak energy expenditure at a
noticeably lower Megajoule per mile (MJ/mile) value, leading to a substantially lower
cumulative energy expenditure. (The small peaks in the 55/45 FTP/HW cycle correspond to
accelerations of positive and negative 3.3 mph/sec, accelerations at which these cycles are
truncated.)
8-20
-------
Energy Distribution (10,000 miles)
0.025 MJ/mile bins
Cumulative Energy Use per 10,000 miles
0.025 MJ/mile bins
Figure 8-9: Energy distribution (top) and cumulative energy use (bottom) over 10,000 miles
for the MOVES onroad data, compared to FTP/HW regulatory cycles, weighted 55%/45%.
8-21
-------
To develop a better mix of cycles to represent onroad operation, EPA looked primarily at the
energy distribution, but also factored in the distribution of speeds and the PKE. At the end, a mix
of cycles was chosen that best matched these multiple optimization criteria.
EPA evaluated reweighting the bags of the FTP and incorporating portions of the US06
cycle.249 Reweighting the FTP did not improve the energy distribution match between vehicle
operation across cycles and representative onroad operation used to estimate energy use and fuel
consumption. However, incorporating the high acceleration and high-speed portions of the US06
did improve the energy distribution match with the MOVES data. Moreover, with the inclusion
of the US06 cycle, incorporating the HW cycle conferred no benefit, and this cycle was dropped.
After considering the effects of various cycle mixes, EPA selected a mix of cycles where the
weighting was 27% FTP, 6% US06 bag 1 (a high acceleration "city" bag), and 67% US06 bag 2
(a high speed "highway" bag). The energy expenditure distribution for this new cycle mix is
shown in Figure 8-10, again compared to the MOVES data. As can be seen, the energy
distribution of this cycle mix is much better aligned with the MOVES data, and the total positive
energy expended is nearly identical.
249 The FTP dynamometer cycle is divided into three (or, for hybrids, four) sequential sections, known as "bags."
The bags represent different operational characteristics, both for powertrain warmup and cycle speeds and
accelerations. Emissions from each bag are recorded separately. Bag results are weighted to represent on-road
driving behavior more closely, and the weighted results combined to produce the final FTP emissions values.
8-22
-------
Energy Distribution (10,000 miles)
0.025 MJ/mile bins
Cumulative Energy Use per 10,000 miles
0.025 MJ/mile bins
Figure 8-10: Energy distribution (top) and cumulative energy use (bottom) over 10,000
miles for the new cycle mix (27% FTP, 6% US06 bag 1, 67% US06 bag 2) compared to the
MOVES onroad data.
8-23
-------
In choosing this new cycle mix, EPA also considered the speed distribution of the mix and the
PKE. This mix of cycles had a PKE of 4300 km/hr2 (slightly higher than the MOVES or REAL
data) and an average speed of 40.6 mph. This average speed is higher than that of the MOVES
data, and closer to (but lower than) the REAL data. The speed distribution for this mix is shown
in Figure 8-11.
As can be inferred from Figure 8-11, the FTP and US06 cycles have substantial periods of
operation within a small speed window, giving the speed distribution the clear double-humped
shape. However, the overall speed profile remains similar that from the MOVES data.
8-24
-------
Speed Distribution
C
^ 20%
JZ
u
CO
QJ
-O 15%
QJ
>
CD
CJ
£
!£?
Q
<4—
o
c
o
*4—'
u
CO
10%
5%
0%
—•—MOVES data
—•—NewCycle Mix
20
40
5mph bins
60
80
Cumulative Speed Distribution
100%
T3
ro
U
£
VI
O
s—
O
c
o
4—'
U
TO
i_
M—
4-»
_ru
=3
E
u
80%
70%
60%
50%
40%
30%
20%
10%
0%
MOVES data
•New Cycle Mix
20
40
5mph bins
60
80
Figure 8-11: Speed distribution for the new cycle mix (27% FTP, 6% US06 bag 1, 67%
US06 bag 2) compared to the MOVES onroad data.
8-25
-------
To estimate fuel consumption impacts, OMEGA considers both the fuel(s) used by a given
vehicle and the share of miles driven by the vehicle on that fuel or fuels. For a fossil fuel-only
vehicle (including HEVs) or a BEV, the share of miles driven on the primary fuel would be 100
percent. For a PHEV, the share of miles driven on each fuel, the primary fuel (presumably liquid
fuel), and the secondary fuel (presumably electricity), are considered.
First, the vehicle miles traveled for the given vehicle on each fuel is calculated as below.
VMTvehicle.liquid fuei = VMTvehicle x Onroad EngineON fraction
VWTVeftjCJe;eJectrjCj£y — VMTyghlf-lg — V'MTVeJliCle.liqUi(l fuel
Where,
VMTvehicle = the VMT of the vehicle
VMTvehicie;eiectricity = the VMT of the vehicle on electricity
VMT\eHicie,iiq\ii& fuel = the VMT of the vehicle on liquid fuel
Onroad EngineONfraction is the fraction of operation with the internal combustion engine
running which is calculated for each vehicle in the OMEGA compliance calculations.
8.5.2 Electricity Consumption
To estimate BEV energy consumption, the VMT is multiplied by the rate of energy
consumption, or kWh/mile during onroad operation.
Electricity consumption is then calculated as:
rkWh\
n _ im/iT w ' vehicle-, onroad
Lonsxi7YiptioYiveflicie.eiectricity — vM1 vehicie-,electricity charge efficiency
Where,
Consumptionvehide; electricity = the electricity consumption of the given vehicle
VMTvehide; electricity = the vehicle miles traveled on electricity
(kWh/mile)vehicle-, onroad = the vehicle rate of energy consumption onroad
Charge efficiency = factor to capture losses between the charge point and the vehicle battery,
set via the onroad fuels.csv OMEGA input file
8.5.3 Liquid-Fuel Consumption
For liquid fuel consumption, OMEGA calculates the onroad fuel consumption rate making
use of the onroad C02/mile and the CO2 content of a gallon of the applicable liquid fuel, as
below.
8-26
-------
Gallons
(C°2)
\mile)
vehicle,onroad
mile
vehicle,onroad
(_C0^_\
\gallon)fuel
Where,
((iallons /w/tVveiiicieionroad = the fuel consumption rate of the given vehicle onroad
(^COyw/'/ejvehiciejonroad = the CCh/mile of the given vehicle on the road
(C02 gallon) tud = the CO2 emitted from combustion of a gallon of fuel (8,596 for gasoline,
10,049 for diesel, see below)
Liquid-fuel consumption is then calculated as below.
Consumption^hide; liquid fuel = the liquid fuel consumption of the given vehicle
VMTvShide; liquid fuel = the vehicle miles traveled on liquid fuel
(Gallons/mile)vshide-, onroad = the vehicle rate of liquid-fuel consumption onroad
We use values of 8,596 grams CO2 per gallon of gasoline and 10,049 grams CO2 per gallon of
diesel based on information provided by the Energy Information Administration. (EIA 2014)
That EIA source showed that burning a gallon of petroleum-only gasoline produces 19.64 grams
of CO2 while burning a gallon of E100 (pure ethanol) produces 12.72 pounds of CO2. Similarly,
burning a gallon of petroleum-only diesel produces 22.38 pounds of CO2 while burning a gallon
of B100 (100 percent biodiesel) produces 20.13 pounds of CO2. For retail gasoline we used a
share of 90 percent petroleum-only and 10 percent ethanol and used the same ratios for diesel
fuel to arrive at the values shown above using the equations shown here:
C onsumptionvehicie.iiquid fUei — VMTyghirfg-iiq-uid fuei X
vehicle-, onroad
Where,
gasoline
= (0.9 x 19.64 + 0.1 x 12.72) x 453.6
diesel
= (0.9 x 22.38 + 0.1 x 20.13) x 453.6
Where,
19.64 = pounds of CO2 per gallon of pure gasoline
12.72 = pounds of CChper gallon of E100
8-27
-------
22.38 = pounds of CO2 per gallon of pure diesel
20.13 = pounds of CChper gallon of B100
453.6 = the conversion from pounds to grams
8.5.4 Summary of Fuel and Electricity Consumption in the Analysis
In the tables presented in this summary of liquid fuel and electricity consumption impacts, the
percent changes reflect changes in light- and medium-duty liquid fuel and electricity
consumption relative to the No Action scenario and do not represent percent changes in total
U.S. consumption. Note that according to the Energy Information Administration (EIA), 2022
U.S. electricity consumption was roughly 4,050 TWh. (EIA 2023) This means that the 0.94 TWh
increase shown in 2027 is less than 0.1% of 2022 U.S. electricity consumption and the 360 TWh
increase shown in 2055 is less than 9% of 2022 U.S. electricity consumption.
Table 8-8: Fuel and electricity consumption impacts, final standards.
Calendar
Liquid Fuel
Electricity
Liquid Fuel
Electricity
Year
(billion gallons)
(TWh)
% Change
% Change
2027
-0.07
0.94
-0.049%
; 0.93%
2028
-0.48
4.1
-0.35%
3.2% ""
2029
-1.5
; 13
-1.1%
j 8%
2030
-3
27
-2.4%
15%
2031
-5
47
-4.1%
21%
2032
-7.2
67
-6.1%
1 27%
2033
-io
94
-9.1%
: 35%
2034
-14
120
-12%
43%
2035 i
-17
150
-15%
! 49%
2036 j
-20
180
-19%
55%
2037 '
-23
200
-22%
! 59%
2038
-25
220
-25%
i 63%
2039
-28
240
-27%
' 67%
2040
-30
260
-30%
70%
2041
-32
270
-32%
72%
2042 "
-34
1 290
-34%
74%
2043
-36
310
-36%
76%
2044
-37
320
-38%
78%
2045
-39
T 330
-39%
78%
2046
-40
340
-40%
; 79%
2047
-41
340
-41%
78%
2048
-42
350
-42%
; 78%
2049
-42
350
-42%
77%
2050
-43
350
' -42%
76%
2051
-43
360
-43%
75%
2052
-43
360
-43%
73%
2053
-43
360
-43%
72%
2054
-43
360
-43%
69%
2055
-43
360
-42%
68%
Sum
-780
6,700
Negative values represent decreases, positive values increases; One Terawatt hour (TWh) is equal to 1 billion kilowatt hours (kWh).
8-28
-------
Table 8-9: Fuel and electricity consumption impacts, Alternative A.
Calendar
Liquid Fuel
Electricity
Liquid Fuel
Electricity
Year
(billion gallons)
(TWh)
% Change
% Change
2027
-0.78
; 6.9
-0.54%
I 6.9%
"" 2028
-2.1
'! 18
-1.6%
14%
2029
-3.9
34
' -3%
22%
2030
-6
52
-4.7%
28%
2031
' ' -8.3
72
; -6.8%
33%
2032
-11
92
! -9%
: 37%
2033 ;
' -14
120
-12%
44%
2034
-17
140
-15%
j 51%
2035 i
-20
170
-18%
57%
2036
-23
200
I -21%
61%
2037
-25
220
-24%
65%
2038
-28
240
-27%
: 69%
2039
-30
260
-30%
72%
2040
-32
270
l" -32%
74%
2041
-34
290
-34%
75%
2042
-36
300
f -36%
77%
2043 F
-37
310
-38%
78%
2044
-39
320
-39%
79%
2045
-40
330
-40%
79%
2046
-41
340
-41%
80%
2047
-42
350
-42%
80%
2048
-42
350
-42%
79%
2049
-43
360
-43%
78%
2050
-43
360
' -43%
77%
2051
-44
] 360
-43%
76%
2052
-44
; 360
-43%
; 74%
2053
-44
360
| -43%
72%
2054
-44
360
-43%
70%
2055 V
-44
360
i -43%
68%
Sum
-830
7,000
Negative values represent decreases, positive values increases; One Terawatt hour (TWh) is equal to 1 billion kilowatt hours (kWh)
8-29
-------
Table 8-10: Fuel and electricity consumption impacts, Alternative B.
Calendar
Liquid Fuel
Electricity
Liquid Fuel
Electricity
Year
(billion gallons)
(TWh)
% Change
% Change
2027
-0.052
0.79
-0.036%
0.78%
"" 2028
-0.4
'! 3'.1
-0.29%
2.4%
2029
-1.4
11
: -i%
7%
2030
-2.8
23
-2.2%
12%
2031
' ' -4.5
39
-3.7%
18%
2032
-6.6
58
: -5.6%
23%
2033 ;
-9.4
84
i -8.3%
31%
2034
-12
no
-11%
38%
2035 i
-15
130
i -13%
42%
2036
-17
140
: -16%
46%
2037
-19
160
-18%
48%
2038
-21
180
-21%
51%
2039
-23
190
-23%
; 53%
2040
-25
200
i -25%
55%
2041
-27
210
i -27%
; 56%
2042
-28
220
-28%
57%
2043 F
-30
230
-30%
'i 59%
2044
-31
" 240
-31%
59%
2045
-32
250
! -33%
1 60%
2046
-33
i 260
-34%
60%
2047
-34
260
-34%
] 60%
2048
-35
270
-35%
60%
2049
-35
270
-35%
r ' 59%
2050
-36
270
; -36%
58%
2051
-36
270
-36%
57%
2052
-36
270
! -36%
55%
2053
-36
270
-36%
; 54%
2054
-36
270
-35%
52%
2055 V
-36
270
-35%
1 50%
Sum
-660
5,200
Negative values represent decreases, positive values increases; One Terawatt hour (TWh) is equal to 1 billion kilowatt hours (kWh)
8-30
-------
8.6 Estimating Emission Inventories in OMEGA
To estimate emission inventory effects due to a potential policy, OMEGA uses, as inputs, a
set of vehicle and electricity generating unit (EGU) emission inventories. In a circular process,
we first generate emission inventories using very detailed emissions models that estimate
inventories from vehicles (EPA's MOVES model), EGUs (EPA's Power Sector Modeling
Platform, v.6.21), and refineries (EPA's 2016v3 emissions modeling platform). The generation of
those inventories is described in Chapter 7 and Chapter 5. However, upstream inventories
(EGUs) made use of a set of bounding runs that looked at two possible futures—one with a low
level of fleet electrification and another with a higher level of electrification. These bounding
runs represented our best estimate of these two possible futures—the continuation of the 2021
FRM (lower) and our proposed Alternative 3 (upper)—at the time that those model runs were
conducted. With those bounded sets of inventories, and the associated fuel and electricity
demands within them (i.e., electricity demands for EGUs), we can calculate emission rates for
the two ends of these bounds. Using those rates, we can interpolate, using the given OMEGA
policy scenario's fuel demands, to generate a unique set of emission rates for that OMEGA
policy scenario. Using those unique rates, OMEGA then generates emission inventories for any
future OMEGA policy scenario depending on the liquid fuel and electricity demands of that
specific policy.
For vehicle emissions, EPA made use of two sets of MOVES emission inventory runs—one
assuming no future use of gasoline particulate filters and one assuming such use. Using the miles
traveled (for tailpipe, tire wear, and brake wear emissions) and liquid fuel consumed (for
evaporative and fuel spillage emissions), we can then generate sets of emission rates for use in
OMEGA. Using those rates, which are specific to fuel types and vehicle types (car vs. truck,
etc.), we can then generate unique emission inventories for the given OMEGA policy scenario.
This is important given the changing nature of the transportation fleet (BEV vs ICE, car vs CUV
vs pickup) and the way those change for any possible policy scenario and the many factors
within that impact the future fleet composition and the very different vehicle emission rates for
BEV vs ICE vehicles. This is especially true given the consumer choice elements within
OMEGA and the wide variety of input parameters that can have significant impacts on the
projected future fleet.
8.6.1 Calculating EGU Emission Rates in OMEGA
As described in Chapter 5 and presented in Chapter 5.2.3, EPA has generated EGU
inventories for the no-action case and the proposed Alternative 3. Those inventories are
presented in Tables 5-2 and 5-3 and are shown graphically in the accompanying charts. To
generate those inventories, EPA first ran OMEGA to estimate PEV energy demands into the
future. Those energy demands were used in the modeling of EGU inventories presented in
Chapter 5. EPA then uses the resultant inventories along with the associated "Generation" values
shown in Tables 5-2 and 5-3, appropriately, and the estimated PEV energy demands from
OMEGA used in generating the EGU inventory results, and linear interpolation between years to
generate a set of emission rates as a function of years from 2028. The resultant EGU emission
rates by scenario and pollutant are shown in Figure 8-12 and Figure 8-13.
8-31
-------
—No Action
—Final
—AltA
AltB
2027 2030 2040 2050
o 0.0025
2
« 0.0020
J? 0.0015
|
< 0.0010
£ 0.0005
0.0000
— No Action
—Final
-AltA
AltB
Figure 8-12: EGU GHG emission rates in the no action, final and alternative scenarios
(grams/kWh of US generation).
PM2.5
o 0.1000
'¦H
re
No Action
o 0.0800
Final
OJ
J? 0.0600
AltA
3
< 0.0400
E
2 0.0200
TO
AltB
2027
2030
2040
2050
voc
0.0080
c 0.0070
0
'% 0.0060
c 0.0050
O
J? 0.0040
1 0.0030
E 0.0020
ra
M 0.0010
0.0000
-No Action
-Final
-AltA
AltB
2027
2030
2040
2050
NOx
0.1200
0.0000
-No Action
-Final
-AltA
AltB
2027
2030
2040
2050
SOx
-No Action
-Final
-AltA
AltB
0.0000
2027
2030 2040
2050
Figure 8-13: EGU criteria pollutant emission rates in the no action, final and alternative
scenarios (grams/kWh of US generation).
Using these curves, OMEGA can calculate the estimated U.S. electricity generation in any
year of the analysis under any scenario and then calculate the EGU inventories unique to that
scenario.
To estimate the unique EGU emission rates for any given OMEGA scenario, OMEGA first
determines the PEV consumption estimate for a given year, which is driven by the level of the
standards and the expected PEV penetration rate, among other impacts (consumer acceptance,
critical materials, etc.). OMEGA then calculates a base U.S. generation as the IPM modeled low
demand generation less the light- and medium-duty fleet demand used in generating the IPM
inventories. To that result, OMEGA adds the estimated scenario demand to arrive at a new U.S.
generation for the policy in any given year. With that estimated scenario demand, OMEGA then
8-32
-------
interpolates the scenario emission rate using the IPM modeled demands and the rates calculated
using the IPM modeled inventories.
Rate
scenario
(iGeneration
US:scenario
— Generation
USdow demand
)
('GeYieVQ-tiOYlus-high demand
~l~ Rateiow demand
— Generation,
USdow demand
x (Ratehigh
demand demand)
Where, for a given pollutant in a given year of a given OMEGA scenario,
Rate scenario = the EGU emission rate in the scenario
RateXovi demand = the EGU emission rate calculated using the low demand inventory
Ratehigi demand = the EGU emission rate calculated using the high demand inventory
Generation^ low demand US electricity generation using the low demand IPM results
Generation^, high demand = US electricity generation using the high demand IPM results
Generation^; scenario = US estimated electricity generation in the scenario
8.6.2 Calculating Refinery Emission Rates in OMEGA
To estimate refinery emission inventories, OMEGA needs refinery emission rates, i.e., grams
of pollutant per gallon of fuel refined. These refinery emission rates can then be applied to fuel
consumption or, more specifically, changes in fuel consumption to estimate refinery emission
impacts associated with the liquid fuel consumption impacts expected from the light-duty and
medium-duty fleets.
The starting point for estimating the refinery emission rates was the refinery emission
inventory estimates generated in support of the air quality modeling (U.S. EPA 2024). Those
refinery inventories were generated for select calendar years and reflect estimates associated with
the air quality modeling no action, or reference, case. These are shown in Table 8-11.
Table 8-11: Emissions from Refineries that Refine Onroad Liquid Fuels (US tons per year).
Calendar Year
CO
C02
ch4
N20
NOx
PM2.5
SOx
voc
2030
50,463
179,019,970
9,608
1,529
75,350
17,738
22,955
57,274
2035
50,498
179,497,795
9,583
1,533
75,484
17,759
22,996
57,298
2040
50,829
180,908,447
9,621
1,545
76,169
17,883
23,134
57,416
2045
51,266
183,618,188
9,662
1,568
76,945
18,054
23,316
57,608
2050
51,794
186,521,729
9,743
1,593
77,830
18,253
23,501
57,829
Knowing the gallons of fuel refined in generating those emissions inventories, we could use
them to generate the desired emissions per gallon refined. But the number of gallons refined by
U.S. refiners, especially the gallons refined specifically for domestic consumption, is not readily
known. Refineries in the U.S. refine onroad liquid fuels not only for consumption on domestic
roads, but also for export and consumption elsewhere. Further, refineries in the U.S. refine more
products than onroad gasoline and diesel fuels. As a result, not all their emissions are the result
of refining onroad liquid fuel for domestic consumption.
8-33
-------
To account for the former effect, that of exports, we start with the level of domestic fuel use,
or consumption. For this, we can use AEO 2023 as presented in Table 8-12. This gives us the
gasoline and diesel fuel use in the U.S., but it does not give us the fuel refined by U.S. refiners.
We note that the U.S. is now a net exporter of petroleum products. AEO 2023 does not include
estimated exports of liquid fuels. Instead, it presents estimates of net product imports along with
projections of liquid fuel use in the U.S. We present those projections in Table 8-13 (note that
AEO 2023 projects net imports and shows them as negative values, i.e., exports; we show them
as positive net exports for clarity here).
Table 8-12: AEO 2023 Projections of Domestic Liquid Fuel Use (Million Barrels per Day,
(U.S. EIA 2023) see Table 11, Reference case).
Calendar Year Motor Gasoline Diesel Fuel Other Petroleum Products
2023 8.75 3.66 7.61
2030 8.12 3.31 8.15
2035 7.61 3.22 8.47
2040 7.30 3.19 8.79
2045 7.23 3.20 9.16
2050 7.43 3.20 9.53
Note the presence of a value for net exports in calendar year 2022. We present this value
because we use it to calculate a net export scaler that we then apply to subsequent projections of
domestic fuel use. Remember that the emission inventories we started with and the AEO
projections of fuel use do not reflect the projected light-duty and medium-duty fleets that we
project with each run of the OMEGA model and the different policies considered in each. Those
scaling factors are shown in Table 8-13 and are simple ratios of each calendar year projection to
the 2022 value for net exports.
Table 8-13: Net Exports and Export Scaler Used to Project Future Net Exports Associated
with any OMEGA Policy Scenario (Million Barrels per Day, (U.S. EIA 2023) see Table 11,
Reference case).
Calendar Year Net Exports Export Scaler
2022 3.99 1.00
2023 4.28 1.07
2030 6.08 1.53
2035 6.56 1.65
2040 6.80 1.70
2045 6.86 1.72
2050 6.43 1.61
However, our goal was to estimate the share of the net exports shown in Table 8-13 that are
gasoline versus diesel versus other products. To this end, we consulted EIA's database of past
exports. This database showed that, in 2022, the U.S. exported motor gasoline, low sulfur diesel
fuel, and finished petroleum products in the amounts shown in Table 8-14.
8-34
-------
Table 8-14: EIA Petroleum Product Export Data for 2022 (EIA Imports by Area of Entry
2023)
Product Million Barrels per Day
Motor Gasoline 0.867
Low Sulfur Diesel Fuel 1.011
Finished Petroleum Products 3.087
Combining the export scalers presented in Table 8-13 with the data shown in Table 8-14, we
then generated projections of future exports of gasoline, diesel, and other products that comprise
the net exports shown in Table 8-13. Our projections are shown in Table 8-15. Note that the
projections for "Other Petroleum Products" are simply the AEO 2023 net exports value less our
projections for motor gasoline and diesel.
Table 8-15: EPA Projections of Net Exports of Petroleum Products (Million Barrels per
Day).
Calendar Year Motor Gasoline Diesel Fuel Other Petroleum Products Net Exports
2023 0.93 1.08 2.26 4.28
2030 1.32 1.54 3.22 6.08
2035 1.43 1.66 3.47 6.56
2040 1.48 1.72 3.60 6.80
2045 1.49 1.74 3.63 6.86
2050 1.40 1.63 3.40 6.43
Combining the domestic fuel use data presented in Table 8-12 with our projected exports of
each fuel as shown in Table 8-15, we then estimated the amount of each refined by U.S. refiners.
Those results are shown in Table 8-16.
Table 8-16: EPA Estimated Domestic Refining (Million Barrels per Day).
Calendar Year Motor Gasoline Diesel Fuel Other Petroleum Products Total
2023 9.68 4.75 9.87 24.30
2030 9.44 4.85 11.37 25.67
2035 9.04 4.89 11.94 25.86
2040 8.78 4.92 12.39 26.08
2045 8.73 4.94 12.79 26.45
2050 8.83 4.84 12.93 26.59
To account for the latter effect impacting the refinery emissions attributable to refining of
liquid fuel for domestic consumption, that being that refineries refine more products than onroad
liquid fuels, we calculated scaling factors to apportion emission inventories specifically to the
refining of gasoline and diesel fuels versus other refined products. The scaling factors are based
on the relative energy demand of refining various fuels calculated by Wang et al. (Wang 2004).
Wang et al. expressed the energy demand of refining fuels in terms of mass and included outputs
that are not refinery products, so we removed non-refinery products and adjusted the energy
demand factors to be based on volume instead of mass.
Refinery emissions for refined products are related to the energy needed to refine those
products, but also depend on the emissions of other pollutants specific to refining those products.
8-35
-------
For example, the refining of gasoline causes higher methane emissions than an equivalent
volume of diesel. We developed pollutant-specific apportionment factors based on relative
emissions of refining gasoline, diesel, and other products using emission factors from DOE's
Greenhouse gases, Regulated Emissions, and Energy use in Technologies model (GREET 2021).
Final apportionment factors for each pollutant we modeled in our refinery analysis appear in
Table 8-17.
Table 8-17: Refinery Emission Apportionment by Fuel Type (unitless).
Pollutant
Gasoline
Diesel
Carbon Monoxide (CO)
0.602
0.057
Carbon Dioxide (CO2)
0.591
0.061
Methane (CH4)
0.640
0.053
Nitrous Oxide (N2O)
0.583
0.063
Nitrogen Oxides (NOx)
0.610
0.056
Particulate Matter (PM2.5)
0.620
0.054
Sulfur Dioxide (SO2)
0.596
0.058
fitile Organic Compounds (VOC)
0.570
0.058
OMEGA uses the inventory data presented in Table 8-11, the estimated domestic refining
data presented in Table 8-16, and the apportionment data presented in Table 8-17 to internally
calculate refinery emission rates. That is, the rates are not inputs to OMEGA, but rather the raw
data outlined above are the inputs. OMEGA also uses linear interpolations in years between the
years where data are available. Further, for years prior to 2030, the 2030 inventory data are used,
and beyond 2050, the 2050 data are used. In other words, we do not extrapolate data outside the
bounds of data availability. When OMEGA runs, the refinery emission rates that are internally
calculated are saved as one of the output files. Those results are shown in Table 8-18 for gasoline
and Table 8-19 for diesel.
Table 8-18: Refinery Emission Rates Calculated in OMEGA for Gasoline
(US tons per billion gallons refined).
Calendar Year
CO
CO2
ch4
N2O
NOx
PM2.5
SOx
VOC
2030
210
731.000
42.5
6.16
317
76.1
94.6
226
2035"
220
766.000
44.3
6.45
332
79.5
99
236
2040
228
795.000
45.8
6.69
345
82.4
103
243
2045
231
811.000
46.2
6.83
351
83.7
104
246
2050
231
814.000
46.1
6.86
350
83.7
104
244
8-36
-------
Table 8-19: Refinery Emission Rates Calculated in OMEGA for Diesel Fuel
(US tons per billion gallons refined).
Calendar Year
CO
C02
ch4
N2O
NOx
PM2.5
SOx
VOC
2030
38.7
147.000
6.81
1.29
56.4
12.7
18
44.9
2035
38.5
146.000
6.75
1.29
56.1
12.7
17.9
44.6
2040
38.5
146.000
: 6.73
1.29
56.3
12.7
17.9
44.4
2045
38.7
148.000
6.73
1.30
56.7
12.8
18
44.4
2050
39.9
154.000
6.93
1.35
58.5
13.2
18.5
45.5
After calculating the refinery emission rates shown, OMEGA uses them to estimate refinery
emissions associated with refining the fuels at issue, namely gasoline and diesel fuel meant for
domestic onroad consumption. This is done using the methodology described in 8.6.4 where we
also describe how our estimated reductions in fuel consumption are expected to impact refining
in the U.S.
8.6.3 Vehicle Emission Rates in OMEGA
For this analysis, EPA used an updated regulatory version of MOVES4, MOVES4.R2, to
create criteria pollutant and air toxic emission rate inputs for OMEGA. As described in Chapter 7
and further detailed in a memo to the docket (Mo 2024), MOVES4.R2 was developed to
represent our understanding of expected emissions under the rule at the time of our air quality
modeling analysis. Months later, when we ran MOVES to develop OMEGA emission rate
inputs, we had more information about how the final rule would regulate emissions of NMOG
and NOx emissions and the implications for ICE vehicles. Given the form and values of the
standards in the final rule, we now expect "backsliding" of hydrocarbon and NOx emissions
among LD and MD ICE vehicles will be negligible. Thus, we created new default input
databases for MOVES4.R2: MOVES4.R2a and MOVES4.R2b. These lack the LD and MD
changes to the "emissionrateadjustment" table described in the docket memo.
To create inputs for OMEGA, EPA ran MOVES for two scenarios: gasoline engines with
gasoline particulate filters (GPFs) (MOVES4.R2a) and without GPFs (MOVES4.R2b). The
emission rates for these scenarios differed in that in the scenario with GPFs, the emission rates
for exhaust PM were calculated by applying the GPF reduction factors and phase-in described in
Chapter 7 for MY 2027 and later. In the scenario without GPFs, the emission rates remain at
MOVES4 levels. We ran MOVES in inventory mode to create inventory and activity output by
calendar year, model year, fuel type, source type, and regulatory class for brake wear, tire wear,
start, running, and evaporative emissions for criteria emission precursors and air toxics. In these
runs, the only air toxics affected by GPFs were particle-phase PAHs, which are chained to
exhaust PM in MOVES. We consolidated the polycyclic aromatic hydrocarbons (PAH) output to
separately report emissions for naphthalene and for a potency-weighted (U.S. EPA 2021) sum of
the 15 other PAHs estimated by MOVES.
These two sets of MOVES output were then used to generate vehicle emission rates for use as
OMEGA inputs. Since MOVES generates emission inventories, and the applicable miles
traveled or gallons consumed attributes associated with those inventories, we can calculate
nationwide vehicle emission rates from them (grams per mile or grams per gallon). These
emission rates are then multiplied by the applicable miles driven or gallons consumed to estimate
8-37
-------
vehicle exhaust, evaporative, tire wear, and brake wear emissions for all vehicles in both the
analysis and legacy fleets and for each age in their lifetimes.
8.6.4 Calculating Upstream Emission Inventories
8.6.4.1 Electric Generating Units
To calculate upstream emission inventories, OMEGA operates on individual vehicles making
use of the VMTpoiiCy on each applicable fuel in the given OMEGA scenario.
For upstream emissions from EGUs, OMEGA first calculates the given vehicle's electricity
consumption according to the Consumptionvehicie ;electricity equation shown in Chapter 8.5.2.
OMEGA then estimates the required EGU generation by accounting for grid losses as below.
„ . CoTisumptioTipgfiiQig.gigQft-iQity
GenerationVghiclg.glectricity = transmission efficiency
Where,
Generation^hicie;eiectricity = the estimated EGU generation requirement to satisfy the electricity
consumption of the vehicle
Consumption^hicie;eiectricity = the electricity consumption of the given vehicle (described
above), inclusive of estimated charging losses
transmission efficiency = factor to account for the estimated efficiency of grid transmission,
set via the onroad fuels.csv OMEGA input file
OMEGA then calculates the annual electricity demand for the light- and medium-duty fleet.
Using that value, OMEGA calculates a U.S. electricity generation for the given year as:
US G6Tl6TCLtiOTlscenari0 US G6Y16VCLtiOYliow demand Demand,pM [ow demand DemandSCenario
where,
US (jeneralioihccwAXM = the estimated U.S. electricity generation in the given year for the given
scenario or policy
US Generation\owdemand = the U.S. generation used in the IPM modeling discussed in Chapter
5
Demandivu low demand = the light- and medium-duty fleet demand used in the IPM modeling
discussed in Chapter 5
DemandscavMU) = the light- and medium-duty flet demand calculated in OMEGA for the given
year and scenario
OMEGA then uses the estimated U.S. generation to interpolate an emission rate for each
pollutant using the IPM modeling results and the emission rates generated from them as
described in Chapter 8.6.1. With the applicable emission rate, OMEGA then calculates an
inventory for the given year as:
8-38
-------
„ r> j R®t€pollutant;scenario
Tonspollutant - Demandscenario x ^rams per ton
Where,
Zow^poiiutant = The inventory tons (U.S. or metric) of the given pollutant
DemandscavMU) = the estimated EGU generation requirement to satisfy the light- and medium-
duty fleet in the given year
_/?£?fepollutant;scneario — the EGU emission rate for the given pollutant in the given scenario and
year
grams per ton = 1,000,000 for metric tons (GHGs) or 907,185 for US (short) tons (criteria air
pollutants)
Importantly, the EGU inventories calculated in OMEGA represent EGU emissions associated
with generating electricity for the light- and medium-duty fleet. The EGU inventories are not
meant to reflect total US inventories. As such, any reductions or increases and percentage
changes reflect changes in emissions associated only with generating electricity for the light- and
medium-duty fleet, not the entire U.S.
8.6.4.2 Refineries
We have made several updates to the calculation of refinery emissions relative to what was
done in the DRIA. Those are:
1) Refinery impacts make use of refinery inventories generated as part of our AQM as
described in 8.6.2. We use those inventories to generate refinery emission rates and to establish a
"context" inventory from which any impacts can be calculated; and,
2) We have updated our estimates of the impacts on U.S. refining resulting from our projected
changes in domestic fuel demand.
Regarding number 2 above, while our NPRM analysis assumed that most of the reduced
refined product demand caused by the proposed rulemaking would result in a similar reduction in
U.S. refinery operations (93 percent), for reasons explained below, we also conducted a
sensitivity case in which U.S. refineries would continue to operate at current crude oil capacities.
If refineries continue to operate at their current capacities while demand for U.S. refined
products is decreasing, it means that there would be reduced U.S. product imports and increased
exports and U.S. refiners would continue to produce refined products, including coproducts such
as asphalt and tires.
There are good economic reasons why U.S. refineries might continue to operate despite
reduced U.S. product demand. In addition to coproducts as mentioned, the generally lower
natural gas and crude oil prices available in the U.S. allows U.S. refineries to have lower
production costs compared to other refinery regions around the world. The lower refinery
production costs are attributed to the lower feedstock costs. (EIA Today in Energy 2014)
The higher profit margins experienced by U.S. refineries starting after 2005 would be
expected to result in lower imports and higher exports and this is in fact is what has occurred.
Figure 8-14 shows US gasoline and diesel fuel net imports over time and shows a decrease in
8-39
-------
gasoline and diesel fuel nret imports starting in 2006 associated with improved U.S. refinery
margins. (EIA Imports by Area of Entry 2023) (EIA Spot Prices 2024) Note that the decrease in
net imports could either be a decrease in imports or increase in exports.
Net Imports and Crude Oil Spot Prices
>¦
ro
"C
15
-D
t
o
a.
£
a>
2
•Gasoline
~ Distillate
• Gasoline & Distillate
• Brent Crude Oil Prices
Figure 8-14: US Net Imports and Crude Oil Spot Prices.
As Figure 8-14 also shows, the decrease in net imports corresponded with the increase in
crude oil prices. Although not shown in the figure, the decrease in net gasoline imports also
corresponded with the increase in blending more corn ethanol into gasoline, which on a
volumetric basis could account for a large portion of the decline in net gasoline imports.
Beginning in 2014, the rate of decreased net exports began to level off and perhaps was
associated with simultaneous decrease in lower crude oil prices and leveling of ethanol demand.
Despite the favorable economic conditions for refiners here in the U.S., there were some
refinery closures or conversions. There were a couple refinery closures in the last several years
and they seem to be at least partially due to impacts from the C OVID-19 pandemic when product
demand, crude oil prices, and refinery margins plummeted. In addition to the pandemic, damage
from a hurricane and a desire to pivot toward lower carbon fuel options were reasons provided
by Shell for why it closed its Convent, Louisiana, refinery at the end of 2020—the Convent
refinery had a crude oil refining capacity of 211,000 barrels per day (bbd). (Mosbrucker, Without
a buyer, Shell may convert shuttered Convent refinery into alternative fuels facility 2021) Also
citing significant hurricane damage in 2021, Phillips 66 decided to shutter its Belle Chasse,
Louisiana, refinery—the Belle Chasse refinery had a crude oil refining capacity of 255,000 bpd.
(1012 Industry Report 2021) Additionally, several refiners have opted to fully or partially
convert their petroleum refineries to produce renewable diesel in recent years, including full
conversions of the Marathon refinery in Dickinson, North Dakota, and the Holly Frontier
8-40
-------
refineries in Artesia, New Mexico, and Cheyenne, Wyoming, and a partial conversion of the
CVR refinery in Wynnewood, Oklahoma.
Despite better refinery margins in the U.S. overall, the simultaneous closure or conversion of
some U.S. refineries in recent years makes the case that there is likely to continue to be the
closure or conversion of some U.S. refineries that have lower margins or face other issues as
demand for gasoline and diesel fuel declines in the U.S. The extent that U.S. refineries keep
operating, shutdown, or are converted, is difficult to project since it depends on the economics of
each particular refinery, the economic condition of the parent company, and strategy pursued by
each company's board for providing a return to its shareholders.
After careful consideration of the above, the history of decreased U.S. refinery net imports
associated with the more desirable economic conditions for refiners in the U.S., and weighing
this against the closure/conversion of some U.S. refineries over the past several years, we have
changed our projection of how refinery emissions would be impacted by this rulemaking.
Instead of estimating that U.S. refineries would largely reduce their production in response to
reduced refined product demand as done in the NPRM, we are now estimating that U.S.
refineries would respond at the midpoint of our range. Thus, of a certain reduction in U.S.
refined fuel demand, U.S. refinery output would account for half of that reduced demand, while
reduced net imports would account for the other half of that reduced demand.
A recently issued refining industry study seems to also support choosing a midpoint of the
range. McKinsey and Company projected how increased electrification of transportation vehicles
would affect refinery production in different refining production regions. (Cherry Ding 2022) As
shown in Figure 8-15, the McKinsey study analyzed three different demand outlook scenarios
which estimated three different rates of divestment from petroleum fuels. These demand outlook
scenarios estimate reductions in crude oil refined by the world's refineries by nine distinct
refinery sectors over time.
8-41
-------
Exhibit 1
Global liquids demand peaks before 2035 across scenarios.
Global liquids demand outlook by scenario, mbd
Peak demand
40
20
2025 2027 2033
0
1990 2000 2010 2020 2030 2040 2050
Fading Momentum (FM)
Slower uptake of EVs and less plastic
recycling, or avoidance due to technology,
or supply delays and a lack of regulation
enforcement
Current Trajectory (CT)
Current trajectory of battery decline and
increasing recycling continues, however,
currently active policies remain
insufficient to close the gap to ambition
Further Acceleration (FA)
Further acceleration of transition driven
by country-specific commitments, though
financial and technological restraints
remain
Disclaimer: Analysis conducted before the invasion of Ukraine in February 2022.
Source: McKinsey Energy Insights' Global Energy Perspective 2022
Figure 8-15: Global liquids demand; from (Cherry Ding 2022).
It is necessary to link the demand scenarios in the above figure to the impacts on refinery
throughput for the various world refining sectors. Per Figure 8-16, the study estimated refining
impacts for 2030 and 2040 based on these demand scenarios.
8-42
-------
Distillation capacity change by region, % of 2019 capacity
Fading momentum
Current trajectory
Further acceleration
2030
2040
2030
2040
2030
2040
Change, %
North America
o
O
o
>10
Latin America
o
o
0
o
(0-10)
Europe
o
o
O
(-10-0)
FSU'
€»
o
o
o
<-10
Africa
12
12
12
12
&
Middle East
25
25
21
21
19
South and SEAsia
16
tiffin
&
o
China
dp
Northeast Asia exc China
O
<5J»>
e>
Global, %
«
o
O
Global, mbd
(6jD
©i
fbj) j
| CO |
\g/ 1
Disclaimer Analysis conducted before the invasion of Ukraine in February 2022.
1Former Soviet Union.
Source: McKinsey Energy Insights' Global Downstream Model 2022
Figure 8-16: Distillation capacity change by region; from (Cherry Ding 2022).
The report (Cherry Ding 2022) estimates that North American refining capacity would
decrease by 41 percent in 2040 based on the Further Acceleration scenario, and this scenario
estimates a 19 percent reduction in crude oil demand as shown in Figure 8-15. This final
rulemaking is estimated to reduce gasoline and diesel fuel demand by about 40 billion gallons
per year toward the end of the analysis period, which equates to 2.6 million barrels per day.
Assuming that each barrel of gasoline/diesel fuel reduced equates to a barrel of crude oil
reduced, this amounts to 2.5 percent of the world crude oil demand which was about 100 million
barrels per day in 2019, the baseline year of the McKinsey analysis. We can estimate this
rulemaking's impact on North American refining production based on this McKinsey study by
comparing the 2.5 percent impact on crude oil production to the 19 percent estimated by the
study—doing so projects that this rulemaking would have 2.5/19 impact on crude oil demand
estimated in the study. Assuming that this smaller impact in world crude demand would impact
North American refineries at the same rate in the study, then North American refineries would
experience a 2.5/19*41 percent decrease in crude oil throughput, which equates to 5.6 percent
decrease in North American refinery throughput volume. North American refineries are
estimated to refine around 22 to 25 million barrels of crude oil per day, so the reduction in crude
oil refining would amount to 1.3 million barrels per day of decreased crude oil throughput at
North American refineries. The 1.3 MMbbl/day decreased crude oil throughput is half of the
estimated 2.6 MMbbl/day decrease in gasoline and diesel demand caused by the rulemaking.
Despite some simplifying assumptions used in this analysis using the McKinsey study, it does
seem to support our premise that of the decrease in U.S. refined product demand caused by this
rulemaking, about half would be due to reduced U.S. refinery production, while the balance
would be reduced net imports. As a sensitivity, EPA also estimated that 20 percent of reduced
8-43
-------
domestic liquid fuel demand would result in reduced domestic refining. We chose this sensitivity
as an estimate that falls between our central case where 50 percent of reduced demand would
result in reduced domestic refining and a possible case in which this final rule would have no
impact on domestic refining.
Regarding establishing a "context" inventory from which to estimate refinery impacts, we
start with the EPA estimates of domestic refining as shown in Table 8-16. The refinery estimates
shown in Table 8-16 reflect our estimates of refining throughput for all onroad gasoline and
diesel fuel. But we want to estimate the refinery impacts associated with passenger cars, light
trucks, and medium-duty vehicles. Therefore, we need context refinery throughput at a more
granular level than simply gasoline and diesel. To do this, we first ran OMEGA using a no-action
set of inputs to determine the gasoline and diesel fuel consumed by passenger cars, light trucks,
and medium-duty vehicles. With those results, we could apportion the liquid fuel consumption as
shown in Table 8-20.
Table 8-20: Share of Gasoline and Diesel Fuel Consumed by Regulatory Class (unitless).
Calendar Year
2023
2030
2035
2040
2045
2050
Passenger Car
0.42
0.32
0.25
0.21
0.20
0.19
Gasoline
Light Truck
0.53
0.62
0.68
0.71
0.72
0.72
Medium-Duty
0.05
07
08
08
09
09
Passenger Car
0.C
o.c
o.c
o.c
o.c
o.c
Diesel
Light Truck
0.06
0.11
0.14
0.15
0.15
0.15
Medium-Duty
0.90
0.87
0.86
0.85
0.85
0.85
We then estimated the share of onroad gasoline and diesel that are consumed by light- and
medium-duty vehicles versus heavy-duty vehicles using data generated in support of our heavy-
duty phase 3 final rule (Sherwood, OMEGA Refinery Data Inputs 2024). Those shares are shown
in Table 8-21.
Table 8-21: Share of Gasoline and Diesel fuel Consumed by Weight Classes (unitless).
Light- and Medium-Duty
Heavy
-Duly
Calendar Year
Gasoline
Diesel
Gasoline
Diesel
2023
0.98
0.15
0.02
0.85
2030
0.97
0.14
0.03
0.86
2035
0.97
0.13
0.03
0.87
2040
0.97
0.13
0.03
0.87
2045
0.97
0.14
0.03
0.86
By applying the shares shown in Table 8-20 and Table 8-21 to the estimated refinery
throughput shown in Table 8-16, we can get a better estimate of the context refinery throughput
from which to measure changes in fuel consumption within any OMEGA session. Note that
billion barrels per day are converted to gallons per year using 42 gallons per barrel and 365 days
per year.
The refinery inventories are estimated on an annual basis by regulatory class and fuel type and
by powertrain type. This latter element matters in the case of plug-in hybrid vehicles which
consume both liquid fuel and electricity. This is discussed in more detail below.
8-44
-------
For pure ICE, HEV, and PHEV powertrains that burn liquid fuel, we first calculate the
context gallons as described above using the values in Table 8-16 multiplied by the appropriate
apportionment shown in Table 8-20 and Table 8-21. We then determine the session fuel
consumption each year for the given regulatory class and fuel type. The inventory in the given
year for that regulatory class and fuel type is calculated as:
EmissionRate x [GallonsContext — Impact x (GallonsContext — GallonsSession)]
Inventory =
ConversionF actor
Where,
EmissionRate = the specific pollutant rate as shown in Table 8-18 or Table 8-19 converted to
grams per gallon
Gallonscontext = the context gallons for the given regulatory class and fuel type by combining
Table 8-16 with Table 8-20 and Table 8-21
Gallonssession = the session gallons for the given regulatory class and fuel type
Impact = the share of fuel savings that lead to reductions in domestic refining as described
above.
CornersionFactor = a conversion to U.S. tons for criteria pollutants or metric tons for GHGs
Note that the calculations for refinery inventories are done for every session in an OMEGA
run, including the No Action session. We then estimate the refinery impacts associated with any
action set of standards relative to the no action session and not relative to the context session.
Regarding refinery impacts of pure ICE and HEVs versus PHEVs, we calculate the annual
share of fuel consumed by PHEVs within a given regulatory class and fuel type relative to the
share consumed by pure ICE and HEVs. We then apportion the refinery impacts according to
those shares. For example, if PHEV gasoline cars consume 10 percent of the gasoline consumed
by gasoline cars in a given year, then we apportion 10 percent of the inventory impacts to
PHEVs and the remaining 90 percent to the pure ICE and HEV gasoline cars.
8.6.5 Calculating Vehicle Emission Inventories
A similar process to that described above for upstream emissions is used for vehicle emissions
with the exception that exhaust emission rates and both brake wear and tire wear emission rates
are multiplied by the VMTpoiiCy value, while evaporative, spillage, and leakage emission rates are
multiplied by the liquid-fuel consumption values described in Chapter 8.5.3. Exhaust emission
inventories are then added to evaporative, spillage, and leakage emission inventories to arrive at
vehicle emission inventories.
8-45
-------
8.6.6 Summary of Inventories and Inventory Impacts
8.6.6.1 Greenhouse Gas Inventory Impacts
Note that, in the tables presented in this section, CO2 equivalent (CChe) values use 100-year
global warming potential values of 28 and 265 for CH4 and N2O, respectively. (IPCC 2014) Note
also that any percent changes reflect changes in light- and medium-duty inventories relative to
the No Action scenario and do not represent percent changes in total U.S. inventories.
Table 8-22: Greenhouse gas emission inventory impacts, Final standards
(million metric tons)*.
Calendar
Year
C< >2
CH4
N20
C02e
Vehicle
EGU ;
Refinery
Vehicle
EGU !
Refinery
Vehicle
EGU
Refinery
Vehicle, EGU i Refinery
2027
-0.66
: 0.27 :
-0.022
-0.0000059
0.000018 ;
-0.0000013
; -0.0000087:
0.0000025
-0.0000002
: -0.67
; 0.27:
-0.022
2028
-4.5
! 1.2 :
-0.16
' -0.000045 ;
0.000078 ;
-0.000009
! -0.000051 ;
0.000011
-0.0000013
-1.5 "
1.2 ;
-0.16
2029
-14
; 2.8 ;
-0.48
-0.00016 1
0.00018 1
-0.000028
-0.00019 j
0.000024
-0.000004
; -14
r 2.81'
-0.48
2030
-29
j' 5.6 '
-0.97
35
0.00035 "
-0.000056
: -0.00043 1
0.000048
-0.0000082
;' "-29
! 5.6 V
-0.97
2031
-48
J' 9,1
-1.6
-0.00059
0.00059 :
-0.000094
i -0.00071 ;
0.000079
-0.000014
: -48
9.4 T
-1.6
2032
-68
I 13 "
-2.4
i -0.00087 1
0.00079 1
-0.00014
r -0.0011 ;
0.00011
-0.00002
: -69
113 *i
-2.4
2033
-99
' 17 ¦
-3.4
-0.0013 ]
0.001 1
-0.0002
: -0.0016 ;
0.00013
-0.000029
7 -99
; 17 '
-3,1
2034
-130
19
-4.5
r -0.0018 :
0.0012 ;
-0.00026
f -0.0021 ;
0.00015
-0.000038
f -130
20 :
-4.6
2035
-160
i 22 I
-5.7
' -0.0023
0.0013 1
-0.00033
r -0.0026
0.00017
-0.000048
-160
: 22'";"
-5.7
2036
-190
['24 !
-6.8
-0.0028 i
0.0014
-0.00039
r -0.0031
0.00018
-0.000057
-190
r 24""
-6.8
2037
-210
" 24
-7.8
! -11.111133
0.0014 1
-0.00045
1 ""-0.0036 -
0.00018
-0.000065
: -220
! "24":
-7.8
2038
-240
: 24 1
-8.8
-0.0038 '
0.0014 i
-0.0005
-0.004
0.00018
-0.000074
r -240
! 24 V
-8.8
2039
"' -260
24 1
-9.7
P -0.0043 1
0.0014 I
-0.00056
I -0.0044 :
0.00018
-0.000082
; -260
:"24']
-9.7
2040
-290
; 23
-11
: -0.0047 '
0.0014 ;
-0.00061
r-0.0048 1
0.00018
-0.000089
; -290
= 23"!"
-11
" 2041
-300
"24 :
-11
; -0.0052 ;
0.0013 I
-0.00065
-0.0051
0.00017
-0.000095
-310
: 24 V
-11
2042
-320
: '"24"i
-12
! -0.0056 ;
0.0013 f
-0.00069
1 -0.0055 j
0.00016
-0.0001
1 -320
i 24'V
-12
2043
-340
1 24 1
-13
: -0.006
0.0012 f
-0.00073
-0.0058 i
0.00015
-0.00011
-340
""24 7
-13
2044
-360
24 *
-13
i -0.0063 i
0.0011 |
-0.00076
: -0.006 :
0.00014
-0.00011
i -360
r 24'V
-13
2045
-370
r "24 :
-14
! -0.0066
0.001
-0.00079
-0.0063
0.00013
-0.00012
1-370
s 24 T
-14
" 2046
-380
; 24 *
-14
-0.0069 j
0.00098 :
-0.00082
' -0.0065 1
0.00012
-0.00012
! -380
24'";"
-14
2047
-390
; 23 '
-15
r -0.0072 i
0.00095 :
-0.00083
r -0.0066 I
0.00012
-0.00012
-390
! 23
-15
2048
-400
23 '
-15
"'-0.0074
0.00091
-0.00085
-0.0068 s
0.00011
-0.00013
j -400
: 23 T
-15
2049
-400
! 22 s
-15
" -0.0075
0.00086 ;
-0.00086
i' -0.0069 "
0.0001
-0.00013
-400
; 22 1
""-15
2050
-410
: 21 "I
-15
-0.0076
0.00082 ;
-0.00087
-0.0069 i
0.0001
-0.00013
: -410
: 21T
-15
2051
-410
21
-16
: -0.0077
0.00083
-0.00088
: -0.007
0.0001
-0.00013
-410
i 22'T
-16
2052
-410
I '22
-16
; -0.0078 '"i
0.00083 :
-0.00088
' -0.0071
0.0001
-0.00013
; -410
f 22 "T
-16
2053
-410
f 22'"'
-16
j -0.0078
0.00083 !
-0.00088
; -0.0071
0.0001
-0.00013
: -410
! 22 T
-16
2054
-410
r 21 :
-16
: -0.0078 ]
0.00083 5
-0.00088
f -0.0071 j
0.0001
-0.00013
-410
r 22";
-16
2055 ....
-410
r "21 *
-16
J' -0.0079 :
0.00083
-0.00088
r -0.0071 ;
0.0001
-0.00013
! -410
; "21
-16
Sum
-7,500
f 550 ;
-280
-0.13
0.027 i
-0.016
f -0.13 :
0.0034
-0.0023
-7,500
:550"'
-280
* Negative values reflect reductions; positive values increases.
8-46
-------
Table 8-23: Greenhouse gas emission inventory impacts, Alternative A
(million metric tons)*.
Calendar
Year
C< >2
CH4
N20
C02e
V ehicle, EGU ; Refinery
Vehicle
EGU : Refinery
Vehicle
EGU
Refinery
Vehicle, EGU; Refinery
2027
-7.3
2.
-0.25 :
-0.000078; 0.00013;-0.000015;
-0.000093 0.0000181
-0.0000021
¦ -7.4
2.
-0.25
2028
-20
; 5.3
-0.7 :
-0.00021
0.00035 -0.000041;
-0.00024
0.000048
-0.0000059
; -20
5.4
-0.7
2029
-37
7.3
-1.3
-0.0004
0.00046 -0.000075
-0.00046
0.000063:
-0.000011
= -37
7.3
-1.3
2030
-57
11 ;
-2.
-0.00065
0.00067 -0.00011 ;
-0.00077
; 0.00009 ;
-0.000017
; -58
11
-2.
2031
-79
14
-2.7 i
-0.00093
; 0.0009 ; -0.00016
-0.0011
0.00012 ;
-0.000023
: -79
15
-2.7
2032
-100
17
-3.5 :
-0.0012
; 0.0011 ; -0.0002
-0.0015
0.00014
-0.000029
; -loo
18
-3.5
2033
-130
21
-4.5 :
-0.0017
0.0013 ; -0.00026 :
-0.002
0.00017
-0.000038
: -130
21
-4.5
2034
-160
23 ,
-5.6
-0.0022
: 0.0014 ; -0.00033 ;
-0.0025
0.00018 ;
-0.000047
-160
23
-5.7
2035
-190
25
-6.7
-0.0027
! 0.0015 -0.00039 :
-0.003
0.0002
-0.000057
-190
25 ,
-6.8
2036
-220
26 :
-7.7
-0.0032
0.0015 -0.00045
-0.0035
0.0002
-0.000065
: -220
: 26
-7.8
2037
-240
27
-8.7
-0.0036
0.0016 ; -0.0005 :
-0.0039
0.0002
-0.000073
-240
27
-8.7
2038
-260
26
-9.6
-0.0041
; 0.0015 : -0.00055 ;
-0.0043
0.0002
-0.000081
-270
26
-9.7
2039
-290
25
-11
-0.0046
; 0.0015 ; -0.00061 :
-0.0047
0.00019;
-0.000089
-290
26
-11
2040
-300
24
-11
-0.0051
0.0014 ; -0.00065 ;
-0.005
0.00019
-0.000095
; -310
24
-11
2041
-320
25
-12 !
-0.0054
: 0.0014 ; -0.00069 ;
-0.0053
0.00018;
-0.0001
-320
, 25 ¦
-12
2042
-340
25
-13
-0.0058
: 0.0013 ; -0.00072 ;
-0.0056
0.00017
-0.00011
-340
. 25 ,
-13
2043
-350
25 :
-13
-0.0062
; 0.0012 : -0.00076 :
-0.0059
0.00016
-0.00011
; -360
25
-13
2044
-370
25
-14
-0.0065
; 0.0011 : -0.00079 ;
-0.0062
0.00014
-0.00012
; -370
25 !
-14
2045
-380
25 :
-14
-0.0068
; 0.001 i -0.00081 :
-0.0063
0.00013
-0.00012
; -380
25 :
-14
2046
-390
24
-15 l
-0.0071
0.00099 -0.00083
-0.0065
0.00012
-0.00012
; -390
24
-15
2047
-400
24
-15 ;
-0.0073
0.00096; -0.00085 ;
-0.0067
0.00012
-0.00013
-400
24
-15
2048
-400
23
-15
-0.0075
0.00092; -0.00087
-0.0068
0.00011 ;
-0.00013
; -400
23
-15
2049
-410
22
-16
-0.0076
0.00087; -0.00088 ;
-0.0069
; 0.00011 ;
-0.00013
-410
22
-16
2050
-410
22 !
-16 i
-0.0077
;0.00084 -0.00089 ;
-0.007
0.0001 ;
-0.00013
: -420
22
-16
2051
-420
22
-16
-0.0078
0.00084; -0.00089
-0.0071
0.0001 ;
-0.00013
-420
22
-16
2052
-420
22 5
-16 i
-0.0079
0.00084; -0.0009 :
-0.0071
0.0001
-0.00013
-420
22
-16
2053
-420
22 ;
-16 *
-0.0079
-0.0009
-0.0072
0.0001
-0.00013
; -420
22
-16
2054
-420
22 :
-16
-0.0079
0.00084; -0.0009 ;
-0.0072
0.0001 ,
-0.00013
-420
22
-16
2055
-410
22
-16 i
-0.0079
0.00084: -0.00089 :
-0.0072
0.0001
-0.00013
-420
22
-16
Sum
-7.900
600 i
-300
-0.14
; 0.03 ; -0.017 :
-0.13
0.0039
-0.0025
; -8.000
600
-300
* Negative values reflect reductions: positive values increases.
8-47
-------
Table 8-24: Greenhouse gas emission inventory impacts, Alternative B
(million metric tons) *.
Calendar
Year
C< >2
CH4
N20
C02e
: Vehicle=EGU • Refinery
i Vehicle
EGU
Refinery
Vehicle
EGU
Refinery
Vehicle i EGU: Refinery
2027
: -0.49
0.23
-0.017
-0.000004 0.0000151
-0.000001 =
-0.0000069:0.0000021:
-0.0000001
-0.49
0.23
-0.017
2028
-3.8
0.9
-0.13
35 59
-0.0000076;
-0.00004
0.0000082:
-0.0000011
-3.8
0.9
-0.13
2029
-13
2.4
-0.45
; -0.00015
0.00015
-0.000026
-0.00017
: 0.000021 ;
-0.0000038
-13
2.4
-0.45
2030
* -27
4.8
-0.9
i -0.00033
; 0.0003
-0.000052
-0.0004
: o.oooo4i
-0.0000076
-27
4.8
-0.91
2031
-43
7.9
-1.5
-0.00054
: 0.00049
-0.000085 :
-0.00066
; 0.000066 :
-0.000012
-43
8. :
-1.5
2032
-63
11 ;
-2.2
: -0.0008
; 0.00067
-0.00012 :
-0.00099
0.00009 ;
-0.000018
-63
11
-2.2
2033
: -90
15
-3.1
-0.0012
i 0.00089
-0.00018
-0.0014
0.00012
-0.000026
-90
15
-3.1
2034
-120
17
-4.1
l -0.0016
0.001
-0.00023 <
-0.0018
0.00013 :
-0.000034
-120
17
-4.1
2035
-140
18
-4.9
l -0.0019
0.0011
-0.00029 ;
-0.0022
0.00014 ,
-0.000042
-140
18
-5.
2036
; -160
19 :
-5.8
-0.0023
0.0011
-0.00033 ;
-0.0026
0.00015 :
-0.000049
-160
19 ,
-5.8
2037
-180
20
-6.6
; -0.0027
: 0.0011 ,
-0.00038 :
-0.0029
0.00015 :
-0.000055
-180
20
-6.6
2038
-200
19
-7.3
-0.0031
0.0011
-0.00042 i
-0.0033
0.00015
-0.000062
-200
19
-7.4
2039
-220
19
-8.1
-0.0035
0.0011
-0.00047 :
-0.0036
0.00014 :
-0.000068
-220
19
-8.1
2040
-240
18
-8.8
-0.0038
: 0.0011
-0.0005 :
-0.0039
0.00014
-0.000074
-240
18
-8.8
2041
-250
18 i
-9.4
; -0.0041
; 0.001 :
-0.00054
-0.0042
0.00013
-0.000079
-250
18 ,
-9.4
2042
-270
19
-10
-0.0045
0.00098
-0.00057 !
-0.0045
, 0.00013
-0.000084
-270
19
-10
2043
-280
19
-11
-0.0048
0.00092
-0.0006 i
-0.0047
0.00012 :
-0.000089
-280
19 :
-11
2044
-300
19 ;
-11
i -0.0051
: 0.00085 ;
-0.00063
-0.0049
! 0.00011 :
-0.000093
-300
19
-11
2045
-310
18
-12
-0.0054
0.00078 ;
-0.00066 :
-0.0051
0.000096 ,
-0.000097
-310
19 :
-12
2046
-320
18
-12
: -0.0056
0.00076
-0.00068 :
-0.0053
0.000093 ;
-0.0001
-320
18 :
-12
2047
-330
18
-12
-0.0058
0.00073
-0.0007
-0.0055
: 0.00009
-0.0001
-330
18 ;
-12
2048
-330
17
-13
: -0.006
: 0.00069 :
-0.00071
-0.0056
0.000085
-0.00011
-330
17
-13
2049
-340
17
-13
i -0.0061
0.00066 :
-0.00072 ,
-0.0057
0.00008
-0.00011
-340
17
-13
2050
-340
16
-13
-0.0062
0.00063
-0.00073 :
-0.0058
0.000076
-0.00011
-340
16
-13
2051
-340
16
-13
-0.0063
0.00063
-0.00073 i
-0.0059
, 0.000076
-0.00011
-340
16
-13
2052
-340
16
-13
: -0.0064
0.00063
-0.00074 s
-0.0059
0.000076 :
-0.00011
-350
16
-13
2053
-340
, 16 :
-13
-0.0064
0.00063
-0.00073 :
-0.0059
0.000076
-0.00011
-350
16
-13
2054
-340
16
-13
-0.0064
0.00062
-0.00073
-0.006
, 0.000075
-0.00011
-350
16
-13
2055
-340
16
-13
: -0.0064
0.00062
-0.00073
-0.006
; 0.000075 :
-0.00011
-340
16
-13
Sum
-6.300
430
-230
-0.11
0.021
-0.013
-0.11
0.0027
-0.002
-6.300
430
-230
* Negative values reflect reductions; positive values increases.
8-48
-------
Table 8-25: Net Greenhouse gas emission inventory impacts, Final standards*
Calendar
Year
Vehicle, EGU, Refinery
(Million metric tons per year)
% Change
C< >2
CH4
N20
C02e
C<>2
CH4
\2<>
C02e
2027
-0.41
1 1
-0.0000064
: -0.41
-0.027%
0.022%
-0.028%
-0.027%
2028
-3.5
0.000024
-0.000042
-3.5
-0.24%
0.052%
-0.19%
-0.24%
2029
-12
-0.000011
-0.00017
-12
-0.83%
-0.026%
-0.77%
-0.83%
2030
V" -24 ;
-0.000057
-0.00039
-24
-1.8%
-0.14% i
-1.9%
-1.8%
2031
-40
-0.0001
-0.00064
-40
-3% ;
-0.27% ;
-3.2%
-3%
2032
-58
-0.00023
-0.00097
-58
-4.6%
-0.64%
" -5%
-4.6%
2033
-85
-0.00054
-0.0015
-86
-7%
-1.6%
-7.8%
-7%
2034
1 -no ;
-0.00092
-0.002
-110
-9.5% ;
-2.9%
-11%
-9.5%
2035
-140
-0.0013 •
-0.0025
-140
-12%
-4.5%
-14%
-12%
2036
-170
-0.0018
-0.003
r -no
'-15%
-6.3% :
-17%
" -15%
2037
-200
-0.0023
-0.0035
; -200
-18%
-8.4%
-19%
-18%
2038
-220
-0.0029
-0.0039
-230
-20%
-11% j
-22%
-20%
2039
-250
-0.0034
-0.0043
-250
-23%
-13% !
-24%
-23%
2040
-270
-0.004
-0.0047
-270
-25%
-16%
-27%
-25%"
2041
-290
-0.0045
-0.0051
-290 I
-27%
-18%
-29% j
-27%
2042
-310
-0.005
-0.0054
-310
-29% :
-21%
-31%
-29%
2043
-330
-0.0055
-0.0057
-330 f
-31%
-23%
-33%
-31%
2044
-340
-0.006
-0.006
-350
-32%
-26%
-34%
-32%
2045
T -360
-0.0064
-0.0063
-360
-34%
-28%
-35%
-34%
2046
-370
-0.0068
-0.0065
-370
-35%
-30% ;
-36%"" j
-35%
2047
-380
-0.007
-0.0066
-380
-36%
-31%
-37%
-36%
2048
-390
-0.0073 ^
-0.0068
-390
-36% !
-32% 1
-37% 1
-36%
2049
r -390
-0.0075
-0.0069
-400
-37%
-33%
-38% ;
-37%
2050
-400
-0.0077
-0.007
-400
-37% !
-34%
-38%
-37%
2051
! -400
-0.0078
-0.0071
-410
-37% =
-34%
-38%
-37%
2052
-410
-0.0078
-0.0071
-410
-38%
-34%
-38%
-38%
2053
-410
-0.0079
-0.0071
-410 '
-38%
-35% ;
-38%
-38%
2054
-410
-0.0079
-0.0072
-410
-37% ;
-34%
-38%
-37%
2055
-410 ;
-0.0079
-0.0072
: -4io |
-37% :
-34%
-38%
-37%
Sum
; -7,200
-0.12
-0.13
; -7,200
""-21%
-15%
-23% !
-21%
* Negative values reflect reductions; positive values increases. Percent change reflects emissions associated with the light- and medium-duty
fleet only, not total US EGU emissions.
8-49
-------
Table 8-26: Net Greenhouse gas emission inventory impacts, Alternative A*
Calendar
Year
Vehicle, EGU, Refinery
(Million metric tons per year)
% Change
C()2
CH4
N20
I try
C()2
CH4
N20
C02e
2027
-5.6
0.00004
-0.000076
-5.6
-0.37%
0.083%
-0.33%
-0.37%
2028
-16
0.0001
-0.00019
-16
-1.1%
0.23%
-0.87%
-1.1%
2029
-31
-0.000012
-0.00041
-31
-2.2%
-0.029%
-1.9%
-2.2%
2030
-49
-0.000096 ;
-0.0007
-49
-3.6%
-0.24% '
-3.3% ;
-3.6%
2031
-67
-0.00019
-0.00099
-67
-5.1% i
-0.49%
; -4.9% i
-5.1%
2032
-86 i
-0.00037
-0.0013
-87
-6.8%
-1%
r -6.9%
-6.8%
2033
-no
-0.00071
-0.0018
-no
-9.2%
-2.1%
-9.6% j
-9.2%
2034
r -i4o
-0.0011
-0.0023
f -140
-12% !
-3.6%
-13% '
-12%
2035
-170
-0.0016
-0.0028
; -170
-15%
-5.3%
-16%
" -15%
2036
-200
-0.0021
-0.0033
-200
-17% i
-7.2%
-18%
-17%
2037
-220 f
-0.0026
-0.0038
; -220
-20%
-9.4%
f -21%
-20%
2038
-250
-0.0032
-0.0042
-250
-22%
' -12%
-24%
-22%
2039
T" -270
-0.0037
-0.0046
-270
-25%
-15%
; -26%
-25%
2040
-290
-0.0043
-0.0049
" -290
-27%
-17%
-28%
-27%
2041
' -310
-0.0048
-0.0053
-310
-29%
-20%
! -30%
-29%
2042
-330
-0.0052
-0.0056
-330 1
-31% ;
-22%
-32% !
' -31%
2043
-340
-0.0057
-0.0059
1 -340
-32%
-24%
1 -33%
-32%
2044
-360
-0.0062
-0.0061
-360
-34% ;
-27%
: -35% {
-34%
2045
-370 |
-0.0066
-0.0063
-370
-35% !
-29%
-36%
-35%
2046
-380
-0.0069
-0.0065
-380
-36% ;
-30%
-37% ]
-36%
2047
-390 j
-0.0072
-0.0067
-390
-36%
-32%
! " -37%
-36%
2048
-400
-0.0074
-0.0068
-400
-37% j
-33%
i -38% ;
-37%
2049
-400
-0.0076
-0.007
-400
-38%
-34%
-38%
-38%
2050
; -410
-0.0078
-0.0071
-410 '
-38%
-34%
-39% :
-38%
2051
-410
-0.0078 j
-0.0071
-410
-38%
-35%
-39% ;
-38%
2052
-410
-0.0079
-0.0072
-410
-38% :
-35%
r -39%
-38%
2053
-410
-0.008
-0.0072
-420
-38%
' -35%
-39% ,
-38%
2054
-410
-0.008
-0.0072
-410
-38%
-35%
-39%
-38%
2055 ^
-410
-0.008
-0.0072
-410
-37% ;
-35%
-38%
-37%
Sum
-7,600
-0.12
-0.13
i' -7,700
-23%
-15%
-24%
-23%
* Negative values reflect reductions; positive values increases. Percent change reflects emissions associated with the light- and medium-duty
fleet only, not total US EGU emissions.
8-50
-------
Table 8-27: Net Greenhouse gas emission inventory impacts, Alternative Bs
Calendar
Year
Vehicle, EGU, Refinery
(Million metric tons per year)
% Change
C()2
CH4
N20
C02e
C()2
till
N20
C02e
2027
-0.28
0.00001
19
-0.28
-0.018%
0.021% ]
-0.022% ;
-0.018%
2028
-3
0.000016
-0.000033
-3
-0.21%
0.036%
-0.15%
-0.21%
2029
-11
-0.000017
-0.00015
-11
-0.78%
-0.04%
-0.71%
-0.78%
2030
'-23
-0.000078
-0.00037
-23
-1.7%
r -0.2%
-1.8% |
-1.7%
2031
-37
-0.00013
-0.0006
-37
-2.8%
-0.35%
-3%
-2.8%
2032
7 -54
-0.00026
-0.00092
-54
-4.3%
j -0.73% |
-4.7%
-4.3%
2033
-78
-0.00048
-0.0013
-79
-6.4%
-1.4%
-7%
-6.4%
2034
-100 i
-0.00078
-0.0017
-100
-8.6%
j -2.4% i
' -9.3% 5
-8.6%
2035
-130 "
-0.0011 ;
-0.0021
F -130
-11%
* -3.7% J
-12%
-11%
2036
-150
-0.0015
-0.0025
-150
-13%
-5.2%
-14%
-13%
2037
-170
-0.0019
-0.0029
-170
-15%
i -6.9% i
-16%
-15%
2038
-190
-0.0023
-0.0032
-190
-17%
-8.9%
-18%
-17%
2039
-210 !
-0.0028
-0.0035
-210
-19%
-11%
-20%
-19%
2040
' -230
-0.0033
-0.0039
-230
-21%
; -13% !
-22%
-21%
2041
-240
-0.0037
-0.0041
-240
-23%
-15%
-23%
-23%
2042
-260 :
-0.0041 I
-0.0044
-260
-24%
-17%
-25% i
-24%
2043
-270
-0.0045
-0.0047
-280 [
-26%
; -19% ;
-27% ;
-26%
2044
-290
-0.0049
-0.0049
-290
-27%
-21%
-28%
-27%
2045
T -300
-0.0053
-0.0051
-300
-28%
-23%
-29%
-28%
2046
-310
-0.0055
-0.0053
-310
-29%
! -24%
-30%
-29%
2047
-320
-0.0058
-0.0055
; -320
-30%
-25%
-31%
-30%
2048
1 -330
-0.006
-0.0056
-330 ;
-31%
-26%
-31%
-31%
2049
-330
-0.0062 T
-0.0057
-330
-31%
: -27% j
-32%
-31%
2050
-340 1
-0.0063
-0.0058
-340
-31%
-28%
-32%
-31%
2051
' -340
-0.0064
-0.0059
-340
-32%
-28%
-32%
-32%
2052
-340
-0.0065
-0.006
f" -340
-32%
-28% i
-32%
-32%
2053
-340
-0.0065
-0.006
-340
-31%
-29% i
-32% i
-31%
2054
-340 ;
-0.0065 :
-0.006
-340
-31%
l -29%
-32%
-31%
2055
-340
-0.0066
-0.006
f -340
-31%
-29% ;
-32% I
-31%
Sum
1 -6,100
-0.099
-0.1
-6,100
-18%
-12%
-19%
-18%
* Negative values reflect reductions; positive values increases. Percent change reflects emissions associated with the light- and medium-duty
fleet only, not total US EGU emissions.
8-51
-------
8.6.6.2 Criteria Air Pollutant Inventory Impacts
Table 8-28: Criteria air pollutant impacts from vehicles, Final standards.
(US tons per year)
Calendar
Year
PM2.5
NOX
NMOG
SOX
CO
2027
-110
14
-37
-2.9
-410
2028
-290
-88
-470
-21
-6.700
2029
; -510
-580
-1.700
-66
-25.000
2030
; -860
-1.600 ;
-3.700
-130
-54.000
2031
: -1.200
-2.700
-6.400
-220
-91.000
2032
i -1.600
-4.300 :
-9.400
-320
-130.000
2033
: -2.000
-6.400
-14.000
-460
-210.000
2034
-2.500
-8.500 i
-19.000
-600
-290.000
2035
-2.900
-11.000
-25.000
-750
-380.000
2036
-3.300
-13.000 ;
-31.000
-890
-470.000
2037
-3.800
-15.000
-37.000
-1.000
-570.000
2038
-4.300
-17.000 i
-43.000
-1.100
-670.000
2039
-4.800
-19.000 ;
-48.000
-1.200
-770.000
2040
; -5.300
-22.000 ;
-54.000
-1.300
-870.000
2041
; -5.700
-23.000 :
-60.000
-1.400
-960.000
2042
: -6.100
-25.000
-67.000
-1.500
-1.100.000
2043
-6.400
-27.000
-73.000
-1.600
-1.200.000
2044
-6.700
-28.000 :
-80.000
-1.700
-1.300.000
2045
i -7.000
-30.000 ;
-85.000
-1.700
-1.300.000
2046
-7.300
-31.000 :
-92.000
-1.800
-1.400.000
2047
-7.500
-32.000 ;
-99.000
-1.800
-1.500.000
2048
-7.700
-32.000 :
-110.000
-1.900
-1.600.000
2049
i -7.900
-33.000 ;
-110.000
-1.900
-1.600.000
2050
; -8.000
-33.000 ;
-120.000
-1.900
-1.600.000
2051
-8.200
-34.000 !
-120.000
-1.900
-1.700.000
2052
* -8.300
-34.000 :
-130.000
-1.900
-1.700.000
2053
-8.300
-34.000 ,
-130.000
-1.900
-1.700.000
2054
-8.400
-35.000 :
-140.000
-1.900
-1.700.000
2055
: -8.500
-35.000 !
-140.000
-1.900
-1.700.000
* Negative values reflect reductions; positive values increases.
8-52
-------
Table 8-29: Criteria air pollutant impacts from vehicles, Alternative A
(US tons per year).
Calendar
Year
PM2.5
NOX
NMOG
SOX
CO
2027
; -130
-190 :
-830
-34
-13.000
2028
i -330
-550 :
-2.300
-95
-34.000
2029
: -560
-1.300 ;
-4.400
-180
-65.000
2030
! -910
-2.500 i
-7.200
-270
-100.000
2031
-1.300
-3.800 ;
-10.000
-370
-150.000
2032
; -1.600
-5.500 ,
-14.000
-470
-200.000
2033
: -2.100
-7.600 i
-19.000
-600
-280.000
2034
: -2.600
-9.800
-24.000
-750
-370.000
2035
; -3.000
-12.000 i
-29.000
-880
-450.000
2036
; -3.400
-14.000
-35.000
-1.000
-550.000
2037
-3.900
-16.000 :
-42.000
-1.100
-650.000
2038
-4.400
-19.000 :
-48.000
-1.200
-750.000
2039
: -4.900
-21.000 :
-54.000
-1.300
-850.000
2040
-5.300
-23.000 i
-59.000
-1.400
-940.000
2041
-5.700
-24.000 :
-65.000
-1.500
-1.000.000
2042
: -6.100
-26.000 :
-71.000
-1.600
-1.100.000
2043
i -6.400
-27.000 ,
-78.000
-1.700
-1.200.000
2044
; -6.800
-29.000 :
-84.000
-1.700
-1.300.000
2045
: -7.000
-30.000 :
-89.000
-1.800
-1.400.000
2046
; -7.300
-31.000 ;
-97.000
-1.800
-1.500.000
2047
: -7.500
-32.000 :
-100.000
-1.900
-1.500.000
2048
: -7.700
-33.000
-110.000
-1.900
-1.600.000
2049
-7.900
-33.000 i
-120.000
-1.900
-1.600.000
2050
i -8.100
-34.000 :
-120.000
-1.900
-1.700.000
2051
: -8.200
-34.000 :
-130.000
-1.900
-1.700.000
2052
; -8.300
-34.000 s
-130.000
-2.000
-1.700.000
2053
i -8.400
-35.000
-140.000
-2.000
-1.700.000
2054
-8.400
-35.000 i
-140.000
-2.000
-1.800.000
2055
-8.500
-35.000 :
-140.000
-1.900
-1.800.000
* Negative values reflect reductions; positive values increases.
8-53
-------
Table 8-30: Criteria air pollutant impacts from vehicles, Alternative B
(US tons per year).
Calendar
Year
PM2.5
NOX
NMOG
SOX
CO
2027
; -110
19
-17
: -2.1
-110
2028
i -290
-65
-380
= -18
-5.100
2029
: -500
-540 ;
-1.600
; -61
-23.000
2030
! -860
-1.500
-3.500
: -120
-50.000
2031
-1.200
-2.500
-5.800
-200
-83.000
2032
; -1.600
-4.000
-8.600
; -290
-120.000
2033
: -2.000
-5.900
-13.000
-420
-190.000
2034
: -2.500
-7.600 i
-17.000
-540
-250.000
2035
; -2.900
-9.300
-21.000
-650
-310.000
2036
; -3.300
-11.000 i
-25.000
-760
-380.000
2037
-3.800
-13.000
-30.000
* -850
-450.000
2038
; -4.300
-15.000 ;
-35.000
: -950
-530.000
2039
: -4.800
-16.000 :
-39.000
l -1.000
-610.000
2040
-5.300
-18.000 i
-44.000
-1.100
-690.000
2041
-5.700
-20.000 :
-48.000
-1.200
-760.000
2042
: -6.100
-21.000 :
-54.000
-1.300
-830.000
2043
i -6.400
-22.000 ;
-59.000
-1.300
-910.000
2044
-6.700
-24.000 i
-64.000
-1.400
-980.000
2045
: -7.000
-25.000
-69.000
: -1.400
-1.100.000
2046
; -7.300
-26.000 i
-75.000
-1.500
-1.100.000
2047
: -7.500
-27.000 ;
-80.000
: -1.500
-1.200.000
2048
: -7.700
-27.000 :
-85.000
: -1.600
-1.200.000
2049
-7.900
-28.000 ;
-90.000
-1.600
-1.300.000
2050
i -8.000
-28.000 :
-94.000
i -1.600
-1.300.000
2051
: -8.100
-29.000
-98.000
-1.600
-1.300.000
2052
; -8.200
-29.000
-100.000
: -1.600
-1.400.000
2053
i -8.300
-29.000 i
-110.000
* -1.600
-1.400.000
2054
-8.400
-30.000 ;
-110.000
: -1.600
-1.400.000
2055
-8.400
-30.000 :
-110.000
; -1.600
-1.400.000
* Negative values reflect reductions; positive values increases.
8-54
-------
Table 8-31: Criteria air pollutant impacts from EGUs and refineries, Final standards
(US tons per year) *.
Calendar
EGU
Refinery
Year
PM2.5
i NOX
i NMOG
SOX
PM2.5
NOX
NMOG
SOX
CO
2027
17
110
7.8
110
-2.6
-11
-7.6
-3.2
-7.1
2028
73
500
34
490
-18
-74
-53
-22 "
-49
2029
180
; 1,200
92
1,000
-55
-230
-160
-68
-150
2030
370
f 2,200
190
1,700
-110
-460
-330
-140
-310
2031
630
; 3,700
310
2,800
-190
-780
-550
-230
-520
2032
860
: 4,900
430
3,700
-270
-1,100
-800
-340
-740
2033
1,100
i 6,200
" 570
4,600
-390
-1,600
-1,200
-490
-1,100
2034
1,400 '
: 7,300
700
5,100
-520
-2,200
-1,500
-650
-1,400
2035
1,600
j 8,000
820
5,300
-650
-2,700
-1.91 in
-810
-1,800
2036
1,700
i 8,500
900
5,500
-780
-3,200
-2,300
-970
-2,100
2037
1,800
8,600
950
5,400
-890
-3,700
-2,600
-1,100
' -2,500
2038
1,800
; 8,500
980
5,200
-1,000
-4,200
-3,000
-1,200
-2,800
2039
1,800
f 8,200
1,000
4,800
-1,100
-4,600
-3,300
-1,400
-3,100
2040
1,800
7,900
1.
4,300
-1,200
-5,100
-3,600
-1,500
-3,300
2041
1,800
; 7,800
1,000
4,100
-1,300
-5,400
-3,800
-1,600
-3,600
2042
1,800
i 7,600
1,100
3,800
-1,400
-5,800
-4,100
' -1,700
-3,800
2043
1,800
i 7,400
1,100
3,500
-1,500
-6,100
-4,300
-1,800
-4,000
2044
1,800
: 7,000
1,100
3,000
-1,500
-6,400
-4,500
-1,900
-4,200
2045
1,700
; 6,600
1,100
2,600
-1,600
-6,600
-4,600
-2,000
-4,400
2046
1,700
6,500
1,000
2,400
-1,600
-6,800
-4,800
-2,000
-4,500
2047
1,600
; 6,300
r 1,000
2,100
-1,700
-7,000
-4,900
-2,100
-4,600
2048
1,600
: 6,000
1,000
1,800
" -1,700
-7,100
-5,000
-2,100
-4,700
2049
1,500
I 5,700
960
1,500
-1,700
-7,200
-5,000
-2,100
-4,800
2050
1,500
f" 5,500
940
1,300
-1,700
-7,300
-5,100
-2,200
-4,800
2051
1,500
: 5,600
" ] 940
1,300
-1,800
-7,400
-5,100
-2,200
-4,800
2052
1,500
i 5,600
T 950
1,300
-1,800
-7,400
-5,200
-2,200
-4,900
2053
1,500
! 5,600
950
1,300
-1,800
-7,400
-5,200
-2,200
-4,900
2054
1,500
: 5,600
940
1,300
-1,800
-7,400
-5,100
-2,200
-4,900
2055
1,500
; 5,500
930
1,300
-1,800
-7,400
-5,100
-2,200
-4,900
* Negative values reflect reductions; positive values increases. Data were not available for calculating CO inventories from EGUs.
8-55
-------
Table 8-32: Criteria air pollutant impacts from EGUs and refineries, Alternative A
(US tons per year) *.
Calendar
EGU
Refinery
Year
PM2.5
i NOX
i NMOG
SOX
PM2.5
NOX
NMOG
SOX
CO
2027
120
850
57
840
-29
-120
-86
-36
-80
2028
330
; 2,200
150
2,200
-80
-340
-240
-100
-220
2029
480
: 3,100
240
2,600
-150
-610
-440
-180
-410
2030
700
4,200
360
3,300
-230
-940
-670
-280
-620
2031
960
! 5,600
480
4,400
-310
-I.3I in
-920
-390
-860
2032
l.2i in
! 6,700
590
5,100
-400
-1,700
-1,200
-500
-1,100
2033
1,400
j 7,900
710
5,800
-520
-2,200
-I.5I in
-640
-1,400
2034
1,600
! 8,700
830
6,100
-640
-2,700
-1,900
-800
-1,800
2035
1,800
j 9,300
9111
6,200
-770
-3,200
-2,300
-960
-2,100
2036
1.91 in
: 9,500
1,000
6,200
-890
-3,700
-2,600
-1,100
-2,400
2037
1,900
| 9,500
1,000
6,000
-990
-4,200
-2,900
-I.2I in
-2,700
2038
2,000
9,200
1,100
5,600
-1,100
-4,600
-3,300
-1,400
-3,000
2039
1.91 in
i 8,800
1,100
5,200
-1,200
-5,000
-3,600
-I.5I in
-3,300
2040
1.91 III
: 8,300
1,100
4,600
-1,300
-5,400
-3,800
-1,600
-3,600
2041
1,900
;8,100
1,100
4,300
-1,400
-5,700
-4,000
-1,700
-3,800
2042
1,900
P 7,900
1,100
4,000
-1,400
-6,000
-4,200
-1,800
-4,000
2043
1,800
i 7,600
1,100
3,600
-I.5Ihi
-6,300
-4,500
-1,900
-I.2I in
2044
1,800
f 7,200
1,100
3,100
-1,600
-6,600
-4,600
'-2,000
-4,300
2045
1,700
1 6,700
1,100
2,600
-1,600
-6,800
-4,800
-2,000
-4,500
2046
1,700
P 6,500
1,100
2,400
-1,700
-7,000
-4,900
-2,100
-4,600
2047
1,700
j 6,300
1,000
2,100
-1,700
-7,100
-5,000
-2,100
-4,700
2048
1,600
6,100
1,000
1.91 hi
-1,700
-7,300
-5,100
' -2,100
-4,800
2049
1,600
5,800
980
1,600
-1,800
-7,400
-5,100
-2,200
-4,800
2050
l.5i in
| 5,600
950
l.3ihi
-1,800
-7,400
-5,200
-2,200
-4,900
2051
1,500
f 5,600
950
1,300
-1,800
-7,500
-5,200
-2,200
-1.91 in
2052
1,500
5,700
960
l.3ihi
-1,800
-7,500
-5,200
-2,200
-4,900
2053
1,500
P 5,700
960
l.3ihi
-1,800
-7,500
-5,200
-2,200
-5,000
2054
1,500
f 5,600
950
1,300
-1,800
-7,500
-5,200
-2,200
-4,900
2055
1,500
5,600
!' 940
1,300
-1,800
-7,400
"-5,200
-2,200
-1.91 in
* Negative values reflect reductions; positive values increases. Data were not available for calculating CO inventories from EGUs.
8-56
-------
Table 8-33: Criteria air pollutant impacts from EGUs and refineries, Alternative B
(US tons per year) *
Calendar
EGU
Refinery
Year
PM2.5
i NOX
i NMOG
SOX
PM2.5
NOX
NMOG
SOX
CO
2027
14
96
6.5
96
-1.9
-7.9
-5.6
-2.4
-5.2
2028
56
380
26 ""
380
-15 "
-63
-45
-19
-42 '
2029
160
; 1,000
80
880
-51
-210
-150
-64
-140
2030
320
!"" 1,900
160
1,500
-100
-430
-310
-130
-290
2031
530
! 3,100
260
2,400
-170
-700
-500
-210
-470
2032
740
4,200
370
3,200
-250
-1,000
-730
-310
-680
2033
1,000
: 5,600
510
4,100
-360
-1,500
-1,100
-440
-990
2034
1,200
: 6,400
610
" 4,400
-460
-1,900
-1,400
-580
-1,300
2035
1,300
i 6,700
690
4,400
-560
-2,400
-1,700
-700
-1,600
2036
1,400
! 7,000
740
4,500
-660
-2,800
-2,000
-820
-1,800
2037
1,400
; 7,000
770
4,300
-750
-3,100
-2,200
-940
-2,100
2038
1,500
6,800
790
4,100
-840
-3,500
-2,500
-1,000
" -2,300
2039
1,400
6,600
800
3,800
-930
-3,900
-2,700
-1,200
-2,600
2040
1,400
: 6,200
800
3,400
-1,000
-4,200
-3,000
-I.2I in
-2,800
2041
1,400
: 6,100
810
3,200
-1,100
-4,500
-3,100
-1,300
-2,900
2042
1,400
i 5,900
820
3,000
-1,100
-4,800
-3,300
-1,400
-3,100
2043
1,400
i 5,700
830
2,700
-1,200
-5,000
-3,500
-I.5I in
-3,300
2044
1,400
; 5,400
820
2,300
-1,300
-5,300
-3,700
-1,600
-3,500
2045
1,300
! 5,100
820
2,000
-1,300
-5,500
-3.91 III
-1,600
-3,600
2046
1,300
i 5,000
810
1,800
-1,400
-5,700
-4,000
-1,700
-3,700
2047
1,300
f 4,800
790
1,600
-1,400
-5,800
-4,100
-1,700
-3,800
2048
1,200
! 4,600
T 770
1,400
-1,400
-5.91 in
-4,100
-1,800
-3,900
2049
1,200
; 4,400
740
1,200
-1,400
-6,000
-I.2I in
-1,800
-4,000
2050
1,200
j 4,200
r 730
950
-1,500
-6,100
-4,300
-1,800
-4,000
2051
1,200
f 4,200
730
950
-1,500
-6,100
-I.3I in
-1,800
-4,000
2052
1,200
f 4,200
730
960
-1,500
-6,200
-4,300
-1,800
-4,100
2053
1,100
: ' 4,200
720
950
-1,500
-6,200
-4,300
-1,800
-4,100
2054
1,100
: 4,200
720
950
-1,500
-6,200
-4,300
-1,800
-4,000
2055
1,100
! 4,200
710
950
-1,500
-6,100
-4,300
-1,800
-4,000
* Negative values reflect reductions; positive values increases. Data were not available for calculating CO inventories from EGUs.
8-57
-------
Table 8-34: Net criteria air pollutant impacts from vehicles, EGUs and refineries, Final
standards *.
Calendar Vehicle, EGU, Refinery % Change
Year (US tons per year)
PM2.5
NOX
NMOG
SOX
CO
PM2.5
NOX
NMOG
SOx 1
CO
2027
-93
120
-37
110
-420
-0.22%
; 0.023%
-0.0054%
0.32% :
-0.0039%
2028
-230
330
-490
450
-6,700
-0.55%
; 0.072%
-0.079%
1.3% 1
-0.068%
2029
-380
350
-1,800
880
-25,000
-0.92%
1 0.085%
-0.31%
2.6%
-0.28%
2030
-600
170
-3,900
1,500
-54,000
-1.5%
' 0.045%
-0.72%
4.7% !
-0.64%
2031
-770
170
-6,600
2,400
-92,000
-1.9%
; 0.049%
-13%
7.7% :
""-1.2%
2032
-970
-480
-9,800
3,100
-140,000
-2.4%
= -0.16%
-2%
10% I
-1.9%
2033
-1,300
-1,700
-15,000
3,600
-210,000
-3.3%
1 -0.63%
-3.2%
12%
-3.2%
2034
-1,600
-3,400
-20,000
3,800
-300,000
" -4.2%
I -1.3%
-4.4%
14%
-4.7%
2035
-2,000
-5,400
-26,000
3,800
-380,000
-5.2%
r -2.3%
-6.1%
15%
-6.6%
2036
-2,400
-7,500
-32,000
3,700
-470,000
-6.3%
i -3.5%
-7.9%
"" 15% [
-8.9%
2037
-2.91 hi
-10,000
-38,000
3,300
-570,000
-7.7%
f -5.1%
-10%
13%
-12%
2038
-3,500
-13,000
-45,000
2,800
-680,000
-9.3%
; -7%
-12%
12%
-15%
2039
-4,100
-16,000
-51,000
2,200
-780,000
-11%
-9.1%
-14%
' 9.4% r
-18%
2040
-4,700
-19,000
-57,000
1,500
-870,000
-13%
-11%
-17%
6.7%
-21%
2041
-5,200
-21,000
-63,000
1,100
-970,000
-14%
-13%
-19%
4.9%
-25%
2042
-5,600
-23,000
-70,000
600
-1,100,000
-15%
-15%
-22%
2.8% ;
-29%
2043
-6,100
-25,000
-77,000
78
-1,200,000
-16%
f -17%
-24%
0.37% !
"" -32%
2044
-6,500
-28,000
-83,000
-510
-1,300,000
-18%
-19%
-27%
-2.5% j
-36%
2045
-6,900
-30,000
-89,000
-1,100
-1,300,000
-19%
r -20%
-29%
-5.7%
-39%
2046
-7,200
-31,000
-96,000
-1,400
-1,400,000
-19%
-22%
-32%
-7.5%
-42%
2047
-7,500
-32,000
-100,000
-1,800
-I.5I hi.
-20%
-23%
-34%
-9.5%
-44%
2048
-7,800
-34,000
-110,000
-2,100
-1.61 III.
" -21%
; -23%
-36%
-12% :
-46%
2049
-8,100
-34,000
-120,000
-2,500
-1.61 III.
" -21%
-24%
-38%
-14% ;
"" -48%
2050
-8,300
-35,000
-120,000
-2,800
-1,700,000
-22%
." -25%
-40%
-16% T
-49%
2051
-8,400
-36,000
-130,000
-2,800
-1,700,000
-22%
I -25%
-41%
-16% f
-50%
2052
-8,500
-36,000
-130,000
-2,800
-1,700,000
-22%
-25%
-43%
"16% ;
-51%
2053
-8,600
-36,000
-140,000
-2,800
-1,700,000
-22%
-25%
-44%
-16% i
-51%
2054
-8,700
-36,000
-140,000
-2,800
-1,700,000
-22%
-25%
-45%
-16% r
-51%
2055
-8,700
-36,000
-150,000
-2,800
-1,700,000
-22%
-25%
-46%
'-16%
" -52%
* Negative values reflect reductions; positive values increases. Data were not available for calculating CO inventories from EGUs. Percent
change reflects emissions associated with the light- and medium-duty fleet only, not total US emissions.
8-58
-------
Table 8-35: Net criteria air pollutant impacts from vehicles, EGUs and refineries,
Alternative A *.
alendar
Vehicle, EGU, Refinery
% Change
Year
(US tons per year)
PM2.5
NOX
; NMOG
SOx
CO
PM2.5
NOX
; NMOG
SOx
CO
2027
-37
530
-860
" 770
-13,000
-0.087%
0.1%
; ' -0.13%
2.3%
-0.12%
2028
-81
1,400
-2,300
2,000
-34,000
-0.19%
0.29%
f -0.38%
5.8%
-0.35%
2029
-230
1,200
f -4,600
" 2,300
-66,000
-0.55%
0.28%
-0.81%
6.8%
-0.73%
2030
-440
750
-7,500
" 2,700
-110,000
"" -1.1%
0.2%
-1.4%
8.8%
-1.3%
2031
-620
520
: -11,000
3,600
-150,000
-1.5%
0.15%
: -2.1%
12%
-1.9%
2032
-850
-440
! -14,000
4,100
-200,000
-2.1%
-0.14%
1 -3%
14%
-2.8%
2033
-I.2I in
-1,900
; -19,000
4,500
-280,000
-3%
-0.69%
-4.2%
16%
-4.2%
2034
-1,600
-3,800
1 -25,000
4,600
-370,000
-4%
-1.5%
-5.5%
17%
-5.9%
2035
-2,000
-6,000
: -31,000
4,400
-450,000
-5.1%
-2.6%
-7.2%
17%
-7.9%
2036
-2,400
-8,300
: -37,000
4,100
-550,000
-6.3%
-3.9%
-9.1%
16%
-10%
2037
-2.91 in
-11,000
-44,000
3,600
-650,000
-7.7%
-5.6%
-12%
15%
-13%
2038
-3.51 in
-14,000
I -50,000
3,000
-760,000
-9.4%
-7.5%
-14%
13%
-17%
2039
-4,100
-17,000
-56,000
2,300
-860,000
-11%
-9.7%
! -16%
10%
-20%
2040
-4,700
-20,000
-62,000
1,500
-950,000
-13%
-12%
-18%
6.8%
" -23%
2041
-5,200
-22,000
I -68,000
1,100
-1,000,000
-14%
-14%
-21%
4.9%
-27%
2042
-5,700
-24,000
-75,000
570
-1,100,000
-15%
-16%
i -23%
2.7%
-30%
2043
-6,100
-26,000
-81,000
31
-I.2I in.
-17%
-17%
; -26%
0.15%
""-34%
2044
-6,500
-28,000
; -88,000
-560
-I.3i in.
-18%
-19%
: -28%
"-2.8%
-37%
2045
-6,900
-30,000
1 -93,000
-I.2Ihi
-1. inn.
' -19%
-21%
-31%
-6%
-40%
2046
-7,300
-31,000
-100,000
-I.5Ihi
-1.51 in.
-20%
-22%
T" -33%
-7.8%
-43%
2047
-7,600
-33,000
r -110,000
-1,800
-I.5i in.
-20%
""-23%
i -35%
-9.8%
-45%
2048
-7,800
-34,000
"-110,000
-2,200
-1.61 III.
-21%
-24%
r -37%
-12%
-47%
2049
-8,100
-35,000
-120,000
-2,500
-1,600,000
-22%
-24%
I -39%
-14%
-48%
2050
-8,300
-36,000
T -120,000
-2,800
-1,700,000
-22%
-25%
-41%
-16%
"-50%
2051
-8,400
-36,000
-130,000
-2.91 hi
-1,700,000
-22%
-25%
-42%
-16%
-50%
2052
-8,500
-36,000
1 -130,000
-2,900
-1,700,000
-22%
-25%
; -44%
-16%
-51%
2053
-8,600
-37,000
T -140,000
-2,900
-1,700,000
-22%
-25%
-45%
-16%
-52%
2054
-8,700
-37,000
T -140,000
-2,900
-1,800,000
-22%
-25%""
r -46%
-16%
-52%
2055
-8,700
-37,000
j -150,000
-2,800
-1,800,000
-22%
-25%
-46%
-16%
-52%
* Negative values reflect reductions; positive values increases. Data were not available for calculating CO inventories from EGUs. Percent
change reflects emissions associated with the light- and medium-duty fleet only, not total US emissions.
8-59
-------
Calendar
Year
Table 8-36: Net criteria air pollutant impacts from vehicles, EGUs and refineries,
Alternative B *.
Vehicle, EGU, Refinery
% Change
(US tons per year)
: PM2.5
j NOX
NMOG
i SOX ;
CO
PM2.5 i
NOX
NMOG
SOX
CO
2027
-91
110
-16
f" 91 "1
-120
-0.22% ;
0.021% :
-0.0024%
^ 0.27% ;
-0.0011°
2028
-240
250
-400
r 340
-5.21 hi
-0.58%
0.054%
-0.064%
" 0.96% :
-0.053°/
2029
-390
260
-1,600
760
-23,000
-11.96",,
0.063% 1
-0.29%
2.3% :
-0.25%
2030
f -640
-34
-3,600
| 1,200 :
-50,000
-1.6% :
-0.009% i
-0.67%
"" 4%
-0.6%
2031
: -850
-160
-6,000
2,000 1
-83,000
-2.1%
-0.047% !
-1.2%
r 6.4% *
-1.1%
2032
; -1,100
-880
-9,000
; 2,600
-120,000
-2.7% |
-0.29%
-1.9%
; 8.4% " 1
-1.7%
2033
i -1,400
: -1,800 :
-13,000
;" 3,200"""
-190,000
-3.5% i
""-0.64%
-2.8%
11%
-2.8%
2034
: -i,8oo
: -3,200 :
-17,000
i ' 3,300
-250,000
-4.5% !
-1.3%
-3.8%
12%
-4%
2035
; -2,200
; -1.91 III j
-22,000
¦ 3,100 |
-310,000
-5.6% :
-2.1% ;
-5.1%
12%
-5.4%
2036
; -2,600
r -6,800 1
-26,000
i 2.91 III
-380,000
-6.8% 1
-3.2%""
-6.5%
12%
-7.2%
2037
! -3,100
5 -8,900
-31,000
j 2,600
-460,000
-8.3% !
-4.5% :
-8.3%
i 10%
-9.4%
2038
¦" -3,700
j -11,000 j
-36,000
! 2. Inn
-540,000
-9.9% !
-6.1% ;
-10%
! 8.9% !
-12%
2039
f -4,300
-14,000 j
-41,000
1,600
-620,000
-11%'" i
-7.9%
-12%
7% i
-14%
2040
1-4,800
| -16,000
-46,000
f 1,100 1
-690,000
-13% 1
-9.7% ;
-14%
: 4.7% j
-17%
2041
i -5,300
-18,000 1
-51,000
! 690 |
-760,000
-14% ;
-11%
-16%
: 3.1% ";
-20%
2042
r -5,800
-20,000
-56,000
? 300
-840,000
-16% ;
-13%
-18%
1.4% ;
-23%
2043
i -6,200
; -22,000 :
-62,000
1-130i
-910,000
-17% ;
-14%
-20%
: "-0.64% '
-25%
2044
: -6,600
-23,000
-67,000
! -600 ;
-990,000
-18% I
-16%
-22%
i " -3%
-28%
2045
i -7,000
! -25,000 1
-72,000
! -i,ioo ;
-1,100,000
-19% ;
-17%
-24%
-5.5%" :
-31%
2046
: "-7,300
f -27,000 i
-78,000
' -1,400 1
-1,100,000
-20% |
-18%
-26%
"7.1% ' :
-33%
2047
: -7,600
! -28,000
-84,000
-1,600 ;
-1,200,000
-20% !
-19%
-27%
= -8.7% :
-35%
2048
; -7,900
f -29,000 ]
-89,000
-1,900
-1,200,000
-21% i
-20% r
-29%
f" -10% I
-36%
2049
; -8,100
F -30,000 i
-94,000
r -2,200
-1,300,000
-22% :
-21% j
-31%
-12% !
-38%
2050
f -8,300
i -30,000
-98,000
i -2,500 I
-1,300,000
-22%
-21%
-32%
i' "-14% j
-39%
2051
= -8,400
: -31,000 1
-100,000
i -2,500 ""i
-I.3I in.
"" -22% ;
-21%
-33%
; -14% !
-40%
2052
; -8,600
-31,000 ".
-110,000
: -2,500
-1. inn.
-22% '.
-22% !
-35%
: '-14% !
-40%
2053
j -8,600
-31,000 1
-110,000
; -2,500 1
-1. inn.
-22% 1
-22%
-35%
-14% !
-41%
2054
! -8,700
-31,000 \
-110,000
f -2,500 1
-1,400,000
-22% :
-22%
-36%
; '"14% '
-41%
2055
: -8,800
1-32,000
-120,000
: -2,500
-1,400,000
-22% !
-22%
-37%
: ' -14%"" j
-42%
* Negative values reflect reductions; positive values increases. Data were not available for calculating CO inventories from EGUs. Percent
change reflects emissions associated with the light- and medium-duty fleet only, not total US emissions.
8-60
-------
8.7 Estimating Energy Security Effects
The energy security premia (the energy security savings in dollars per barrel of reduced
imported oil) and the process used to estimate those values are described in Chapter 10. The
discussion here focuses on how OMEGA estimates the oil consumption impacts to which the
energy security premia can be multiplied to estimate monetized benefits.
8.7.1 Calculating Oil Consumption from Fuel Consumption
Chapter 8.5.3 describes how OMEGA estimates liquid-fuel consumption. This is done for
every vehicle that operates any miles on a liquid-fuel, whether that fuel be gasoline or diesel.
Chapter 8.5.4 presents the estimated impacts of the final standards and alternatives on overall
fuel consumption.
8.7.2 Calculating Oil Imports from Oil Consumption
To estimate energy security benefits, OMEGA converts fuel consumption impacts to oil
import impacts. This is done using the values shown in Table 8-37: Parameters used in
estimating oil import impacts.
Table 8-37: Parameters used in estimating oil import impacts.
Item Value
Share of pure gasoline in retail gasoline 0.9
Share of pure diesel in retail diesel 1.0
Energy density ratio of pure gasoline to crude oil 0.881
Energy density ratio of diesel to crude oil 0.998
Gallons per barrel of crude oil 42
Oil import reduction as percent of total oil demand reduction 0.948
The barrels of oil consumed in a given scenario are estimated as shown below.
Energy Density Ratio
Barrels = FuelConsumptionvehicle.liquid x Share x GallonsPerBarrel
Where,
Barrels = the barrels of oil associated with the fuel consumption value
FuelConsumptioriMQhicie;iiquid = the liquid-fuel consumption of the given vehicle (see Chapter
8.5.3)
Share = the applicable "pure share" shown in Table 8-37
EnergyDensityRatio = the applicable energy density ratio shown in Table 8-37
GallonsPerBarrel = 42 as shown in Table 8-37
8-61
-------
Table 8-37 shows an "Oil import reduction as percent of total oil demand reduction" factor
equal to 0.948. In Chapter 8.6.4.2, we described our new estimate of the impact of reduced
domestic liquid fuel demand on U.S. refinery throughput. That estimate also impacts our
estimate of the oil import factor used in the energy security analysis. As explained in Chapter
11.4 of the draft LMDV RIA:
"For this energy security analysis, we undertake a detailed analysis of differences in U.S. fuel
consumption, crude oil imports/exports, and exports ofpetroleum products for the time frame
2027-2050 using the AEO 2021 (Reference Case) in comparison with an alternative AEO 2021
sensitivity case, Low Economic Growth. The Low Economic Growth Case is used since oil
demand decreases in comparison to the Reference Case. EPA estimates that approximately 90.7
percent of the change in fuel consumption resulting from these proposed standards is likely to be
reflected in reduced U.S. imports of crude oil over the time frame of analysis of this proposed
rule. The 90.7 percent oil import reduction factor is calculated by taking the ratio of the changes
in U.S. net crude oil and refined petroleum product imports. We are using AEO 2021, as
opposed to the more recent AEO 2022, for the quantitative analysis of this proposed rule to
maintain consistency with other parts of the analysis (i.e., air quality modeling) of this proposed
rule. The AEO 2021 projects oil market trends through 2050. The time frame for EPA's analysis
of this proposed rule is from 2027 to 2055. Thus, we report oil market trends to 2050 based upon
AEO 2021 in Table 11-1. We also report U.S. oil import reductions through 2055 in Table 11-1
as well. Calculated using series "Petroleum Consumption (Excluding Biofuels) Annual" (Table
1.3) and "Petroleum Consumption Total Heat Content Annual" (Table A3). We looked at
changes in U.S. crude oil imports/exports and net petroleum products in the AEO 2021
Reference Case, Table 11. Petroleum and Other Liquids Supply and Disposition, in comparison
to an alternative case, the Low Economic Growth Case. See the spreadsheet in the Docket,
"AEO2021 Change in product demand on imports" divided by the change in U.S. oil
consumption in the two different AEO cases considered. "
The docketed spreadsheet contains a summary table for the analysis contained in the
spreadsheet, and Table 8-38 shows that for the reduction in refined product estimated by AEO
2021 Low Economic Growth Case relative to the Reference Case, 83.7 percent of the reduced
product demand is attributed to reduced imported crude oil, while 7.0 percent is attributed to
reduced net imports—resulting in the 90.7 percent import factor.
Table 8-38 Oil Import Factor based on AEO 2021
Average over the years 2027 to 2050
83.7 Percent of imported crude oil
9.3 Percent reduction in domestic crude oil
7.0 Percent reduction in net imported product
100.0
90.7 Total percentage of imported petroleum
For the final rule, the same methodology based on the AEO 2023 would result in a 89.6
percent oil import factor—84.8 percent of which would be due to reduced imported crude oil and
4.8 percent would be due to reduced net imports.
8-62
-------
Table 8-39 Oil Import Factor based on AEO 2023
Average over the years 2027 to 2050
84.8 Percent of imported crude oil
10.3 Percent reduction in domestic crude oil
4.8 Percent reduction in net imported product
100.0
89.6 Total percentage of imported petroleum
Use of the two AEO cases cited above estimates a large reduction in U.S. refinery
throughput—AEO2021 estimates that 93 percent (83.7+9.3) of the reduced product demand
would be attributed to reduced throughput at U.S. refineries, while based on AEO2023, the
reduction in U.S. refinery throughput would be 95.1 percent. However, for the final rulemaking
we are estimating that refineries would not reduce their throughput to the same extent. Instead, of
a given volume reduction of gasoline and diesel fuel demand caused by this rulemaking, we are
estimating that 50 percent of that reduced demand would be due to reduced production by U.S.
refineries, while the other 50 percent would be from reduced net imports (see Chapter 8.6.4.2 for
an explanation on the basis for this assumption). Thus, we needed a way to estimate the energy
security impacts assuming that U.S. refiners would continue producing domestic fuels at a much
higher level associated with the 50/50 assumption.
Since we are now estimating that in response to reduced refined product demand, half of that
reduced demand would be reduced production from U.S. refineries and the other half would be
decreased net imports, two different methods for estimating the oil import factor can both be
used. The portion of reduced refinery demand projected to result in reduced refinery throughput
can be represented by the oil import factor estimated by the two AEO cases. However, since
reduced refinery throughput is estimated to comprise all of the reduced demand, we instead
assumed that the percent reduction in net imported product would also be reduced imported
crude oil—thus, all of the 89.6 percent reduced imported petroleum would be imported crude oil.
Conversely, the balance of reduced refinery demand which US refineries keep operating can be
represented by the oil import factor, which by definition would be 100 percent (since net imports
would have to decrease at the same rate that refinery demand decreases). Thus, the oil import
factor is estimated by the following equation:
Oil import reduction as percent of total oil demand reduction = 89.6% x 0.5 + 100% x 0.5 =
94.8%
The reduced barrels of imported oil are then calculated as shown below.
BcLTTCls^angg jn imports (.B &TT 6 ISno Action BCLTTGlSj^c^ion) X Oil ilTipOTt fCLCtOT
Where,
Oil impact factor = the Oil import reduction as percent of total oil demand reduction
8-63
-------
8.7.3 Summary of Energy Security Effects
Table 8-40: Impacts on oil consumption and oil imports, Final standards
(millions).
Calendar
Barrels
Barrels Imported
Barrels Imported per Day
Year
2027
-1.3
-1.3
-0.0035
2028
-9.
-8.5
-0.023
2029
-28
-27
-0.073
2030
-57
-54
-0.15
2031
-95
-90
-0.25
2032
-140
-130
-0.36
2033
-200
-190
-0.51
2034
-260
-240
-0.67
2035
-320
-300
-0.83
2036
-380
-360
-0.98
2037
-430
-410
-1.1
2038
-480
-450
-1.2
2039
-530
-500
-1.4
2040
-570
-540
-1.5
2041
-610
-570
-1.6
' 2042
-640
-610
-1.7
2043
-680
-640
-1.8
2044
-710
-670
-1.8
2045
-730
-700
-1.9
2046
-760
-720
-2.
2047
-770
-730
-2.
2048
-790
-750
-2.1
2049
-800
-760
-2.1
2050
-810
-770
-2.1
2051
-820
-770
-2.1
2052
-820
-780
-2.1
2053
-820
-780
-2.1
2054
-820
-780
-2.1
2055
-820
-780
-2.1
Sum
-15,000
-14,000
* Negative values reflect reductions; positive values increases.
8-64
-------
Table 8-41: Impacts on oil consumption and oil imports, Alternative A
(millions).
Calendar
Barrels
Barrels Imported
Barrels Imported per Day
Year
2027
-15
-14
-0.038
' 2028
-41
-39
-0.11
2029
-75
-71
-0.19
2030
-110
-no
-03
2031
-160
-150
-0.41
2032
-200
-190
-0.52
2033
-260
-240
-0.67
2034
-320
-300
-0.83
2035
-380
-360
-0.98
2036
-430
-410
-1.1
2037
-480
-450
-1.2
2038
-530
-500
-1.4
2039
-570
-540
-1.5
2040
-610
-580
-1.6
2041
-640
-610
-1.7
2042
-680
-640
-1.8 "
2043
-710
-670
-1.8
2044
-730
-700
-1.9
2045
-750
-710
-2.
' 2046
-770
-730
-2.
2047
-790
-750
-2.1
2048
-800
-760
-2.1
2049
-820
-770
-2.1
2050
-820
-780
' -2.1
2051
-830
-790
-2.2
2052
-830
-790
' -2.2
2053
-840
-790
-2.2
2054
-840
-790
-2.2
2055
-830
-780
-2.1
Sum
-16,000
-15,000
* Negative values reflect reductions; positive values increases.
8-65
-------
Table 8-42: Impacts on oil consumption and oil imports, Alternative B
(millions).
Calendar
Barrels
Barrels Imported
Barrels Imported per Day
Year
2027
-0.99
-0.94
-0.0026
' 2028
-7.6
-7.2
-0.02
2029
-26
-25
-0.068
2030
-53
-50
-0.14
2031
-86
-82
-0.22
2032
-130
-120
-0.33
2033
-180
-170
-0.47
2034
-230
-220
-0.6
2035
-280
-260
-0.72
2036
-320
-310
-0.84
2037
-360
-350
-0.95
2038
-400
-380
-1.
2039
-440
-420
-1.1
2040
-470
-450
-1.2
2041
-500
-480
-1.3
2042
-530
-510
-1.4
2043
-560
-540
-1.5
2044
-590
-560
-1.5
2045
-610
-580
-1.6
' 2046
-630
-600
-1.6
2047
-650
-620
-1.7
2048
-660
-630
-1.7
2049
-670
-640
-1.7
2050
-680
-650
-1.8
2051
-680
-650
-1.8
2052
-690
-650
-1.8
2053
-690
-650
-1.8
2054
-690
-650
-1.8
2055
-680
-650
-1.8
Sum
-13,000
-12,000
* Negative values reflect reductions; positive values increases.
8-66
-------
Chapter 8 References
1012 Industry Report. 2021. "Belle Chasse oil refinery closing after being damaged by Hurricane
Ida." 1012 Industry Report. November 16. Accessed 3 7, 2024.
https://www.1012industryreport.com/industry/belle-chasse-oil-refinery-closing-after-being-
damaged-by-hurricane-ida/.
Cherry Ding, Alexandre Ferro, Tim Fitzgibbon, and Piotr Szabat. 2022. "Refining in the energy
transition through 2040." McKinsey & Company, October.
EIA. 2014. "How much carbon dioxide is produced by burning gasoline and diesel fuel?" Energy
Information Administration, May 21. Accessed February 1, 2024.
EIA Imports by Area of Entry. 2023. "Imports by Area of Entry, Petroleum and Other Liquids."
Petroleum and Other Liquids. Accessed 3 7, 2024.
https://www.eia.gov/dnav/pet/pet_move_imp_dc_NUS-ZOO_mbblpd_a.htm.
EIA Spot Prices. 2024. "Spot Prices; Petroleum and Other Liquids." Petroleum and Other
Liquids. Accessed 3 7, 2024. https://www.eia.gov/dnav/pet/pet_pri_spt_sl_a.htm.
EIA Today in Energy. 2014. "Lower crude feedstock costs contribute to North American refinery
profitability." Today in Energy. June 5. Accessed 3 7, 2024.
https://www.eia.gov/todayinenergy/detail.php?id=16571.
2021. "GREET." Greenhouse gases, Regulated Emissions, and Energy use in Technologies
Model ® (2021 Excel). Computer Software, Web. doi:10.11578/GREET-Excel-
2021/dc.20210902.1. U.S. Department of Energy (DOE) Office of Energy Efficiency and
Renewable Energy (EERE), October 11.
IPCC. 2014. "Climate Change 2014: Synthesis Report. Contribution of Working Groups I, II and
III to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change." Climate
Change 2014. IPCC. 87.
Mo, Tiffany. 2024. "Revisions to MOVES for Air Quality Modeling to support the FRM for the
Multi-Pollutant Emissions Standards for Model Years 2027 and Later Light-Duty and Medium-
Duty Vehicles ." Memorandum to Docket.
Mosbrucker, Kristen. 2021. "Without a buyer, Shell may convert shuttered Convent refinery into
alternative fuels facility." The Advocate. October 14. Accessed 3 7, 2024.
https://www.theadvocate.com/baton_rouge/news/business/without-a-buyer-shell-may-convert-
shuttered-convent-refinery-into-alternative-fuel-facility/article_54ff85f2-2dl8-llec-af75-
13fba5943b71.html.
NHTSA. 2023. "88 FR 56128." Corporate Average Fuel Economy Standards for Passenger Cars
and Light Trucks for Model Years 2027-2032 and Fuel Efficiency Standards for Heavy-Duty
Pickup Trucks and Vans for Model Years 2030-2035. NHTSA, August 17.
—. 2022. "Corporate Average Fuel Economy Standards for Model Years 2024-2026 Passenger
Cars and Light Trucks." May 2.
8-67
-------
—. 2023. "Draft Technical Support Document." Corporate Average Fuel Economy Standards for
Passenger Cars and Light Trucks for Model Years 2027 and Beyond and Fuel Efficiency
Standards for Heavy-Duty Pickup Trucks and Vans for Model Years 2030 and Beyond. July.
Sherwood, Todd. 2024. "OMEGA Refinery Data Inputs." Memo to Docket No. EPA-HQ-OAR-
2022-0829.
U.S. EIA. 2023. Annual Energy Outlook. U.S. Energy Information Administration.
U.S. EPA. 2021. Dose-Response Assessment for Assessing Health Risks Associated With
Exposure to Hazardous Air Pollutants. September 29. Accessed November 15, 2022.
https://www.epa.gov/fera/dose-response-assessment-assessing-health-risks-associated-exposure-
hazardous-air-pollutants.
U.S. EPA. 2024. "Memo to the Docket: Air Quality Modeling Analysis for the Light- and
Medium- Duty 2027 Multipollutant Final Rule." EPA-HQ-OAR-2022-0829.
U.S. NHTSA. 2022. "2022 CAFE FRIA." Final Regulatory Impact Analysis: Final Rulemaking
for Model Years 2024-2026 Light-Duty Vehicle Corporate Average Fuel Economy Standards.
U.S. Department of Transportation, National Highway Traffic Safety Administration, March.
Wang, M., Lee, H. & Molburg, J. 2004. Allocation of energy use in petroleum refineries to
petroleum products. https://doi.org/10.1007/BF02978534, Int J LCA 9, 34-44 (2004).
8-68
-------
Chapter 9: Costs and Benefits of the Final Standards in OMEGA
This chapter presents the costs and benefits calculated within OMEGA. The results presented
here show the estimated annual costs, fuel savings, and benefits of the program for the indicated
calendar years (CY). Costs and benefits presented here are calculated relative to No Action
unless stated otherwise. The results also show the present-values (PV) and the equivalent
annualized values (AV) of costs and benefits for the calendar years 2027-2055 using 2, 3, and 7
percent discount rates. For the estimation of the stream of costs and benefits, we assume that the
MY 2032 standards apply to each year thereafter.
9.1 Costs
Vehicle technology costs are estimated in OMEGA using the technology cost inputs presented
in Chapter 2 of this RIA. Insurance, repair, maintenance, congestion, and noise costs are
estimated in OMEGA using the approaches described in Chapter 4 of this RIA. The resultant
costs associated with the final standards are presented in Table 9-1. Table 9-2 and Table 9-3
show the analogous results for Alternatives A and B, respectively.
Table 9-1: Costs associated with the final standards (billions of 2022 dollars)*.
Calendar
Year
Vehicle
Technology
Costs**
Insurance
Costs
Repair
Costs
Maintenance
Costs
Congestion
Costs
Noise
Costs
Sum of
Costs
2027
$2.6
$0.02
$0,027
$0,042
$0.0013
$0.000015
$2.7
2028
$7.3
$0.06
$0,081
$0,096
$0,027
$0.00041
$7.6
2029
$16
$0.15
$0.16
$0,089
$0.05
$0.00077
$17
2030
$23
$0.27
$0.26
-$0,027
$0,073
$0.0011
$24
2031
$29
$0.41
$0.35
-$0.35
$0,094
$0.0015
$29
2032
$30
$0.55
$0.38
-$0.9
$0.11
$0.0017
$30
2035
$55
$1.5
$0.7
-$3.3
$0.59
$0.0095
$54
2040
$50
$2.1
-$0.81
-$13
$1.3
$0,021
$40
2045
$46
$2.3
-$3.4
-$24
$1.9
$0.03
$23
2050
$42
$2.1
-$5.7
-$32
$2.3
$0,037
$9.4
2055
$38
$1.9
-$7.1
-$35
$2.4
$0.04
$0.59
PV2
$870
$33
-$40
-$300
$25
$0.41
$590
PV3
$760
$28
-$32
-$250
$21
$0.34
$530
PV7
$450
$15
-$12
-$110
$10
$0.17
$350
AV2
$40
$1.5
-$1.8
-$14
$1.2
$0,019
$27
AV3
$39
$1.4
-$1.6
-$13
$1.1
$0,018
$28
AV7
$37
$1.2
-$0.99
-$9.3
$0.83
$0,014
$29
* Negative values reflect decreased costs, or savings; positive values reflect increased costs.
** Costs exclude consideration of IRA battery tax credits (IRS 45X) and IRA purchase tax credits (IRS 30D and
45W).
As shown, estimated repair and maintenance costs, or reductions in those costs, are
significant. BEVs have considerably less maintenance needs than do ICE vehicles (see Chapter
4.3 of this RIA which shows BEVs having 30 to 40 percent less maintenance than ICE vehicles).
Congestion and noise costs are associated with rebound VMT as discussed in Chapter 4.3.8.
Insurance costs were not included in the NPRM so are new for the final analysis and are
discussed in Chapter 4.3.6 of this RIA.
9-1
-------
Table 9-2: Costs associated with Alternative A (billions of 2022 dollars)*.
Calendar
Year
Vehicle
Technology
Costs**
Insurance
Costs
Repair
Costs
Maintenance
Costs
Congestion
Costs
Noise
Costs
Sum of
Costs
2027
$16
$0,099
$0,091
$0,097
$0.0034
$0.000031
$16
2028
$25
$0.24
$0.23
$0.14
$0,066
$0,001
$26
2029
$32
$0.4
$0.36
-$0.0079
$0.12
$0.0019
$33
2030
$36
$0.56
$0.48
-$0.34
$0.18
$0.0029
$37
2031
$35
$0.7
$0.55
-$0.91
$0.24
$0.0038
$36
2032
$33
$0.82
$0.57
-$1.7
$0.27
$0.0043
$33
2035
$54
$1.6
$0.75
-$4.9
$0.73
$0,012
$52
2040
$49
$2.2
-$0.88
-$15
$1.4
$0,023
$37
2045
$45
$2.3
-$3.4
-$25
$1.9
$0,032
$21
2050
$43
$2.1
-$5.7
-$32
$2.3
$0,038
$9.5
2055
$39
$1.9
-$7.3
-$35
$2.4
$0.04
$0.2
PV2
$940
$35
-$40
-$320
$27
$0.44
$640
PV3
$820
$30
-$31
-$270
$23
$0.37
$580
PV7
$510
$17
-$12
-$130
$11
$0.18
$400
AV2
$43
$1.6
-$1.8
-$15
$1.2
$0.02
$29
AV3
$43
$1.6
-$1.6
-$14
$1.2
$0,019
$30
AV7
$41
$1.4
-$0.94
-$10
$0.92
$0,015
$33
* Negative values reflect decreased costs, or savings; positive values reflect increased costs.
** Costs exclude consideration of IRA battery tax credits (IRS 45X) and IRA purchase tax credits (IRS 30D and
45W).
Table 9-3: Costs associated with Alternative B (billions of 2022 dollars)*.
Calendar
Year
Vehicle
Technology
Costs**
Insurance
Costs
Repair
Costs
Maintenance
Costs
Congestion
Costs
Noise
Costs
Sum of
Costs
2027
$2.3
$0,018
$0,026
$0,042
$0.0016
$0.000019
$2.4
2028
$5.9
$0,047
$0,067
$0,083
$0,025
$0.00039
$6.1
2029
$15
$0.13
$0.14
$0.09
$0,046
$0.00074
$16
2030
$21
$0.24
$0.24
-$0.0077
$0,078
$0.0012
$22
2031
$24
$0.35
$0.33
-$0.29
$0.1
$0.0016
$24
2032
$27
$0.48
$0.36
-$0.79
$0.11
$0.0018
$27
2035
$42
$1.1
$0.35
-$3.2
$0.47
$0.0077
$41
2040
$40
$1.7
-$1.1
-$11
$1.1
$0,018
$30
2045
$39
$1.8
-$3.3
-$20
$1.7
$0,027
$19
2050
$35
$1.7
-$5.3
-$27
$2
$0,033
$6.3
2055
$30
$1.5
-$6.6
-$30
$2.2
$0,036
-$2.8
PV2
$710
$26
-$41
-$260
$22
$0.36
$450
PV3
$610
$22
-$32
-$210
$18
$0.3
$410
PV7
$360
$12
-$13
-$98
$8.9
$0.15
$270
AV2
$32
$1.2
-$1.9
-$12
$1
$0,017
$21
AV3
$32
$1.2
-$1.7
-$11
$0.96
$0,016
$21
AV7
$30
$0.98
-$1.1
-$8
$0.73
$0,012
$22
* Negative values reflect decreased costs, or savings; positive values reflect increased costs.
** Costs exclude consideration of IRA battery tax credits (IRS 45X) and IRA purchase tax credits (IRS 30D and
45W).
9-2
-------
9.2 Fuel Savings
The final standards are projected to reduce liquid fuel consumption (e.g., gasoline) while
simultaneously increasing electricity consumption. The estimated impacts on fuel and electricity
consumption are shown in Chapter 8.5 of this RIA.
The net effect of these changes in consumption for consumers is decreased liquid-fuel
expenditures or fuel savings and increased electricity expenditures. For more information of fuel
and electricity consumption, including other considerations like rebound driving, see RIA
Chapter 4.
Table 9-4 shows the undiscounted annual monetized fuel savings associated with the final
standards as well as the present value (PV) of those costs and equivalent annualized value (AV)
for the calendar years 2027-2055 using 2, 3, and 7 percent discount rates. We include here the
social costs associated with EVSE ports, as discussed in detail in Chapter 5.3. These reflect the
upfront costs associated with procuring and installing PEV charging infrastructure needed to
meet the anticipated electricity demand relative to the no action case. We include these EVSE
port costs in the net benefits presented in Chapter 9.6. Net benefits are determined using pre-tax
fuel savings since fuel taxes do not contribute to the value of the fuel and the EVSE port costs.
We present fuel taxes and other transfers below in Chapter 9.7.
Table 9-4: Pretax fuel savings and EVSE port costs associated with the final standards
(billions of 2022 dollars)*.
Calendar Year
Gasoline
Diesel
Eleetrieitv
EVSE Port
Sum
Savings
Savings
Savings
Costs
2027
$0.14
$0.0079
$0.02
$1.3
-$1.2
2028
$1.1
$0,013
-$0.24
$0.55 ;
$0.34
2029
$3.5;
$0,095
-$1.1
$2.3 f
$0.25
2030
$7.1 i
$0.3 1
-$2.5 !
$2.3 1
$2.7
203 1
$12
$0.52 T
-$4.3
$10 :
-$2.5
2032
$17 T
$0.86
-$6.4
$10 :
$0.93
2035
$39
$1.7 [
-$13
$10 :
$17
2040
$72 r
$2.6 :
-$21
$9 |
$44
2045
$94
$3.3 T
-$26 T
$12
$60
2050
$110
$3.9
-$27 :
$13
$74
2055
$120
$4.3
-$27 1
$8.6
$86
PV2
$1,300
$49
-$360
$190
$820
PV3
$1,100
$41
-$300
$160
$680
PV7
$560
$21
-$160
$96 :
$330
AV2
$61
$2.3 f
-SI 7
$9 j
$37
AV3
$58 I
$2.2 f
-$16
$8.8
$35
AV7
$46
$1.7 I '
-$13
$7.9 :
$26
* Negative electricity savings represent increased costs: the Sum column is the sum of Savings columns less the
EVSE Port Costs column.
9-3
-------
Table 9-5: Pretax fuel savings and EVSE port costs associated with Alternative A
(billions of 2022 dollars) *.
Calendar Year
Gasoline
Diesel
Electricity
EVSE Port Costs
Sum
2027
$2 r
$0,025
-$0.73 r
$1.3
-$0,052
2028
$5.2 ;
$0,078
-$1.8
$0.55 :
$3
2029
$9.4
$0.2 ]"'
-$3.2 I"
$2.3 r
$4.2
2030
$14
$0.42
-$4.8
$2.3
$7.7
203 1
$20 !
$0.64
-$6.7 1"
$10 :
$3.2
2032
$25 '
$0.97
-$8.8
$10 :
$7
2035
$47 j
$1.8
-$15
$10
$23
2040
$77 !
$2.8 1
-$22 ;
$9 !
$48
2045
$97 i
$3.5
-$27 i
$12
$62
2050
$110
$4
-$28 T
$13 :
$75
2055
$120
$4.3
-$28
$8.6
$86
PV2
$1,400
$52 !
-$390
$190
$900
PV3
$1,200
$43
-$330 :
$160
$750
PV7
$630 :
$23 :
-$180
$96 ;
$370
AV2
$66 ;
$2.4 :
-$18
$9
$41
AV3
$63 ]"
$2.3:
-$17 Y
$8.8
$39
AV7
$51
$1.8
-$15
$7.9 Y
$30
* Negative electricity savings represent increased costs: the Sum column is the sum of Savings columns less the
EVSE Port Costs column.
Table 9-6: Pretax fuel savings and EVSE port costs associated with Alternative B
(billions of 2022 dollars) *.
Calendar Year
Gasoline
Diesel
Electricity
EVSE Port Costs
Sum
2027
$0,094
$0.0063 i
$()"()41
$1.3
-$1.2
2028
$0.98
$0,011
-$0.17
$0.55 :
$0.26
2029
$3.4 :
$0,089
-$1.1
$2.3 1"
$0.16
2030
$6.7 ;
$0.29
-$2.3 |
$2.3 1"
$2.5
203 1
$11
$0.5 :
-$3.8 ;
$10
-$2.9
2032
$16
$0.81
-$5.9
$10 :
$0.41
2035
$34 ;
$1.6
-$12
$10 :
$14
2040
$61
$2.5 V"
-$18
$9 :
$37
2045
$80
$3.2 :
-$22 1
$12
$50
2050
$94
$3.7 T"
-$23 7"
$13 !
$62
2055
$100 :
$4
-$23 T
$8.6
$73
PV2
$1,100
$47:
-$310
$190
$690
PV3
$950
$39
-$260
$160
$570
PV7
$480
$20 T"
-$140
$96
$270
AV2
$52:
$2.1
-$14
$9 ;
$31
AV3
$50 :
$2*r
-$14
$8.8
$29
AV7
$39
$1.6
-$11
$7.9 ]
$22
* Negative electricity savings represent increased costs: the Sum column is the sum of Savings columns less the
EVSE Port Costs column.
9-4
-------
9.3 Non-Emission Benefits
Non-emission benefits are shown in Table 9-7 through Table 9-9 for the final standards,
Alternative A, and Alternative B, respectively. The drive value represents the value that
consumers place on the additional driving they may do resulting from the rebound effect. The
value is positive here which represents a benefit to consumers because we have estimated a small
amount of rebound driving relative to the no action case. The value of time spent refueling is
shown as a negative benefit, or disbenefit, because we estimate additional time spent refueling
relative to the no-action scenario. This is due to the additional BEV stock in the fleet and the
additional time required, using current estimates, to refuel a BEV relative to the refueling time
involved for an ICE vehicle. Energy security benefits (which are discussed further in Chapter 10)
are shown as positive because we estimate reductions in liquid-fuel consumption and
corresponding reductions in imported oil. Note that any benefits shown as negative values
represent disbenefits.
Table 9-7: Non-emission benefits associated with the final standards
(billions of 2022 dollars) *.
Calendar Year Drive Value Value of Time Ener^v Total
Spent Seeuritv
Refueling
2027
$0,002
$0.0022
$0.0047
$0.0089
2028
$0,042
$0,026
$0,032
$0.1
2029
$0,081
-$0,012
$0.1 ;
$0.17
2030
$0.12
-$(). 11
$0.21
$0.22
203 1
$0.16 :
-$0.27
$0.36 :
$0.26
2032
$0.2 1
-$0.47
$0.53 f
$0.26
2035
$1
-$0.59
$1.3
$1.7
2040
$2.3 ;
-$0.86
$2.5 i
$3.9
2045
$3.3:
-$1.1
$3.4
$5.6
2050
$4.2
-$1.4
$4
$6.8
2055
$4.7 1
-$1.7 j
$4.1
$7
PV2
$46
-$17 !
$47 i
$75
PV3
$38
-$15
$39
$62
PV7
$18
-$7.5 ;
$20 ]
$30
AV2
$2.1 :
-$0.8
$2.1 ;
$3.4
AV3
$2]
-$0.76 :
$21
$3.2
AV7
$1.5
-$0.61
$1.6
$2.5
* Positive values represent benefits while negative values represent disbenefits.
9-5
-------
Table 9-8: Non-emission benefits associated with Alternative A
(billions of 2022 dollars) *.
Calendar Year Drive Value Value of Time Ener^v Seeuritv Total
Spent Refueling
2027
$0.0052
$0,023
$0,052
$0.08
2028
$0.11
$0,099 :
$0.15
$0.35
2029
$0.21
$0.11
$0.27 1
$0.59
2030
SO.32
$0,039 :
$0.43
$0.79
203 1
$0.42
-$0,098
$0.59
$0.92
2032
SO.5
-$0.28
SO.77
$0.98
2035
$1.3
-$0.43
$1.5
$2.4
2040
$2.5 ;
-$0.75 1
$2.7 ;
$4.4
2045
$3.4
-$1
$3.5 !
$5.9
2050
$4.2
-$1.4
$4.1
$7
2055
$4.7 7"
-$1.7 !
$4.1
$7
PV2
$49
-$15
$50 I
$84
PV3
$41
-$13
$42
$70
PV7
$20;
-$6.2 ;
$21 :
$36
AV2
$2.2 ]
-SO.7
$2.3 ;
$3.8
AV3
$2.1
-$0.66
$2.2
$3.6
AV7
$1.71
-$0.5 '!
$1.8
$2.9
* Positive values represent benefits while negative values represent disbenefits.
Table 9-9: Non-emission benefits associated with Alternative B
(billions of 2022 dollars) *.
Calendar Year
Drive Value
Value of Time
Energy Seeuritv
Total
Spent Refueling
2027
$0.0024
$0.0015
$0.0035
$0.0074
2028
$0,043
$0,023
33 $0,027 ]' '
$0,093
2029
$0,082
-$0,018
$0,096 i
$0.16
2030
$0.14
-$0.12
$0.2 :
$0.22
203 1
$0.19
-$0.29
$0.33 1
$0.23
2032
$0.22 ;
' -$o.5 :
$0.48
$0.2
2035
SO.87
-$0.76 i
$1.1
$1.2
2040
$2 1
-$i.2 ;
$2.1
$2.9
2045
$3 !
-$1.5
$2.9 :
$4.3
2050
$3.8 i
-$1.8
$3.4
$5.4
2055
$4.3
-$2.2 1
$3.4
$5.5
PV2
$41
-$23 1
$39
$58
PV3
$34
-$19
$33 :
$48
PV7
$17 ;
-$9.8
$17 j
$23
AV2
$1.9
-$1.1
$1.8
$2.6
AV3
$1.8
-$1
$1.7 1 '
$2.5
AV7
$1.3
-$0.8
$1.3
$1.9
* Positive values represent benefits while negative values represent disbenefits.
9-6
-------
9.4 Benefits of GHG Reductions
Table 9-10 through Table 9-13 present the estimated annual, undiscounted climate benefits of
reduced CO2, CH4, N2O emissions under the final rule, and the annual total monetized climate
benefits (i.e., from all GHG reductions), using the three SC-CO2, SC-CH4, SC-N2O estimates
presented in U.S. EPA (EPA 2023f) for the stream of years beginning with the first year of rule
implementation, 2027, through 2055. Also shown are the present values (PV) and equivalent
annualized values (AV) associated with each of the three SC-GHG values. In this analysis, to
calculate the present and annualized values of climate benefits, EPA uses the same discount rate
as the near-term target Ramsey rate used to discount the climate benefits from future GHG
reductions. That is, future climate benefits estimated with the SC-GHG at the near-term 2 percent
Ramsey rate are discounted to 2027 using the same 2 percent rate.250 Table 9-14 and Table 9-15
present the benefits of reduced GHG emissions associated with Alternatives A and B,
respectively.
250 As discussed in U.S. EPA (EPA 2023f), the error associated with using a constant discount rate rather than the
certainty-equivalent rate path to calculate the present value of a future stream of monetized climate benefits is small
for analyses with moderate time frames (e.g., 30 years or less). U.S. EPA (EPA 2023f) also provides an illustration
of the amount that climate benefits from reductions in future emissions will be underestimated by using a constant
discount rate relative to the more complicated certainty-equivalent rate path.
9-7
-------
Table 9-10: Benefits of reduced CO2 emissions from the final standards.
(billions of 2022 dollars)
Calendar Year Ncar-lcrm Ramscv Discount Rale
2.5%
2.0%
1.5%
2027
$0,063
$0.1
$0.17
2028
$0.54
$0.87
$1.5
2029
$1.8
$3 ~
$5
2030
$3.9
$6.2
$10
203 1
$6.5
$10
17 $17 *
2032
$9.7
$15
$26
2033
$15
$23
$38
2034
$20
$31
$51
2035
$25
$40
$66
2036
$31
$49
$79
2037
$36
$57
$93
2038
$42
$65
$110
2039
$47
$73 .7
$120
2040
$53
$81
$130
2041
$57
$88
$140
2042
$62
$96
$150
2043
$67
$100
$160
2044
$72
$110
$170
2045
$76
$110
$180
2046
$79
$120
$190
2047
$83
$130
$200
2048
$86
$130
$200
2049
$89
$130
$210
2050
$92
$140
$220
2051
$94
$140
$220
2052
$96
$140
$220
2053
$97
$150
$230
2054
$98
$150
$230
2055
$100
$150
$230
PV
$940
$1,600
$2,800
AV
$46
$72
$120
Note: Climate benefits are based on changes (reductions) in C02 emissions and are calculated using updated
estimates of the SC- C02 from U.S. EPA (EPA 2023f). Climate benefits include changes in vehicle, EGU and
refinery C02 emissions.
9-8
-------
Table 9-11: Benefits of reduced CH4 emissions from the final standards.
(billions of 2022 dollars)
Calendar Year Ncar-lcrm Ramscv Discount Rale
2.5%
2.0%
1.5%
2027
-$0.00002 1
-$0.000026
-$0.000035
2028
-$0.000048
-$0.00006
-$0.00008
2029
$0.000023
$0.000028
$0.000038
2030
$0.00012
$0.00015
$0.0002
203 1
$0.00023
$0.00028
$0.00037
2032
$0.00053
$0.00065
$0.00085
2033
$0.0013
$0.0016
$0.0021
2034
$0.0023
$0.0028
$0.0037
2035
$0.0035
$0.0043
$0.0055
2036
$0.0048
$0.0059
$0.0076
2037
$0.0064
$0.0078
$0.01
2038
$0.0082
$0.01
$0,013
2039
$0.01
$0,012
$0,016
2040
$0,012
$0,015
$0,019
2041
$0,014
$0,017
$0,022
2042
$0,016
$0,019
$0,025
2043
$0,018
$0,022
$0,028
2044
$0.02
$0,024
$0.03 1
2045
$0,022
$0,027
$0,034
2046
$0,024
$0,029
$0,036
2047
$0,026
$0.03 1
$0,039
2048
$0,028
$0,033
$0,041
2049
$0,029
$0,035
$0,043
2050
$0.03
$0,036
$0,045
2051
$0.031
$0,038
$0,047
2052
$0,032
$0,039
$0,048
2053
$0,033
$0.04
$0,049
2054
$0,034
$0.04
$0.05
2055
$0,035
$0,041
$0.051
PV
$0.26
$0.35
$0.48
AV
$0,013
$0,016
$0,021
Note: Climate benefits arc based on changes (reductions) in CH4 emissions and arc
calculated using updated estimates of the SC-CH4 from U.S. EPA (EPA 20230. Climate
benefits include changes in vehicle. EGU and refinery CH4 emissions.
9-9
-------
Table 9-12: Benefits of reduced N2O emissions from the final standards
(billions of 2022 dollars).
Calendar Year Ncar-lcrm Ramscv Discount Rale
2.5%
2.0%
1.5%
2027
$0.0003
$0.00045
$0.0007
2028
$0,002
$0,003
$0.0047
2029
$0.0081
$0,012
$0,019
2030
$0,019
$0,029
$0,045
203 1
$0,033
$0,049
$0,075
2032
$0.051
$0,075
$0.12
2033
$0,079
$0.12
$0.18
2034
$0.11
$0.16
$0.24
2035
$0.14
$0.2
$0.3 1
2036
$0.17
$0.25
$0.38
2037
$0.2
$0.29
$0.44
2038
$0.23
$0.33
$0.51
2039
$0.26
$0.38
$0.57
2040
$0.29
$0.42
$0.63
2041
$0.32
$0.46
$0.69
2042
$0.34
$0.49
$0.74
2043
$0.37
$0.53
$0.8
2044
$0.4
$0.57
$0.85
2045
$0.42
$0.6
$0.9
2046
$0.44
$0.63
$0.94
2047
$0.46
$0.66
$0.97
2048
$0.48
$0.68
$1
2049
$0.5
$0.7
$1
2050
$0.51
$0.73
$1.1
2051
$0.53
$0.74
$1.1
2052
$0.54
$0.76
$1.1
2053
$0.55
SO.78
$1.1
2054
$0.56
$0.79
$1.1
2055
$0.57
$0.8
$1.2
PV
$5.2
$8.2
$13
AV
$0.26
$0.38
$0.58
Note: Climate benefits arc based on changes (reductions) in N2O emissions and arc
calculated using updated estimates of the SC-N2O from U.S. EPA (EPA 2023f). Climate
benefits include changes in vehicle. EGU and refinery N2O emissions.
9-10
-------
Table 9-13: : Benefits of reduced GHG emissions from the final standards.
Calendar Year Ncar-lcrm Ramscv Discount Rale
2.5%
2.0%
1.5%
2027
$0,063
$0.1
$0.17
2028
$0.54
$0.87
$1.5
2029
$1.9
[I $3 j
$5
2030
$3.9
$6.2
$10
203 1
$6.6
$10
$17 '
2032
$9.8
$15
$26
2033
$15
$23
$38
2034
$20
$31
$52
2035
$26
$40
$66
2036
$31
$49
$80
2037
$37
$57
$93
2038
$42
$65
$110
2039
$48
III I' $74 I_
$120
2040
$53
$82
$130
2041
$58
$89
$140
2042
$63
$96
$150
2043
$67
$100
$160
2044
$72
$110
$170
2045
$76
$120
$180
2046
$80
$120
$190
2047
$83
$130
$200
2048
$87
$130
$210
2049
$90
$130
$210
2050
$92
$140
$220
2051
$94
$140
$220
2052
$96
$140
$220
2053
$98
$150
$230
2054
$99
$150
$230
2055
$100
$150
$230
PV
$950
$1,600
$2,800
AV
$46
$72
$120
Note: Climate benefits arc based on changes (reductions) in GHG emissions and arc
calculated using updated estimates of the SC-GHG from U.S. EPA (EPA 2023f). Climate
benefits include changes in vehicle. EGU and refinery GHG emissions.
9-11
-------
Table 9-14: Benefits of reduced GHG emissions from Alternative A.
Calendar Year Ncar-lcrm Ramscv Discount Rale
2.5%
2.0%
1.5%
2027
$0.85
$1.4
$2.3
2028
$2.5
$3.9
$6.6
2029
$5
$8
$13
2030
$7.9
$13
$21
203 1
$11
$18
$29
2032
$15
$23
$38
2033
$19
$31
$50
2034
$25
$39
$64
2035
$30
$47llZ.
$78
2036
$36
$56
$91
2037
$41
$64
$100
2038
$46
$72
$120
2039
$52
$80
$130
2040
$57
$88
$140
2041
$61
$94
$150
2042
$66
$100
$160
2043
$70
$110
$170
2044
$75
$110
$180
2045
$78
$120
$190
2046
$82
$120
$200
2047
$85
$130
$200
2048
$88
$130
$210
2049
$91
$140
$220
2050
$94
$140
$220
2051
$96
$140
$220
2052
$98
$150
$230
2053
$99
$150
$230
2054
$100
$150
$230
2055
$100
$150
$230
PV
$1,000
$1,700
$2,900
AV
$49
$77
$130
Note: Climate benefits arc based on changes (reductions) in GHG emissions and arc
calculated using updated estimates of the SC-GHG from U.S. EPA (EPA 20230- Climate
benefits include changes in vehicle. EGU and refinery GHG emissions.
9-12
-------
Table 9-15: Benefits of reduced GHG emissions from Alternative B.
Calendar Year Ncar-lcrm Ramscv Discount Rale
2.5%
2.0%
1.5%
2027
$0,043
$0,068
$0.12
2028
$0.47
$0.76
$1.3
2029
$1.7
$2.8
$4.7
2030
$3.7
$5.9
$9.8
203 1
$6.1
$9.6
$16
2032
$9.1
$14
$24
2033
$13
$21
$35
2034
$18
$28
$46
2035
$22
$35
$58
2036
$27
$42
$69
2037
$31
$49
$80
2038
$36
$55
$90
2039
$40
$62
$100
2040
$44
$68
$110
2041
$48
! 2 $74
$120
2042
$52
$80
$130
2043
$56
$86
$140
2044
$60
$92
$150
2045
$64
$97
$150
2046
$67
$100
$160
2047
$70
$110
$170
2048
$73
$110
$170
2049
$76
$110
$180
2050
$78
$120
$180
2051
$79
$120
$190
2052
$81
$120
$190
2053
$82
$120
$190
2054
$83
$120
$190
2055
$84
$120
$190
PV
$800
$1,300
$2,300
AV
$39
$61
$100
Note: Climate benefits arc based on changes (reductions) in GHG emissions and arc
calculated using updated estimates of the SC-GHG from U.S. EPA (EPA 20230- Climate
benefits include changes in vehicle. EGU and refinery GHG emissions.
Unlike many environmental problems where the causes and impacts are distributed more
locally, GHG emissions are a global externality making climate change a true global challenge.
GHG emissions contribute to damages around the world regardless of where they are emitted.
Because of the distinctive global nature of climate change, in the RIA for this final rule the EPA
centers attention on a global measure of climate benefits from GHG reductions. Consistent with
all IWG recommended SC-GHG estimates to date, the SC-GHG values presented in Section 6
provide a global measure of monetized damages from CO2, CH4,and N2O and Table 9-10
through Table 9-12 present the monetized global climate benefits of the CO2, CH4, and N2O
emission reductions expected from the final rule. This approach is the same as that taken in EPA
9-13
-------
regulatory analyses from 2009 through 2016 and since 2021. It is also consistent with OMB
Circular A-4 guidance that states when a regulation is likely to have international effects, "these
effects should be reported."251 EPA also notes that EPA's cost estimates in RIAs, including the
cost estimates contained in this RIA, regularly do not differentiate between the share of
compliance costs expected to accrue to U.S. firms versus foreign interests, such as to foreign
investors in regulated entities.252 A global perspective on climate effects is therefore consistent
with the approach EPA takes on costs. There are many reasons, as summarized in this section—
and as articulated by OMB and in IWG assessments (IWG 2010) (IWG 2013) (IWG 2016a)
(IWG 2016b) (IWG 2021), the 2015 Response to Comments (IWG 2015), and in detail in U.S.
EPA (EPA 2023f) and in Appendix A of the Response to Comments document for the December
2023 Final Oil and Gas NSPS/EG Rulemaking—why the EPA focuses on the global value of
climate change impacts when analyzing policies that affect GHG emissions.
International cooperation and reciprocity are essential to successfully addressing climate
change, as the global nature of greenhouse gases means that a ton of GHGs emitted in any other
country harms individuals in the U.S. just as much as a ton emitted within the territorial U.S.
Assessing the benefits of U.S. GHG mitigation activities requires consideration of how those
actions may affect mitigation activities by other countries, as those international mitigation
actions will provide a benefit to U.S. citizens and residents by mitigating climate impacts that
affect U.S. citizens and residents. This is a classic public goods problem because each country's
reductions benefit everyone else, and no country can be excluded from enjoying the benefits of
other countries' reductions. The only way to achieve an efficient allocation of resources for
emissions reduction on a global basis—and so benefit the U.S. and its citizens and residents—is
for all countries to base their policies on global estimates of damages. A wide range of scientific
and economic experts have emphasized the issue of international cooperation and reciprocity as
support for assessing global damages of GHG emission in domestic policy analysis. Using a
global estimate of damages in U.S. analyses of regulatory actions allows the U.S. to continue to
actively encourage other nations, including emerging major economies, to also assess global
climate damages of their policies and to take steps to reduce emissions. For example, many
251 While OMB Circular A-4 recommends that international effects be reported separately, the guidance also
explains that "[d]ifferent regulations may call for different emphases in the analysis, depending on the nature and
complexity of the regulatory issues." Circular A-4 (2023) states that "[i]n certain contexts, it may be particularly
appropriate to include effects experienced by noncitizens residing abroad in your primary analysis. Such contexts
include, for example, when:
• assessing effects on noncitizens residing abroad provides a useful proxy for effects on U. S. citizens and residents
that are difficult to otherwise estimate;
• assessing effects on noncitizens residing abroad provides a useful proxy for effects on U.S. national interests that
are not otherwise fully captured by effects experienced by particular U.S. citizens and residents (e.g., national
security interests, diplomatic interests, etc.);
• regulating an externality on the basis of its global effects supports a cooperative international approach to the
regulation of the externality by potentially inducing other countries to follow suit or maintain existing efforts; or
• international or domestic legal obligations require or support a global calculation of regulatory effects".
252 For example, in the RIA for the 2018 Proposed Reconsideration of the Oil and Natural Gas Sector Emission
Standards for New, Reconstructed, and Modified Sources, the EPA acknowledged that some portion of regulatory
costs will likely "accru[e] to entities outside U.S. borders" through foreign ownership, employment, or consumption
(EPA 2018). In general, a significant share of U.S. corporate debt and equities are foreign-owned, including in the
oil and gas industry.
9-14
-------
countries and international institutions have already explicitly adapted the global SC-GHG
estimates used by EPA in their domestic analyses (e.g., Canada, Israel) or developed their own
estimates of global damages (e.g., Germany), and recently, there has been renewed interest by
other countries to update their estimates since the draft release of the updated SC-GHG estimates
presented in the December 2022 Oil and Gas NSPS/EG Supplemental Proposal RIA.253 Several
recent studies have empirically examined the evidence on international GHG mitigation
reciprocity, through both policy diffusion and technology diffusion effects. See U.S. EPA (EPA
2023f) for more discussion.
For all of these reasons, the EPA believes that a global metric is appropriate for assessing the
climate benefits of avoided GHG emissions in this final RIA. In addition, as emphasized in the
(National Academies 2017) recommendations, "[i]t is important to consider what constitutes a
domestic impact in the case of a global pollutant that could have international implications that
impact the United States." The global nature of GHG pollution and its impacts means that U.S.
interests are affected by climate change impacts through a multitude of pathways and these need
to be considered when evaluating the benefits of GHG mitigation to U.S. citizens and residents.
The increasing interconnectedness of global economies and populations means that impacts
occuring outside of U.S. borders can have significant impacts on U.S. interests. Examples of
affected interests include direct effects on U.S. citizens and assets located abroad, international
trade, and tourism, and spillover pathways such as economic and political destabilization and
global migration that can lead to adverse impacts on U.S. national security, public health, and
humanitarian concerns. Those impacts point to the global nature of the climate change problem
and are better captured within global measures of the social cost of greenhouse gases.
In the case of these global pollutants, for the reasons articulated in this section, the assessment
of global net damages of GHG emissions allows EPA to fully disclose and contextualize the net
climate benefits of GHG emission reductions expected from this final rule. As EPA explained in
the final Oil and Gas NSPS/EG rule, EPA disagrees with public comments received on the
December 2022 Oil and Gas NSPS/EG Supplemental Proposal that suggested that the EPA can
or should use a metric focused on benefits resulting solely from changes in climate impacts
occuring within U.S. borders.254 The global models used in the SC-GHG modeling described
above do not lend themselves to be disaggregated in a way that could provide sufficiently robust
information about the distribution of the rule's climate benefits to citizens and residents of
particular countries, or population groups across the globe and within the U.S. Two of the
models used to inform the damage module, the GIVE and DSCIM models, have spatial
resolution that allows for some geographic disaggregation of future climate impacts across the
world. This permits the calculation of a partial GIVE and DSCIM-based SC-GHG measuring the
damages from four or five climate impact categories projected to physically occur within the
253 In April 2023, the government of Canada announced the publication of an interim update to their SC-GHG
guidance, recommending SC-GHG estimates identical to the EPA's updated estimates presented in the December
2022 Supplemental Proposal RIA. The Canadian interim guidance will be used across all Canadian federal
departments and agencies, with the values expected to be finalized by the end of the year. See more at
https://www.canada.ca/en/environment-climate-change/services/climate-change/science-research-data/social-cost-
ghg.html
254 EPA is noting this for informational purposes only. We are not reopening the Oil and Gas NSPS/EG Rule in this
proceeding.
9-15
-------
U.S., respectively, subject to caveats. As discussed at length in U.S. EPA (EPA 2023f), these
damage modules are only a partial accounting and do not capture all of the pathways through
which climate change affects public health and welfare. For example, this modeling omits most
of the consequences of changes in precipitation, damages from extreme weather events (e.g.,
wildfires), the potential for nongradual damages from passing critical thresholds (e.g., tipping
elements) in natural or socioeconomic systems, and non-climate mediated effects of GHG
emissions other than CO2 fertilization (e.g., tropospheric ozone formation due to CH4 emissions).
Thus, they only cover a subset of potential climate change impacts. Furthermore, as discussed at
length in U.S. EPA (EPA, 2023f), the damage modules do not capture spillover or indirect
effects whereby climate impacts in one country or region can affect the welfare of residents in
other countries or regions—such as through the effect of climate change on international
markets, trade, tourism, and other activities. Supply chain disruptions are a prominent pathway
through which U.S. business and consumers can be affected by climate change impacts abroad.
Additional climate change-induced international spillovers can occur through pathways such as
damages across transboundary resources, economic and political destabilization, and global
migration that can lead to adverse impacts on U.S. national security, public health, and
humanitarian concerns.
Additional modeling efforts can and have shed further light on some omitted damage
categories. For example, the Framework for Evaluating Damages and Impacts (FrEDI) is an
open-source modeling framework developed by the EPA255 to facilitate the characterization of
net annual climate change impacts in numerous impact categories within the contiguous U.S. and
monetize the associated distribution of modeled damages (Sarofim, et al. 2021) (EPA 2021). The
additional impact categories included in FrEDI reflect the availability of U.S.-specific data and
research on climate change effects. As discussed in U.S. EPA (EPA 2023f) results from FrEDI
show that annual damages resulting from climate change impacts within the contiguous U.S.
(CONUS) (i.e., excluding Hawaii, Alaska, and U.S. territories) and for impact categories not
represented in GIVE and DSCIM are expected to be substantial. As discussed in U.S. EPA (EPA
2021), results from FrEDI show that annual damages resulting from climate change impacts
within the contiguous U.S. (CONUS) (i.e., excluding Hawaii, Alaska, and U.S. territories) and
for impact categories not represented in GIVE and DSCIM are expected to be substantial. For
example, FrEDI estimates a partial SC-CO2 of $41/mtC02 for damages physically occurring
within CONUS for 2030 emissions (under a 2 percent near-term Ramsey discount rate) (Hartin,
McDuffie, et al. 2023), compared to a GIVE and DSCIM-based U.S.-specific SC-CO2 of
$18/mtC02 and $16/mtC02, respectively, for 2030 emissions (2022 USD). While the FrEDI
results help to illustrate how monetized damages physically occurring within CONUS increase as
more impacts are reflected in the modeling framework, they are still subject to many of the same
limitations associated with the DSCIM and GIVE damage modules, including the omission or
255 The FrEDI framework and Technical Documentation have been subject to a public review comment period and
an independent external peer review, following guidance in the EPA Peer-Review Handbook for Influential
Scientific Information (ISI). Information on the FrEDI peer-review is available at the EPA Science Inventory.
9-16
-------
partial modeling of important damage categories.256'257 Finally, none of these modeling efforts—
GIVE, DSCIM, and FrEDI—reflect non-climate mediated effects of GHG emissions experienced
by U.S. populations (other than CO2 fertilization effects on agriculture). In addition to its climate
impacts, methane also contributes to the chemical formation of tropospheric ozone, which
contributes to mortality. One recent paper on this effect, (McDuffie, et al. 2023) estimated the
monetized increase in respiratory-related human mortality risk from the ozone produced from a
marginal pulse of methane emissions. Using the socioeconomics from the RFF-SPs and the 2
percent near-term Ramsey discounting approach, this additional health risk to U.S. populations is
on the order of approximately $360/mtCH4 (2022 USD) for 2030 emissions.
Applying the U.S.-specific partial SC-GHG estimates derived from the multiple lines of
evidence described above to the GHG emissions reduction expected under the final rule would
yield substantial benefits. For example, the present value of the climate benefits of the final rule
as measured by FrEDI from climate change impacts in CONUS are estimated to be $231 billion
(under a 2 percent near-term Ramsey discount rate).258 However, the numerous explicitly omitted
damage categories and other modeling limitations discussed above and throughout U.S. EPA
(EPA 2023f) make it likely that these estimates underestimate the benefits to U.S. citizens and
residents of the GHG reductions from the final rule; the limitations in developing a U.S.-specific
estimate that accurately captures direct and spillover effects on U.S. citizens and residents further
demonstrates that it is more appropriate to use a global measure of climate benefits from GHG
reductions. The EPA will continue to review developments in the literature, including more
robust methodologies for estimating the magnitude of the various damages to U.S. populations
from climate impacts and reciprocal international mitigation activities, and explore ways to
better inform the public of the full range of GHG impacts.
256 Another method that has produced estimates of the effect of climate change on U.S.-specific outcomes uses a top-
down approach to estimate aggregate damage functions. Published research using this approach include total-
economy empirical studies that econometrically estimate the relationship between GDP and a climate variable,
usually temperature. As discussed in U.S. EPA (EPA 2023f) the modeling framework used in the existing published
studies using this approach differ in important ways from the inputs underlying the SC-GHG estimates described
above (e.g., discounting, risk aversion, and scenario uncertainty) and focus solely on SC-CO2. Hence, we do not
consider this line of evidence in the analysis for this RIA. Updating the framework of total-economy empirical
damage functions to be consistent with the methods described in this RIA and ibid, would require new analysis.
Finally, because total-economy empirical studies estimate market impacts, they do not include any non-market
impacts of climate change (e.g., heat related mortality) and therefore are also only a partial estimate. The EPA will
continue to review developments in the literature and explore ways to better inform the public of the full range of
GHG impacts.
257 FrEDI estimates a partial SC-CH4 (N2O) of $660/mtCH4 ($12,000/mtN20) for damages physically occurring
within CONUS for 2030 emissions (under a 2 percent near-term Ramsey discount rate) (Hartin, et al. 2023)
compared to a GIVE and DSCIM-based U.S.-specific SC-CH4 of $310/mtCH4 ($5,600/mtN20) and $84/mtCH4
($4,300/mtN20), respectively, for 2030 emissions (2022 USD).
258 DSCIM and GIVE use global damage functions. Damage functions based on only U.S.-data and research, but not
for other parts of the world, were not included in those models. FrEDI does make use of some of this U.S.-specific
data and research and as a result has a broader coverage of climate impact categories.
9-17
-------
9.5 Criteria Air Pollutant Benefits
For the analysis of the final standards, we use the same reduced-form "benefit-per-ton" (BPT)
approach used in the proposal to estimate the monetized PIVh.s-related health benefits of the final
standards, except the constant dollar year has been updated from year 2020 dollars to year 2022
dollars. As described in RIA Chapter 6.4, the BPT approach monetizes avoided premature deaths
and illnesses that are expected to occur as a result of reductions in directly-emitted PM2.5 and
PM2.5 precursors attributable to the standards. The upstream BPT estimates used in this final rule
are the same as those used in the proposal and were also updated to year 2022 dollars. A chief
limitation to using PM2.5-related BPT values is that they do not reflect benefits associated with
reducing ambient concentrations of ozone, direct exposure to NO2, or exposure to mobile source
air toxics, nor do they account for improved ecosystem effects or visibility. The estimated
benefits of the final standards would be larger if we were able to monetize these unquantified
benefits at this time.
Using the BPT approach, we estimate the annualized value of PM2.5-related benefits of the
final standards to be $5.3 to $10 billion at a 3% discount rate and $3.6 to $7.2 billion at a 7%
discount rate. Benefits are reported in year 2022 dollars and reflect the PM2.5-related benefits
associated with reductions in NOx, SO2, and direct PM2.5 emissions. Because premature
mortality typically constitutes the vast majority of monetized benefits in a PM2.5 benefits
assessment, we present PM benefits based on risk estimates reported from two different long-
term exposure studies using different cohorts to account for uncertainty in the benefits associated
with avoiding PM-related premature deaths: the National Health Interview Survey (NHIS) cohort
study (Pope III et al. 2019) and an extended analysis of the Medicare cohort (Wu et al. 2020).
Table 9-16 presents the annual, undiscounted PM2.5-related health benefits estimated for the
stream of years beginning with the first year of rule implementation, 2027, through 2055 for the
final standards. Benefits are presented by source (onroad and upstream) and are estimated using
either a 3 percent or 7 percent discount rate to account for annual avoided health outcomes that
are expected to accrue over more than a single year (the "cessation" lag between the change in
PM exposures and the total realization of changes in health effects). Table 9-16 also shows the
present and annualized values of PM2.5-related benefits for the final program between 2027 and
2055 (discounted back to 2027). Table 9-17 and Table 9-18 present the results for each of the
alternatives.
We use a constant 3-percent and 7-pecent discount rate to calculate present and annualized
values in Table 9-17, consistent with current applicable OMB Circular A-4 guidance. For the
purposes of presenting total net benefits (see RIA Chapter 9.6), we also use a constant 2-percent
discount rate to calculate present and annualized values. We note that we do not currently have
BPT estimates that use a 2-percent discount rate to account for cessation lag. If we discount the
stream of annual benefits in Table 9-17 based on the 3-percent cessation lag BPT using a
constant 2-percent discount rate, the present value of total PM2.5-related benefits would be $120
to $240 billion and the annualized value of total PM2.5-related benefits would be $6.4 to $13
billion, depending on the assumed long-term exposure study of PM2.5-related premature
mortality risk.
This analysis includes many data sources that are each subject to uncertainty, including
projected emission inventories, air quality data from models, population data, population
estimates, health effect estimates from epidemiology studies, economic data, and assumptions
9-18
-------
regarding the future state of the world (i.e., regulations, technology, and human behavior). When
compounded, even small uncertainties can greatly influence the size of the total quantified
benefits. There are also inherent limitations associated with using the BPT approach. Despite
these uncertainties, we believe the criteria pollutant benefits presented here are our best estimate
of benefits absent air quality modeling and we have confidence in the BPT approach and the
appropriateness of relying on BPT health estimates for this rulemaking. Please refer to RIA
Chapter 6 for more information on the uncertainty associated with the benefits presented here.
Table 9-16: Monetized PM2.5 health benefits of onroad and upstream emissions reductions
associated with the final standards (billions of 2022 dollars).
Calendar Total Onroad Total Upstream Total Benefits
Year
3% Diseount
7% Diseount
3% Diseount
7% Diseount
3% Diseount
7%
Rate
Rate
Rate
Rate
Rate
Diseount
Rate
2027
0.078 to 0.17
0.07 to 0.15
-0.0087 to-0.019
; -0.0078 to-0.017 :
0.069 to 0.15
f 0.062 to 0.13
2028
0.21 to 0.45
0.19 to 0.41
-0.034 to-0.072
-0.03 to -0.064
0.18 to 0.38
f 0.16 to 0.34
2029
0.38 to 0.81
0.34 to 0.73
-0.064 to-0.14
-0.057 to-0.12
0.31 to 0.67
: 0.28 to 0.61
2030
0.74 to 1.5
f 0.66 to L4
"" -0.12 to-0.25
: -0.11 to-0.23
0.61 to 1.3
! 0.55 to 1.1
2031
1 to 2.1
0.93 to 1.9
-0.2 to-0.42
-0.18 to-0.38
0.84 to 1.7
• 0.75 to 1.6
2032 ;
1.3 to 2.8
1.2 to 2.5
-0.26 to-0.53
-0.23 to -0.47
1.1 to 2.2
0.98 to 2
2033
1.8 to 3.6
1.6 to 3.3
-0.29 to-0.59
-0.26 to-0.53
1.5 to 3
1.3 to 2.7
2034
2.2 to 4.5
1 1.9 to 4
-0.29 to-0.59
-0.26 to-0.53
1.9 to 3.9
: 1.7 to 3.5
2035 T
2.9 to 5.9
! 2.6 to 5.3
-0.28 to-0.55
-0.25 to-0.5
2.6 to 5.3
2.4 to 4.8
2036
3.3 to 6.7
3 to 6.1 i
-0.23 to-0.45
-0.21 to-0.4
3.1 to 6.3
2.8 to 5.7
2037
3.8 to 7.7
3.4 to 6.9
-0.15 to-0.29
-0.13 to -0.26
3.7 to 7.4
3.3 to 6.7
2038 :
4.3 to 8.7
: 3.9 to 7.8
-0.053 to-0.096
-0.049 to-0.085
4.3 to 8.6
3.8 to 7.8
2039
4.8 to 9.7
4.3 to 8.7
0.058 to 0.13
IIH5I ion.12
4.9 to 9.8
: 4.4 to 8.9
2040
6 to 12
5.4 to 11
' 0.21 to 0.43
0.19 to 0.38
6.2 to 12
5.5 to 11
2041
6.4 to 13
5.8 to 11
0.3 to 0.6
0.27 to 0.54
6.7 to 13
6.1 to 12
2042
6.9 to 14
; 6.2 to 12
0.39 to 0.8
0.35 to 0.71
7.3 to 14
6.5 to 13
2043
' 7.3 to 14
i 6.5 to 13
0.49 to 1
0.45 to 0.89
7.7 to 15
7 to 14
2044
7.6 to 15
6.8 to 14 ;
11.6 io 1.2
0.54 to 1.1
8.2 to 16
7.4 to 15
2045
8.7 to 17
] 7.8 to 15
0.7 to 1.4
0.63 to 1.3
9.4 to 18
8.5 to 17
2046
9.1 to 18
8.1 to 16
0.77 to 1.5
0.69 to 1.4
9.8 to 19
8.8 to 17
2047
9.4 to 18
8.4 to 16
0.83 to 1.7
0.75 to 1.5
10 to 20
f 9.1 to 18
2048
9.6 to 19
8.6 to 17 1
0.89 to 1.8
0.8 to 1.6
10 to 21
" 9.4 to 19
2049 i
9.8 to 19
8.8 to 17
0.95 to 1.9
: 0.85 to 1.7
11 to 21
9.7 to 19
2050
11 to 21
9.7 to 19
0.99 to 2
0.9 to 1.8
12 to 23
1 1 to 21
2051
11 to 21
9.8 to 19
1 to 2
0.9 to 1.8
12 to 23
11 to 21
2052
11 to 21
10 to 19
1 to 2
0.91 to 1.8
12 to 23
11 to 21
2053
11 to 22
10 to 19
1 to 2
0.91 to 1.8
12 to 24
11 to 21
2054
11 to 22
10 to 20 i
1 to 2
0.91 to 1.8
12 to 24
11 to 21
2055
12 to 23
11 to 21
1 to 2
0.91 to 1.8
13 to 25
12 to 23
Present Value
97 to 190
13 io 86
4.6 to 9.3
f 1.3 to 2.6 *
100 to 200
L. 1° 88
Annualized
Value
5.1 to 10
3.5 to 7
0.24 to 0.49
0.11 to 0.22
5.3 to 10
3.6 to 7.2
The benefits in this table reflect two separate but equally plausible premature mortality estimates derived from the Medicare study (Wu et al.
2020) or the NHIS study (Pope III et al. 2019), respectively. All benefits estimates are rounded to two significant figures. Annual benefit
values presented here are not discounted. Negative values are health disbenefits related to increases in estimated emissions. The present value
of benefits is the total aggregated value of the series of discounted annual benefits that occur between 2027-2055 (in 2022 dollars) using either
a 3 percent or 7 percent discount rate. The upstream impacts associated with the standards presented here include health benefits associated
with reduced criteria pollutant emissions from refineries and health disbenefits associated with increased criteria pollutant emissions from
EGUs. The benefits in this table also do not include the full complement of health and environmental benefits (such as health benefits related
to reduced ozone exposure) that, it quantified and monetized, would increase the total monetized benefits.
9-19
-------
Table 9-17: Monetized PM2.5 health benefits of onroad and upstream emissions reductions
associated with Alternative A (billions of 2022 dollars).
Calendar Total Onroad Total Upstream Total Benefits
Year
3% Diseount
7%
3% Diseount
7% Diseount
3% Diseount
7% Diseount
Rate
Diseount
Rate
Rate
Rate
Rate
Rate
2027
0.1 to 0.21
0.09 to 0.19
I -0.059 to-0.12
:' -0.053 to-0.11
; 0.042 to 0.09
0.037 to 0.081
2028
0.25 to 0.54
0.23 to 0.49
-0.15 to -0.33
-0.14tD-0.29
0.1 to 0.21
0.089 to 0.19
2029
0.43 to 0.93
0.39 to 0.83
-0.17 to-0.35
J -0.15 to-0.31
0.27 to 0.58
0.24 to 0.52
2030
0.8 to 1.7
0.72 to 1.5
-0.23 to-0.46
f -0.2 to -0.42
0.57 to 1.2
(1.52 to I.I
2031
1.1 to 2.3
1 to 2 1
-0.3 to -0.61
-0.27 to-0.55
0.81 to 1.7
0.73 to 1.5
2032
1.4 to 2.9
1.3 to 2.6
-0.33 to -0.68
-0.3 to-0.61
1.1 to 2.3
0.98 to 2
2033
1.8 to 3.8
1.6 to 3.4
-0.35 to-0.72
; -0.32to-0.65
1.5 to 3 1
1.3 to 2.8
2034
1 2.2 to 4.6
2 to 4.2
-0.33 to-0.67
-0.3 to -0.6
1.9 to 4
1.7 to 3.6
2035
3 to 6.1
2.7 to 5.5
-0.31 to-0.62
-0.28 to-0.56
2.7 to 5.5
2.4 to 4.9
2036
3.4 to 6.9
3.1 to 6.2
-0.24 to-0.48
-0.22 to-0.43
3.2 to 6.5
2.9 to 5.8
2037
3.9 to 7.9
3.5 to 7.1
-0.16 to-0.3
-0.14 to-0.27
! " 3.8 to 7.6
3.4 to 6.8
2038
4.4 to 8.9
4 to 8
; -0.047 to -0.083
T" -0.044 to -0.073
4.4 to 8.8
3.9 to 7.9
2039
4.9 to 9.9
4.4 to 8.9
] 0.074 to 0.16
0.065 to 0.15
5 to 10
4.5 to 9
2040
6.1 to 12
5.4 to 11
[ 0.24 to 0.49
0.21 to 0.44
6.3 to 12
5.7 to 11
2041
6.5 to 13 j'
5.9 to 12
0.33 to 0.66
0.29 to 0.6
6.8 to 14
6.2 to 12
2042
6.9 to 14
6.2 to 12
0.42 to 0.86
0.38 to 0.77
7.4 to 15
6.6 to 13
2043
7.3 to 14
6.6 to 13
' 0.52 to 1.1
0.47 to 0.95
7.8 to 16
7 to 14
2044
7.7 to 15
6.9 to 14
; 0.63 to 1.3
0.56 to 1.1
8.3 to 16
7.4 to 15
2045
8.8 to 17
7.9 to 15
0.73 to 1.5
' 0.65 to 1.3
9.5 to 19
8.5 to 17
2046
9.1 to 18
8.2 to 16
0.79 to 1.6
0.71 to 1.4
9.9 to 19
8.9 to 17
2047
9.4 to 18
8.4 to 17
0.85 to 1.7
0.76 to 1.5
10 to 20
9.2 to 18
2048
9.6 to 19 I
8.7 to 17
0.91 to 1.8
0.82 to 1.6
11 to 21
9.5 to 19
2049
9.8 to 19
8.8 to 17
0.97 to 1.9
! 0.87 to 1.7
1 1 to 21
9.7 to 19
2050
1 1 to 21
9.7 to 19
1 to 2
r 0.91 to 1.8
12 to 23
11 to 21
2051
11 tO 21 :
9.9 to 19
1 to 2.1
0.92 to 1.8
12 to 23
11 to 21
2052
11 to 21
10 to 19
1 to 2.1
0.92 to 1.8
12 to 24
11 to 21
2053
' 11 to 22
10 to 19
1 to 2.1
0.93 to 1.9
12 to 24
11 to 21
2054
11 to 22
10 to 20
1 to 2.1
0.92 to 1.8
12 to 24
11 to 21
12055'."'.'
f 12 to 23
11 to 21
1 to 2
[ 0.91 to 1.8
13 to 26
12 to 23
Present
Value
98 to 190
44 to 87
4.2 to 8.5
0.95 to 1.9
100 to 200
45 to 89
Annualized
Value
5.1 to 10
3.6 to 7.1
0.22 to 0.44
0.077 to 0.15
5.4 to 11
3.7 to 7.3
The benefits in this table reflect two separate but equally plausible premature mortality estimates derived from the Medicare study (Wu et
al. 2020) or the NHIS study (Pope III et al. 2019), respectively. All benefits estimates are rounded to two significant figures. Annual benefit
values presented here are not discounted. Negative values are health disbenefits related to increases in estimated emissions. The present
value of benefits is the total aggregated value of the series of discounted annual benefits that occur between 2027-2055 (in 2022 dollars)
using either a 3 percent or 7 percent discount rate. The upstream impacts associated with the standards presented here include health
benefits associated with reduced criteria pollutant emissions from refineries and health disbenefits associated with increased criteria
pollutant emissions from EGUs. The benefits in this table also do not include the full complement of health and environmental benefits
(such as health benefits related to reduced ozone exposure) that if quantified and monetized, would increase the total monetized benefits.
9-20
-------
Table 9-18: Monetized PM2.5 health benefits of onroad and upstream emissions reductions
associated with Alternative B (billions of 2022 dollars).
Calendar Total Onroad Total Upstream Total Benefits
Year
3% Diseount
7%
3% Diseount
7% Diseount
3% Diseount
7% Diseount
Rate
Diseount
Rate
Rate
Rate
Rate
Rate
2027
0.077 to 0.16
0.069 to 0.15
: -0.0074 to-0.016
: -0.0067 to-0.014
0.07 to 0.15
0.063 to 0.13
2028
0.21 to 0.45
0.19 to 0.4
: -0.025 to -0.053
-0.022 to-0.047
0.18 to 0.39
0.17 to 0.36
2029
0.38 to 0.8
0.34 to 0.72
-0.054 to-0.11
-0.049 to-0.1
0.32 to 0.69
0.29 to 0.62
2030
0.73 to 1.5
0.66 to 1.4
-0.1 to-0.21
i -0.091 to-0.19
0.63 to 1.3
0.57 to 1.2
2031
1 to 2.1
0.93 to 1.9
-0.16 to-0.34
-0.15 to-0.3
0.87 to 1.8
0.78 to 1.6
2032
1.3 to 2.8
1.2 to 2.5
-0.21 to -0.43
-0.19 to-0.38
1.1 to 2.3
1 to 2.1
2033
1.7 to 3.6
1.6 to 3.3
-0.25 to -0.52
-0.23 to -0.46
1.5 to 3.1
1.3 to 2.8
2034
2.1 to 4.4
1.9 to 4
-0.24 to-0.5
-0.22 to-0.45
1.9 to 3.9
1.7 to 3.5
2035
2.9 to 5.8
2.6 to 5.3
-0.22 to-0.43
-0.2 to-0.39
2.7 to 5.4
2.4 to 4.9
2036
3.3 to 6.7
3 to 6
-0.16 to-0.32
-0.15 to-0.29
3.1 to 6.4
2.8 to 5.7
2037
3.8 to 7.6
3.4 to 6.8
-0.095 to-0.18
: -0.087 to-0.16
3.7 to 7.4
3.3 to 6.7
2038
4.3 to 8.6
3.8 to 7.7
-0.014 to-0.017
-0.013 to-0.015
4.3 to 8.6
3.8 to 7.7
2039
4.8 to 9.6
4.3 to 8.6
0.079 to 0.17
0.07 to 0.15
4.8 to 9.8
4.3 to 8.8
2040
5.9 to 12
5.3 to 10
0.21 to 0.42
0.19 to 0.38
6.1 to 12
5.5 to 11
2041
6.3 to 13
5.7 to 11
0.28 to 0.57
0.25 to 0.51
6.6 to 13
6 to 12
2042
6.8 to 13
6.1 to 12
0.36 to 0.73
0.33 to 0.66
7.1 to 14
6.4 to 13
2043
7.1 to 14
6.4 to 13
0.45 to 0.9
0.4 to 0.81
7.6 to 15
6.8 to 14
2044
7.5 to 15
6.7 to 13
0.53 to 1.1
0.48 to 0.96
8 to 16
7.2 to 14
2045
8.6 to 17
7.7 to 15
0.62 to 1.2
0.56 to 1.1
9.2 to 18
8.3 to 16
2046
8.9 to 17
8 to 16
0.67 to 1.4
0.61 to 1.2
9.6 to 19
8.6 to 17
2047
9.2 to 18
8.3 to 16
0.73 to 1.5
0.66 to 1.3
9.9 to 19
8.9 to 18
2048
9.5 to 19
8.5 to 17
0.78 to 1.6
0.7 to 1.4
10 to 20
9.2 to 18
2049
9.7 to 19
8.7 to 17
0.83 to 1.7
0.74 to 1.5
11 to 21
9.4 to 19
2050
11 to 21
9.6 to 19
0.87 to 1.7
0.78 to 1.6
12 to 22
10 to 20
2051
11 to 21
9.7 to 19
0.87 to 1.7
0.79 to 1.6
12 to 23
10 to 20
2052
11 to 21
9.8 to 19
0.88 to 1.8
0.79 to 1.6
12 to 23
11 to 21
2053
11 to 21
9.9 to 19
0.88 to 1.8
0.79 to 1.6
12 to 23
11 to 21
2054
11 to 21
10 to 19
0.88 to 1.8
0.79 to 1.6
12 to 23
11 to 21
2055
12 to 23
11 to 21
0.87 to 1.8
0.79 to 1.6
13 to 25
12 to 22
Present
Value
96 to 190
43 to 85
4.3 to 8.6
1.3 to 2.6
100 to 200
44 to 87
Annualized
Value
5 to 9.8
3.5 to 6.9
0.22 to 0.45
0.1 to 0.21
5.2 to 10
3.6 to 7.1
i The benefits in this table reflect two separate but equally plausible premature mortality estimates derived from the Medicare study (Wu et al.
: 2020) or the NHIS study (Pope III et al. 2019), respectively. All benefits estimates are rounded to two significant figures. Annual benefit
i values presented here are not discounted. Negative values are health disbenefits related to increases in estimated emissions. The present value
: of benefits is the total aggregated value of the series of discounted annual benefits that occur between 2027-2055 (in 2022 dollars) using either
i a 3 percent or 7 percent discount rate. The upstream impacts associated with the standards presented here include health benefits associated
: with reduced criteria pollutant emissions from refineries and health disbenefits associated with increased criteria pollutant emissions from
i EGUs. The benefits in this table also do not include the full complement of health and environmental benefits (such as health benefits related
: to reduced ozone exposure) that if quantified and monetized, would increase the total monetized benefits.
9.6 Summary and Net Benefits
We summarize the costs, savings, and benefits of the final rule, as shown in Table 9-19. Table 9-
19 reproduces the final rule's costs from Table 9-1, fuel savings less EVSE port costs from Table
9-4, non-emission benefits from Table 9-7, climate benefits from Table 9-13, and criteria air
pollutant benefits from Table 9-16, in a single table. We summarize the costs, savings, and
benefits of Alternatives A and B in Table 9-20 and Table 9-21, respectively.
9-21
-------
Table 9-19: Summary of costs, fuel savings and benefits of the final standards
(billions of 2022 dollars)a'b'c'd.
CY 2055
PV, 2%
; PV, 3%
; PV, 7%
; AV, 2%
; AV, 3%
; AV, 7%
Vehicle Technology
Costs
$38
$870
$760
$450
$40
$39
$37
Insurance Costs
$1.9
$33
$28
$15
$1.5
$1.4
$1.2
Repair Costs
-$7.1
-$40
-$32
-$12
-$1.8
-$L6
-$0.99
Maintenance Costs
-$35
-$300
! ' -$250
-$110
-$14
-$13
-$9.3
Congestion Costs
" "'$2.4
$25
$21
$10
$1.2
$1.1
$0.83
Noise Costs
$0.04
$0.41
$0.34
$0.17
$0,019
; $0,018
$0,014
Sum of Costs
$0.59
$590
$530
$350
$27
$28
$29
Pre-tax Fuel Savings
$94
$1,000
$840
$420
$46
'$44
$34
EVSE Port Costs
'$8.6'
......r...$i9o.z
;;;;$i6oi"
$96
I $9 " "
[2.
L.I $7-9l
Sum of Fuel
Savings less
$86
$820
$680
$330
$37
$35
$26
EVSE Port Costs
Drive Value Benefits
$4.7 ;
$46
$38
$18
[ $2.1
$2
$1.5
Refueling Time
Benefits
-$1.7
-$17
-$15
-$7.5
-$0.8
-$0.76
-$0.61
Energy Security
Benefits
$4.1
$47
$39
$20
$2.1
$2
$1.6
Sum of Non-
Emission
$7
$75
$62
$30
$3.4
$3.2
$2.5
Benefits
Climate Benefits,
$150
$1,600
$1,600
$1,600
$72
$72
$72
2% Near-term Ramsey
PM2.5 Health Benefits :
$25
$240
$200
$88
$13
$10
$7.2
Sum of Emission
Benefits
$170
$1,800
$1,800
$1,700
$85
$83
$80
Net Benefits
" $270
$2,100
$2,000
$1,700
$99
$94
$80
a Net benefits are emission benefits, non-emission benefits, and fuel savings (less EVSE port costs) minus the costs of the program. Values
rounded to two significant figures; totals may not sum due to rounding. Present and annualized values are based on the stream of annual
calendar year costs and benefits included in the analysis (2027 - 2055) and discounted back to year 2027. Climate benefits are based on
reductions in GHG emissions and are calculated using three different SC-GHG estimates that assume either a 1.5 percent, 2.0 percent, or 2.5
percent near-term Ramsey discount rate. See EPA's Report on the Social Cost of Greenhouse Gases: Estimates Incorporating Recent
Scientific Advances (EPA, 2023). For presentational purposes in this table, we use the climate benefits associated with the SC-GHG estimates
under the 2-percent near-term Ramsey discount rate. See Chapter 9.4 of this RIA for the full range of monetized climate benefit estimates. All
other costs and benefits are discounted using either a 2-percent, 3-percent, or 7-percent constant discount rate. For further discussion of the
SC-GHGs and how EPA accounted for these estimates, please refer to Chapter 6.2 of the RIA.
k To calculate net benefits, we use the monetized suite of total avoided PM2.5-related health effects that includes avoided deaths based on the
Pope 111 et al., 2019 study, which is the larger of the two PM2.5 health benefits estimates presented in Chapter 9.5 of this RIA.
° The annual PM2.5 health benefits estimate presented in the CY 2055 column reflects the value of certain avoided health outcomes, such as
avoided deaths, that are expected to accrue over more than a single year discounted using a 3-percent discount rate.
^ We do not currently have year-over-year estimates of PM2.5 benefits that discount such annual health outcomes using a 2-percent discount
rate. We have therefore discounted the annual stream of health benefits that reflect a 3-percent discount rate lag adjustment using a 2-percent
discount rate to populate the PV, 2% and AV, 2% columns. The annual stream of PM2.5-related health benefits that reflect a 3-percent and 7-
percent discount rate lag adjustment were used to populate the PV/AV 3% and PV/AV 7% columns, respectively. See Chapter 9.5 of this RIA
for more details on the annual stream of PM2.5-related benefits associated with this rule.
9-22
-------
Table 9-20: Summary of costs, fuel savings and benefits of Alternative A
(billions of 2022 dollars)a'b'c.
CY 2055
PV, 2%
; PV, 3%
; PV, 7%
; AV, 2%
; AV, 3%
; AV, 7%
Vehicle Technology
Costs
$39
$940
$820
$510
$43
$43
$41
Insurance Costs
$1.9
$35
$30
$17
$1.6
$1.6
$1.4
Repair Costs
-$7.3
-$40
-$31
-$12
-$1.8
-$1.6
-$0.94
Maintenance Costs
-$35
-$320
! -$270
-$130
-$15
'" -$14
-$10
Congestion Costs
" "'$2.4
$27
$23
$11
$1.2
$1.2
$0.92
Noise Costs
$0.04 '
$0.44
$0.37
$0.18
$0.02
$0,019
; $0,015
Sum of Costs
$0.2 "
$640
$580
$400
$29
$30
$33
Pre-tax Fuel Savings
$95
$1,100
$910
$470
$50
"$47""
$38
EVSE Port Costs
$8.6"i
$190
r* $i6o.;:
$96
l:::$8.8;:
.$7.9""
Sum of Fuel
Savings less
$86
$900
$750
$370
$41
$39
$30
EVSE Port Costs
Drive Value Benefits
$4.7 :
$49
$41
$20
$2.2
$2.1
; $1.7
Refueling Time
Benefits
-$1.7
-$15
-$13
-$6.2
-$0.7
-$0.66
-$0.5
Energy Security
Benefits
$4.1
$50
$42
$21
$2.3
$2.2
$1.8
Sum of Non-
Emission
$7
$84
$70
$36
$3.8
$3.6
$2.9
Benefits
Climate Benefits,
2% Near-term Ramsey ;
$150
$1,700
$1,700
$1,700
$77
$77
$77
PM2.5 Health Benefits j
$26
$250
$200
$89
$13
$11
$7.3
Sum of Emission
Benefits
$180
$1,900
$1,900
$1,800
$90
$88
$84
Net Benefits
$270
$2,300
$2,100
$1,800
$110
$100
$85
a Net benefits are emission benefits, non-emission benefits, and fuel savings (less EVSE port costs) minus the costs of the program. Values
rounded to two significant figures; totals may not sum due to rounding. Present and annualized values are based on the stream of annual
calendar year costs and benefits included in the analysis (2027 - 2055) and discounted back to year 2027. Climate benefits are based on
reductions in GHG emissions and are calculated using three different SC-GHG estimates that assume either a 1.5 percent, 2.0 percent, or 2.5
percent near-term Ramsey discount rate. See EPA's Report on the Social Cost of Greenhouse Gases: Estimates Incorporating Recent
Scientific Advances (EPA, 2023). For presentational purposes in this table, we use the climate benefits associated with the SC-GHG estimates
under the 2-percent near-term Ramsey discount rate. See Chapter 9.4 of this RIA for the full range of monetized climate benefit estimates. All
other costs and benefits are discounted using either a 2-percent, 3-percent, or 7-percent constant discount rate. For further discussion of the
SC-GHGs and how EPA accounted for these estimates, please refer to Chapter 6.2 of the RIA.
k To calculate net benefits, we use the monetized suite of total avoided PM2.5-related health effects that includes avoided deaths based on the
Pope 111 et al., 2019 study, which is the larger of the two PM2.5 health benefits estimates presented in Chapter 9.5 of this RIA.
° The annual PM2.5 health benefits estimate presented in the CY 2055 column reflects the value of certain avoided health outcomes, such as
avoided deaths, that are expected to accrue over more than a single year discounted using a 3-percent discount rate.
^ We do not currently have year-over-year estimates of PM2.5 benefits that discount such annual health outcomes using a 2-percent discount
rate. We have therefore discounted the annual stream of health benefits that reflect a 3-percent discount rate lag adjustment using a 2-percent
discount rate to populate the PV, 2% and AV, 2% columns. The annual stream of PM2.5-related health benefits that reflect a 3-percent and 7-
percent discount rate lag adjustment were used to populate the PV/AV 3% and PV/AV 7% columns, respectively. See Chapter 9.5 of this RIA
for more details on the annual stream of PM2.5-related benefits associated with this rule.
9-23
-------
Table 9-21 Summary of costs, fuel savings and benefits of Alternative B
(billions of 2022 dollars)a'b'c.
CY 2055
PV, 2%
PV, 3%
PV, 7%
AV, 2%
AV, 3%
AV, 7%
Vehicle
Technology Costs
$30
$710
$610
$360
$32
$32
$30
Insurance Costs
$1.5
$26
$22
$12
$1.2
$1.2
$0.98
Repair Costs
-$6.6
-$41
-$32
-$13
-$1.9
-$1.7
-$1.1
Maintenance
Costs
-$30
-$260
-$210
-$98
-$12
-$11
-$8
Congestion Costs
$2.2
$22
$18
$8.9
$1
$0.96
$0.73
Noise Costs
$0,036
$0.36
$0.3
$0.15
$0,017
$0,016
$0,012
Sum of Costs
-$2.8
$450
$410
$270
$21
$21
$22
Pre-tax Fuel
Savings
$81
$880
$730
$370
$40
$38
$30
EVSE Port Costs
$8.6
$190
$160
$96
$9
$8.8
$7.9
Sum of Fuel
Savings less
EVSE Port Costs
$73
$690
$570
$270
$31
$29
$22
Drive Value
Benefits
$4.3
$41
$34
$17
$1.9
$1.8
$1.3
Refueling Time
Benefits
-$2.2
-$23
-$19
1
&
VO
bo
-$1.1
-$1
-$0.8
Energy Security
Benefits
$3.4
$39
$33
$17
$1.8
$1.7
$1.3
Sum of Non-
Emission
Benefits
$5.5
$58
$48
$23
$2.6
$2.5
$1.9
Climate Benefits,
2% Near-term
Ramsey
$120
$1,300
$1,300
$1,300
$61
$61
$61
PM2.5 Health
Benefits
$25
$240
$200
$87
$12
$10
$7.1
Sum of Emission
Benefits
$150
$1,600
$1,500
$1,400
$74
$72
$68
Net Benefits
$230
$1,900
$1,700
$1,400
$87
$82
$70
a Net benefits are emission benefits, non-emission benefits, and fuel savings (less EVSE port costs) minus the costs of the program. Values
rounded to two significant figures; totals may not sum due to rounding. Present and annualized values are based on the stream of annual
calendar year costs and benefits included in the analysis (2027 - 2055) and discounted back to year 2027. Climate benefits are based on
reductions in GHG emissions and are calculated using three different SC-GHG estimates that assume either a 1.5 percent, 2.0 percent, or 2.5
percent near-term Ramsey discount rate. See EPA's Report on the Social Cost of Greenhouse Gases: Estimates Incorporating Recent
Scientific Advances (EPA, 2023). For presentational purposes in this table, we use the climate benefits associated with the SC-GHG estimates
under the 2-percent near-term Ramsey discount rate. See Chapter 9.4 of this RIA for the full range of monetized climate benefit estimates. All
other costs and benefits are discounted using either a 2-percent, 3-percent, or 7-percent constant discount rate. For further discussion of the
SC-GFIGs and how EPA accounted for these estimates, please refer to Chapter 6.2 of the RIA.
k To calculate net benefits, we use the monetized suite of total avoided PM2.5-related health effects that includes avoided deaths based on the
Pope 111 et al., 2019 study, which is the larger of the two PM2.5 health benefits estimates presented in Chapter 9.5 of this RIA.
° The annual PM2.5 health benefits estimate presented in the CY 2055 column reflects the value of certain avoided health outcomes, such as
avoided deaths, that are expected to accrue over more than a single year discounted using a 3-percent discount rate.
^ We do not currently have year-over-year estimates of PM2.5 benefits that discount such annual health outcomes using a 2-percent discount
rate. We have therefore discounted the annual stream of health benefits that reflect a 3-percent discount rate lag adjustment using a 2-percent
discount rate to populate the PV, 2% and AV, 2% columns. The annual stream of PM2.5-related health benefits that reflect a 3-percent and 7-
percent discount rate lag adjustment were used to populate the PV/AV 3% and PV/AV 7% columns, respectively. See Chapter 9.5 of this RIA
for more details on the annual stream of PM2.5-related benefits associated with this rule.
9-24
-------
9.7 Transfers
There are four types of transfers included in our analysis. Two of these transfers come in the
form of tax credits arising from the Inflation Reduction Act to encourage investment in battery
technology and the purchase of electrified vehicles. These are transfers from the government to
producers of vehicles (the 45X battery production tax credits), or to purchasers of vehicles (the
30D tax credit) or to lessors or commercial purchasers (the 45W tax credit). There are also
transfers from the government to residents and businesses who install EVSE (the 30C tax
credit)259 though we don't quantify these transfers as part of our analysis. The third, new for the
final rule, is state taxes on the purchase of new, higher cost vehicles which represents transfers
from purchasers to government. The fourth is fuel and electricity taxes which are transfers from
purchasers of fuel to the government. The final standards result in less liquid-fuel consumed and,
therefore, less money transferred from purchasers of liquid fuel to the government while the
reverse is true for electricity consumption where the increase associated with PEVs results in
more money transferred from purchasers to the government. For more detail on the IRS Section
45X, 30D, and 45W tax credits please see Section IV of the preamble and Chapter 2.6.8 of this
RIA. Table 9-22 presents transfers associated with the final standards. Table 9-23 and Table 9-
24 present transfers associated with Alternatives A and B, respectively.
Table 9-22: Transfers associated with the final standards
(billions of 2020 dollars).
Calendar Battery Tax Vehiele State Sales Liquid Fuel and Sum of
Year Credits PurehaseTax Taxes Eleetrieitv Transfers
Credits Taxes
2027
$0.25 ;
$0.4
-$0.12
$0,036 i
$0.56
2028
$1.4
$21
-$0.27l
$0.23 j
$3.4
2029
$4.1
$5.4
-$0.61
$0.69
$9.5
2030
$5.1
$9.2
-$0.9
$1.4
$15
203 1
$5.4
$15
-$1.2 i
$2.2 i
$22
2032
$3.6
$20;
-$1.3
$3.2 1
$25
2035
So
So
-$2.7 ;
$7.3!
$4.5
2040
$0 !
$0 !
-$2.5 1
$13
$10
2045
$o;
$0 !
-$2.3 :
$16
$13
2050
$0 1
$0;
-$2.1 I
$18
$16
2055
$0 i
$0 :
-$ 1.9
$18
$16
PV2
$18
$47 !
-$43
2 $230 ]
$250
PV3
$17 1
$45
-$37 !
$190
$220
PV7
$15
$38:
-$22 ;
$98
$130
AV2
$0.83
$2.2
-$2",
$10
$11
AV3
$0.91
$2.4 ;
-$ 1.9
$9.9 i
$11
AV7
$1.2 :
$3.1
-$1.8
$7.9 !
$10
* Negative values reflect transfers from taxpayers to governments: positive values reflect transfers
from government to taxpayers.
259 The IRA extends the Internal Revenue Code 30C Alternative Fuel Refueling Property Tax Credit through Dec
31, 2032, with modifications. See Preamble Section IV.C.4 and RIA Chapter 5 for more details.
9-25
-------
Table 9-23: Transfers associated with Alternative A
(billions of 2022 dollars).
Calendar
Battery Tax
Vehiele
State
Liquid Fuel
Sum of
Year
Credits
Purchase
Sales
and
T ransfers
Tax Credits
Taxes
Electricity
Taxes
2027
$3.3 I
$3.9
-$0.65
$0.37 ;
$7
2028
$5;
$6.7
-$1
$1
$12
2029
$7.2 |
$9.8
-$1.3
$1.8
$18
2030
$6.9
$12 !
-$1.4
$2.8 T
$21
203 1
$5.9
$17 !
-$1.5
$3.7 ;
$25
2032
$3.7:
$20 :
-$1.5
$4.7 ;
$27
2035
$0 1
So
-$2.7 ¦
$8.6
$5.9
2040
$0 ;
$0 I
-$2.5 ]
$13
$11
2045
$0:
$0 :
-$2.3 ;
$16
$14
2050
$0:
$0 1
-$2.2 :
$19
$17
2055
$0 :
$0 I
-$ 1.9
$18
$16
PV2
$30 i
$64
-$46
$250 1
$290
PV3
$29 i
$62 f
-$40
$210
$260
PV7
$25 ]
$52 f
-$24 T
$110
$160
AV2
$1.4
$2.9
-$2.1 :
$11
$13
AV3
$1.5
$3.2 f
-$2.1 ;
$11
$13
AV7
$2.1
$4.3
-$2 T
$8.9
$13
* Negative values reflect transfers from taxpayers to governments: positive values reflect
transfers from government to taxpayers.
Table 9-24: Transfers associated with Alternative B
(billions of 2022 dollars).
Calendar
Battery Tax
Vehicle
State Sales
Liquid Fuel and
Sum of
Year
Credits
Purchase Tax
Taxes
Electricity
T ransfers
Credits
Taxes
2027
$0.17
$0.33 T
-$(). 11
$0,028
$0.42
2028
$1.2 i
$1.6
-SO.231
$0.2]
$2.7
2029
$4
$5.1
-$0.56
$0.64
$9.2
2030
$4.5
$8 |
-$0.85
$1.3
$13
203 1
$4.5
$13
-$0.98
$2;
$18
2032
$3.4
$17 j
-$1.2 ;
$2.9
$23
2035
$0 |
$0 |
-$2.1 :
$6.4
$4.2
2040
$0 !
$0 !
-$2 |
$u
$8.5
2045
$o i
$0 !
-$2 I
$13
$11
2050
$o !
$0 :
-$1.8
$16
$14
2055
$0 !
$0 !
-$1.5
$15 ;
$14
PV2
$16
$41
-$35 :
$190
$220
PV3
$16
$39
-$30 ;
$160
$190
PV7
$13 !
$33 1
-$18
** 1 $83 J
$110
AV2
$0.75 ;
$1.9
-$ 1.6
$8.9
$9.9
AV3
$0.82
$2 ]
-$ 1.6
$8.4
$9.7
AV7
$1.1
$2.7 1
-$1.4
$6.8
$9.1
* Negative values reflect transfers from taxpayers to governments: positive values reflect transfers
from government to taxpayers.
9-26
-------
Chapter 9 References
EPA. 2023f. "Supplementary Material for the Regulatory Impact Analysis for the Final
Rulemaking: Standards of Performance for New, Reconstructed, and Modified Sources and
Emissions Guidelines for Existing Sources: Oil and Natural Gas Sector Climate Review." EPA
Report on the Social Cost of Greenhouse Gases: Estimates Incorporating Recent Scientific
Advances, Washington, DC. doi:Docket ID No. EPA-HQ-OAR-2021-0317.
—. 2021. "Technical Documentation on the framework for evaluating damages and impacts
(FrEDI)." EPA Science Inventory.
https://cfpub.epa.gov/si/si_public_record_report.cfm?dirEntryId=351316&Lab=OAP&simplesea
rch=0&showcriteria=2&sortby=pubDate&searchall=fredi&timstype=&datebeginpublishedprese
nted=02/l 4/2021.
Hartin, C, EE McDuffie, K Noiva, M Sarofim, B Parthum, J Martinich, S Barr, J Neumann, J
Wilwerth, and A Fawcett. 2023. "Advancing the estimation of future climate impacts within the
United States." Earth System Dynamics 14(5): 1015-1037.
IWG. 2016a. Addendum to Technical Support Document on Social Cost of Carbon for
Regulatory Impact Analysis under Executive Order 12866: Application of the Methodology to
Estimate the Social Cost of Methane and the Social Cost of Nitrous Oxide.
https://www.epa.gov/sites/default/files/2016-12/documents/addendum_to_sc-
ghg_tsd_august_2016. pdf.
IWG. 2015. Response to comments: social cost of carbon for regulatory impact analysis under
executive order 12866. Response to Comments, United States Government.
https://obamawhitehouse.archives.gov/sites/default/files/omb/inforeg/scc-responseto-comments-
final-july-2015 .pdf.
—. 2010. Technical Support Document: Social Cost of Carbon for Regulatory Impact Analysis
under Executive Order 12866. Accessed 2023. https://www.epa.gov/sites/default/files/2016-
12/documents/ scc_tsd_2010.pdf.
—. 2021. Technical Support Document: Social Cost of Carbon, Methane, and Nitrous Oxide
Interim Estimates under Executive Order 13990. Accessed 2023.
https://www.whitehouse.gov/wp-
content/uploads/2021/02/TechnicalSupportDocument_SocialCostofCarbonMethaneNitrouSOxid
e.pdf.
—. 2013. Technical Support Document: Technical Update of the Social Cost of Carbon for
Regulatory Impact Analysis Under Executive Order 12866.
https://www.ourenergypolicy.org/wp-
content/uploads/2013/06/social_cost_of_carbon_for_ria_2013_update.pdf.
—. 2016b. Technical Support Document: Technical Update of the Social Cost of Carbon for
Regulatory Impact Analysis Under Executive Order 12866. Accessed 2023.
https://www.epa.gov/sites/default/files/2016- 12/documents/sc_CO2_tsd_august_2016.pdf.
9-27
-------
McDuffie, EE, MC Sarofim, W Raich, M Jackson, H Roman, K Seltzer, BH Henderson, et al.
2023. "The social cost of ozone-related mortality impacts from methane emissions." Earth's
Future 11(9). doi:https://doi.org/10.1029/2023EF003853.
National Academies. 2017. Valuing Climate Damages: Updating Estimation of the Social Cost
of Carbon Dioxide. Washington, D.C.: The National Academies Press.
Pope III et al., CA, Lefler, JS, Ezzati, M, Higbee, JD, Marshall, JD, Kim, S-Y, Bechle, M,
Gilliat, KS, Vernon, SE and Robinson, AL. 2019. "Mortality risk and fine particulate air
pollution in a large, representative cohort of US adults." Environmental Health Perspectives 127
(7): 077007.
Sarofim, MC, J Martinich, JE Neumann, J Willwerth, Z Kerrich, M Kolian, C Fant, and C
Hartin. 2021. "A temperature binning approach for multi-sector climate impact analysis."
Climate Change 165(22). https://doi.org/10.1007/sl0584-021-03048-6.
Wu et al., X, Braun, D, Schwartz, J, Kioumourtzoglou, M and Dominici, F. 2020. "Evaluating
the impact of long-term exposure to fine particulate matter on mortality among the elderly."
Science Advances 6 (29): eaba5692.
9-28
-------
Appendix to Chapter 9
This appendix presents the climate benefits of the final standards using the interim Social
Cost of Greenhouse Gas (SC-GHG) values used in the NPRM. We have updated the interim
values to 2022 dollars for the analysis in this RIA. The updated interim SC-GHG values are
presented in Table 9-25. The climate benefits using these values are presented in Table 9-26
through Table 9-29 for the reductions in CO2, CH4, N2O and all GHGs, respectively. Table 9-30
presents the summary of cost and benefits of the final standards using the 3% average benefits
across the GHGs. Costs and benefits presented here are calculated relative to the No Action case
unless stated otherwise.
Table 9-25 Interim Social Cost of GHG Values, 2027-2055 (2022 $/metric ton)
(iik-iulur (;()2 (M4 N2()
V en i'
= 5%
3%
2.5%
3%
, 5%
3%
2.5%
3%
5%
3%
2.5%
3%
Avji
Avji
Avji
95th
pet ill'
Avg
Avg
Avg
95th
pet ill'
Avg
Avg
Avg
95th
petile
2027
; $20
$66
; $96
$197
: $959
$2,030
: $2,621
$5,379
: $8,053
; $24,029
: $34,734
$63,484
2028
$21
$67
$97
$201
: $989
$2,083
, $2,683
$5,523
, $8,279
$24,518
; $35,358
$64,836
2029
; $21
$68
; $99
$205
$1,020
$2,135
$2,745
$5,667
$8,505
$25,008
$35,981
$66,188
2030
$22
$69
$100
$209
! $1,050
$2,188
i $2,807
$5,810
$8,731
$25,497
$36,604
$67,540
2031
= $22
$70
$102
$213
$1,089
$2,250
, $2,879
$5,983
$9,008
$26,048
: $37,288
$69,062
2032
: $23
$72
$103
$218
$1,127
$2,312
$2,950
$6,155
, $9,285
$26,598
; $37,973
$70,583
2033
$24
$73
$105
$222
: $1,165
$2,374
: $3,022
$6,327
: $9,563
$27,149
i $38,657
$72,105
2034
: $24
$74
$106
$226
; $1,204
$2,436
; $3,093
$6,499
: $9,840
$27,700
, $39,342
$73,626
2035
$25
$76
$108
$230
$1,242
$2,498
: $3,165
$6,671
i $10,117
l $28,250
$40,026
$75,148
2036
! $26
$77
$109
$235
$1,281
$2,560
$3,236
$6,843
* $10,395
: $28,801
: $40,711
$76,669
2037
; $26
$78
, $111
$239
: $1,319
$2,622
$3,308
$7,015
; $10,672
. $29,352
$41,395
$78,191
2038
: $27
$79
$112
$243
$1,358
$2,684
, $3,379
$7,188
: $10,949
: $29,902
: $42,079
$79,712
2039
$28
$81
$114
$248
; $1,396
$2,746
; $3,451
$7,360
: $11,227
$30,453
; $42,764
$81,234
2040
: $28
$82
$115
$252
$1,435
$2,808
, $3,522
$7,532
, $11,504
; $31,004
: $43,448
$82,755
2041
: $29
$83
; $117
$256
* $1,477
$2,870
$3,593
$7,694
i $11,829
5 $31,596
: $44,169
$84,349
2042
: $30
$85
$118
$260
i $1,519
$2,933
: $3,663
$7,856
: $12,154
$32,189
$44,891
$85,944
2043
; $30
$86
$120
$264
$1,561
$2,996
: $3,734
$8,018
$12,479
> $32,781
; $45,612
$87,538
2044
$31
$87
: $121
$267
: $1,603
$3,058
, $3,804
$8,180
$12,803
$33,374
$46,333
$89,132
2045
: $32
$88
$123
$271
$1,645
$3,121
; $3,875
$8,342
! $13,128
: $33,967
: $47,054
$90,727
2046
$33
$90
$124
$275
$1,687
$3,183
$3,946
$8,504
; $13,453
: $34,559
: $47,775
$92,321
2047
; $33
$91
$126
$279
; $1,729
$3,246
: $4,016
$8,666
; $13,778
; $35,152
s $48,496
$93,915
2048
$34
$92
$127
$283
$1,771
$3,309
$4,087
$8,828
i $14,103
; $35,745
: $49,217
$95,510
2049
: $35
$93
: $129
$287
, $1,813
$3,371
; $4,157
$8,990
: $14,428
$36,337
: $49,939
$97,104
2050
: $35
$95
$130
$291
, $1,855
$3,434
$4,228
$9,152
i $14,753
, $36,930
i $50,660
$98,698
2051
$36
$95
$132
$292
$1,887
$3,478
$4,276
$9,204
$15,141
, $37,548
, $51,366
$99,533
2052
$37
$96
$133
$293
i $1,913
$3,513
$4,314
$9,243
: $15,500
$38,141
, $52,071
$101,081
2053
: $38
$97
$135
$294
! $1,939
$3,548
: $4,352
$9,282
: $15,859
i $38,735
$52,775
$102,629
2054
: $38
$99
$136
$295
$1,965
$3,584
$4,390
$9,320
, $16,219
; $39,329
! $53,480
$104,177
2055
; $39
$100
$137
$298
; $1,991
$3,619
. $4,428
$9,359
i $16,578
: $39,922
: $54,184
$105,724
Note: The 2027-2055 values are identical to those reported in the 2016 TSD (IWG 2016a) adjusted to 2022 dollars using the annual GDP
Implicit Price Deflator values used elsewhere in the analysis presented in this RIA. This table displays the values rounded to the nearest
dollar; the annual unrounded values used in the calculations in this analysis are available on OMB's website:
https://www.whitehouse.gOv/omb/information-regulatory-affairs/regulatory-matters/#scghgs. The estimates were extended for the period
2051 to 2055 using methods, assumptions, and parameters identical to the 2020-2050 estimates. The values are stated in $/metric ton and
vary depending on the year.
A9-1
-------
Table 9-26 Benefits of reduced CO2 emissions from the final standards using the interim
SC-GHG values
(billions of 2022 dollars)
Calendar Discount Rale
Year
5%
3%
2.5%
-3 0/
J /O
95th percentile
2027
$0.0083
$0,027
$0.04
$0,081
2028
$0,072
$0.23
$0.34
$0.7
2029
$0.25
$0.8
$1.2
$2.4
2030
$0.52
$1.7
$2.4
! $5
203 1
$0.89
$2.8
$4
$8.5
2032
$1.3
$4.2
$6
$13
2033
$2
$6.2
$9
$19
2034
$2.8
$8.5
$12
$26
2035
$3.6
$11
$16
$33
2036
$4.4
$13
$19
$40
2037
$5.2
$16
$22
$47
2038
$6
$18
$25
$55
2039
$6.9
$20
$28
$62
2040
$7.7
$22
$31
$69
2041
$8.4
$24
$34
$75
2042
$9.2
$26
$37
$81
2043
$10
$28
$39
$87
2044
$11
$30 !
$42
$92
2045
$11
$32
$44
$97
2046
$12
$33 "" |
$46
$100
2047
$13
$34
$48
$110
2048
$13
$36
$49
$110
2049
$14
$37
$51
$110
2050
$14
$38
$52
$120
2051
$15
"$38
$53
$120
2052
$15
$39
$54
$120
2053
$15
$40
$55
$120
2054
$16
$40
$55
$120
2055
$16
$40
$56
$120
PV
$89
$360
$550
$1,100
AV
$5.9
$19
" $27
$57
Note: Climate benefits arc based on changes (reductions) in CO2 emissions and arc calculated using the IWG
interim SC-GHG estimates from (IWG 2021). Climate benefits include changes in vehicle, EGU, and refinery
CO2 emissions.
A9-2
-------
Table 9-27 Benefits of reduced CH4 emissions from the final standards using the interim
SC-GHG values
(billions of 2022 dollars)
Calendar Discount Rale
Year
5%
3%
2.5%
-3 0/
J /O
95th percentile
2027
-$0.00001
-$0.000022
-$0.000028
-$0.000058
2028
-$0.000024
-$0.00005
-$0.000064
-$0.00013
2029
$0.000011
$0.000023
$0.00003
$0.000062
2030
$0.00006
$0.00013
$0.00016
$0.00033
203 1
$0.00011
$0.00023
$0.00029
$0.00061
2032
$0.00025
$0.00052
$0.00067
$0.0014
2033
$0.00063
$0.0013
$0.0016
$0.0034
2034
$0.0011
$0.0022
$0.0028
$0,006
2035
$0.0017
$0.0034
$0.0042
$0,009
2036
$0.0023
$0.0046
$0.0058
$0,012
2037
$0,003
$0.0061
$0.0076
$0,016
2038
$0.0039
$0.0077
$0.0097
$0,021
2039
$0.0048
$0.0095
$0,012
$0,025
2040
$0.0057
$0,011
$0,014
$0.03
2041
$0.0066
$0,013
$0,016
$0,035
2042
$0.0076
$0,015
$0,018
$0,039
2043
$0.0086
$0,016
$0.02
$0,044
2044
$0.0096
$0,018
$0,023
$0,049
2045
$0,011
$0.02
$0,025
$0,054
2046
$0,011
$0,022
$0,027 ""
$0,057
2047
$0,012
$0,023
$0,028
$0,061
2048
$0,013
$0,024
$0.03
$0,064
2049
$0,014
$0,025
$0.031
$0,067
2050
$0,014
$0,026
$0,032
$0.07
2051
$0,015
$0,027
$0,033
$0,071
2052
$0,015
$0,028
$0,034
$0,072
2053
$0,015
$0,028
$0,034
$0,073
2054
$0,016
$0,028
$0,035
$0,074
2055
$0,016
$0,029
$0,035
$0,074
PV
$0,074
$0.21
$0.29
$0.55
AV
$0.0049
$0,011
$0,014
$0,029
Note: Climate benefits arc based on changes (reductions) in CH4 emissions and arc calculated using the
IWG interim SC-GHG estimates from (IWG 2021). Climate benefits include changes in vehicle, EGU, and
refinery CH4 emissions.
A9-3
-------
Table 9-28 Benefits of reduced N2O emissions from the final standards using the interim
SC-GHG values
(billions of 2022 dollars)
Calendar Discount Rale
Year
5%
3%
2.5%
-3 0/
J /O
95th percentile
2027
$0.000051
$0.00015
$0.00022
$0.00041
2028
$0.00035
$0,001
$0.0015
$0.0027
2029
$0.0014
$0.0041
$0.0059
$0,011
2030
$0.0034
$0.0099
$0,014
$0,026
203 1
$0.0058
$0,017
$0,024
$0,044
2032
$0,009
$0,026
$0,037
$0,069
2033
$0,014
$0.04
$0,057
$0.11
2034
$0.02
$0,055
$0,079
$0.15
2035
$0,025
$0,071
$0.1
$0.19
2036
$0.031
$0,087
$0.12
$0.23
2037 ""
$0,037
$0.1
$0.14
$0.27
2038
$0,043
$0.12
$0.16
$0.3 1
2039
$0,049
$0.13
$0.19
SO.35
2040
$0,054
$0.15
$0.21
$0.39
2041
$0.06
$0.16
$0.22
$0.43
2042
$0,066
$0.17
$0.24
$0.47
2043
$0,072
$0.19
$0.26
$0.5
2044
;;:;;:;;:;;$o.o77
$0.2
$0.28
$0.54
2045
$0,082
$0.21
$0.29
$0.57
2046
$0,087
$0.22
$0.3 1
$0.6
2047
$0,091
$0.23
$0.32
$0.62
2048
$0,096
$0.24
$0.33
$0.65
2049
$0,099
$0.25
$0.34
$0.67
2050
$0.1
$0.26
$0.35
$0.69
2051
$0.11
$0.26
$0.36
$0.7
2052
$0.11
$0.27
$0.37
$0.72
2053
$0.11
$0.28
$0.38
$0.73
2054
$0.12
$0.28
$0.38
$0.75
2055
$0.12
$0.29
$0.39
$0.76
PV
$0.64
$2.4
$3.7
$6.4
AV
$0,042
$0.13
$0.18
$0.33
Note: Climate benefits arc based on changes (reductions) in N2O emissions and arc calculated using the IWG
interim SC-GHG estimates from (IWG 2021). Climate benefits include changes in vehicle. EGU. and refinery
N2O emissions.
A9-4
-------
Table 9-29 Benefits of reduced GHG emissions from the final standards using the interim
SC-GHG values
(billions of 2022 dollars)
Calendar Discount Rale
Year
5%
3%
2.5%
-3 0/
J /O
95th percentile
2027
$0.0083
$0,027
$0.04
$0,082
2028
$0,072
$0.23
$0.34
$0.7
2029
$0.25
$0.8
$1.2
$2.4
2030
$0.52
$1.7
$2.4
: $5
203 1
$0.89
$2.8
$4.1
$8.5
2032
$1.3
$4.2
$6
$13
2033
$2
$6.3
$9
$19
2034
$2.8
$8.5
$12
$26
2035
$3.6
$11
$16
$33
2036
$4.4
$13
$19
$41
2037
$5.3
$16
$22
$48
2038
$6.1
$18
$25
$55
2039
$6.9
$20
$29
$62
2040
$7.7
$22
$32
$69
2041
$8.5
$24
$34
$75
2042
$9.3
$26
$37
$81
2043
$10
$28
$40
$87
2044
$11
$30
$42
$93
2045
$11
$32
$44
$98
2046
$12
$33
$46
$100
2047
$13
$35
$48
$110
2048
$13
$36
$50
$110
2049
$14
$37
$51
$110
2050
$14
$38
$53
$120
2051
$15
$39
$54
$120
2052
$15
$39
$55
$120
2053
$15
$40
$55
$120
2054
$16
$40
$56
$120
2055
$16
$41
$56
$120
PV
$90
$360
$550
$1,100
AV
$5.9
$19
$27
$57
Note: Climate benefits arc based on changes (reductions) in GHG emissions and arc calculated using the IWG
interim SC-GHG estimates from (IWG 2021). Climate benefits include changes in vehicle, EGU, and refinery
GHG emissions.
A9-5
-------
Table 9-30 Summary of costs, fuel savings and benefits of the final standards
(billions of 2022 dollars)*
CY 2055
PV, 2%
PV, 3%
PV, 7%
AV, 2%
AV, 3%
AV, 7%
Vehicle
Technology
Costs
$38
$870
$760
$450
$40
$39
$37
Insurance Costs
$1.9
$33
$28
$15
$1.5
$1.4
$1.2
Repair Costs
-$7.1
-$40
-$32
-$12
-$1.8
-$1.6
-$0.99
Maintenance
Costs
-$35
-$300
-$250
-$110
-$14
-$13
-$9.3
Congestion Costs
$2.4
$25
$21
$10
$1.2
$1.1
$0.83
Noise Costs
$0.04
$0.41
$0.34
$0.17
$0,019
$0,018
$0,014
Sum of Costs
$0.59
$590
$530
$350
$27
$28
$29
Pre-tax Fuel
Savings
$94
$1,000
$840
$420
$46
$44
$34
EVSE Port Costs
$8.6
$190
$160
$96
$9
$8.8
$7.9
Sum of Fuel
Savings less
EVSE Port
Costs
$86
$820
$680
$330
$37
$35
$26
Drive Value
Benefits
$4.7
$46
$38
$18
$2.1
$2
$1.5
Refueling Time
Benefits
-$1.7
-$17
-$15
-$7.5
-$0.8
-$0.76
-$0.61
Energy Security
Benefits
$4.1
$47
$39
$20
$2.1
$2
$1.6
Sum of Non-
Emission
Benefits
$7
$75
$62
$30
$3.4
$3.2
$2.5
Climate Benefits,
3% Average
$41
$360
$360
$360
$19
$19
$19
PM2.5 Health
Benefits
$25
$240
$200
$88
$13
$10
$7.2
Sum of
Emission
Benefits
$66
$600
$560
$450
$31
$29
$26
Net Benefits
$160
$910
$770
$450
$45
$40
$26
* Please see the footnotes to Table 9-21 of this RIA. Climate benefits are based on changes (reductions) in GHG emissions and are
calculated using the IWG interim SC-GHG estimates from (IWG 2021).
A9-6
-------
Chapter 10: Energy Security Impacts
In this Chapter of the RIA, we evaluate the energy security impacts of this final light- and
medium-duty vehicle (LMDV) GHG rule (2027-2032). Energy security is broadly defined as the
uninterrupted availability of energy sources at affordable prices (IEA 2019). Most discussions of
U.S. energy security revolve around the topic of the economic costs of U.S. dependence on oil
imports.260 Energy independence and energy security are distinct but related concepts, and an
analysis of energy independence informs our assessment of energy security. The goal of U.S.
energy independence is generally the elimination of all U.S. imports of petroleum and other
foreign sources of energy, or more broadly, reducing the sensitivity of the U.S. economy to
energy imports and foreign energy markets (Greene 2010).
The U.S.'s oil consumption had been gradually increasing in recent years (2015-2019) before
the COVID-19 pandemic in 2020 dramatically decreased U.S. and global oil consumption (U.S.
EIA 2022). By July 2021, U.S. oil consumption had returned to pre-pandemic levels and has
remained fairly stable since then (U.S. EIA 2022). The U.S. has increased its production of oil,
particularly "tight" (i.e., shale) oil, over the last decade (U.S. EIA 2022). As a result of the recent
increase in U.S. oil production, the U.S. became a net exporter of crude oil and refined petroleum
products in 2020 and is projected to be a net exporter of crude oil and refined petroleum products
for the foreseeable future (U.S. EIA 2023). This is a significant reversal of the U.S.'s net export
position since the U.S. has been a substantial net importer of crude oil and refined petroleum
products starting in the early 1950s (U.S. EIA 2022).
Oil is a commodity that is globally traded and, as a result, an oil price shock is transmitted
globally. Given that the U.S. is projected to be a net exporter of crude oil and refined petroleum
products for the timeframe of this analysis (2027-2055) for this rule, one could reason that the
U.S. no longer has a significant energy security problem. However, U.S. refineries still rely on
significant imports of heavy crude oil which could be subject to supply disruptions. Also, oil
exporters with a large share of global production have the ability to raise or lower the price of oil
by exerting the market power associated with the Organization of Petroleum Exporting Countries
(OPEC) to alter oil supply relative to demand. These factors contribute to the vulnerability of the
U.S. economy to episodic oil supply shocks and price spikes, even when the U.S. is projected to
be an overall net exporter of crude oil and refined petroleum products. Reducing U.S. net oil
imports and use reduces the U.S.'s exposure to oil price volatility.
EPA estimates that U.S. consumption and net imports of petroleum will be reduced as a result
of this final rule, both from an increase in fuel efficiency of LMDVs using petroleum-based fuels
and from the greater use of plug-in electric vehicles (PEVs), which are fueled with electricity. A
reduction of U.S. net petroleum imports reduces both financial and strategic risks caused by
potential sudden disruptions in the supply of petroleum to the U.S. and global market, thus
increasing U.S. energy security. In other words, reduced U.S. oil imports act as a "shock
absorber" when there is a supply disruption in world oil markets.
260 The issue of cyberattacks is another energy security issue that could grow in significance over time. For example,
in 2021, one of the U.S.'s largest pipeline operators, Colonial Pipeline, was forced to shut down after being hit by a
ransomware attack. The pipeline carries refined gasoline and jet fuel from Texas to New York. (New York Times
2021).
10-1
-------
It is anticipated that manufacturers will choose to comply with this final standard with
significant increases in PEVs in the light- and medium-duty vehicle fleet. The wider use of
electricity to power vehicles in the U.S. will likely result in the use of a generally more
affordable fuel that has less price volatility compared to the current widespread use of gasoline in
light- and medium-duty vehicles. Furthermore, the U.S. supply and demand of electricity is
almost entirely domestic, and largely independent of electricity markets outside of North
America. Over time, the wider penetration of PEVs into the U.S. vehicle fleet will likely provide
significant energy security benefits, principally by reducing the overall U.S. demand for oil. As
new PEVs enter the vehicle market and the stock of PEVs becomes an increasingly larger
fraction of the total stock of vehicles on the road, high oil prices and oil price shocks will have a
diminishing impact on the overall U.S. economy, leading to greater energy security. The wider
use of electricity to power LMDVs will also move the U.S. toward energy independence; that is
independence of foreign markets, since the electricity to power PEVs will almost exclusively be
produced in the U.S.
This Chapter of the RIA first reviews the historical and recent energy security literature
relevant in the context of this final rule. This review provides a discussion of recent oil security
literature, recent studies on tight oil and recent electricity security studies on the wider use of
PEVs. Second, this Chapter also provides an assessment of the electricity security implications
of this final rule. Third, in the last section of this Chapter, the agency's estimates of U.S. oil
import reductions of the final standards are presented. The military cost impacts of this final rule
are discussed as well. However, due to methodological limitations, we do not quantify the
military costs savings from reduced U.S. oil imports.
10.1 Review of Historical Energy Security Literature
Energy security discussions are typically based around the concept of the oil import premium,
sometimes also labeled the oil security premium. The oil import premium is the extra
cost/impacts of importing oil beyond the price of the oil itself as a result of: (1) potential
macroeconomic disruption and increased oil import costs to the economy from oil price spikes or
"shocks"; and (2) monopsony impacts. Monopsony impacts stem from changes in the demand
for imported oil, which changes the price of all imported oil. Oil import premia are used to
quantify decreases in vulnerability to oil supply shocks resulting from a policy which reduces
U.S. net imports of oil.
The so-called oil import premium gained attention as a guiding concept for energy policy in
the aftermath of the oil shocks of the 1970s. (Bohi and Montgomery 1982), (EMF 1982), and
(Plummer, et al. 1982) provided valuable discussion of many of the key issues related to the oil
import premium as well as the analogous oil stockpiling premium. (Bohi and Montgomery 1982)
detailed the theoretical foundations of the oil import premium and established many of the
critical analytic relationships. Broadman and Hogan revised and extended the established
analytical framework to estimate optimal oil import premia with a more detailed accounting of
macroeconomic effects (Broadman and Hogan 1988) (Broadman 1986) (Hogan 1981). Since the
original work on energy security was undertaken in the 1980s, there have been a couple of
reviews on this topic: (Leiby, Jones, et al. 1997), (Parry and Darmstadter 2003).
The economics literature on whether oil shocks are the same level of threat to economic
stability as they once were is mixed. Some of the literature asserts that the macroeconomic
component of the energy security externality is small. For example, (National Research Council
10-2
-------
2010) argued that the non-environmental externalities associated with dependence on foreign oil
are small, and potentially trivial. (Nordhaus 2007) and (Blanchard and Gall 2010) question the
impact of oil price shocks on the economy in the early 2000s timeframe. They were motivated by
attempts to explain why the economy actually expanded during the oil shock in the early-2000s
timeframe, and why there was no evidence of higher energy prices being passed on through
higher wage inflation. One reason, according to Nordhaus and Blanchard and Gali, is that
monetary policy has become more accommodating to the price impacts of oil shocks. Another
reason is that consumers have simply decided that such movements are temporary and have
noted that price impacts are not passed on as inflation in other parts of the economy.
One study, by (Hamilton 2012), reviews the empirical literature on oil shocks and suggests
that the results are mixed, noting that some work finds less evidence for economic effects of oil
shocks or declining effects of shocks (Rasmussen and Roitman 2011) (Blanchard and Gali 2010),
while other work continues to find evidence regarding the economic importance of oil shocks.
For example, (Baumeister and Peersman 2013) find that an "oil price increase of a given size
seems to have a decreasing effect over time, but noted that the declining price-elasticity of
demand meant that a given physical disruption had a bigger effect on price and turned out to
have a similar effect on output as in the earlier data." Hamilton observes that "a negative effect
of oil prices on real output has also been reported for a number of other countries, particularly
when non-linear functional forms have been employed" (citing as examples (Kim 2012) and
(Engemann, Kliesen and Owyang 2011)). Alternatively, rather than a declining effect, (Ramey
and Vine 2010) find "remarkable stability in the response of aggregate real variables to oil
shocks once we account for the extra costs imposed on the economy in the 1970s by price
controls and a complex system of entitlements that led to some rationing and shortages."
Some of the literature on oil price shocks emphasizes that economic impacts depend on the
nature of the oil shock, with differences between price increases caused by a sudden supply loss
and those caused by rapidly growing demand. Recent analyses of oil price shocks have
confirmed that "demand-driven" oil price shocks have greater effects on oil prices and tend to
have positive effects on the economy while "supply-driven" oil shocks still have negative
economic impacts (Baumeister, Peersman and Van Robays 2010). (Kilian and Vigfusson 2014),
for example, assigns a more prominent role to the effects of price increases that are unusual, in
the sense of being beyond the range of recent experience. Kilian and Vigfusson also conclude
that the difference in response to oil shocks may well stem from the different effects of demand-
and supply-based price increases: "One explanation is that oil price shocks are associated with a
range of oil demand and oil supply shocks, some of which stimulate the U.S. economy in the
short-run and some of which slow down U.S. growth" (Kilian 2009).
The general conclusion that oil supply-driven shocks reduce economic output is also reached
in (Cashin, et al. 2014), which focused on 38 countries from 1979 to 2011. They state: "The
results indicate that the economic consequences of a supply-driven oil-price shock are very
different from those of an oil-demand shock driven by global economic activity and vary for oil-
importing countries compared to energy exporters." Cashin et al. continues ".. .oil importers
(including the U.S.) typically face a long-lived fall in economic activity in response to a supply-
driven surge in oil prices." But almost all countries see an increase in real output caused by an
oil-demand disturbance.
10-3
-------
Considering all of the recent energy security literature, EPA's assessment concludes that there
are benefits to the U.S. from reductions in its oil imports. There is some debate as to the
magnitude of energy security benefits from U.S. oil import reductions. However, differences in
economic impacts from oil demand and oil supply shocks have been distinguished, with oil
supply shocks resulting in economic losses in oil importing countries. The oil import premium
calculations in this analysis (described in Chapter 10.4.2) are based on price shocks from
potential future supply events. Oil supply shocks, which reduce economic activity, have been the
predominant focus of oil security issues since the oil price shocks/oil embargoes of the 1970s.
10.2 Review of Recent Energy Security Literature
There have also been a handful of recent studies that are relevant for the issue of oil security:
one by Resources for the Future (RFF), a study by Brown, two studies by Oak Ridge National
Laboratory (ORNL), and three studies by Newell and Prest, Bj0rnland et al. and Walls and
Zheng, on the responsiveness of U.S. tight oil to world oil price changes. We provide a review
and high-level summary of each of these studies below. In addition, we review the recent
literature on electricity security in the context of the wider use of PEVs.
10.2.1 Recent Oil Security Studies
The first studies on the energy security impacts of oil that we review are by Resources for the
Future (RFF), a study by Brown and two studies by Oak Ridge National Laboratory (ORNL).
The RFF study (Krupnick, et al. 2017) attempts to develop updated estimates of the
relationship among gross domestic product (GDP), oil supply and oil price shocks, and world oil
demand and supply elasticities. In a follow-on study, (Brown 2018) summarized the RFF study
results as well. The RFF work argues that there have been major changes that have occurred in
recent years that have reduced the impacts of oil shocks on the U.S. economy. First, the U.S. is
less dependent on imported oil than in the early 2000s due in part to the 'Tracking revolution"
(i.e., tight/shale oil), and to a lesser extent, increased production of renewable fuels such as
ethanol and biodiesel. In addition, RFF argues that the U.S. economy is more resilient to oil
shocks than in the earlier 2000s timeframe. Some of the factors that make the U.S. more resilient
to oil shocks include increased global financial integration and greater flexibility of the U.S.
economy (especially labor and financial markets), many of the same factors thatNordhaus and
Blanchard and Gali pointed to as discussed above.
In the RFF effort, a number of comparative modeling scenarios are conducted by several
economic modeling teams using three different types of energy-economic models to examine the
impacts of oil shocks on U.S. GDP. The first is a dynamic stochastic general equilibrium model
developed by (Balke and Brown 2018). The second set of modeling frameworks use alternative
structural vector autoregressive models of the global crude oil market (Kilian 2009), (Kilian and
Murphy 2014), (Baumeister and Hamilton 2019). The last of the models utilized is the U.S.
Energy Information Administration's National Energy Modeling System (NEMS).
Two key parameters are focused upon to estimate the impacts of oil shock simulations on U.S.
GDP: oil price responsiveness (i.e., the short-run price elasticity of demand for oil) and GDP
sensitivity (i.e., the elasticity of GDP to an oil price shock). The more inelastic (i.e., the less
responsive) short-run oil demand is to changes in the price of oil, the higher will be the price
impacts of a future oil shock. Higher price impacts from an oil shock result in higher GDP
10-4
-------
losses. The more inelastic (i.e., less sensitive) GDP is to an oil price change, the less the loss of
U.S. GDP with future oil price shocks.
For oil price responsiveness, RFF reports three different values: a short-run price elasticity of
oil demand from their assessment of the "new literature," -0.17; a "blended" elasticity estimate,
-0.05; and short-run oil price elasticities from the "new models" RFF uses, ranging from -0.20
to -0.35. The "blended" elasticity is characterized by RFF in the following way: "Recognizing
that these two sets of literature [old and new] represent an evolution in thinking and modeling,
but that the older literature has not been wholly overtaken by the new, Benchmark-E [the
blended elasticity] allows for a range of estimates to better capture the uncertainty involved in
calculating the oil security premiums."
The second parameter that RFF examines is the GDP sensitivity. For this parameter, RFF's
assessment of the "new literature" finds a value of-0.018, a "blended elasticity" estimate of-
0.028, and a range of GDP elasticities from the "new models" that RFF uses that range from -
0.007 to -0.027. One of the limitations of the RFF study is that the large variations in oil price
over the last fifteen years are believed to be predominantly "demand shocks": for example, a
rapid growth in global oil demand followed by the Great Recession and then the post-recession
recovery.
There have only been two recent situations where events have led to a potential significant
supply-side oil shock in the last several years. The first event was the attack on the Saudi
Aramco Abqaiq oil processing facility and the Khurais oil field. On September 14, 2019, a drone
and cruise missile attack damaged the Saudi Aramco Abqaiq oil processing facility and the
Khurais oil field in eastern Saudi Arabia. The Abqaiq oil processing facility is the largest crude
oil processing and stabilization plant in the world, with a capacity of roughly 7 MMBD or about
7 percent of global crude oil production capacity (U.S. EIA 2019). On September 16, the first
full day of commodity trading after the attack, both Brent and WTI crude oil prices surged by
$7.17/barrel and $8.34/barrel, respectively, in response to the attack, the largest price increase in
roughly a decade.
However, by September 17, Saudi Aramco reported that the Abqaiq plant was producing 2
MMBD, and they expected its entire output capacity to be fully restored by the end of September
(U.S. EIA 2019). Tanker loading estimates from third-party data sources indicated that loadings
at two Saudi Arabian export facilities were restored to the pre-attack levels (U.S. EIA 2019). As
a result, both Brent and WTI crude oil prices fell on September 17th, but not back to their
original levels. The oil price spike from the attack on the Abqaiq plant and Khurais oil field was
prominent and unusual, as Kilian and Vigfusson (2014) describe. While pointing to possible
risks to world oil supply, the oil shock was short-lived, and generally viewed by market
participants as being transitory, so it did not influence oil markets over a sustained time period.
The second situation is the set of events leading to the recent world oil price spike
experienced in 2022. World oil prices rose fairly rapidly at the beginning of 2022. For example,
as of January 3, 2022, the WTI crude oil price was roughly $76 per barrel (U.S. EIA 2023). The
WTI oil price increased to roughly $123 per barrel on March 8, 2022, a 62 percent increase (U.S.
EIA 2023). High and volatile oil prices in the first part of 2022 were a result of supply concerns
with Russia's invasion of Ukraine on February 24 contributing to crude oil price increases (U.S.
EIA 2022). Russia's invasion of Ukraine came after eight consecutive quarters of global crude
oil inventory decreases. The lower inventory of crude oil stocks was the result of rising economic
10-5
-------
activity after COVID-19 pandemic restrictions were eased. Oil prices drifted downwards
throughout the second half of 2022 and in the early part of 2023. Since both significant demand
and supply factors were influencing world oil prices in 2022, it is not clear how to evaluate
unfolding oil market price trends from an energy security standpoint. Thus, the attack of the
Abqaiq oil processing facility in Saudi Arabia and the unfolding events in the world oil market in
2022 do not currently provide enough empirical evidence to undertake an updated estimate of the
response of the U.S. economy to an oil supply shock of a significant magnitude.261
More recently, in its November 2023 Short-term Energy Outlook, EIA is forecasting global
oil production will increase by 1.0 million barrels per day in 2024 (U. EIA 2023). Ongoing
OPEC+ production cuts will offset production growth from non-OPEC countries and help
maintain a relatively balanced global oil market in 2024. The surprise attack by Hamas on Israel
on October 7, 2023, leading to the Hamas-Israel War, is leaving oil markets on edge, increasing
fears that fighting between Israel and Hamas may affect oil production in the Middle East.
Although the conflict between Israel and Hamas has not affected physical oil supply at this point,
uncertainties surrounding the conflict and other global oil supply conditions could put upward
pressure on crude oil prices in the coming months. EIA is forecasting the average price of Brent
crude oil will be $93/barrel in 2024.
A second set of recent studies related to energy security are from ORNL. In the first study,
(Uria-Martinez, et al. 2018) undertake a quantitative meta-analysis of world oil demand
elasticities based upon the recent economics literature. The ORNL study estimates oil demand
elasticities for two sectors (transportation and non-transportation) and by world regions (OECD
and Non-OECD) by meta-regression. To establish the dataset for the meta-analysis, the authors
undertake a literature search of peer-reviewed journal articles and working papers between 2000
and 2015 that contain estimates of oil demand elasticities. The dataset consisted of 1,983
elasticity estimates from 75 published studies. The study finds a short-run price elasticity of
world oil demand of-0.07 and a long-run price elasticity of world oil demand of-0.26.
The second relevant ORNL study from the standpoint of energy security is a meta-analysis
that examines the impacts of oil price shocks on the U.S. economy as well as many other net oil-
importing economies (Oladosu, et al. 2018). Nineteen studies after 2000 were identified that
contain quantitative/accessible estimates of the economic impacts of oil price shocks. Almost all
studies included in the review were published since 2008. The key result that the study finds is a
short-run oil price elasticity of U.S. GDP, roughly one year after an oil shock, of-0.021, with a
68 percent confidence interval of-0.006 to -0.036.
261 The Hurricanes Katrina/Rita in 2005 primarily caused a disruption in U.S. oil refinery production, with a more
limited disruption of some crude supply in the U.S. Gulf Coast area. Thus, the loss of refined petroleum products
exceeded the loss of crude oil, and the regional impact varied even within the U.S. The Katrina/Rita hurricanes were
a different type of oil disruption event than is quantified in the Stanford EMF risk analysis framework, which
provides the oil disruption probabilities that ORNL is using.
10-6
-------
10.2.2 Recent Tight (i.e.. Shale) Oil Studies
The discovery and development of U.S. tight (i.e., shale) oil reserves that started in the mid-
2000s could affect U.S. energy security in at least a couple of ways.262 First, the increased
availability of domestic supplies has resulted in a reduction of U.S. oil imports and an increasing
role of the U.S. as exporter of crude oil and petroleum-based products. In December 2015, the
40-year ban on the export of domestically produced crude oil was lifted as part of the
Consolidated Appropriations Act, 2016. Pub. L. 114-113 (Dec. 18, 2015). According to the
GAO, the ban was lifted in part due to increases in tight (i.e., shale) oil (U.S. GAO 2020).263
Second, due to differences in development cycle characteristics and average well productivity,
tight oil producers could be more price responsive than most other oil producers. However, the
oil price level that triggers a substantial increase in tight oil production appears to be higher in
2021-2022 relative to the 2010s as tight oil producers seek higher profit margins per barrel in
order to reduce the debt burden accumulated in previous cycles of production growth (Kemp
2021). Other factors such as cost inflation and supply chain constraints have contributed to the
slow pace of tight oil production growth in the early 2020s, despite high world oil prices.
Although some of those factors may be transitory, the muted production response of 2021-2022
suggests that tight oil producers (and their investors) are not likely to increase drilling in a quick,
coordinated manner in response to future potential world oil price spikes. For that reason, the
short-run price responsiveness assumed for U.S. tight oil for the estimation of the oil security
benefits of this final rule is the same as for other non-OPEC oil supplies.
U.S. crude oil production increased from 5.0 Million Barrels a Day (MMBD) in 2008 to an
all-time peak of 12.7 MMBD in 2023 (January through July) and tight oil wells have been
responsible for most of the increase (U.S. EIA 2023). Figure 10-1 below shows tight oil
production changes from various tight oil producing regions (i.e., Eagle Ford, Bakken etc.) in the
U.S. and the West Texas Intermediate (WTI) crude oil spot price. As illustrated in Figure 10-1,
the annual average U.S. tight oil production grew from 0.6 MMBD in 2008 to 7.8 MMBD in
2019 (U.S. EIA 2023). Growth in U.S. tight oil production during this period was only
interrupted in 2015-2016 following the world oil price downturn which began in mid-2014. The
second growth phase started in late 2016 and continued until 2020. The sharp decrease in
demand that followed the onset of the COVID-19 pandemic resulted in a 25 percent decrease in
tight oil production in the period from December 2019 to May 2020. U.S. tight oil production in
2020 and 2021 averaged 7.4 MMBD and 7.2 MMBD, respectively. More recently, in March
2023, tight oil production surpassed the previous historical maximum (8.37 MMBD in
November 2019) with 8.43 MMBD. Growth in tight oil production continued over the following
262 The Union of Concerned Scientist define tight oil as follows: "Tight oil is a type of oil found in impermeable
shale and limestone rock deposits. Also known as "shale oil," tight oil is processed into gasoline, diesel, and jet fuels
-just like conventional oil - but is extracted using hydraulic fracturing, or 'Tracking." (Union of Concerned
Scientists 2015).
263 According to the GAO, "Between 1975 and the end of 2015, the Energy Policy and Conservation Act directed a
ban on nearly all exports of U.S. crude oil. This ban was not considered a significant policy issue when U.S. oil
production was declining and import volumes were increasing. However, U.S. crude oil production roughly doubled
from 2009 to 2015, due in part to a boom in shale oil production made possible by advancements in drilling
technologies. In December 2015, Congress effectively repealed the ban, allowing the free export of U.S. crude oil
worldwide."
10-7
-------
months, reaching 8.57 MMBD in July 2023. Most of the 2023 growth has come from two
Permian producing regions: Spraberry and Bonespring.
Importantly, U.S. tight oil is considered the most price-elastic component of non-OPEC
supply due to differences between its development and production cycle and that of conventional
oil wells. Unlike conventional wells where oil starts flowing naturally after drilling, tight oil
wells require the additional step of fracking to complete the well and release the oil.264 Tight oil
producers keep a stock of drilled but uncompleted wells and can optimize the timing of the
completion operation depending on oil price expectations. Combining this decoupling between
drilling and production with the "front-loaded" production profile of tight oil - the fraction of
total output from a well that is extracted in the first year of production is higher for tight oil wells
than conventional oil wells - tight oil producers have a clear incentive to be responsive to prices
in order to maximize their revenues (Bjornland, Nordvik and Rohrer 2020).
10 T
-150
200S
2010
2012
2014
2016
2018
2020
2022
WTI
Producing Regions Price
Bakken Niobrara-Cod el I Bonespring Eagle Ford
(ND&MT) Hi (CO & WY) (TX & NM Permian) (TX)
H Spraberry I1|| Wolfcamp R ,.
(TX Permian) (TX & NM Permian)
Figure 10-1: U.S. tight oil production by producing regions (in MMBD) and West Texas
Intermediate (WTI) crude oil spot price (in U.S. Dollars per Barrel) Source: (U.S. EIA
2023) (U.S. EIA 2023).
Only in recent years have the implications of the "tight/shale oil revolution" been felt in the
international market where U.S. production of oil is rising to be roughly on par with Saudi
Arabia and Russia. Recent economics literature of the tight oil expansion in the U.S. has a
bearing on the issue of energy security as well. It could be that the large expansion in tight oil
has eroded the ability of OPEC to set world oil prices to some degree, since OPEC cannot
directly influence tight oil production decisions. Also, by affecting the percentage of global oil
supply controlled by OPEC, the growth in U.S. oil production may be influencing OPEC's
264 Hydraulic fracturing ("fracking") involves injecting water, chemicals, and sand at high pressure to open fractures
in low-permeability rock formations and release the oil that is trapped in them.
10-8
-------
degree of market power. But given that the tight oil expansion is a relatively recent trend, it is
difficult to know how much of an impact the increase in tight oil is having, or will have, on
OPEC behavior.
Three recent studies have examined the characteristics of tight oil supply that have relevance
for the topic of energy security. In the context of energy security, the question that arises is: can
tight oil respond to an oil price shock more quickly and substantially than conventional oil? If so,
then tight oil could potentially lessen the impacts of future oil shocks on the U.S. economy by
moderating the price increases from a future oil supply shock.
The first study considered, (Newell and Prest 2019), looks at differences in the price
responsiveness of conventional versus tight oil wells, using a detailed dataset of 150,000 oil
wells, during the timeframe of 2005-2017 in five major oil-producing states: Texas, North
Dakota, California, Oklahoma, and Colorado. For both conventional oil wells and tight oil wells,
Newell and Prest estimate the elasticities of drilling operations and well completion operations
with respect to expected revenues and the elasticity of supply from wells already in operation
with respect to spot prices. Combining the three elasticities and accounting for the increased
share of tight oil in total U.S. oil production during the period of analysis, they conclude that
U.S. oil supply responsiveness to prices increased more than tenfold from 2006 to 2017. They
find that tight oil wells are more price responsive than conventional oil wells, mostly due to their
much higher productivity, but the estimated oil supply elasticity is still relatively small. Newell
and Prest note that the tight oil supply response still takes more time to arise than is typically
considered for a "swing producer," referring to a supplier able to increase production quickly,
within 30-90 days. In the past, only Saudi Arabia and possibly one or two other oil producers in
the Middle East have been able to ramp up oil production in such a short period of time.
Another study, (Bj0rnland, Nordvik and Rohrer 2020), uses a well-level monthly production
data set covering more than 16,000 crude oil wells in North Dakota from February 1990 to June
2017 to examine differences in supply responses between conventional and tight oil. They find a
short-run (i.e., one-month) supply elasticity with respect to oil price for tight oil wells of 0.71,
whereas the one-month response of conventional oil supply is not statistically different from
zero. It should be noted that the elasticity value estimated by Bj0rnland et al. combines the
supply response to changes in the spot price of oil as well as changes in the spread between the
spot price and the 3-month futures price. (Walls and Zheng 2022) explore the change in U.S. oil
supply elasticity that resulted from the tight oil revolution using monthly, state-level data on oil
production and crude oil prices from January 1986 to February 2019 for North Dakota, Texas,
New Mexico, and Colorado. They conduct statistical tests that reveal an increase in the supply
price elasticities starting between 2008 and 2011 coinciding with the times in which tight oil
production increased sharply in each of these states. Walls and Zheng also find that supply
responsiveness in the tight oil era is greater with respect to price increases than price decreases.
The short-run (one-month) supply elasticity with respect to price increases during the tight oil
area ranges from zero in Colorado to 0.076 in New Mexico; pre-tight oil, it ranged from zero to
0.021.
The results from (Newell and Prest 2019), (Bj0rnland, Nordvik and Rohrer 2020), and (Walls
and Zheng 2022) all suggest that tight oil may have a larger supply response to oil prices in the
short-run than conventional oil, although the estimated short-run elasticity is still relatively
small. The three studies use datasets that end in 2019 or earlier. The responsiveness of U.S. tight
10-9
-------
oil production to recent price increases does not appear to be consistent with that observed
during the episodes of crude oil price increases in the 2010s captured in these three studies.
Despite an 80 percent increase in the WTI crude oil spot price from October 2020 to the end of
2021, Figure 10-1 shows that U.S. tight oil production has increased by only 8 percent in the
same period. It is a somewhat challenging period in which to examine the supply response of
tight oil to its price to some degree, given that the 2020-2021 time period coincided with the
COVID-19 pandemic. Previous tight oil production growth cycles were financed predominantly
with debt, at very low interest rates (McLean 2018). Most U.S. tight oil producers did not
generate positive cashflow (McLean 2018). As of 2021, U.S. tight oil producers have pledged to
repay their debt and reward shareholders through dividends and stock buybacks (Crowley and
Wethe 2021). These pledges translate into higher prices that need to be reached (or sustained for
a longer period) than in the past decade to trigger large increases in drilling activity.
In its first quarter 2022 energy survey, the Dallas Fed (Federal Reserve Bank of Dallas 2022)
asked oil exploration and production firms about the WTI price levels needed to cover operating
expenses for existing wells or to profitably drill a new well. The average breakeven price to
continue operating existing wells in the tight oil regions ranged from $23/barrel (bbl) to $35/bbl.
To profitably drill new wells, the required average WTI prices ranged from $48/bbl to $69/bbl.
For both types of breakeven prices, there was substantial variation across companies, even within
the same region. The actual WTI price level observed in the first quarter of 2022 was roughly
$95/bbl, substantially larger than the breakeven price to drill new wells. However, the median
production growth expected by the respondents to the Dallas Fed Energy Survey from the fourth
quarter of 2021 to the fourth quarter of 2022 is modest (6 percent among large firms and 15
percent among small firms). Investor pressure to maintain capital discipline was cited by 59
percent of respondents as the primary reason why publicly traded oil producers are restraining
growth despite high oil prices. The other reasons cited included supply chain constraints,
difficulty in hiring workers, environmental, social, and governance concerns, lack of access to
financing, and government regulations. Given the recent behavior of tight oil producers, we do
not believe that tight oil will provide additional significant energy security benefits in the
timeframe of this analysis, 2027-2055, due to its muted price responsiveness. The ORNL model
still accounts for the effect of U.S. tight oil production increases on U.S. oil imports and, in turn,
the U.S.'s energy security position.
Finally, despite continuing uncertainty about oil market behavior and outcomes and the
sensitivity of the U.S. economy to oil shocks, it is generally agreed that it is beneficial to reduce
petroleum fuel consumption from an energy security standpoint. The relative significance of
petroleum consumption and import levels for the macroeconomic disturbances that follow from
oil price shocks is not fully understood. Recognizing that changing petroleum consumption will
change U.S. imports, our quantitative assessment of the oil security costs of this rule focuses on
those incremental social costs that follow from the resulting changes in net imports, employing
the usual oil import premium measure used in the energy security literature.
10.2.3 Recent Electricity Security Studies
The International Energy Agency (IEA) defines energy security as the uninterrupted
availability of energy sources at affordable prices (IEA 2019). The energy security literature, first
developed in response to the oil shocks of the 1970s, is extensive. This literature mainly focuses
on the energy security benefits of reduced oil use, particularly oil imports. However, even though
10-10
-------
there is likely to be a substantial increase in the use of electricity from PEVs in the U.S., the
literature on the topic of the energy security implications of wider use of PEVs is somewhat
limited. We have not been able to identify any study that systematically quantifies the
differential energy security risks of using electricity versus petroleum-based fuels to power
vehicles in the U.S. Nonetheless, a review of existing, published studies provides information to
help assess the implications of the use of electricity as transportation fuel in light- and medium-
duty vehicles in the U.S. across multiple dimensions of energy security - affordability, price
stability, and resilience/reliability - as well as energy independence.265
Since the energy security literature has largely focused on the economic and national security
risks associated with oil imports, early studies considering the energy security benefits of PEVs
focus on the reduction in oil imports that results from widespread PEV adoption. (Michalek, et
al. 2011) quantifies this aspect of the energy security impacts of PEVs. The study focuses on the
benefits associated with a reduction in U.S. oil imports from the wider use of PEVs and provides
a direct estimate of the energy security benefits of using PEVs in the U.S. based on the amount
of oil PEVs displace over the lifetime of a typical PEV. They use a $0.34/gal (2010 dollars)
estimate of the avoided macroeconomic disruption costs/monopsony/military cost savings for oil
to calculate an energy security benefit of roughly $1,000 over the lifetime of a PEV. (Michalek,
et al. 2011) is similar to the approach used by EPA in past vehicle rulemakings: estimate the
displaced petroleum use and apply a security cost premium that draws on some of the same
studies that EPA uses. But EPA does not include monopsony impacts or quantify military cost
savings as benefits. The Michalek et al. study also does not account for electricity supply
stability.
10.2.3.1 Fuel Costs
Most of the cost comparisons of PEVs versus gasoline-powered vehicles in the literature are
total cost of ownership (TCO) studies, which compare the total cost of purchasing, owning, and
operating each type of vehicle for a specified number of years. They include the vehicle purchase
costs as well as annual operation (fees, fuel, and insurance) and maintenance costs.
Vehicles are refueled fairly frequently and increased fueling costs due to energy price spikes
are felt almost immediately by consumers, whereas the impact of price changes in components
and materials used to produce vehicles (e.g., alloys, batteries, etc.), which are also considered in
a TCO analysis, only impact consumers when purchasing a vehicle. Our focus in this Chapter is
on energy markets. Critical materials and the supply chains necessary for PEV production are,
therefore, outside of our intended scope in this discussion of energy security. See preamble
IV.C.7 and Chapter 3 of the RIA for a discussion of critical materials and PEV supply chains.
TCO studies of vehicles in the U.S. find that fuel costs are lower for PEVs than internal
combustion engine (ICE) vehicles. See, for examples, (Slowik, Isenstadt, et al., Assessment of
Light-duty Electric Vehicle Costs and Consumer Benefits in the United States in the 2022-2035
Time Frame 2022), (Liu, et al. 2021), (Lutsey and Nicholas 2019), and (Breetz and Salon 2018).
TCO studies tend to not explore in great detail the heterogeneity in fuel costs for PEV owners
265 Our discussion of "affordability" in this Chapter only considers fuel costs, including gasoline prices and charging
costs for PEVs. Vehicle purchase costs are not considered within the scope of our evaluation of energy security.
More discussion of consumer impacts in the context of PEVs is presented in Chapter 4 of this RIA.
10-11
-------
depending on geography and charging location or strategy, but other studies focus on the sources
of PEV fuel cost variability. For example, a 2017 brief by the Union of Concerned Scientists
examines the rates offered by electric utilities in the 50 largest U.S. cities and finds that all of
them offered at least one electricity rate that results in fuel savings for PEV owners compared to
a gasoline-powered vehicle, with median annual savings of $770 (Union of Concerned Scientists
2015). Clearly these savings depend on the prevailing price of petroleum fuels, which varies
widely over location and the assumed efficiency of the comparable gasoline vehicle.
One study, by (Borlaug, Salisbury, et al., Levelized Cost of Charging Electric Vehicles in the
United States 2020), performed a detailed analysis of PEV charging costs in the U.S. that takes
into consideration the type of charging equipment, a range of real-world electricity rates, and
frequency of charging at home versus workplace or public stations. They find that PEV fuel cost
savings over a 15-year period ranged from $3,000 to $10,500 (2019 dollars) for average U.S.
electricity and gasoline price projections, respectively, with additional variability across states
and depending on PEV lifetimes. The percentage of battery charging done at home versus using
public chargers is an important source of variability in the fuel costs of individual PEV owners.
Extracting charging rate information from a commercial database that includes records for more
than 30,000 U.S. public chargers, (Trinko, et al. 2021) reports mean rates of 28 cents/kWh for
Level 2 chargers and 32 cents/kWh for faster Direct Current Fast Charging chargers; in contrast,
the study reports a lower mean residential electric rate of 13 cents/kWh as of March 2021.
To date, residential charging access has been prevalent among PEV owners. (Y. Ge, C.
Simeone, et al., There's No Place Like Home: Residential Parking, Electrical Access, and
Implications for the Future of Electric Vehicle Charging Infrastructure 2021) find that the
percentage of PEVs with residential charging access is likely to become more uncertain as the
PEV market share of light-duty vehicles increases. They conducted a survey to gather detailed
information on residential parking availability, parking behavior, and electrical access by parking
location. Combining public data on housing stock and light-duty vehicle stock characteristics
with the survey results, the authors develop estimates of residential charging access percentages
for each housing type and a PEV adoption likelihood model using housing type, housing tenure
(owning versus renting), income, population density, and presence/absence of zero emission
vehicle incentives in the state of residence as explanatory variables. For PEV shares no greater
than 10 percent of total light-duty vehicles, residential charging access is estimated to range from
78 percent to 98 percent. For a 90 percent PEV share, the estimated residential charging access
percentage ranges from 35 percent to 75 percent. The higher end of the ranges represents a
scenario that requires modifications in parking behavior (e.g., parking in garage rather than
driveway) and installation of electrical access whenever possible, if not already available at the
residential parking location.
In a study for the California Public Utilities Commission, (Sieren-Smith, et al. 2021) projects
future fuel costs of PEVs in California for the 2020-2030 timeframe in comparison to gasoline-
powered vehicles. This study finds that there is wide spatial variability in fuel costs for PEVs and
there are substantial differences across individual electric utilities within California alone. The
study also finds that for customers with Time of Use (TOU) tariffs, charging a PEV regularly at
the off-peak rates (i.e., "managed charging" as opposed to "unmanaged charging") results in
significant fuel cost savings. With TOU tariffs or Time Variable Pricing, electricity prices
depend on the time of use, and change at set times and amounts through the day - generally with
higher prices in an afternoon peak period and lower prices in overnight off-peak hours (U.S.
10-12
-------
DOE 2022). The study also finds that PEV fuel costs are likely to be lower than gasoline-
powered vehicles' fuel costs across a variety of assumptions about projected gasoline and
electricity prices and managed/unmanaged PEV charging rates in California over the timeframe
of the analysis.
In the U.S., according to (Hardman, et al. 2021), the lowest income households spend 11.2
percent of their annual income on fuel, maintenance, and repairs of vehicles compared to all
other households that spend 4.5 percent of their annual income on these expenses. For the most
common use case in terms of PEV charging equipment (i.e., at-home charging), fuel costs in the
U.S. are lower for PEVs than gasoline-powered vehicles. Therefore, owning a PEV results in a
lower percentage of household income going toward that expense category. However, (Hardman,
et al. 2021) find that lower income households are less able to afford installation of residential
charging equipment and more likely to live in multi-unit dwellings without a designated parking
space and charging equipment. Thus, low-income households that purchase a PEV and have no
residential charging and, thus, rely primarily on public chargers, could potentially face higher
fuel costs and a larger overall energy burden (i.e., fraction of household income directed toward
energy costs) with a PEV than a gasoline-powered vehicle. The Inflation Reduction Act (IRA)
signed into law on August 16, 2022, can help reduce the costs for deploying charging
infrastructure (Inflation Reduction Act 2022). The IRA extends the Alternative Fuel Refueling
Property Tax Credit (Section 13404) through Dec 31, 2032, with modifications. Under the new
provisions, residents in low-income and rural areas would be eligible for a 30 percent credit for
the cost of installing residential charging equipment up to a $1,000 cap.
10.2.3.2 Fuel Price Stability/Volatility
One study, by (Melodia and Karlsson 2022), shows that the rate of inflation and volatility of
U.S. retail electricity prices have been historically much lower than for gasoline. Using consumer
price data from the Bureau of Labor Statistics from 1968 to 2022, the authors report that gasoline
was almost four times more volatile than electricity during that period. The diversity of the fuel
mix used to produce electricity and the stronger regulatory oversight of the U.S. electricity
sector, where residential electricity rates must meet a "just and reasonable" standard, are among
the reasons for the lower volatility of electricity prices versus gasoline prices. The authors also
discuss how renewable electricity generation can contribute to electricity price stability. First, the
cost profile of renewable resources such as wind and solar involves an initial large fixed-capital
investment but have no fuel costs once they are in operation, removing a key source of the price
volatility experienced by electricity generation plants that use fossil fuels. Moreover, wind and
solar resources are available much more widely across the globe than oil and gas resulting in
lower geopolitical supply risk - although some risk is still present through the critical materials
needed to produce renewable energy infrastructure components such as wind turbines, solar
panels, and electric batteries (Melodia and Karlsson 2022).
While (Melodia and Karlsson 2022) discuss the positive contribution that increased use of
renewables can make to electricity price stability, other authors consider how the process of
decarbonization in the energy sector might affect oil price stability. (Bordoff and O'Sullivan
2022) suggest that a smooth transition to clean energy in response to climate change may be
challenging and may result in more price volatility in oil markets. In other words, they suggest
that the transition to clean energy may be "jagged." According to the authors, the combination of
pressure on investors to divest from fossil fuels and uncertainty about the future of oil demand
10-13
-------
may raise concerns that oil investment levels may decrease in the future, leading to oil supplies
declining at a faster rate than oil demand falls - or declining even as oil demand continues to
rise. This outcome could produce more volatile oil prices. Also, in the early stages of the
transition to clean energy before oil demand declines significantly, the power of OPEC and other
non-competitive suppliers could be boosted by increasing their revenues, while giving OPEC
extra clout as a "swing producer" when world oil markets are tight.
10.2.3.3 Electricity Reliability/Resiliency
Reliability and resilience of electricity service are needed to ensure the "continuous
availability" of service that is required for a fuel to be considered secure. (U.S. DOE 2017)
defines the two terms as follows. Reliability is "the ability of the electric power sector to provide
a stable source of electricity to consumers, both households and businesses, under normal
operating conditions." Resilience is "the ability of the electric power sector to withstand and
recover from any disruptions created by extreme weather, cyberattack, terrorism, or other
unanticipated events." A reliable and resilient electricity sector is crucial for the U.S.'s national
security. The Department of Defense is the largest customer of the electricity grid in the U.S.
(U.S. DOE 2017). Also, the electricity sector is interconnected with many other types of critical
infrastructure - water systems, oil, natural gas, communications, information technology, and
financial services - crucial for the U.S. economy to function (U.S. DOE 2017). Standards and
metrics to track reliability are better established than those for resilience, which is concerned
with lower probability, high-consequence events (U.S. DOE 2017).
Electricity, while generally reliably provided in the U.S., is subject to periodic supply
disruptions (i.e., "electricity outages") due to a variety of factors including (but not limited to):
weather-related events such as hurricanes, heat waves/storms, wildfires; cybersecurity risks; and
system/equipment failures. On average, U.S. electricity customers experienced eight hours of
power outages in 2020, the most since the DOE's Energy Information Administration (EIA)
began collecting electricity reliability data in 2013 (U.S. EIA 2021). The Fourth National
Climate Assessment, released in 2018, concludes that "climate change will increasingly threaten
the U.S. energy supply via more frequent and long-lasting power outages that will broadly affect
critical energy infrastructure" (Zamuda, et al. 2018). It also states that extreme weather is already
the most frequent cause of electricity grid outages in the U.S. Electricity in the U.S. is provided
by a set of local and regional interconnected electric grids. Thus, electricity supply disruptions
are likely to result in electricity outages that are more local or regional in their nature in
comparison to petroleum disruptions, which commonly have national or, oftentimes, global
impacts. See Chapter 5.4 of the RIA for further discussion of grid reliability.
U.S. electric utilities follow long-term plans to ensure electricity reliability. These plans,
typically known as integrated resource plans, set out an investment roadmap to ensure sufficient
regional generation capacity and power purchases to meet the projected demand in their
electricity service areas. According to (Bistline 2021), although these long-term plans contribute
to electricity supply reliability, both resource planning and electric grid operation are becoming
more difficult due to overlapping layers of increased variability in electricity supply and demand.
For example, climate change is leading to an increase in the frequency and severity of extreme
weather events which affects both supply (e.g., droughts reducing hydropower generation) and
demand (e.g., record peak loads due to heat waves). Increased penetration of wind and solar also
results in significant fluctuations in electricity production at different time scales that need to be
10-14
-------
managed by electric grid operators and planners. Maintaining reliability of supply and price
stability under this new set of evolving conditions requires a range of technology, analysis, and
policy solutions (Bistline 2021).
As auto manufacturers respond to this final rule with increased sales of PEVs, U.S. electricity
demand is anticipated to increase. Overall, U.S. electricity demand is projected to increase by 12
Terawatt-hours (TWh) in 2028 (a 0.26 percent increase), 45 TWh in 2030 (a 0.94 percent
increase), 178 TWh in 2035 (a 3.43 percent increase), 322 TWh in 2040 (a 5.73 percent
increase), and 475 TWh in 2050 (a 7.3 percent increase). See Chapter 5 of this RIA for more
discussion of these estimates. Projections of PEV uptake will need to be accounted for by U.S.
electric utilities and transmission system operators in their resource planning processes.
At early levels of PEV adoption, the investments needed to shore up electric grid reliability
might first appear at the local distribution level. Early PEV sales to date have often happened in
clusters such that some neighborhoods have achieved large PEV penetrations even as PEV
market share remained lower at the regional or national level. The extent of distribution level
reliability impacts will depend on multiple factors: number of PEVs, PEV mix (BEVs/PHEVs),
type of charger used (Level 1, Level 2), and most importantly, whether charging is managed or
unmanaged. (Muratori 2018) evaluates the effect of uncoordinated PEV charging on residential
demand. The author finds that uncoordinated PEV charging leads to more pronounced and abrupt
load (i.e., electricity demand) peaks which shorten the life of distribution transformers. Using
detailed datasets of charging events at homes and public chargers in California to simulate future
PEV charging behavior (timing of charging and duration), (Jenn and Highleyman 2022)
conclude that in a scenario with 6 million PEVs in California (compared to approximately 1
million in 2021), more than 20 percent of distribution feeder circuits would experience loads
greater than their capacity, resulting in accelerated degradation of the distribution network
equipment and requiring upgrades to maintain adequate electricity grid reliability.
Another study, (Powell, et al. 2022), explores electric grid impacts in the U.S. portion of the
Western Interconnection grid in 2035 under scenarios with high penetration (greater or equal to
50 percent adoption) of light-duty PEVs. They find that the timing of the extra electricity
demand brought about by PEVs depends on charging behavior and is crucial to the magnitude of
the electric grid impacts. The authors develop a detailed model of charging behavior where
drivers are assigned to clusters based on combinations of the battery capacity of their PEVs,
number of miles driven per year, and access to charging infrastructure. The aggregated PEV
charging demand is then used as an input in a generation dispatch model that represents the
Western Interconnection 2035 grid by accounting for planned generation unit
additions/retirements, increasing baseline demand to reflect electrification of other sectors, and
multiplying solar generation by a factor of 3.5 and wind generation by a factor of 3 relative to
2019 levels.
The authors calculate the electric grid impacts for various scenarios regarding charging
controls and access to home and workplace charging infrastructure. All charging scenarios
assume unidirectional charging (i.e., no vehicle-to-grid flows). For the Western Interconnection,
given the high level of penetration of solar generation expected by 2035, daytime charging leads
to lower costs and emissions because it aligns better with the solar generation profile. Investing
in widespread access to workplace charging leads to lower peak net demand (i.e., peak demand
net of solar and wind generation), lower electricity grid storage capacity investment needs, less
10-15
-------
ramping-related costs from the operation of fossil fuel generators, and lower CO2 emissions per
mile driven by PEVs. Since the U.S. electricity grid is composed of a set of regional electricity
grids with different fuel mixes, the charging infrastructure and charging schedules that will best
match and balance the extra electricity demand from PEVs with electricity supply will vary on a
region-by-region basis.
Large and abrupt electricity demand peaks due to PEV charging deserve special attention
when they are linked to extreme weather events that can also disrupt the demand and supply of
electricity. (Feng, et al. 2020) explore the mobility implications of vehicle fleets with high PEV
penetration rates during extreme weather events triggering evacuation orders. They simulate the
evacuation traffic flow during Hurricane Irma and compare electricity demand if all evacuating
vehicles were PEVs with the transmission capacity in the Florida electric grid. They conclude
that up to a fleet-wide PEV penetration rate of 45 percent could have been supported by the
existing transmission network during that evacuation scenario. The more general insights from
the analysis include: 1) fleetwide PEV penetration levels of up to 45 percent can be helpful
during an evacuation scenario to alleviate gasoline shortages, 2) PHEVs are especially valuable
during those events as drivers can start the evacuation trip using their battery and fill their
gasoline tanks away from the population centers that experience gasoline shortages when an
evacuation order is announced, and 3) development of disaster-optimized charging schedules
would be crucial to avoid surges of power during an extreme event such as a hurricane as PEV
penetration increases. Under the final standards, the penetration rates of PEVs in the stock of
U.S. light- and medium-duty vehicles are projected to remain below a 45 percent rate until the
2040s. By the 2040s, there should be sufficient lead time for the U.S. electricity grid to expand
and accommodate increasingly higher penetration rates of PEVs.
With PEVs becoming an increasingly significant portion of vehicles on the road in the U.S.,
some losses in overall U.S. output, measured in terms of a loss in U.S. gross domestic product
(GDP), will likely result from electricity supply disruptions. The losses in U.S. output will be
determined by the extent and duration of the future electricity supply disruptions, the flexibility
of the additional electricity demand from PEVs, and whether PEVs can help avoid or ameliorate
electricity supply disruptions. Given the local and regional nature of electricity supply
disruptions and noting that the U.S. is projected to produce almost all of its own electricity (see
discussion below), the losses in U.S. output from future electricity supply disruptions will likely
be lower than output losses that have resulted from world oil supply disruptions with the
widespread use of gasoline-powered vehicles. Higher electricity payments in the event of a U.S.
electricity supply disruption will be transferred to other electricity producers in the U.S., not to
foreign suppliers, as was the case in past oil supply disruptions, which will reduce the effective
cost to the U.S. economy. However, more analysis is needed to make a definitive statement
about the net effect of this final rule on expected GDP losses from future electricity and oil
supply disruptions or price spikes. Estimates of disruption probabilities and associated U.S.
macroeconomic disruption costs are available for oil but not for electricity. Without an estimate
of electricity disruption probabilities and expected U.S. output losses, it is difficult to conduct
assessments of the size and types of potential investments, or initiatives in the U.S. electricity
sector, that could mitigate or adapt to those losses.
Although PEVs can pose challenges for electricity supply reliability if PEV charging is not
coordinated, PEVs can also potentially provide an important source of electricity storage, which
could help to improve the overall functioning of the U.S. electricity grid in terms of the
10-16
-------
reliability and availability of electricity over time. See Chapter 5.4 of the RIA for more
discussion on this topic. With a bidirectional connection to the electricity grid that enables
vehicle-to-grid (V2G) flows, PEVs can act as a storage resource that provides energy during
electric peak demand hours by discharging their batteries while parked. PEVs can also provide
services to the electrical grid such as frequency and voltage regulation or act as electricity
reserves, ready to supply energy in response to an outage at an electricity generation facility. In
addition, PEVs can be used to provide electricity to home residences in the event of an electricity
supply disruption. Managed bidirectional flows of energy from a large PEV fleet could also be
particularly valuable to integrate higher levels of variable renewables (wind and solar) into the
electricity generation mix (Yilmaz and Krein 2013).
The wider use of electricity in U.S. vehicles also provides both short- and long-run fuel
substitution opportunities for vehicle owners facing high and volatile world oil prices. For
example, drivers of PHEVs can switch to using more electricity during an oil price shock
(Lemoine 2010). Also, during an oil shock, a wider penetration of PEVs will allow for a short-
run reduction in oil use by multi-vehicle households that can drive their PEVs more, rather than
using their gasoline-powered vehicles. Flexibility is achieved when drivers have options to shift
to electricity, and the responsiveness of oil demand to the oil price (i.e., the elasticity of demand
for oil) increases. These benefits occur because there is more substitutability between electricity
and oil in end-use fuel use. With electricity supply disruptions, on the other hand, multi-vehicle
households could also switch to driving their gasoline-powered vehicles more. Households with
only one vehicle, dedicated to gasoline or electricity, are likely to be the most affected by volatile
oil prices and electricity outages, since they cannot substitute among vehicles or fuels in
response to changing oil prices and the availability of electricity, as multi-vehicle households or
owners of PHEVs can.
10.2.3.4 Energy Independence
The goal of U.S. energy independence is generally equated with the elimination of all U.S.
imports of petroleum and other foreign sources of energy, but more broadly, it is the elimination
of U.S. sensitivity to the variations in the price and supply of foreign sources of energy (Greene
2010). (Grove 2008) and (Stein 2013) promote the idea that the wider use of PEVs can bring
about U.S. energy independence by substituting electricity for oil to power vehicles in the U.S.
As Grove/Stein note, the physical characteristics of oil and electricity can have very different
consequences for energy independence. Oil is a commodity that is globally traded. In
comparison, Grove labels electricity as "sticky": in other words, "it stays in the continent where
it is produced." As a result, global electricity markets are not nearly as linked or interconnected
as global oil markets. The interconnectedness of the oil market means that price shocks are
transmitted globally, but it also contributes to its resilience. Oil tankers can be redirected to those
destinations where price signals reveal that their value is highest. In contrast, the volume of
electricity that can be rerouted across regions in response to an emergency is strictly limited by
the number and configuration of electricity transmission interconnections.
The wider use of PEVs in U.S. light-duty and medium-duty vehicles will likely result in the
substitution of one fuel, oil, with significant imports and which is subject to global price shocks,
for another fuel, electricity, which is almost exclusively produced in the U.S. and has different
and an independent set of local and regional factors influencing its reliability and resiliency. As
(Bordoff and O'Sullivan 2022) point out, electricity is much more likely to be produced locally
10-17
-------
and regionally; less than three percent of global electricity was traded across international
borders in 2018, compared with two-thirds of global oil supplies in 2014. As a result, the greater
use of electricity as transportation fuel will move the U.S. toward the goal of energy
independence.
U.S. energy security analysis has traditionally focused on the benefits of the reduction of U.S.
oil imports. However, even when oil imports get close to zero, energy security concerns remain
for oil because of the global, integrated nature of the oil market. Unless the U.S. entirely
disengages from international oil trade, oil price shocks starting anywhere in the world will
continue to be transmitted to oil prices in the U.S. and those price shocks will continue to have
adverse impacts on U.S. households. An increased movement toward electrification does not
eliminate energy security concerns, but it does reduce vulnerability to them. Supply shocks for
electricity also happen, but they are typically of a different nature than oil shocks: they are local
or regional instead of global, and they may involve a combination of electricity outages and/or
retail electricity price increases.
Recent geopolitical events are an example of how the energy price and energy price stability
attributes in U.S. energy security remain an important concern even after the U.S. has become a
net exporter of crude oil and petroleum products. (Bordoff and O'Sullivan 2022) suggest that
energy security will join climate change as a top concern for policymakers as a result of the
Russian-Ukrainian war, which has disrupted energy supplies and increased global energy prices.
They argue that these dual priorities - energy security and climate change - are poised to reshape
national energy planning, energy trade flows, and the broader global economy. One consequence
of the Russian-Ukrainian War, according to Bordoff and O'Sullivan, is that countries across the
world will increasingly be looking inward, prioritizing domestic energy production and regional
cooperation even as they transition to net-zero carbon emissions. These changes will likely be
defined by greater, not less, government intervention in the world's energy sector.
10.3 Electricity Security Impacts
Addressing the issue of U.S. energy security, this section offers comparisons of electricity and
gasoline as transportation fuels in terms of cost per mile driven and their relative price stability
and volatility. In the U.S. during the past decade, the cost per mile driven for a new PEV
charging at home has been consistently lower than that for a new gasoline-powered vehicle using
regular gasoline. This result is robust to the spatial variation in relative electricity and gasoline
prices in different U.S. states. The impact of fuel costs on consumers is not only about average
fuel cost levels but also fuel cost stability. On the metric of fuel cost stability, retail electricity
also has fared better than gasoline because retail electricity prices have been more predictable
and less volatile for vehicle owners than gasoline prices. The predictability is partly a result of
the electricity rate setting process - most consumers pay a set tariff (i.e., electricity price) that
only changes at monthly or annual intervals. The section also presents data to support the idea
that an increased use of electricity as a transportation fuel in U.S. LMDVs moves the U.S.
toward greater energy independence.
10-18
-------
10.3.1 Recent Fuel Costs for Gasoline-Powered Vehicles Compared to PEVs in the
U.S.
10.3.1.1 National (i.e., U.S.) Analysis
To compare fuel costs of PEVs versus gasoline-powered vehicles, the relevant units are
dollars per mile instead of dollars per gallon of gasoline equivalent (or other energy content unit)
because of the higher end-use efficiency of the electric motor relative to the internal combustion
engine (ICE). This is a central feature of the comparison between PEVs and gasoline-powered
vehicles. The relative cost of gasoline and electric fuel in the U.S. will depend on three main
factors: the efficiency of the vehicle; the prevailing prices of gasoline and electricity (electricity
prices being more stable over time), and the market and charger location in which the PEV is
recharged (electricity costs tend to vary significantly across states to a greater degree than
gasoline prices, and commercial recharging costs are higher than residential charging costs).
Most PEV charging to date in the U.S. uses at-home chargers, and thus EPA's analysis of fuel
costs hinges on prices observed by U.S. households: retail, regular gasoline prices (in dollars per
gallon) and retail residential electricity rates (in cents per kilowatt-hour). As PEV adoption
extends to drivers without at-home charging capabilities in the future, commercial charging rates
will play a larger role in a national analysis of PEV fuel costs, but home recharging is expected
to continue to play a dominant role.
Comparing fuel costs for PEVs and gasoline-powered light-duty vehicles requires converting
retail prices into a common unit (U.S. cents per mile driven) that accounts for the differences in
energy content between gasoline and electricity as well as the higher efficiency of electric
drivetrains relative to internal combustion engines, expressed as fuel economies (miles per gallon
of gasoline equivalent (gge)).266 The fuel economy data used to compute fuel costs per mile
driven are on-road new vehicle values by model year (i.e., the average fuel economy across all
sold new gasoline-powered light-duty vehicles or PEVs of a same model year). The data for
PEVs includes only battery electric vehicles (BEVs), but also applies to plug-in hybrid electric
vehicles (PHEVs) for the miles driven in electric vehicle mode, and the data for gasoline vehicles
includes conventional hybrids.267 On-road fuel economy increased from 22.2 miles per gge in
2011 to 24.6 miles per gge in 2021 for new gasoline-powered light-duty vehicles and from 97
miles per gge to 112.8 miles per gge for new PEVs.
Figure 10-2 shows the average U.S. fuel cost per mile driven for two vehicle-fuel
combinations, gasoline-powered light-duty vehicles using regular gasoline and PEVs charging
at-home at the residential retail rate, and Figure 10-3 presents the same information for a subset
of individual states in the U.S.
266 The conversion factor from kilowatt-hours to gasoline gallon equivalents (gge) is 33.705 kWh/gge
(https ://www3. epa.gov/otaq/gvg/learn-more-technology. htm).
267 It should be noted that the time unit for the fuel economy data, the "model year", does not coincide exactly with a
calendar year.
10-19
-------
gasoline LDV, PEV, 100% at-home charging
regular gasoline at U.S. average residential retail price
Figure 10-2: Average U.S. fuel cost per vehicle mile driven of gasoline-powered vehicles
and PEVs from 2011 to 2021 Sources: Electricity prices: (U.S. EIA 2022); Gasoline prices:
(U.S. EIA 2022); Fuel economies: (U.S. EPA 2022).
Monthly fuel cost per mile driven has been consistently and substantially lower for new PEVs
than new gasoline-powered light-duty vehicles. The average fuel cost per mile driven from
January 2011 to December 2021 was 13.7 cents per mile for new gasoline vehicles using regular
gasoline and 4.6 cents per mile for a new PEV charged at-home 100 percent of the time at the
average residential retail rate. The average annual fuel savings of new PEVs in comparison to a
new gasoline vehicle using regular gasoline over the ten-year timeframe of 2011 to 2021 was
$1,260. We recognize that, to date, the bulk of PEVs sold tend to be in the small or mid-size car
segments and, thus, more energy efficient. This is evolving as more PEV models are offered. For
Model Year 2022, an analysis of fuel costs for every light-duty vehicle model shows that most
PEV models have lower fuel costs than most gasoline-powered models regardless of vehicle
class and size (U.S. DOE 2022).
While vehicle size and prevailing oil prices matter, the lower fuel cost per mile driven for
PEVs is largely a result of the much higher efficiency of electric drivetrains relative to internal
combustion engines. Comparing U.S. electricity and gasoline prices on a dollar per unit-energy
basis, residential electricity has actually been more expensive than retail gasoline over the last
10-20
-------
decade: the 2011—2021 averages were 2.6 cents per megajoule (MJ) for regular gasoline and 4.0
cents per MJ for residential retail electricity.268
10.3.1.2 State-Level Analysis
The fuel cost per mile driven for new PEVs was lower than the fuel cost for new gasoline
light-duty vehicles in all the states shown in Figure 10-3 (see below) and in every month from
2011 to the end of 2021. However, as stated above, the fuel cost savings do vary significantly
across states. The average savings in fuel cost per mile driven for a new PEV versus a new
gasoline vehicle ranged from 6.7 cents in Massachusetts to 10.2 cents in California and 11.9
cents in Washington. For the other three states depicted in Figure 10-3, Texas, Ohio and Florida,
the fuel savings averaged 8-9 cents per mile. Both California and Massachusetts have some of
the highest electricity residential retail rates in the U.S. The large savings afforded by PEVs in
California result from that state having higher retail gasoline prices than the rest of the states in
Figure 10-3. The savings are even larger for Washington because of a combination of high
gasoline prices and low electricity rates due to Washington's relative abundance of hydroelectric
power resources (U.S. EIA 2022). Assuming that new gasoline-powered cars and new PEVs are
both driven -14,000 miles per year, the annual average fuel cost savings in the first year of
vehicle operation during this period would have ranged from $933 in Massachusetts to $1,643 in
Washington (Davis and Boundy 2022). While vehicle use typically declines with age, the decline
is slow, and 15 years later the average car would still provide 62 percent of these annual savings
(Davis and Boundy 2022).
o
CM
c
0>
>
25-
Califomia
A k i
20-
-iWvVvv— —
J
15-
—
Sr^\
10-
* T
V
5 -
o-
Florida
Massachusetts
^A I
" i i \ i 'i mi
Ohio
Texas
Washington
K fV V <0 Co \
t-C ^ ^ ^
^ <9
-------
Figure 10-3: Fuel cost per mile driven by gasoline-powered vehicles and PEVs for six
states from 2011 to 2021 Sources: Electricity prices: (U.S. EIA 2022); Gasoline prices: (U.S.
EIA 2022); Fuel economies: (U.S. EPA 2022).
10.3.2 Fuel Price Stability/Volatility
Absolute differences in fuel costs between PEVs and ICE vehicles, discussed above, are an
important aspect of the "affordability" component of IEA's definition of energy security, but fuel
price stability is another important consideration from the consumer's perspective.269 While U.S.
retail electricity prices vary widely with location, charging equipment, and charging behavior,
they are generally more stable over time than U.S. gasoline prices. Figure 10-4 displays the
monthly percentage price changes for U.S. retail gasoline and residential electricity. The monthly
change in U.S. average residential electricity prices was less than 5 percent (in absolute value) in
every month during the 2011-2021 period. For regular gasoline, prices changed up or down by
more than 5 percent in 30 percent of months over that period. The volatility of monthly U.S.
retail prices from January 2011 to December 2021 was 21 percent for residential electricity
prices and 60 percent for regular gasoline prices.270
269 The International Energy Agency (IEA) defines energy security as the uninterrupted availability of energy
sources at affordable prices. (IEA 2019).
270 Volatility is calculated as the standard deviation of the monthly price returns multiplied by the square root of the
number of periods (months).
10-22
-------
15%
C
E -15%
2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022
— electricity, residential sector — regular gasoline
Figure 10-4: Monthly percentage changes in U.S. retail electricity and gasoline prices from
2011 to 2021 Source: (U.S. EIA 2022).
Another desirable attribute of PEVs for fuel cost stability is that PEVs diversify and thereby
help stabilize total road-vehicle fuel costs. Diversification benefits are gained when the prices of
the two fuels do not move together. In fact, historically when oil prices increased, electricity
prices have tended to decrease, and vice-versa. Looking at fuel price trends over roughly the last
decade, from January 2011 to December 2021, monthly U.S. residential electricity prices have
been negatively correlated, -0.37, with monthly U.S. average gasoline prices.211 A negative
correlation helps plug-in hybrid electric vehicle (PHEV) owners and multi-vehicle households
with access to gasoline light-duty vehicles and PEVs, and the nation as a whole, by diversifying
transportation fuel cost risk. During all of the 2011-2021 period, the cost of at-home PEV
charging resulted in lower fuel costs than gasoline refilling. The value of a household being able
to switch between PEVs and gasoline-powered vehicles depending upon prevailing fuel prices
(or between electricity and gasoline for a PHEV), is sometimes labeled the "real option value."
Lemoine (2010) finds that using a real option value framework instead of a di scounted cash flow
analysis raises the retail price at which the extra battery capacity of a PHEV (relative to a HEV)
pays for itself by $50/kWh (15 percent). The extra value results from the real option value
approach accounting for (1) gasoline price uncertainty and (2) the nonlinearity in payoffs that the
PHEV fuel flexibility provides - the payoff is either 0 or positive, but never negative. Real
option value could increase if the residential electricity costs of PEVs increase, or commercial
211 The estimated correlation coefficient is a Pearson correlation coefficient with a p-value of 0.0069.
10-23
-------
recharging costs decrease, and the relative ranking of home or commercial PEV charging versus
gasoline refueling costs changes more frequently in the future as oil prices fluctuate.272
10.3.3 Energy Independence
The substitution of electricity for oil for powering U.S. vehicles will reduce U.S reliance on
fuel imports. Although the U.S. has become a net exporter of crude oil and petroleum liquids, it
still imports significant volumes of crude oil to meet the preferred barrel specifications of
domestic refineries. See Table 10-2 below for estimates of U.S. oil import reductions from this
final LMDV GHG (2027-2032) rule. Figure 10-5 shows that the U.S. has been a very small net
importer of electricity over the most recent decade: net U.S. imports accounted for an average of
only 1.2 percent of total U.S. electricity use from 2011 to 2022. The EIA projects net U.S.
imports of electricity to decrease further from that average percentage in the next decades across
all the Annual Energy Outlook (AEO) 2023 scenarios. By 2050, the AEO scenarios project net
U.S. electricity imports to range from 0.6 percent in the Low Zero-Carbon Technology Cost
scenario to 0.9 percent in the High Zero-Carbon Technology Cost scenario.
272 "Real option value" analysis applies the concepts used to value the financial assets called "options" to
investments in certain real/physical assets. Unlike traditional discounted cashflow analysis which states that
investment in a project/asset should only happen if its expected net present value is greater than zero, real option
analysis takes into account the extra value that can be realized when cashflows are uncertain and the asset holder can
choose between the different options. In the LMDV case considered here with PEVs and gasoline-powered vehicles,
real option value results when households can switch between PEVs and gasoline-powered vehicles when fuel costs
fluctuate. Vehicle switching in this case allows households to purchase the least-costly fuel.
10-24
-------
1.8
Historical
Projections
1.6-
1.4-
1.2-
a>
O)
CO
c 1.0-
-------
10.4 Oil Security Impacts
10.4.1 U.S. Oil Import Reductions
In this section, we compare oil import reductions from this final rule with an assessment of
overall U.S. oil market trends. Over the timeframe of analysis of this final rule, 2027-2055, the
U.S. Department of Energy's (DOE) Energy Information Administration's (EIA) Annual Energy
Outlook (AEO) 2023 (Reference Case) projects that the U.S. will be both an exporter and an
importer of crude oil273 (U.S. EIA 2023). The U.S. produces more light crude oil than its
refineries can refine. Thus, the U.S. exports lighter crude oil and imports heavier crude oils to
satisfy the needs of U.S. refineries, which are configured to efficiently refine heavy crude oil.
U.S. crude oil exports are projected to be relatively stable, between 2.9 and 3.4 MMBD, from
2027 through 2050. See Table 10-2 below. U.S. crude oil imports, meanwhile, are projected to
range between 6.6 and 7.2 MMBD between 2027 and 2050. The AEO 2023 also projects that
U.S. net oil refined product exports will grow from 5.8 MMBD in 2027 to 6.7 MMBD in 2045
before dropping off somewhat to 6.2 MMBD in 2050.
U.S. oil consumption is estimated to have decreased from 19.8 MMBD in 2019 to 17.5
MMBD in 2020 and 19.1 MMBD in 2021 due to social distancing and quarantines that limited
personal mobility as a result of the COVID-19 pandemic (U.S. EIA 2022).274 AEO 2023 projects
that U.S. oil consumption will continue to increase from 18.6 MMBD in 2027 to 18.9 MMBD in
2050 (U.S. EIA 2023). It is not just U.S. crude oil imports alone, but both imports and
consumption of petroleum from all sources and their role in economic activity, that exposes the
U.S. to risk from price shocks in the world oil price. During the 2027-2055 timeframe, the U.S.
is projected to continue to consume significant quantities of oil and to rely on significant
quantities of crude oil imports. As a result, U.S. oil markets are expected to remain tightly linked
to trends in the world crude oil market.
In Chapter 8, EPA estimates changes in U.S. petroleum consumption as a result of this final
rule. EPA uses an oil import reduction factor to estimate how changes in U.S. refined product
demand from this rule (i.e., changes in U.S. oil consumption) influences U.S. net oil imports
(i.e., changes in U.S. oil imports). After carefully reviewing comments on refinery throughput
and in consultation with DOE and NHTSA, EPA is updating its assessment of the impact of this
final rule on U.S. refinery throughput and, in turn, the air quality impacts from refinery
emissions. Instead of estimating that U.S. refineries would largely reduce their production in
response to reduced refined product demand from this rule, we are now estimating that U.S.
refinery output would decline by half (50 percent) of that reduced demand, while increases in
refined product exports (i.e., equivalently a decline in net refined product imports) would
account for the other half (50 percent) of that reduced demand. We also look at an additional
case where U.S. refinery throughput would be maintained by 80 percent as a result of increases
in refined product exports, while 20 percent of the refinery throughput would be reduced. We
chose this sensitivity as an assumption that falls between our central case where U.S refinery
273 The AEO 2023 projects oil market trends through 2050. The timeframe for EPA's analysis of this final rule is
from 2027 to 2055. Thus, we report oil market trends to 2050 based upon AEO 2023 in Table 10-2. We report U.S.
oil import reductions through 2055 in Table 10-2 as well.
274 Calculated using EIA's Monthly Energy Review series "Petroleum Consumption (Excluding Biofuels) Annual"
(Table 1.3) and "Petroleum Consumption Total Heat Content Annual" (Table A3).
10-26
-------
output would be reduced by half (50 percent) and the case where there would be no effect on
U.S. refining output (100 percent). See Chapter 8 of the RIA for more discussion of how EPA is
updating its refinery throughput assumptions and, in turn, air quality impacts from refinery
emissions, as a result of this rule. See Section 21 of the Response to Comment document for
EPA's response to comments on EPA's update of the oil import reduction factor.
Since EPA's refinery throughput assumptions are being updated for this final rule, this will
influence EPA's estimate of the oil import reductions and, in turn, the energy security benefits
estimated in this analysis. For the DRIA, a summary table was docketed that contained the
estimates of the oil import reduction factor. Table 10-1 shows that for a reduction in refined
product estimated by AEO's 2021 Low Economic Growth Case relative to the Reference Case,
83.7 percent of the reduced product demand is attributed to reduced imported crude oil, while 7
percent is attributed to reduced net imported products - resulting in the 90.7 percent oil import
reduction factor. Global (i.e., rest of the world) oil demand is not changed in the Low Economic
Case compared to the Reference Case, so the comparison between the AEO Reference Case and
the Low Economic Growth Case is only in the overall pattern of U.S. oil demand changes.
Table 10-1: Oil Import Reduction Factor, Average Over Years 2027 to 2050.
AEO 2021 AEO 2023
Percent of imported crude oil 83.7 84.8
Pc rcc nl rcducl io n i n do mcsl ic c rude oil 9.3 10.3
Percent reduction in net imported products 7.0 4.8
Total percentage of imported petroleum 90.7 89.6
For the final rule, the same methodology based on the AEO 2023 would result in an 89.6
percent oil import reduction factor - 84.8 percent of which would be due to reduced imported
crude oil and 4.8 percent would be due to reduced net imported products.
Use of the two AEO cases cited above estimates a large reduction in U.S. refinery throughput
- AEO 2021 estimates that 93 percent (83.7+9.3) of the reduced product demand would be
attributed to reduced throughput at U.S. refineries. Based on AEO 2023, the reduction in U.S.
refinery throughput would be 95.1 percent (84.8+10.3).
However, for the final rulemaking, we are estimating that U.S. refineries would not reduce
their throughput to the same extent. Instead, for a given reduction in a volume of gasoline and
diesel fuel demand, 50 percent of that reduced demand will result in reduced production by U.S.
refineries, while for the other 50 percent, refineries will continue to operate and will increase
their refined product exports (i.e., reduce their net refined product imports). Thus, we needed a
way to estimate the energy security impacts of the final rule, assuming that U.S. refiners will
continue producing domestic fuels at a much higher level associated with the 50/50 assumption.
Since we are now estimating that in response to reduced refined product demand, half of that
reduced demand will be reduced production from U.S. refineries and, for the other half,
refineries will continue to operate and increase their refined product exports, two different
methods for estimating the oil import reduction factor are being used. The portion of reduced
refinery demand projected to result in reduced refinery throughput can be represented by the oil
import reduction factor estimated by the two AEO 2023 cases, 89.6 percent. Conversely, the
balance of reduced refinery demand which U.S. refineries keep operating can be represented by
10-27
-------
the oil import reduction factor which, by definition, will be 100 percent, since refineries will
increase their refined product exports. Thus, the oil import reduction factor is estimated by the
following equation:
Oil Import Reduction Factor = 89.6% x 0.5 + 100% x 0.5 = 94.8%
If the 80/20 percent refinery throughput assumption is utilized, the oil import reduction factor
is estimated by the following equation:
Oil Import Reduction Factor = 89.6% x 0.2 + 100% x 0.8 = 97.9%
Based upon the changes in oil consumption estimated by EPA and the 94.8 percent oil import
reduction factor, the reductions in U.S. oil imports as a result of this final rule are estimated in
Table 10-2 below for the 2027-2055 timeframe.275 Included in Table 10-2 are estimates of U.S.
crude oil exports and imports, net oil refined product exports, net crude oil and refined petroleum
product exports and U.S. oil consumption for the years 2027-2050 based on the AEO 2023 (U.S.
EIA 2023).
Table 10-2: Projected trends in U.S. crude oil exports/imports, net refined oil product
exports, net crude oil and refined petroleum product imports, oil consumption and U.S. oil
import reductions resulting from the final rule from 2027 to 2055 (MMBD).a
; 2027
2030
; 2032
; 2035
2040
2045
2050
U.S. Crude Oil Exports
: 3.3
3.4
3.4
3.4
= 3.2
: "3.2
2.9
U.S. Crude Oil Imports
6.9
I 7.0
7.1
7.1
T 7.2
7.1
6.6
U.S. Net Refined Petroleum Product
5.8
i 6.0
6.1
6.4
6.7
6.7
6.2
Exports'1
U.S. Net Crude Oil and Petroleum Product
; 2.3
2.4
; 2.5
2.8
2.8
2.9
; 2.7
Exports
U.S. Oil Consumption1"
18.6
18.4
18.3
18.2
18.2
18.5
18.9
Reduction in U.S. Oil Imports from the
0.0035
0.15
0.36
0.83
1.5
1.9
2.1
Final Standards1'1
Table Notes:
a The AEO 2023 Reference Case, Table A11. Values have been rounded off from the AEO 2023, so the totals may not add up to the AEO estimates,
b Calculated from AEO 2023 Table A11 as Net Product Exports minus Ethanol, Biodiesel, Renewable Diesel, and Other Biomass-derived Liquid Net
Exports.
c Calculated from AEO 2023 Table A11 as "Total Primary Supply" minus "Biofiiels".
d Oil import reductions differ from estimates in Table 8-40, Impacts on oil consumption and oil imports. Final standards, due to rounding.
10.4.2 Oil Security Premiums Used for this Final Rule
The total energy security benefits of this final light- and medium-duty vehicle GHG rule are
calculated based upon U.S. net oil import reductions multiplied by the oil security premiums
estimated for this rule. In the preceding section (Chapter 10.4.1), we presented estimates of U.S.
275 The AEO 2023 projects oil market trends through 2050. The timeframe for EPA's analysis of this final rule is
from 2027 to 2055. Thus, we report oil market trends to 2050 based upon AEO 2023 in Table 10-2. We report U.S.
oil import reductions through 2055 in Table 10-2 as well.
10-28
-------
oil import reductions from this rule. In the section below, we present estimates of the oil security
premiums used for this rule.
In order to understand the energy security implications of reducing U.S. oil imports, EPA has
worked with Oak Ridge National Laboratory (ORNL), which has developed approaches for
evaluating the social costs and energy security implications of oil use. The energy security
estimates provided below are based upon a peer-reviewed methodology developed at ORNL
(Leiby 2008). This ORNL study is an updated version of the approach used for estimating the
energy security benefits of U.S. oil import reductions developed in a 1997 ORNL report (Leiby,
Jones, et al. 1997). This same approach was first used to estimate energy security benefits for the
2010 RFS2 final rule (75 FR 14670) and the 2010 final rulemaking to establish light-duty vehicle
greenhouse gas emission standards and corporate average fuel economy standards for MY 2012-
2016 vehicles (75 FR 25324). ORNL has updated this methodology regularly for EPA to account
for updated projections of future energy market and economic trends reported in the U.S. EIA's
AEO. For this final rule, EPA updated the ORNL methodology using the AEO 2023.
The ORNL methodology is used to compute the oil import premium (concept defined in
Chapter 10.1) per barrel of imported oil. The values ofU.S. oil import premium components
(macroeconomic disruption/adjustment costs and monopsony components) are numerically
estimated with a compact model of the oil market by performing simulations of market outcomes
using probabilistic distributions for the occurrence of oil supply shocks, calculating marginal
changes in economic welfare with respect to changes in U.S. oil import levels in each of the
simulations, and summarizing the results from the individual simulations into a mean and 90
percent confidence intervals for the import premium estimates. The macroeconomic
disruption/adjustment import cost component is the sum of two parts: the marginal change in
expected import costs during disruption events and the marginal change in gross domestic
product due to the disruption. The monopsony component is the long-run change in U.S. oil
import costs as the level of oil import changes.
For this final rule, EPA is using oil import premiums that incorporate the oil price projections
and energy market and economic trends, particularly global regional oil supplies and demands
(i.e., the U.S./OPEC/rest of the world), from the AEO 2023 into its model.276 EPA only considers
the avoided macroeconomic disruption/adjustment oil import premiums (i.e., labeled
macroeconomic oil security premiums below) as costs, since we consider the monopsony
impacts stemming from changes in U.S. oil imports: transfer payments. In previous EPA rules
when the U.S. was projected by EIA to be a net importer of crude oil and petroleum-based
refined products, monopsony impacts represented reduced payments by U.S. consumers to oil
producers outside of the U.S. There was some debate among economists as to whether the U.S.
276 The oil market projection data used for the calculation of the oil import premiums came from AEO 2023,
supplemented by the latest EIA international projections from the Annual Energy Outlook (AEO)/International
Energy Outlook (IEO) 2021. Global oil prices and all variables describing U.S. supply and disposition of petroleum
liquids (domestic supply, tight oil supply fraction, imports, demands) as well as U.S. non-petroleum liquids supply
and demand are from AEO 2023. Global and OECD Europe supply/demand projections as well as OPEC oil
production share are from IEO 2021. The need to combine AEO 2023 and IEO 2021 data arises due to two reasons:
(a) EIA stopped including Table 21 "International Petroleum and Other Liquids Supply, Disposition, and Prices" in
the U.S.-focused Annual Energy Outlook after 2019, (b) EIA does not publish complete updates of the IEO every
year.
10-29
-------
exercise of its monopsony power in oil markets, for example from the implementation of EPA's
rules, was a "transfer payment" or a "benefit." Given the redistributive nature of this monopsony
impact from a global perspective, and since there are no changes in resource costs when the U.S.
exercises its monopsony power, some economists argued that it is a transfer payment. Other
economists argued that monopsony impacts were a benefit since they partially address, and
partially offset, the market power of OPEC. In previous EPA rules, after weighing both
countervailing arguments, EPA concluded that the U.S.'s exercise of its monopsony power was a
transfer payment, and not a benefit (U.S. EPA 2016).
In the timeframe covered by this final rule, the U.S.'s oil trade balance is projected to be quite
a bit different than during the time periods covered in many previous EPA rules. Starting in
2020, the U.S. became a net exporter of crude oil and refined oil products, and the U.S. is
projected to continue to be a net exporter of crude oil and refined petroleum products in the
timeframe of analysis covered by the final standards, 2027-2055. As a result, reductions in U.S.
oil consumption and, in turn, U.S. oil imports, still lower the world oil price modestly. But the
net effect of the lower world oil price is now a decrease in revenue for U.S. exporters of crude oil
and refined petroleum products, instead of a decrease in payments to foreign oil producers. The
argument that monopsony impacts address the market power of OPEC is no longer appropriate.
Thus, we continue to consider the U.S. exercise of monopsony power to be transfer payments.
We also do not consider the effect of this final rule on the costs associated with existing energy
security policies (e.g., maintaining the Strategic Petroleum Reserve or strategic military
deployments), which are discussed below.
In addition, EPA and ORNL have worked together to revise the oil import premiums based
upon recent energy security literature. Based upon EPA and ORNL's review of the recent energy
security literature, EPA is assessing its macroeconomic oil security premiums for this final rule.
The recent economics literature (discussed in Chapter 10.2.1) focuses on three factors that can
influence the macroeconomic oil security premiums: the price elasticity of oil demand, the GDP
elasticity in response to oil price shocks, and the impacts of the U.S. tight oil boom. We discuss
each factor below and provide a rationale for how we are developing estimates for the first two
factors for the macroeconomic oil security premiums being used in this final rule. We are not
accounting for how U.S. tight oil is influencing the macroeconomic oil security premiums in this
final rule, other than how tight oil significantly reduces the need for U.S. oil imports.
First, we assess the price elasticity of demand for oil. In previous EPA light-duty vehicle
rulemakings (i.e., Model Year 2012-2016, Model Year 2017-2025), EPA used a short-run
elasticity of demand for oil of-0.045 (U.S. EPA/NHTSA 2010) (U.S. EPA/NHTSA 2012). In the
most recent EPA rule setting GHG emissions standards for passenger cars and light trucks in
model years 2023 through 2026, we used a short-run elasticity of demand for oil of -0.07, an
update of previously used elasticities based on the below considerations (U.S. EPA 2021). For
this rule, we continue to use the elasticity value of -0.07.
From the RFF study, the "blended" price elasticity of demand for oil is -0.05. The ORNL
meta-analysis estimate of this parameter is -0.07. We find the elasticity estimates from what RFF
characterizes as the "new literature," -0.175, and from the "new models" that RFF uses, -0.20 to
-0.33, somewhat high. We believe it would be surprising if short-run oil demand responsiveness
has changed in a dramatic fashion.
10-30
-------
The ORNL meta-analysis estimate encompasses the full range of the economics literature on
this topic and develops a meta-analysis estimate from the results of many different studies in a
structured way, while the RFF study's "new models" results represent only a small subset of the
economics literature's estimates. We believe using a short-run price elasticity of demand for oil
of -0.07 is more appropriate.277 This increase has the effect of lowering the macroeconomic oil
security premium estimates undertaken by ORNL for EPA.
Second, we consider the elasticity of GDP to an oil price shock. In previous EPA vehicle
rulemakings (i.e., Model Year 2012-2016, Model Year 2017-2025), EPA used an elasticity of
GDP to an oil shock of-0.032 (U.S. EPA/NHTSA 2010) (U.S. EPA/NHTSA 2012). In the most
recent EPA rule setting GHG emissions standards for passenger cars and light trucks through
model years 2023 through 2026, we used an elasticity of GDP of-0.021, an update of previously
used elasticities based on the below considerations (U.S. EPA 2021). For this rule, we continue
to use the elasticity value of-0.021.
The RFF "blended" GDP elasticity is -0.028, the RFF's "new literature" GDP elasticity is -
0.018, while the RFF "new models" GDP elasticities range from -0.007 to -0.027. The ORNL
meta-analysis GDP elasticity is -0.021. We believe that the ORNL meta-analysis value is
representative of the recent literature on this topic since it considers a wider range of recent
studies and does so in a structured way. Also, the ORNL meta-analysis estimate is within the
range of GDP elasticities of RFF's "blended" and "new literature" elasticities. For this final rule,
EPA is using a GDP elasticity of-0.021, a 34 percent reduction from the GDP elasticity used
previously (i.e., the -0.032 value). This GDP elasticity is within the range of RFF's "new
literature" elasticity, -0.018, and the elasticity EPA has used in previous rulemakings, -0.032,
but lower than RFF's "blended" GDP elasticity, -0.028. This decrease has the effect of lowering
the macroeconomic oil security premium estimates. For U.S. tight oil, EPA has not made any
adjustments to the ORNL model, given the limited tight oil production response to rising world
oil prices in the recent 2021-2023 timeframe.278 Increased tight oil production still results in
energy security benefits though, through its impact of reducing U.S. oil imports in the ORNL
model.
Table 10-3 below provides estimates of EPA's macroeconomic oil security premium
estimates for 2027-2055. The macroeconomic oil security premiums are relatively steady over
the time period of this final rule at $3.73/barrel (9 cents/gallon) in 2027 and $3.92/barrel (9
cents/gallon) in 2030, $4.22/barrel (10 cents per gallon) in 2035, $4.62/barrel (11 cents per
gallon) in 2040, and $5.22/barrel (12 cents/gallon) in 2050 and 2055 (in 2022 U.S. dollars).
277 EPA and ORNL have worked together to develop an updated estimate of the short-run elasticity of demand for
oil for use in the ORNL model.
278 The short-run oil supply elasticity assumed in the ORNL model is 0.06 and is applied to production from both
conventional and tight (i.e., shale) oil wells.
10-31
-------
Table 10-3: Macroeconomic oil security premiums for 2027-2055 for final rule
(2022$/barrel).ab
Calendar Year
Macroeconomic Oil
Security Premiums (range)
2027
$3.73
($0.51 -$7.02)
2028
$3.78
($0.51 -$7.15)
2029
$3.87
($0.54-$7.31)
2030
$3.92
($0.51 -$7.46)
2031
$4.00
($0.55 - $7.62)
2032
$4.05
($0.53 - $7.77)
2033
$4.11
($0.47 - $7.93)
2034
$4.16
($0.44 - $8.07)
2035 :
$4.22
($0.45 - $8.20)
2036
$4.28
($0.44 - $8.29)
2037
$4.35
($0.47 - $8.40)
2038
$4.44
($0.52-$8.55)
2039
$4.50
($0.53 - $8.66)
2040
$4.62
($0.65-$8.85)
2041
$4.73
($0.70 - $9.04)
2042
$4.77
($0.69-$9.15)
2043
$4.82
($0.67 - $9.27)
2044
$4.85
($0.66 - $9.35)
2045
$4.91
($0.68 - $9.43)
2046
$4.98
($0.71 - $9.52)
2047 i
$5.09
($0.82 - $9.68)
2048
$5.14
($0.85 - $9.79)
2049 !
$5.16
($0.82-$9.85)
2050
$5.22
($0.91 - $9.89)
2051 b
$5.22
($0.91 - $9.89)
2052 b
$5.22
($0.91 - $9.89)
2053 b
$5.22
($0.91 - $9.89)
2054 b
$5.22
($0.91 - $9.89)
2055 b
$5.22
($0.91 - $9.89)
a The top values in each cell are mean values. Values in parentheses are 90 percent confidence interval.
b The AEO 2023 only provides oil market trend estimates to 2050. We use the same macroeconomic oil security premium for the years from
2050 to 2055 as the value for 2050.
10-32
-------
10.4.3 Cost of Existing U.S. Oil Security Policies
An additional often-identified component of the full economic costs of U.S. oil imports is the
costs to the U.S. taxpayers of existing U.S. energy security policies. The two primary examples
are maintaining the Strategic Petroleum Reserve (SPR) and maintaining a military presence to
help secure a stable oil supply from potentially vulnerable regions of the world.
The SPR is the largest stockpile of government-owned emergency crude oil in the world.
Established in the aftermath of the 1973/1974 oil embargo, the SPR provides the U.S. with a
response option should a disruption in commercial oil supplies threaten the U.S. economy
(Energy Policy and Conservation Act 1975). Emergency SPR drawdowns have taken place in
1991 (Operation Desert Storm), 2005 (Hurricane Katrina), 2011 (Libyan Civil War), and 2022
(War in Ukraine) (U.S. DOE 2022). All of these releases have been in coordination with releases
of strategic stocks from other International Energy Agency (IEA) member countries. In the first
four months of 2022, using the statutory authority under Section 161 of the Energy Policy and
Conservation Act, the U.S. President directed the U.S. DOE to conduct two emergency SPR
drawdowns in response to ongoing oil supply disruptions. The first drawdown resulted in a sale
of 30 million barrels in March 2022 (U.S. DOE 2022). The second drawdown, announced in
April, authorized a total release of approximately one MMBD from May to October 2022 (U.S.
DOE 2022). In 2023, the DOE sold 26 million barrels of oil between April and June (U.S. DOE
2023). A total of 246.6 million barrels were released from the SPR from January 2022 to July
2023. By the end of July 2023, the SPR stock level was 346.8 million barrels (the lowest level
since August 1983). To start replenishing the stock, the SPR office purchased 10.23 million
barrels through competitive solicitations conducted between May and November of 2023, for
deliveries from August 2023 to February 2024. While the costs for building and maintaining the
SPR are more clearly related to U.S. oil use and imports, historically these costs have not varied
in response to changes in U.S. oil import levels. Thus, while the effect of the SPR in moderating
price shocks is factored into the analysis that EPA is using to estimate the macroeconomic oil
security premiums, the cost of maintaining the SPR is excluded.
We have also considered the possibility of quantifying the military benefits components of
energy security but have not done so here for several reasons. The literature on the military
components of energy security has described four broad categories of oil-related military and
national security costs, all of which are hard to quantify. These include possible costs of U.S.
military programs to secure oil supplies from unstable regions of the world, the energy security
costs associated with the U.S. military's reliance on petroleum to fuel its operations, possible
national security costs associated with expanded oil revenues to "rogue states" and relatedly the
foreign policy costs of oil insecurity.
Of these categories listed above, the one that is most clearly connected to petroleum use and
is, in principle, quantifiable is the first: the cost of military programs to secure oil supplies and
stabilize oil supplying regions. There is an ongoing literature on the measurement of this
component of energy security, but methodological and measurement issues - attribution and
incremental analysis - pose two significant challenges to providing a robust estimate of this
component of energy security. The attribution challenge is to determine which military programs
and expenditures can properly be attributed to oil supply protection, rather than some other
objective. The incremental analysis challenge is to estimate how much the petroleum supply
protection costs might vary if U.S. oil use were to be reduced or eliminated. Methods to address
10-33
-------
both of these challenges are necessary for estimating the effect on military costs arising from a
modest reduction (not elimination) in oil use attributable to this final rule.
Since "military forces are, to a great extent, multipurpose and fungible" across theaters and
missions and because the military budget is presented along regional accounts rather than by
mission, according to (Crane, et al. 2009), the allocation to particular missions is not always
clear. Approaches taken usually either allocate "partial" military costs directly associated with
operations in a particular region or allocate a share of total military costs (including some that are
indirect in the sense of supporting military activities overall) (Koplow and Martin 1998).
The challenges of attribution and incremental analysis have led some to conclude that the
mission of oil supply protection cannot be clearly separated from others, and the military cost
component of oil security should be taken as near zero (Moore, Behrens and Blodgett 1997).
(Stern 2010), on the other hand, argues that many of the other policy concerns in the Persian
Gulf follow from oil, and the reaction to U.S. policies taken to protect oil. Stern presents an
estimate of military cost for Persian Gulf force projection, addressing the challenge of cost
allocation with an activity-based cost method. He uses information on actual naval force
deployments rather than budgets, focusing on the costs of carrier deployment. As a result of this
different data set and assumptions regarding allocation, the estimated costs are much higher,
roughly 4 to 10 times, than other estimates. Stern also provides some insight on the analysis of
incremental effects, by estimating that Persian Gulf force projection costs are relatively strongly
correlated to Persian Gulf petroleum export values and volumes. Still, the issue remains of the
marginality of these costs with respect to Persian Gulf oil supply levels, the level of U.S. oil
imports, or U.S. oil consumption levels.
One study, by (Delucchi and Murphy 2008), seeks to deduct from the cost of Persian Gulf
military programs the costs associated with defending U.S. interests other than the objective of
providing more stable oil supply and price to the U.S. economy. Excluding an estimate of cost
for missions unrelated to oil, and for the protection of oil in the interest of other countries,
Delucchi and Murphy estimated military costs for all U.S. domestic oil interests of between $24-
$74 billion annually. Delucchi and Murphy assume that military costs from U.S. oil import
reductions can be scaled proportionally, attempting to address the incremental issue.
Another study, by (Crane, et al. 2009), considers force reductions and cost savings that could
be achieved if oil security were no longer a consideration. Taking two approaches and guided by
post-Cold War force draw downs and by a top-down look at the current U.S. allocation of
defense resources, they concluded that $75—$91 billion, or 12-15 percent of the current U.S.
defense budget, could be reduced. Finally, an Issue Brief by Securing America's Future Energy
(SAFE) (2018) found a conservative estimate of approximately $81 billion per year spent by the
U.S. military protecting global oil supplies (SAFE 2018). This is approximately 16 percent of the
recent U.S. Department of Defense's budget. Spread out over the 19.8 million barrels of oil
consumed daily in the U.S. in 2017, SAFE concludes that the implicit subsidy for all petroleum
consumers is approximately $11.25 per barrel of crude oil, or $0.28 per gallon. According to
SAFE, a more comprehensive estimate suggests the costs could be greater than $30 per barrel, or
over $0.70 per gallon.
As in the examples above, an incremental analysis can estimate how military costs would vary
if the oil security mission is no longer needed, and many studies stop at this point. It is
substantially more difficult to estimate how military costs would vary if U.S. oil use or imports
10-34
-------
are partially reduced, as is projected to be a consequence of this final rule. Partial reduction of
U.S. oil use likely diminishes the magnitude of the security problem, but there is uncertainty that
supply protection forces and their costs could be scaled down in proportion, and there remains
the associated goal of protecting supply and transit for U.S. allies and other importing countries,
if they do not decrease their petroleum use as well. We are unaware of a robust methodology for
assessing the effect on military costs of a partial reduction in U.S. oil use. Therefore, we are
unable to quantify this effect resulting from the projected reduction in U.S. oil use attributable to
this final rule.
10.4.4 Oil Security Benefits of the Final Rule
Estimates of the total annual oil security benefits of the final standards are based on the
ORNL oil import premium methodology with updated oil import premium estimates reflecting
the recent energy security literature and using the AEO 2023. Annual per-gallon benefits are
applied to the reductions in U.S. crude oil and refined petroleum product imports. We do not
consider military cost impacts or the monopsony effect of U.S. crude oil and refined petroleum
product import changes on the energy security benefits of this final rule. The energy security
benefits of this final rule are presented in Table 9-7 of Chapter 9, Non-emission benefits
associated with the final standards. For EPA's assessment of the energy security benefits of a
more and a less stringent alternative for this final rule, see Chapter 9 of the RIA.
10-35
-------
Chapter 10 References
Balke, Nathan S., and Stephen Brown. 2018. "Oil supply shocks and the U.S. economy: An
estimated DSGE model." Energy Policy 116: 357-372.
doi:https://doi.org/10.1016/j.enpol.2018.02.027.
Baumeister, Christiane, and Gert Peersman. 2013. "Time-Varying Effects of Oil Supply Shocks
on the US Economy." American Economic Journal: Macroeconomics 5 (4): 1-28.
Baumeister, Christiane, and James D. Hamilton. 2019. "Structural Interpretation of Vector
Autoregressions with Incomplete Identification: Revisiting the Role of Oil Supply and Demand
Shocks." American Economic Review 109 (5): 1873-1910. doi:10.1257/aer.20151569.
Baumeister, Christiane, Gert Peersman, and Ine Van Robays. 2010. "The Economic
Consequences of Oil Shocks: Differences across Countries and Time." RBA Annual Conference
Papers, Reserve Bank of Australia. Sydney.
https://www.rba.gov.au/publications/confs/2009/baumeister-peersman-vanrobays.html.
Bistline, John. 2021. "Variability in Deeply Decarbonized Electricity Systems." Environmental
Science & Technology 55 (9): 5629-5635. doi:10.1021/acs.est.0c06708.
Bj0rnland, Hilde C., Frode Martin Nordvik, and Maximilian Rohrer. 2020. "Supply flexibility in
the shale patch: Evidence from North Dakota." Journal of Applied Econometrics 36 (3): 273-
292. doi: 10.1002/j ae.2808.
Blanchard, Olivier J, and Jordi Gall. 2010. "The Macroeconomic Effects of Oil Price Shocks:
Why are the 2000s so different from the 1970s?" In International Dimensions of Monetary
Policy, 373-421. University of Chicago Press.
Bohi, Douglas R., and David Montgomery. 1982. "Social Cost of Imported Oil and U.S. Import
Policy." Annual Review of Energy 7: 37-60. doi: 10.1146/annurev.eg.07.110182.000345.
Bordoff, Jason, and Meghan L O'Sullivan. 2022. "Green Upheaval: The New Geopolitics of
Energy." Foreign Affairs, January, https://www.foreignaffairs.com/articles/world/2021-ll-
3 0/geopolitics-energy-green-upheaval.
—. 2022. "The New Energy Order: How Governments Will Transform Energy Markets."
Foreign Affairs, July, https://www.foreignaffairs.eom/articles/energy/2022-06-07/markets-new-
energy-order.
Borlaug, Brennan, Shawn Salisbury, Mindy Gerdes, and Matteo Muratori. 2020. "Levelized Cost
of Charging Electric Vehicles in the United States." Joule 4 (7): 1470-1485.
doi: 10.1016/j.joule.2020.05.013.
Breetz, Hanna L, and Deborah Salon. 2018. "Do electric vehicles need subsidies? Ownership
costs for conventional, hybrid, and electric vehicles in 14 U.S. cities." Energy Policy 120: 238-
249. doi: 10.1016/j.enpol.2018.05.038.
Broadman, Harry G. 1986. The social cost of imported oil, Energy Policy 14 (3): 242-252.
10-36
-------
Broadman, Harry G, and William H Hogan. 1988. The Numbers Say Yes, The Energy Journal 9
(3): 7-29.
Brown, Stephen P.A. 2018. "New estimates of the security costs of U.S. oil consumption."
Energy Policy 113: 171-192. doi:https://doi.org/10.1016/j.enpol.2017.11.003.
Cashin, Paul, Kamiar Mohaddes, Maziar Raissi, and Mehdi Raissi. 2014. "The differential
effects of oil demand and supply shocks on the global economy." Energy Economics 44: 113-
134. doi:https://doi.org/10.1016/j.eneco.2014.03.014.
Crane, Keith, Andreas Goldthau, Michael Toman, Thomas Light, Stuart E. Johnson, Alireza
Nader, Angela Rabasa, and Harun Dogo. 2009. Imported Oil and U.S. National Security. Rand
Corporation.
Crowley, Kevin, and David Wethe. 2021. "Shale Bets on Dividends to Match Supermajors,
Revive Sector." Bloomberg, August 2. https://www.bloomberg.com/news/articles/2021-08-
02/shale-heavyweights-shower-investors-with-dividends-on-oil-rally .
Davis, Stacy C, and Robert G Boundy. 2022. Transportation Energy Data Book Edition 40.
ORNL/TM-2022/2376, Oak Ridge National Laboratory, Table 3.14 Annual Mileage for Cars and
Light Trucks by Vehicle Age.
Delucchi, Mark A., and James J. Murphy. 2008. "US military expenditures to protect the use of
Persian Gulf oil for motor vehicles." Energy Policy (36): 2253-2264.
doi:0.1016/j.enpol.2008.03.006.
EIA, U.S. 2023. Short-term Energy Outlook. November 7th.
EMF. 1982. World Oil: EMF 6 Summary Report. Stanford, CA: Stanford Energy Modeling
Forum.
1975. "Energy Policy and Conservation Act." Public Law No. 94-163. December 22.
Engemann, Kristie M, Kevin L Kliesen, and Michael T Owyang. 2011. "Do Oil Shocks Drive
Business Cycles? Some U.S. and International Evidence." Federal Reserve Bank of St. Louis.
Working Paper Series. No. 2010-007D. doi:10.20955/wp.2010.007.
Federal Reserve Bank of Dallas. 2022. Dallas Fed Energy Survey: Special Questions. March 9.
https://www.dallasfed.Org/research/surveys/des/2022/2201.aspx#tab-questions.
Feng, Kairui, Ning Lin, Siyuan Xian, and Mikhail V Chester. 2020. "Can we evacuate from
hurricanes with electric vehicles?" Transportation Research Part D 86: 102458.
doi: 10.1016/j .trd.2020.10245 8.
Ge, Yanbo, Christina Simeone, Andrew Duvall, and Eric Wood. 2021. There's No Place Like
Home: Residential Parking, Electrical Access, and Implications for the Future of Electric Vehicle
Charging Infrastructure. Technical Report NREL/TP-5400-81065, National Renewable Energy
Laboratory, https://www.nrel.gov/docs/fy22osti/81065.pdf.
Greene, David L. 2010. "Measuring energy security: Can the United States achieve oil
independence?" Energy Policy 38 (4): 1614-1621. doi: 10.1016/j.enpol.2009.01.041.
10-37
-------
Greene, David L. 2010. "Measuring energy security: Can the United States achieve oil
independence?" Energy Policy 38 (4): 1614-1621. doi:10.1016/j.enpol.2009.01.041.
Grove, Andy. 2008. "Our Electric Future." American Enterprise Institute, July 10.
https://www.aei.org/articles/our-electric-future/.
Hamilton, James D. 2012. "Oil Prices, Exhaustible Resources, and Economic Growth." In
Handbook of Energy and Climate Change.
https://econweb.ucsd.edu/~jhamilto/handbook_climate.pdf.
Hardman, Scott, Kelly L Fleming, Eesha Khare, and Mahmoud M Ramadan. 2021. "A
perspective on equity in the transition to electric vehicles." MIT Science Policy Review 2: 46-54.
doi: 10.3 8105/spr. e 1 Ordoaoup.
Hogan, William W. 1981. "Import Management and Oil Emergencies." In Energy and Security,
261-301. Cambridge: Ballinger Publishing Co.
IEA. 2019. Energy security. December 2. Accessed 12 2022. https://www.iea.org/areas-of-
work/ensuring-energy-security.
2022. "Inflation Reduction Act." Public Law No. 117-169. August 16.
ISO New England. 2022. Resource Mix. https://www.iso-ne.com/about/key-stats/resource-mix.
Jenn, Alan, and Jake Highleyman. 2022. "Distribution grid impacts of electric vehicles: A
California case study." iScience25 (1). doi: 10.1016/j.isci.2021.103686.
Kemp, John. 2021. "U.S. Shale Restraint Pushes Oil Prices to Multi-year High." Reuters, June 4.
https://www.reuters.com/business/energy/us-shale-restraint-pushes-oil-prices-multi-year-high-
kemp-2021 -06-04.
Kilian, Lutz. 2009. "Not All Oil Price Shocks Are Alike: Disentangling Demand and Supply
Shocks in the Crude Oil Market." American Economic Review 99 (3): 1053-1069.
https://www.aeaweb.org/articles?id=10.1257/aer.99.3.1053.
Kilian, Lutz, and Daniel P. Murphy. 2014. "The Role of Inventories and Speculative Trading in
the Global Market for Crude Oil." Journal of Applied Econometrics 29: 454-478.
doi:https://doi.org/10.1002/jae.2322.
Kilian, Lutz, and Robert J. Vigfusson. 2014. "The Role of Oil Price Shocks in Causing U.S.
Recessions." Center for Financial Studies. Working Paper Series 460.
Kim, Dong Heon. 2012. "What is an oil shock? Panel data evidence." Empirical Economics 43:
121-143. doi:10.1007/s00181-011-0459-y.
Koplow, Douglas, and Aaron Martin. 1998. Fueling Global Warming: Federal Subsidies to Oil in
the United States. Washington D.C.: Greenpeace.
Krupnick, Alan, Richard Morgenstern, Nathan Balke, Stephen P. A. Brown, Ana Maria Herrera,
and Shashank Mohan. 2017. Oil Supply Shocks, US Gross Domestic Product, and the Oil
Security Premium. Resources for the Future, https://www.rff.org/publications/reports/oil-supply-
shocks-us-gross-domestic-product-and-the-oil-security-premium/.
10-38
-------
Leiby, Paul N. 2008. Estimating the Energy Security Benefits of Reduced U.S. Oil Imports.
ORNL/TM-2007/028, Oak Ridge National Laboratory.
Leiby, Paul N, Donald W Jones, T Randall Curlee, and Russell Lee. 1997. Oil Imports: An
Assessment of Benefits and Costs. Oak Ridge, Tennessee: Oak Ridge National Laboratory.
Lemoine, Derek. 2010. "Valuing Plug-In Hybrid Electric Vehicles' Battery Capacity Using a
Real Options Framework." Energy Journal. doi:10.2139/ssrn.1322164.
Liu, Zhe, Juhyun Song, Joseph Kubal, Naresh Susarla, Kevin W Knehr, Ehsan Islam, Paul
Nelson, and Shabbir Ahmed. 2021. "Comparing total cost of ownership of battery electric
vehicles and internal combustion engine vehicles." Energy Policy 158: 112564-112577.
doi: 10.1016/j.enpol.2021.112564.
Lutsey, Nic, and Michael Nicholas. 2019. Update on electric vehicle costs in the United States
through 2030. Working Paper 2019-06, ICCT. https://theicct.org/publication/update-on-electric-
vehicle-costs-in-the-united-states-through-2030/.
McLean, Bethany. 2018. "The Next Financial Crisis Lurks Underground." The New York Times,
September 1. https://www.nytimes.com/2018/09/01/opinion/the-next-fi nancial-crisis-lurks-
underground. html.
Melodia, Lauren, and Kristina Karlsson. 2022. Energy Price Stability: The Peril of Fossil Fuels
and the Promise of Renewables. Issue Brief, Roosevelt Institute.
https://rooseveltinstitute.org/publications/energy-price-stability/.
Michalek, Jeremy J, Mikhail Chester, Paulina Jaramillo, Constantine Samaras, Ching-Shin
Norman Shiau, and Lester B Lave. 2011. "Valuation of plug-in vehicle life-cycle air emissions
and oil displacement benefits." Proceedings of the National Academy of Sciences 108 (40):
16554-16558. doi: 10.1073/pnas. 1104473108.
Moore, John L, Carl E Behrens, and John E Blodgett. 1997. Oil Imports: An Overview and
Update of Economic and Security Effects. CRS Report 98-1, Congressional Research Service:
Environment and Natural Resources Policy Division.
Muratori, Matteo. 2018. "Impact of uncoordinated plug-in electric vehicle charging on
residential power demand." Nature Energy 3: 193-201. doi:10.1038/s41560-017-0074-z.
National Research Council. 2010. Hidden Costs of Energy: Unpriced Consequences of Energy
Production and Use. Washington DC: National Academies Press.
Newell, Richard G., and Brian C. Prest. 2019. "The Unconventional Oil Supply Boom:
Aggregate Price Response from Microdata." The Energy Journal 40 (3).
doi: 10.5547/01956574.40.3.rnew.
Nordhaus, William D. 2007. "Who's Afraid of a Big Bad Oil Shock?" Brookings Papers on
Economic Activity 2007 (2): 219-238. doi:10.1353/eca.2008.0013.
Oladosu, Gbadebo A., Paul. N. Leiby, David C. Bowman, Rocio Uria-Martinez, and Megan
Johnson. 2018. "Impacts of oil price shocks on the United States economy: A meta-analysis of
the oil price elasticity of GDP for net oil-importing economies." Energy Policy 115 (Virtual
10-39
-------
Special Issue entitled 'Oil Supply Disruptions, U.S. Economic Activity and Oil Security'): 523-
544. 10.1016/j.enpol.2018.01.032.
Parry, Ian W. H., and Joel Darmstadter. 2003. The Cost of U.S. Oil Dependency. Washington,
D.C.: Resources for the Future.
Plummer, James L, Hung-Po Chao, Dermot Gately, and Richard J Gilbert. 1982. "Basic
Concepts, Assumptions, and Numerical Results." In Energy Vulnerability, 13-36. Cambridge:
Ballinger Publishing Co.
Powell, Siobhan, Gustavo Cezar, Liang Min, Ines Azevedo, and Ram Rajagopal. 2022.
"Charging infrastructure access and operation to reduce the grid impacts of deep electric vehicle
adoption." Natural Energy 1-14. doi:0.1038/s41560-022-01105-7.
Ramey, Valerie A, and Daniel J Vine. 2010. "Oil, Automobiles, and the U.S. Economy: How
Much Have Thinks Really Changed?" NBER Macroeconomics Annual 25 (1): 333-368.
doi: 10.1086/657541.
Rasmussen, Tobias, and Agustin Roitman. 2011. "Oil Shocks in a Global Perspective: Are They
Really that Bad?" IMF Working Papers (11/194).
https://papers.ssrn. com/sol3/papers. cfm?abstract_id=1910497.
SAFE. 2018. The Military Cost of Defending the Global Oil Supply. Issue Brief, Securing
America's Energy Future. http://secureenergy.org/wp-content/uploads/2020/03/Military-Cost-of-
Defending-the-Global-Oil-Supply.-Sep.-18.-2018.pdf.
Sieren-Smith, Bridget, Ankit Jain, Alireza Eshraghi, Simon Hurd, Julia Ende, Josh Huneycutt,
Andrew Ngo, MarkLeBel, Carl Linvill, and David Roland-Hoist. 2021. Utility Costs and
Affordability of the Grid of the Future. An Evaluation of Electric Costs, Rates and Equity Issues.
California Public Utilities Commission.
Slowik, Peter, Aaron Isenstadt, Logan Pierce, and Stephanie Searle. 2022. Assessment of Light-
duty Electric Vehicle Costs and Consumer Benefits in the United States in the 2022-2035 Time
Frame. White Paper, ICCT. https://theicct.org/publication/ev-cost-benefits-2035-oct22/.
Stein, Fred. 2013. "Ending America's Energy Insecurity: Why Electric Vehicles Should Drive
the United States to Energy Independence." Homeland Security Affairs 4.
https://www.hsaj.org/articles/236.
Stern, Roger J. 2010. "United States cost of military force projection in the Persian Gulf, 1976-
2007." Energy Policy 38: 2816-2825. doi:10.1016/j.enpol.2010.01.013.
Trinko, David, Emily Porter, Jamie Dunckley, Thomas Bradley, and Timothy Coburn. 2021.
"Combining Ad Hoc Text Mining and Descriptive Analytics to Investigate Public EV Charging
Prices in the United States." Energies 14 (17). doi:10.3390/enl4175240.
U.S. DOE. 2022. Demand Response and Time-Variable Pricing Programs. Accessed September
5, 2022. https://www.energy.gov/eere/femp/demand-response-and-time-variable-pricing-
programs.
10-40
-------
—. 2022. DOE Announces Emergency Notice of Sale of Crude Oil from the Strategic Petroleum
Reserve to Address Oil Supply Disruptions. March 2. https://www.energy.gov/fecm/articles/doe-
announces-emergency-notice-sale-crude-oil-strategic-petrol eum-reserve-address-oil.
—. 2022. DOE Announces Second Emergency Notice of Sale of Crude Oil From The Strategic
Petroleum Reserve to Address Putin's Energy Price Hike. April 1.
https://www.energy.gov/articles/doe-announces-second-emergency-notice-sale-crude-oil-
strategic-petroleum-reserve-address.
—. 2023. DOE Issues Notice of Congressionally Mandated Sale to Purchase Crude Oil from the
Strategic Petroleum Reserve. February 13.
—. 2022. FOTW #1251, August 15, 2022: Electric Vehicles Have the Lowest Annual Fuel Cost
of All Light-Duty Vehicles. August 15. https://www.energy.gov/eere/vehicles/articles/fotw-
1251-august-15-2022-electric-vehicles-have-lowest-annual-fuel-cost-all.
U.S. DOE. 2017. "Valuation of Energy Security for the United States. Report to Congress."
Washington, DC.
U.S. EIA. 2019. Saudi Arabia crude oil production outage affects global crude oil and gasoline
prices. September 23. Accessed 12 2022.
https://www.eia.gov/todayinenergy/detail.php?id=41413.
—. 2023. "Annual Energy Outlook 2021. Table All: Petroleum and Other Liquids Supply and
Disposition (Reference Case)."
—. 2023. "Annual Energy Outlook 2023. Table 8: Electricity Supply, Disposition, Prices, and
Emissions."
—. 2023. "Annual Energy Outlook 2023. Table A11: Petroleum and Other Liquids Supply and
Disposition (Reference Case)."
—. 2023. "Annual Energy Outlook 2023: Table 10: Electricity Trade."
—. 2023. "Electric Power Annual 2022. Table 1.3: Supply and disposition of electricity."
—. 2023. "Electric Power Annual 2022. Table 2.14: Electricity Supply and Disposition."
U.S. EIA. 2022. Monthly Energy Review, https://www.eia.gov/totalenergy/data/monthly/.
—. 2022. "Monthly Energy Review, December 2022. Table 3.1. Petroleum Overview."
—. 2022. Oil and petroleum products explained. November 02. Accessed 12 06, 2022.
https://www.eia.gov/energyexplained/oil-and-petroleum-products/imports-and-exports.php.
U.S. EIA. 2023. Petroleum and Other Liquids: Crude Oil Production.
U.S. EIA. 2023. Petroleum and Other Liquids: Spot Prices.
https://www.eia.gov/dnav/pet/pet_pri_spt_sl_d.htm.
U.S. EIA. 2022. Petroleum and Other Liquids: Weekly Retail Gasoline and Diesel Prices.
https://www.eia.gov/dnav/pet/PET_PRI_GND_DCUS_NUS_W.htm.
10-41
-------
U.S. EIA. 2022. Residential electricity sales and average prices. Electricity sales to ultimate
customers, https://www.eia.gov/opendata/browser/electricity/retail-sales.
U.S. EIA. 2022. "Today in Energy. Crude oil prices increased in first-half of 2022 and declined
in second half of 2022."
—. 2021. U.S. electricity customers experienced eight hours of power interruptions in 2020.
November 10. Accessed December 2022.
https://www.eia.gov/todayinenergy/detail.php?id=50316.
—. 2022. U.S. STATES: State Profiles and Energy Estimates, https://www.eia.gov/state/.
U.S. EIA. 2023. U.S. Tight Oil Production by Play. Accessed 12 14, 2022.
https://www.eia.gov/energyexplained/oil-and-petroleum-products/data/US-tight-oil-
production.xlsx.
U.S. EPA. 2016. "Draft Technical Assessment Report: Midterm Evaluation of Light-Duty
Vehicle Greenhouse Gas Emission Standards and Corporate Average Fuel Economy Standards
for Model Years 2022-2025. EPA-420-D-16-900."
U.S. EPA. 2021. "Revised 2023 and Later Model Year Light Duty Vehicle GHG Emissions
Standards, Regulatory Impact Analysis." EPA-420-R-21-028.
U.S. EPA. 2022. "The 2022 EPA Automotive Trends Report: Greenhouse Gas Emissions, Fuel
Economy, and Technology since 1975." Office of Transportation and Air Quality. EPA-420-R-
22-029.
U.S. EPA/NHTSA. 2010. "Joint Techical Support Document Final Rulemaking to Establish
Light-Duty Vehicle Greenhouse Gas Emission Standards and Corporate Average Fuel Economy
Standards for Model Year 2012-2016 Vehicles." EPA-420-R-10-901.
U.S. EPA/NHTSA. 2012. "Joint Technical Support Document Final Rulemaking to Establish
Light-Duty Vehicle Greenhouse Gas Emissions and Corporate Average Fuel Economy Standards
for Model Year 2017-2025 Vehicles." EPA-420-R-12-901.
U.S. GAO. 2020. Effects of the Repeal of the Crude Oil Export Ban. Report to Congress, United
States Government Accountability Office, https://www.gao.gov/assets/gao-21-118.pdf.
Union of Concerned Scientists. 2015. "What is Tight Oil?"
https://www.ucsusa.org/resources/what-tight-oil.
Uria-Martinez, Rocio, Paul N. Leiby, Gbadebo Oladosu, David C. Bowman, and Megan M.
Johnson. 2018. Using Meta-Analysis to Estimate World Crude Oil Demand Elasticity. Oak
Ridge National Laboratory. https://info.ornl.gov/sites/publications/Files/Publ20229.pdf.
Walls, W.D., and Xiaoli Zheng. 2022. 'Tracking and Structural Shifts in Oil Supply." The
Energy Journal 43 (3). doi: 10.5547/01956574.43.3.wwal.
Yilmaz, Murat, and Philip Krein. 2013. "Review of the Impact of Vehicle-to-Grid Technologies
on Distribution Systems and Utility Interfaces." IEEE Transactions on Power Electronics 28
(12): 5673-5689. doi:10.1109/TPEL.2012.2227500.
10-42
-------
Zamuda, Craig D, Daniel E Bilello, Guenter Conzelmann, Ellen Mecray, Ann Satsangi, Vincent
Tidwell, and Brian J Walker. 2018. Energy Supply, Delivery, and Demand. Vol. II, in Impacts,
Risks, and Adaptation in the United States: Fourth National Climate Assessment, 174-201.
doi:10.7930/NCA4.2018.CH4.
10-43
-------
Chapter 11: Small Business Flexibilities
The Regulatory Flexibility Act, as amended by the Small Business Regulatory Enforcement
Fairness Act of 1996 (SBREFA), generally requires an agency to prepare a regulatory flexibility
analysis of any rule subject to notice-and-comment rulemaking requirements under the
Administrative Procedure Act or any other statute. As a part of this analysis, an agency is
directed to convene a Small Business Advocacy Review Panel (SBAR Panel or 'the Panel'),
unless the agency certifies that the rule will not have a significant economic impact on a
substantial number of small entities. During such a Panel process, the agency would gather
information and recommendations from Small Entity Representatives (SERs) on how to reduce
the impact of the rule on small entities. As discussed below, EPA is certifying that this rule will
not have a significant economic impact on a substantial number of small entities, and thus we
have not conducted an SBAR Panel for this rulemaking. The following discussion provides an
overview of small entities in the vehicle market. Small entities include small businesses, small
organizations, and small governmental jurisdictions. For the purposes of assessing the impacts of
the rule on small entities, a small entity is defined as: (1) a small business that meets the
definition for business based on the Small Business Administration's (SBA) size standards (13
CFR 121.201 2023), (2) a small governmental jurisdiction that is a government of a city, county,
town, school district or special district with a population of less than 50,000; and (3) a small
organization that is any not-for-profit enterprise which is independently owned and operated and
is not dominant in its field.
There are three types of small entities that could potentially be impacted by the GHG
standards: 1) small entity vehicle manufacturers; 2) alternative fuel converters, which are
companies that take a vehicle for which an OEM has already accounted for GHG compliance
and convert it to operate on a cleaner fuel such as natural gas or propane; and 3) independent
commercial importers (ICIs), which are firms that import vehicles from other countries for
individual vehicle purchasers.
EPA is certifying that this rule will have no significant economic impact on a substantial
number of small entities (No SISNOSE). EPA has focused its assessment of potential small
business impacts on three key aspects of the standards, including GHG emissions standards,
criteria pollutants (NMOG+NOx fleet-average standards and PM emissions standards), and
electric vehicle battery warranty and durability.
Under the current light-duty GHG program, small entities are exempt from the GHG
standards. EPA is continuing the current exemption for all three types of small entities, including
small entity manufacturers, alternate fuel converters, and independent commercial importers
(ICIs). However, EPA is finalizing, as proposed, some environmental protections for imported
vehicles, as described below. EPA is also continuing the current provision allowing small entity
manufacturers to opt into the GHG program to earn credits to sell in the credit market. The only
small entity vehicle manufacturers in the market at this time produce only electric vehicles.
On average, historical production data indicates that small entities' annual sales have been
well below this range as shown in Table 11-1. EPA believes that capping the number of vehicles
exempted is an appropriate protection for GHG emissions, while still allowing small entities to
produce vehicles consistent with typical past annual sales. EPA is finalizing a 500 vehicle cap on
11-1
-------
GHG for small entities. EPA notes that for small entities with vehicles exceeding the cap, those
manufacturers could be eligible for the small volume manufacturer standards.
Table 11-1 Small Entity Vehicle Production from Model Year 2017 to 2022
Karma
RUF
Koenigsegg
Pagani
Rimac
2017
0
0
0
46
1
2018
295
2
10
10
0
2019
83
1
12
0
1
2020
153
6
4
0
0
2021
68
7
11
0
0
2022
*
2
0
0
o**
Note: * Karma did not report to EPA for 2022.
**No longer a small business. They are now part of Porsche as Bugatti-Rimac
While ICI's imported vehicles have not been accounted for in an OEM manufacturer's GHG
average, there are typically only a small number of vehicles imported each year. Table 11-2
shows the number of vehicles imported by each of the current ICIs. Since 2014, none of the
current ICIs have imported more than 15 vehicles each year. Under existing EPA regulations,
each ICI is currently limited to importing 50 vehicles per year. EPA is finalizing reducing the
limit to 25 vehicles (excluding BEVs or fuel cell vehicles) per year, as a means of limiting the
potential environmental impact of importing vehicles with potentially high GHG emissions.
Importing of BEVs and fuel cell vehicles will not count against the 25 vehicles cap. EPA
believes this lower vehicle limit is important for capping the potential for high-emitting imported
vehicles, because, unlike with criteria pollutant emissions as discussed below, there are very
limited add-on emissions control options for reducing the GHG emissions of an imported
vehicle. This action will have no financial impact on the ICI businesses, as it is still far above the
average number of vehicles imported by ICIs in recent years.
Table 11-2 ICI Import Records (number of imported vehicles)
2014
2015
2016
2017
2018
2019
2020
2021
2022
Current ICIs
G&K
7
7
6
6
8
12
6
10
8
JK Technologies
13
15
8
10
10
9
3
4
5
Wallace Labs
0
0
15
1
7
5
4
4
10
EPA also has evaluated the potential impacts on small businesses for the final criteria
pollutant emissions standards, including both the NMOG+ NOx standard and the PM standard.
EPA's final NMOG+NOx standards should have no impact on the existing small entity
manufacturers which produce only electric vehicles. EPA is finalizing a delayed phase in for
small entities (as well as small volume manufacturers) such that they will not have to meet the
criteria pollutant standards until the last year of the standards phase-in. The final standards are
expected to have minimal impact on both the alternate fuel converters and ICIs. Alternate fuel
converters acquire vehicles that would already meet the standard on gasoline or diesel fuel and
have the ability to make changes, such as calibration, so the vehicles continue to meet the
standard on an alternate fuel such as propane or natural gas. ICIs take vehicles that were certified
in a foreign country and make the vehicle meet the EPA standard for the year the vehicle was
built. This may require catalyst and calibration changes depending on the vehicle's original
11-2
-------
requirements. EPA believes changes to the NMOG+ NOx standard will require a similar amount
of effort for both alternative fuel converters and ICIs to meet the new standard when compared
to the previous (Tier 3) emissions standard. See Section III.D.9 of the preamble for additional
detail on criteria pollutant standards for small volume manufacturers.
The final PM standard could potentially have a unique impact on each type of small entity.
The current small entity manufacturers all produce only BEVs which have no tailpipe emissions
and therefore would inherently comply with the PM standard. EPA is finalizing a delayed phase-
in for small volume manufacturers. Since the current small entities are also small volume
manufacturers, they will not have to meet the criteria pollutant standards until the last year of the
standards phase-in. Alternative fuel converters buy OEM vehicles that already would need to be
compliant for PM but must test the vehicle on the converted fuel and show that it still meets the
standard. There would be an increased testing burden to measure PM on the cold temperature test
(as discussed further in Section III.D.2 of the preamble), but alternative fuel vehicles are already
exempted from cold testing requirements under existing EPA regulations. EPA is finalizing
continuing this exemption for cold temperature testing, and thus there would be no impact on
alternative fuel converters. ICIs must do a complete set of emissions tests for an imported vehicle
that does not already have an existing certificate (referred to as non-conforming vehicles). ICIs
currently only have to test non-methane hydrocarbons (NMHC) on the cold test; to minimize the
testing burden on ICIs, EPA is finalizing exempting ICIs from measuring PM during cold
testing. ICIs will only need to comply with the new PM levels on the FTP75 and US06. The
stringency of the final PM standard may lead to OEMs choosing to comply by the use of
gasoline particulate filters (GPFs). Most of the ICE vehicles since 2014 have been imported from
Europe where GPFs are mandatory, so EPA estimates that there will be no financial impact to
ICIs based on additional testing or ensuring imported vehicles are compliant with emissions
standards.
The final aspect of the final rule that could have potential impacts on small entities is battery
durability (Section III.G.2 of the preamble). EPA finds it appropriate to exempt small entities
from battery durability requirements at this time while we implement the requirement for larger
manufacturers. Based on our experience with larger manufacturers, we will be in a better
position to judge whether the requirements are appropriate to extend to smaller manufacturers in
a potential future rulemaking.
11-3
-------
Chapter 11 References
13 CFR 121.201. 2023.
11-4
-------
Chapter 12: Compliance Effects
This chapter presents the outputs from OMEGA related to the final light- and medium-duty
vehicle GHG standards, the two alternative standard stringencies, Alternative A (a more stringent
standard) and Alternative B (a less stringent standard) which are presented in Section III.F of the
preamble, and a range of sensitivity scenarios.
In the following sections, we provide detailed modeling results of projected GHG targets,
projected achieved GHG compliance levels, as well as per-vehicle costs and technology
penetrations. These projections are grouped by car and truck regulatory classes, and in select
tables, using EPA's classification of body style in the OMEGA model. Chapter 12.1 presents the
compliance effects for the light-duty vehicle GHG standards and Chapter 12.2 presents the
compliance effects for the medium-duty vehicle GHG standards.
12.1 Light-Duty Vehicles
12.1.1 GHG Targets and Compliance Levels
12.1.1.1 COi g/mi
Shown below are the projected average GHG targets for each manufacturer, as well as their
corresponding average achieved compliance, in g/mile, for cars, trucks and the combined fleet.
12.1.1.1.1 Final standards
OEM-specific GHG emissions targets for the final standards are shown in Table 12-1 for cars,
Table 12-2 for trucks, and the combined fleet in Table 12-3. Similarly, projected achieved GHG
emissions levels are presented for cars, trucks, and the combined fleet in Table 12-4, Table 12-5
and Table 12-6.
12-1
-------
2032
73
73
73
73
72
72
72
73
72
75
72
73
73
72
72 ^
74
72
74
72
73
72
73 "
2032
91
88
94
97
85
84
87
86
80
89
78
88
101
95
80
85
88
86
84
90
Table 12-1: Projected GHG Targets, Final Standards - Cars
2027
140
139
139
140
138
138
138
139
138
145
138
139
140
137
138
2028
127
126
127
126
125
125
125
126
125
130
125
126
127
124
125
2029
i.}
12
11
12
12
12
12
16
12
12
2030
100
100
100
99
98
99
99
99
99
102
98
99
100
98
99
203 1
87
87
87
86
86
86
86
86
86
89
86
86
87
85
86
141
137
143
138
140
138
139
128
124
128
125
126
125
125
14
11
14
12
13
12
112
101
98
101
99
99
99
99
88
86
87
86
86
86
86
Table 12-2: Projected GHG Targets, Final Standards - Trucks
2027 2028 2029 2030 2031
185 166 147 129 110
181 161 142 124 105
192 171 152 133 113
200 178 158 138 117
172 155 137 120 102
171 153 136 119 101
178 159 141 124 105
175 157 139 122 104
162 145 129 114 96
180 162 143 126 107
158 142 127 112 95
178 159 141 124 106
221 196 172 145 122
193 173 153 134 114
161 145 129 113 96
186 159 140 122 103
179 161 142 125 106
177 158 140 123 104
172 153 136 120 102
184 165 146 128 109
12-2
-------
Table 12-3: Projected GHG Targets, Final Standards - Combined
Manufacturer
2027
2028
2029
2030
203 1
2032
Aston Martin
158
143
127
112
97
81
BMW
163
147
130
114
98
82
Ferrari
139
127
113
100
87
73 "
Ford
186
167
148
129
110
91
General Motors
185
166
147
129
110
91
Honda
157
141
126
111
95
79
Hyundai
155
140
124
109
94
78 "
JLR
176
158
140
123
105
87
Kia
158
142
127
112
96
80
Lucid
145
130
116
102
89
75
Ma/.da
158
142
127
112
95
79
McLaren
139
126
112
99
86
73 "
Mercedes Ben/.
167
150
133
117
100
84
Mitsubishi
147
133
118
105
90
75
Nissan
155
140
125
110
95
79
Rivian
221
196
172
145
122
101
Stellantis
187
168
149
131
111
92
Subaru
158
142
126
111
95
79
Tcsla
159
140
124
109
93
78
Toyota
163
147
130
115
98
82
Volvo
169
151
134
118
100
83
VW
161
144
129
113
97
81
TOTAL
170
153
136
119
102
85
Table 12-4:
Achieved GHG Levels,
Final Standards
- Cars
Manufacturer
2027
2028
2029
2030
203 1
2032
Aston Martin
186
166
141
117
117
116
BMW
179
142
114
81
91
73
Ferrari
212
148
149
125
109
96
Ford
118
115
103
98
84
76
General Motors
137
94
82
79
68
59
Honda
116
102
83
63
58
50
Hyundai
124
98
90
80
71
62
JLR
164
147
148
148
146
142
Kia
111
94
84
67
66
53
Lucid
-11
-8
-6
-3
-2
-2
Ma/.da
119
80
69
69
63
56
McLaren
241
171
163
133
103
90
Mercedes Ben/
146
140
129
110
117
86
Mitsubishi
126
112
99
87
79
...... 1Q
Nissan
126
107
89
85
2 77 ^
65
Rivian
-
-
-
-
-
-
Stellantis
188
122
91
74
55
51
Subaru
147
97
81
75
68
56
Tcsla
-11
-8
-6
-3
-2
-2 "
Toyota
107
101
85
74
73
61
Volvo
92
68
54
75 i
55
48
VW
114
121
105
100
96
81
TOTAL
116
97
83
72
67
57
12-3
-------
Table 12-5: Achieved GHG Levels, Final Standards - Trucks
Manufacturer
2027
2028
2029
2030
203 1
2032
Aston Martin
242
220
190
180
146
110
BMW
202
165
163
154
139
112
Ferrari
-
-
-
-
-
-
Ford
199
182
154
139
118
105
General Motors
213
205
175
156
133
108
Honda
167
158
146
137
117
103
Hyundai
165
152
133
118
102
93
JLR
210
190
162
144
127
113
Kia
180
172
154
143
111
102
Lucid
-
-
-
-
-
-
Ma/.da
162
154
135
119
105
93
McLaren
-
-
-
-
-
-
Mercedes Ben/.
184
169
149
125
101
95
Mitsubishi
167
130
116
105
94
85
Nissan
197
184
171
137
113
100
Rivian
-14
-10
-7
-3
-2 ;
-2
Stellantis
207
195
169
150
128
110
Subaru
159
152
134
118
104
93
Tcsla
-14
-10
-7 '
-3
F -2
-2
Toyota
185
163
148
129
109
96
Volvo
143
148
133
115
114
108
VW
177
153
125
114
98
93
TOTAL
186
173
151
135
115
100
Table 12-6:
Achieved GHG Levels,
Final Standards
- Combined
Manufacturer
2027
2028
2029
2030
203 1
2032
Aston Martin
209
188
162
144
129
113
BMW
193
155
142
124
119
97
Ferrari
212
148
149
125
109
96
Ford
191
175
148
135
114
102
General Motors
195
179
153
138
118
96
Honda
144
133
118
104
91
80
Hyundai
145
126
113
100
87
78
JLR
208
189
162
144
128
114
Kia
148
136
122
108
91
80
Lucid
-11
-8
-6
-3
-2
-2
Ma/.da
157
144
127
113
99
88
McLaren
241
171
163
133
103
90
Mercedes Ben/
171
160
142
120
106
92
Mitsubishi
146
121
107
96
86
77
Nissan
157
141
125
108
93
80
Rivian
-14
-10
-7
-3
-2
-2
Stellantis
205
187
161
142
121
104
Subaru
157
144
126
112
99
88
Tcsla
-12
-9
-6
-3
-2 i
-2
Toyota
154
138
123
107
95
82
Volvo
131
130
115
106
101
94
VW
157
143
119
109
98
89
TOTAL
164
149
130
116
100
87
12-4
-------
12.1.1.1.2 Alternative A
Table 12-7 through Table 12-9 show the OEM-specific targets for Alternative A. Achieved
levels are presented, by manufacturer, in Table 12-10 through Table 12-12.
Table 12-7: Projected GHG Targets, Alternative A - Cars
Manufacturer
2027
2028
2029
2030
203 1
2032
Aston Martin
135
118
100
91
83
74
BMW
134
117
99
91
82
73 '
Ferrari
134
117
99
91
83
73
Ford
135
116
98
90
82
73
General Motors
133
115
97
90
81
1 72 '
Honda
133
115
98
90
82
72
Hyundai
134
116
98
90
82
72
JLR
135
117
99
91
82
73
Kia
134
116
98
90
82
72'
Lucid
140
120
101
93
84
75 '
Ma/.da
133
115
97
90
81
72 '
McLaren
134
116
98
91
82
73 '
Mercedes Ben/.
136
118
99
91
83
73
Mitsubishi
132
115
97
90
81
72'
Nissan
134
116
98
90
82
]] 72'
Rivian
-
-
-
-
-
-
Stellantis
137
118
100
92
83
74
Subaru
133
115
97
90
81
! 72
Tcsla
138
118
100
92
83
74
Toyota
134
116
98
90
82
72 '
Volvo
135
117
99
91
82
73 '
VW
133
116
98
90
82
72
TOTAL
134
116
98
90
82
73'
12-5
-------
Table 12-8: Projected GHG Targets, Alternative A
- Trucks
Manufacturer
2027
2028
2029
2030
203 1
2032
Aston Martin
166
143
121
112
103
91
BMW
161
139
118
109
99
88
Ferrari
-
-
-
-
-
-
Ford
171
149
126
116
106
94
General Motors
178
155
131
121
110
97
Honda
154
134
113
105
96
85
Hyundai
153
133
112
104
95
84
JLR
159
138
117
108
99
88
Kia
156
136
115
107
97
86
Lucid
-
-
-
-
-
-
Ma/.da
144
126
107
99
90
80
McLaren
-
-
-
-
-
-
Mercedes Ben/.
162
141
119
110
100
89
Mitsubishi
141
123
105
97
89
79
Nissan
158
138
117
108
98
87
Rivian
195
169
142
126
114
101
Stellantis
172
150
127
117
107
95
Subaru
144
125
107
99
90
80
Tcsla
165
137
115
106
96
85
Toyota
160
139
118
109
99
88
Volvo
158
143
120
110
100
89
VW
154
133
113
104
95
85
TOTAL
164
143
121
112
102
90
Table 12-9: Projected GHG Targets, Alternative A -
Combined
Manufacturer
2027
2028
2029
2030
203 1
2032
Aston Martin
148
128
108
100
91
81
BMW
150
130
110
102
92
82
Ferrari
134
117
99
91
83
73
Ford
167
145
123
114
103
91
General Motors
167
146
123
113
103
91
Honda
145
126
106
98
90
79
Hyundai
143
125
105
97
89
79
JLR
158
138
117
108
98
87
Kia
146
127
107
99
90
80
Lucid
140
120
101
93
84
75
Ma/.da
143
124
105
98
89
79
McLaren
134
116
98
91
82
73
Mercedes Ben/
153
133
113
104
95
84
Mitsubishi
137
119
101
93
85
75
Nissan
144
125
106
98
89
79
Rivian
195
169
142
126
114
101
Stellantis
168
147
124
114
104
92
Subaru
142
124
105
98
89
79
Tcsla
148
126
106
97
88
78 "
Toyota
149
130
110
102
92
82
Volvo
152
137
115
106
96
85
VW
147
128
108
100
91
81
TOTAL
155
135
114
105
%
85
12-6
-------
2032
84
64
94
68
76
45
58
78
42
-2
43
92
69
60
57
101
68
-2
51
65
73
55
2032
107
100
103
108
90
78
94
97
75
92
77
101
-2
103
72
-2
94
86
78
94
Table 12-10: Achieved GHG Levels, Alternative A - Cars
2027 2028 2029 2030 2031
184 139 113 99 92
185 124 107 106 87
227 154 146 122 103
115 106 91 77 76
137 112 94 88 87
114 90 75 66 52
122 104 93 82 73
148 102 117 105 91
112 84 68 58 52
-2 -2 -2 -2 -2
122 76 57 51 33
248 168 159 128 103
135 107 98 91 89
116 95 84 78 70
127 100 90 81 66
168 135 115 121 105
144 112 104 98 90
-2 -2 -2 -2 -2
113 85 74 76 62
81 97 83 88 77
135 105 87 74 72
117 92 79 74 65
Table 12-11: Achieved GHG Levels, Alternative A - Trucks
2027 2028 2029 2030 2031
221 194 190 157 124
192 143 136 112 103
193 160 134 127 114
210 174 144 130 122
155 138 129 115 104
149 123 112 101 87
191 152 134 123 108
163 152 143 129 111
148 119 108 101 89
173 155 137 122 105
150 122 109 101 89
192 164 140 120 113
-2 -2 -2 -2 -2 "
205 166 139 123 116
144 122 102 93 79
-2 -2 ' -2 -2 -2
176 152 137 114 104
151 125 115 104 95
155 130 122 113 90
179 150 130 116 105
12-7
-------
Table 12-12:
Achieved GHG Levels,
Alternative A -
Combined
Manufacturer
2027
2028
2029
2030
203 1
2032
Aston Martin
199
162
145
123
105
93
BMW
189
135
124
110
97
85
Ferrari
227
154
146
122
103
94
Ford
186
154
130
122
110
100
General Motors
192
159
133
121
114
101
Honda
137
117
105
94
81
70
Hyundai
136
114
103
92
80
69
JLR
189
151
134
122
108
94
Kia
140
121
109
97
84
72
Lucid
-2
-2
-2
j -2
-2
-2
Ma/.da
145
114
102
95
82
71
McLaren
248
168
159
128
103
92
Mercedes Ben/.
160
139
124
112
100
85
Mitsubishi
132
108
96
89
79
68
Nissan
155
128
112
98
87
76
Rivian
-2
-2
-2
"" -2
-2
-2
Stellantis
201
163
137
123
115
103
Subaru
144
120
103
94
81
71
Tcsla
" -2
-2
-2
-2 j
-2
-2
Toyota
151
125
112
99
87
77
Volvo
135
119
108
100
91
81
VW
148
122
111
101
85
76
TOTAL
160
132
115
103
93
82
12.1.1.1.3 Alternative B
Table 12-13 through Table 12-15 show the OEM-specific targets for Alternative B. Achieved
levels are presented, by manufacturer, in Table 12-16 through Table 12-18.
12-8
-------
2032
83
83
83
82
82
82
82
82
82
85
82
83
83
82
82
84
82
83
82
82
82
82
2032
102
98
104
108
95
94
97
96
89
99
88
97
113
105
89
95
98
96
94
100
Table 12-13: Projected GHG Targets, Alternative B - Cars
2027
140
139
139
140
138
138
138
139
138
145
138
139
140
137
138
2028
127
126
126
125
125
125
125
126
125
130
125
126
127
124
125
2029
i.}
12
12
12
12
16
12
12
2030
100
100
99
99
99
99
99
99
99
102
98
99
100
98
99
141
137
143
138
140
138
139
128
124
128
125
126
125
125
14
11
14
12
12
12
112
101
98
101
99
99
99
99
Table 12-14: Projected GHG Targets, Alternative B - Trucks
2027 2028 2029 2030 2031
185 165 146 128 115
181 160 141 124 111
192 171 152 133 119
200 178 158 138 123
172 154 137 120 108
171 153 135 119 106
178 159 141 124 111
175 157 139 122 109
162 145 129 114 102
180 161 143 125 112
158 142 127 112 100
178 159 141 123 110
221 196 172 145 129
193 173 153 134 119
161 145 129 113 101
186 159 140 122 109
179 161 142 125 111
177 158 140 123 110
172 153 137 120 107
184 165 146 128 114
12-9
-------
Table 12-15: Projected GHG Targets, Alternative B - Combined
Manufacturer
2027
2028
2029
2030
203 1
2032
Aston Martin
158
143
127
112
102
91
BMW
163
146
129
114
103
92
Ferrari
139
126
113
99
91
83
Ford
186
167
148
130
116
102
General Motors
185
166
147
129
116
102
Honda
157
141
126
111
100
89
Hyundai
155
140
124
109
99
88
JLR
177
158
140
123
110
97
Kia
158
142
127
111
101
90
Lucid
145
130
116
102
93
85
Ma/.da
158
142
127
112
100
88
McLaren
139
126
112
99
91
83
Mercedes Ben/.
167
150
133
117
105
94
Mitsubishi
147
133
118
105
95
85
Nissan
155
140
125
110
99
89
Rivian
221
196
172
145
129
113
Stellantis
187
168
149
130
117
103
Subaru
158
142
126
111
100
88
Tcsla
159
140
124
109
98
88
Toyota
163
147
130
114
103
92
Volvo
169
151
134
117
105
93
VW
161
144
129
113
102
90
TOTAL
170
153
136
119
107
95
Table 12-16:
Achieved GHG Levels,
Alternative B -
Cars
Manufacturer
2027
2028
2029
2030
203 1
2032
Aston Martin
186
162
138
120
105
93
BMW
178
132
98
80
74
68
Ferrari
210
146
135
127
111
87 "
Ford
115
112
97
86
90
80
General Motors
137
94
79
75
71
64
Honda
114
102
86
69
72
54
Hyundai
122
109
89
77
80
63
JLR
168
148
147
146
143
137
Kia
109
94
81
69
72
59
Lucid
-16
-14
-11
-8
-7
-7 "
Ma/.da
117
102
91
90
83
77
McLaren
240
168
153
144
115
88
Mercedes Ben/
143
134
101
85
87
78
Mitsubishi
125
118
102
91
83
67
Nissan
125
113
96
84
66
50
Rivian
-
-
-
-
-
-
Stellantis
186
178
145
143
137
113
Subaru
146
117
109
105
88
84
Tcsla
-16
-14
-11
-8
-7
-7 "
Toyota
106
101
79
79
62
57
Volvo
89
56
50
53
47
41
VW
116
123
115
112
88
74
TOTAL
114
100
83
76
69
58
12-10
-------
2032
112
101
103
124
92
80
109
88
86
95
83
111
-9
111
78
-9
88
96
81
98
2032
101
87
87
100
110
75
72
110
75
-7
85
88
89
74
77
-9
112
79
-8
76
84
79
86
Table 12-17: Achieved GHG Levels, Alternative B - Trucks
2027 2028 2029 2030 2031
242 207 192 181 141
201 165 158 149 130
199 181 154 140 124
212 205 174 154 140
165 160 141 133 107
163 149 135 124 102
203 191 170 149 135
178 165 148 138 113
161 153 132 117 105
183 170 156 142 121
165 146 128 114 102
195 179 157 143 132
-21 -17 -14 -10 -9
206 185 160 142 130
157 151 128 114 104
-21 -17 -14 -10 -9
184 163 148 127 113
142 148 131 116 115
175 150 116 113 107
185 170 148 133 118
Table 12-18: Achieved GHG Levels, Alternative B - Combined
2027
2028
2029
2030
203 1
209
181
160
145
120
191
151
133
121
107
210
146
135
127
111
190
174
149
135
121
195
179
152
135
124
142
134
117
104
91
143
130
113
101
91
202
189
169
149
135
147
133
117
107
95
-16
-14
-11
-8
-7
155
146
127
114
102
240
168
153
144
115
170
158
138
123
110
144
131
114
102
92
156
142
123
110
95
-21
-17
-14
-10
-9
204
184
159
142
131
155
146
125
113
102
-18
-15
-12
-9
-8
152
138
121
108
93
130
127
112
102
99
156
142
116
113
101
163
149
128
116
104
12-11
-------
12.1.1.2 COi Mg
Shown below are the projected average GHG targets for each manufacturer, as well as their
corresponding average achieved compliance, in Mg, for cars, trucks, and the combined fleet.
Total emissions are calculated by multiplying the relevant CO2 emission rate, the production
volume of applicable vehicles, and the expected lifetime vehicle miles traveled (VMT) of those
vehicles. The equation to calculate total Mg (for either total emissions, or credits based on the
difference between target g/mi and achieved g/mi) is:
CO2 (Mg) = (CO2 (g/mi) x VMT x Production) / 1,000,000
In the above equation, "VMT" is in miles, and specified in the regulations as 195,264 miles
for cars and 225,865 for trucks. When using these equations to calculate values for cars and
trucks in aggregate, we use a production weighted average of the car and truck VMT values.
12.1.1.2.1 Final standards
OEM-specific GHG emissions targets for the final standards (in Mg) are shown in Table 12-19,
Table 12-20, and Table 12-21 for cars, trucks, and the combined fleet, respectively. Similarly,
projected achieved GHG emissions (in Mg) are given for cars, trucks, and the combined fleet in
Table 12-22, Table 12-23, and Table 12-24. Finally, overall credits or debits earned are provided
for the combined fleet on a manufacturer-specific basis, in Table 12-25.
Table 12-19: Projected GHG Targets (Mg), Final Standards - Cars
Manufacturer
2027
2028
2029
2030
203 1
2032
Aston Martin
17.704
15.112
12,723 :
11.580
10.414
9.194
BMW
4.811.426
4.066.442
3.427.719
3.106.950
2.788.695
2.457.118
Ferrari
117.462
99.953
83.891
76.418
68.648
60.646
Ford
5.122.916
4.355.612
3.698.790
3.369.716
3.032.682
2.677.299
General Motors
13.284.773
11.284.883
9.566.422
8.707.816
7.843.291
6.910.093
Honda
15.874.797
13.272.593
11.204.051
10.151.855
9.079.013
7.952.824
Hyundai
11.089.685
9.368.3 11
7.926.394
7.214.834
6.469.315
5.675.214
JLR
58.154
49.142
41.335
37,542
33.686
29.607
Kia
8.579.222
7.218.552
6.082.894
5.515.206
4.909.553
4.295.147
Lucid
78.004
64.234
54.044
48.714
43.533
37.941
Ma/.da
895.285
742.414
626.711
569.712
507,736 ;
446.329
McLaren
25.997
22,077 ;
18.565
16.922
15.222
13.420
Mercedes Ben/.
3,233,771 i
2.735.938
2.314.221
2.105.443
1.893.083
1.666.097
Mitsubishi
2.142.649
1.814.980
1.543.394
1.413.520
1.268.148
1.116.610
Nissan
10.791.762
9.050.117
7.621.301
6.886.246
6.169.729
5.415.448
Rivian
-
-
-
-
-
-
Stellantis
5.222.346
4.383.358
3.689.782
3.326.669
2.979.279
2.616.512
Subaru
2.321.298
1.944.516
1.633.929
1.485.516
1,325,755 ;
1.163.655
Tcsla
9.080.496
7.669.414
6.467.854
5.856.666
5.254.182
4.604.033
Toyota
24.889.453
20,822,532 :
17.601.163
15.925.030
14.217.084
12.479.907
Volvo
843.504
708.844
599.614
546.893
493.362
434.544
VW
5.150.246
4.330.856
3.644.818
3.290.000
2.946.308
2.582.385
TOTAL
123,630,948
104,019,879
87,859,616
79,663,247
71,348,717
62,644,024
12-12
-------
Table 12-20: Projected GHG Targets (Mg), Final Standards - Trucks
Manufacturer
2027
2028
2029
2030
203 1
2032
Aston Martin
17.3%
15.041
12.751
11.808
10.786
9.575
BMW
9,277,477
7.954.260
6.767.818
6.229.780
5.642.042
4.984.83 1
Ferrari
-
-
-
-
-
-
Ford
66.572.021
57.647.63 1
49.384.104
45.221.251
40,773,677
36.020.435
General Motors
66.509.393
57.494.012
49.141.761
44.695.577
40.476.847
35,772,236
Honda
25.922.475
22.130.892
18.897.061
17.348.074
15.672.154
13.833.699
Hyundai
15.784.497
13.566.659
11.549.771
10.603.918
9.611.275
8.467.078
JLR
2.691.596
2.334.768
1.992.349
1.830.406
1.661.035
1.468.931
Kia
13.620.640
11.641.709
9.915.831
9.091.180
8.185.434
7.219.243
Lucid
-
-
-
-
-
-
Ma/.da
7.436.302
6.311.034
5.400.534
4.979.559
4.486.060
3.971.235
McLaren
-
-
-
-
-
-
Mercedes Ben/.
8.962.63 1
7.714.081
6.582.712
6.055.126
5.503.215
4.839.005
Mitsubishi
2.445.355
2.096.377
1.793.052
1.655.038
1.492.114
1.320.328
Nissan
11.325.833
9.697.592
8.292.925
7.577.166
6.848.080
6.045.768
Rivian
1.250.064
1.077.654
918.242
807.311
727,744
638.008
Stellantis
62.588.968
53.980.289
46.143.424
41.917.961
37,857,253 ;
33.389.881
Subaru
17.043.371
14.599.846
12.455.656
11.477.299
10.358.520
9.165.862
Tcsla
7.805.005
6.470.334
5.483.939
5.000.067
4.506.321
3.949.565
Toyota
51.119.780
43.672.510
37.312.056
34.072.469
30,673,607 ;
27.055.432
Volvo
3.814.056
3.408.766
2.893.016
2.655.312
2.404.244
2.114.621
VW
14.958.968
12.789.892
10.865.312
9.911.830
9.008.122
7.969.711
TOTAL
389,145,827
334,603,349
285,802,315
261,141,132
235,898,531
208,235,441
Table
12-21: Projected GHG Targets (Mg),
Final Standards - Combined
Manufacturer
2027
2028
2029
2030
203 1
2032
Aston Martin
35.100
30.153
25.474
23.388
21.200
18.768
BMW
14.088.903
12.020.703
10.195.537
9.336.730
8.430.738
7.441.949
Ferrari
117.462
99.953
83.891
76.418
68.648
60.646
Ford
71.694.937
62.003.243
53.082.894
48.590.966
43.806.358
38.697.734
General Motors
79.794.166
68.778.895
58.708.183
53.403.393
48.320.137
42.682.330
Honda
41.797.273
35.403.485
30.101.111
27.499.929
24.751.166
21.786.523
Hyundai
26.874.181
22.934.971
19.476.165
17.818.752
16.080.590
14.142.292
JLR
2.749.749
2.383.910
2.033.684
1.867.948
1.694.722
1.498.537
Kia
22.199.862
18.860.261
15.998.725
14.606.386
13.094.987
11.514.389
Lucid
78.004
64.234
54.044
48.714
43.533
37.941
Ma/.da
8.331.588
7.053.447
6.027.246
5.549.271
4.993.797
4.417.565
McLaren
25.997
22,077 ;
18.565
16.922
15.222
13.420
Mercedes Ben/
12.196.401
10.450.019
8.896.933
8.160.569
7.396.298
6.505.102
Mitsubishi
4.588.004
3.911.357
3.336.446
3.068.558
2.760.262
2.436.938
Nissan
22.117.594
18.747.709
15.914.226
14.463.412
13.017.809
11.461.216
Rivian
1.250.064
1.077.654
918.242
807.3 11
727,744
638.008
Stellantis
67.811.314
58.363.647
49.833.207
45.244.630
40.836.533
36.006.392
Subaru
19.364.669
16.544.362
14.089.585
12.962.814
11.684.275
10.329.517
Tcsla
16.885.501
14.139.748
11.951.794
10.856.733
9.760.503
8.553.598
Toyota
76.009.234
64.495.042
54.913.219
49.997.499
44.890.692
39.535.339
Volvo
4.657.560
4.117.610
3.492.630
3,202,205 ;
2.897.606
2.549.165
VW
20.109.213
17.120.748
14.510.130
13.201.830
11.954.430
10.552.095
TOTAL
512,776,775 :
438,623,227
373,661,931
340,804,379
307,247,248
270,879,465
12-13
-------
Table 12-22: Achieved GHG Levels (Mg), Final Standards - Cars
Manufacturer
2027
2028
2029
2030
203 1
2032
Aston Martin
24.498
21.358
18.068
14.875
14.685
14.482
BMW
6.461.195
4.968.117
3,970,532 i
2.813.789
3.082.550
2.472.437
Ferrari
185.096
126.408
126.587
104.950
90.651
79.963
Ford
4.497.287
4.320.464
3.872.419
3.679.518
3.136.920
2.831.418
General Motors
13.667.178
9.259.628
8.112.180
7.688.383
6.577.994
5.629.758
Honda
13.842.716
11.859.974
9.559.465
7.168.648
6.459.980
5.478.910
Hyundai
10.295.192
7.961.173
7.331.800
6.449.408
5.623.954
4.852.027
JLR
71.023
62.159
62.248
61.645
60.013
57.919
Kia
7.162.874
5.899.605
5.223.460
4.110.088
3.996.119
3.128.986
Lucid
(6,141)
(4.444)
(2,935)
(1,467)
(826)
(812)
Ma/.da
810.721
522.965
447.689
444.338
392.232
349.098
McLaren
46.798
32.396
30,677 :
24.765
19.001
16.527
Mercedes Ben/.
3.484.930
3.286.951
3.019.541
2.544.000
2.691.363
1.951.516
Mitsubishi
2.062.291
1.790.072
1.579.812
1.389.267
1.241.413
1.094.494
Nissan
10.239.445
8.410.859
6.969.758
6.555.763
5.843.185
4.865.082
Rivian
-
-
-
-
-
-
Stellantis
7.220.978
4.543.839
3,393,872 ;
2.701.088
1.959.434
1.794.428
Subaru
2.601.528
1.666.228
1.381.572
1.245.895
1.118.228
913.343
Tcsla
(725.498)
(537.917)
(356.023)
(178.812)
(101.043)
(99.83 1)
Toyota
20.107.302
18.304.953
15.336.214
13.136.200
12.808.146
10.541.485
Volvo
570.465
415.815
330.298
456.691
333.769
289.828
VW
4.438.491
4.587.175
3.941.254
3.670.671
3.464.982
2.898.104
TOTAL
107,058,368
87,497,778 i
74,348,488
64,079,701
58,812,748
49,159,163
Table 12-23: Achieved GHG Levels (Mg), Final Standards - Trucks
Manufacturer
2027
2028
2029
2030
203 1
2032
Aston Martin
25.444
23.132
20.097
18.968
15.393
11.492
BMW
11.704.900
9.468.599
9.398.302
8.845.857
7,923,325
6.387.980
Ferrari
-
-
-
-
-
-
Ford
77.624.297
70.670.795
60.214.482
54.217.512
45.599.501
40.703.728
General Motors
79.878.284
76.180.702
65.579.504
58.005.622
48.963.997
39.669.072
Honda
28.273.952
26.328.688
24.401.552
22.653.942
19.192.441
16.776.753
Hyundai
17.128.917
15.590.153
13.781.418
12.116.250
10.397.326
9.430.151
JLR
3.577.889
3.236.111
2.771.838
2.441.032
2.152.258
1.903.089
Kia
15.810.457
14.862.241
13.297.533
12.226.270
9.386.834
8.551.331
Lucid
-
-
-
-
-
-
Ma/.da
8.452.405
7.860.583
6.940.423
6.055.284
5.235.981
4.615.716
McLaren
-
-
-
-
-
-
Mercedes Ben/
10.229.762
9.358.663
8.272.110
6.910.500
5.564.882
5.204.788
Mitsubishi
2.918.464
2.236.904
1.998.557
1.805.319
1.593.545
1.436.887
Nissan
14.124.693
13.077.052
12.190.645
9.652.389
7.913.675
6.899.578
Rivian
(88.260)
(65.845)
(44.615)
(21.801)
(12.746)
(12.632)
Stellantis
75,738,260 ;
70.704.136
61.630.236
53.994.008
45.741.767
38.859.096
Subaru
18.995.804
17.824.632
15.755.329
13.783.177
12.089.255
10.781.269
Tcsla
(654.749)
(486.740)
(328.085)
(160.255)
(93.673)
(92.808)
Toyota
59.351.190
51.546.443
47.204.764
40.300.061
33,780,576 :
29.626.313
Volvo
3.473.409
3.564.928
3.238.724
2.780.618
2,755,386 i
2.579.669
VW
17.324.288
14.846.142
12.078.304
10.938.364
9.370.015
8.836.461
TOTAL
443,889,405
406,827,321
358,401,119
316,563,117
267,569,739
232,167,934
12-14
-------
Table 12-24: Achieved GHG Levels (Mg), Final Standards - Combined
Manufacturer
2027
2028
2029
2030
203 1
2032
Aston Martin
49.941
44.490
38.165
33.843
30.078
25.974
BMW
18.166.096
14.436.716
13.368.834
11.659.646
11.005.875
8.860.417
Ferrari
185.096
126.408
126.587
104.950
90.651
79.963
Ford
82.121.584
74.991.259
64.086.901
57.897.030
48.736.421
43.535.146
General Motors
93.545.462
85.440.330
73.691.684
65.694.005
55.541.991
45.298.830
Honda
42.116.669
38.188.663
33.961.018
29.822.590
25.652.420
22.255.662
Hyundai
27.424.109
23,551,327
21.113.219
18.565.657
16.021.280
14.282.178
JLR
3.648.911
3.298.270
2.834.086
2.502.678
2.212.271
1.961.007
Kia
22,973,330 ;
20.761.846
18.520.993
16.336.358
13.382.953
11.680.317
Lucid
(6,141)
(4.444)
(2,935)
(1,467)
(826)
(812)
Ma/.da
9.263.127
8.383.548
7.388.112
6.499.622
5.628.213
4.964.814
McLaren
46.798
32.396
30,677 :
24.765
19.001
16.527
Mercedes Ben/.
13.714.692
12.645.614
11.291.651
9.454.499
8.256.246
7.156.305
Mitsubishi
4.980.755
4.026.975
3.578.369
3.194.586
2.834.958
2.531.382
Nissan
24.364.138
21.487.911
19.160.403
16.208.152
13.756.860
11.764.661
Rivian
(88.260)
(65.845)
(44.615)
(21.801)
(12.746)
(12.632)
Stellantis
82.959.238
75,247,975
65.024.109
56.695.095
47.701.201
40.653.524
Subaru
21.597.332
19.490.860
17.136.900
15.029.072
13.207.483
11.694.612
Tcsla
(1.380.247)
(1.024.657)
(684.108)
(339.067)
(194.715)
(192.639)
Toyota
79.458.492
69.851.396
62.540.978
53.436.262
46.588.722
40.167.798
Volvo
4.043.874
3.980.743
3.569.022
3,237,309 :
3.089.155
2.869.497
VW
21.762.779
19.433.317
16.019.558
14.609.035
12.834.996
11.734.565
TOTAL
550,947,774
494,325,098
432,749,607
380,642,818
326,382,487
281,327,097
Table 12-25: GHG Credits/Debits Earned (Mg), Final Standards - Combined
Manufacturer
2027
2028
2029
2030
203 1
2032
Aston Martin
(12.087)
(10.713)
(8.107)
(7,587) ;
(7,623)
(7.200)
BMW
(2.707.269)
(775,332) :
(1.232.020)
(1.068.017)
(2.051.331)
(1.406.323)
Ferrari
(63.315)
(18.198)
(30.721)
(21.221)
(18.579)
(19.188)
Ford
(2.004.876)
(3.621.182)
(281.312)
(2.454.008)
(1.784.047)
(4.690.852)
General Motors
(4.791.101)
(6.713.862)
(3.416.546)
(4.870.901)
(4.037.945)
(2.591.053)
Honda
3.657.510
2.042.151
1.843.918
1.301.140
693.034
(443.257)
Hyundai
1.822.363
2.334.033
1.967.207
1.581.905
1.078.311
(116.741)
JLR
(559.275)
(544.667)
(378.198)
(360.833)
(400.931)
(463.552)
Kia
1.320.899
653,202 :
515.123
213.624
559.212
(149.475)
Lucid
87.001
73.939
64.674
54.859
46.560
38,753
Ma/.da
84.271
(164.194)
(59.762)
(93.431)
(257.639)
(524.815)
McLaren
(19.846)
(8,513)
(9.480)
(6.216)
(3.012)
(3,107)
Mercedes Ben/
(354.066)
(746.882)
(673.315)
(179.237)
(372,265) i
(640.867)
Mitsubishi
28,523 ;
405.984
380.696
278.242
106.926
(84.079)
Nissan
(348.015)
(324.427)
(349.443)
118.930
111.390
(270.919)
Rivian
1.499.464
1.318.253
1.159.273
951.147
791.805
650.639
Stellantis
(6.878.417)
(7.847.274)
(4.842.797)
(4.750.583)
(4.013.766)
(4.643.651)
Subaru
124.019
(359.898)
(59.932)
(114.022)
(669.430)
(1.310.413)
Tcsla
19.606.375
16.842.938
14.729.264
12.514.252
10.538.531
8.746.237
Toyota
4.084.997
3.834.261
2.921.352
3,283,567 ,
1.445.407
(432.744)
Volvo
1.120.959
608.867
518.825
345.891
(55,415)
(363,587)
VW
473.635
169.114
1.380.435
526.313
(56.716)
(1.171.745)
TOTAL
16,171,752
7,147,601
14,139,135
7,243,816
1,642,479
(9,897,938)
12-15
-------
12.1.1.2.2 Alternative A
OEM-specific GHG emissions targets for Alternative A (in Mg) are shown in Table 12-26,
Table 12-27, and Table 12-28 for cars, trucks, and the combined fleet, respectively. Projected
achieved GHG emissions (in Mg) are given for cars, trucks, and the combined fleet in Table
12-29, Table 12-30, and Table 12-31. Overall credits or debits earned are provided for the
combined fleet on a manufacturer-specific basis, in Table 12-31.
Table 12-26: Projected GHG Targets (Mg), Alternative A - Cars
Manufacturer
2027
2028
2029
2030
203 1
2032
Aston Martin
17.704
15.112
12,723 :
11.580
10.414
9.194
BMW
4.811.426
4.066.442
3.427.719
3.106.950
2.788.695
2.457.118
Ferrari
117.462
99.953
83.891
76.418
68.648
60.646
Ford
5.122.916
4.355.612
3.698.790
3.369.716
3.032.682
2.677.299
General Motors
13.284.773
11.284.883
9.566.422
8.707.816
7.843.291
6.910.093
Honda
15.874.797
13.272.593
11.204.051
10.151.855
9.079.013
7.952.824
Hyundai
11.089.685
9.368.3 11
7.926.394
7.214.834
6.469.315
5.675.214
JLR
58.154
49.142
41.335
37,542
33.686
29.607
Kia
8.579.222
7.218.552
6.082.894
5.515.206
4.909.553
4.295.147
Lucid
78.004
64.234
54.044
48.714
43.533
37.941
Ma/.da
895.285
742.414
626.711
569.712
507,736 ;
446.329
McLaren
25.997
22,077 ;
18.565
16.922
15.222
13.420
Mercedes Ben/.
3,233,771 :
2.735.938
2.314.221
2.105.443
1.893.083
1.666.097
Mitsubishi
2.142.649
1.814.980
1.543.394
1.413.520
1.268.148
1.116.610
Nissan
10.791.762
9.050.117
7.621.301
6.886.246
6.169.729
5.415.448
Rivian
-
-
-
-
-
-
Stellantis
5.222.346
4.383.358
3.689.782
3.326.669
2.979.279
2.616.512
Subaru
2.321.298
1.944.516
1.633.929
1.485.516
1,325,755 i
1.163.655
Tcsla
9.080.496
7.669.414
6.467.854
5.856.666
5.254.182
4.604.033
Toyota
24.889.453
20,822,532 i
17.601.163
15.925.030
14.217.084
12.479.907
Volvo
843.504
708.844
599.614
546.893
493.362
434.544
VW
5.150.246
4.330.856
3.644.818
3.290.000
2.946.308
2.582.385
TOTAL
123,630,948
104,019,879
87,859,616
79,663,247
71,348,717
62,644,024
12-16
-------
Table 12-27: Projected GHG Targets (Mg), Alternative A - Trucks
Manufacturer
2027
2028
2029
2030
203 1
2032
Aston Martin
17.3%
15.041
12.751
11.808
10.786
9.575
BMW
9,277,477
7.954.260
6.767.818
6.229.780
5.642.042
4.984.83 1
Ferrari
-
-
-
-
-
-
Ford
66.572.021
57.647.63 1
49.384.104
45.221.251
40,773,677 ^
36.020.435
General Motors
66.509.393
57.494.012
49.141.761
44.695.577
40.476.847
35,772,236
Honda
25.922.475
22.130.892
18.897.061
17.348.074
15.672.154
13.833.699
Hyundai
15.784.497
13.566.659
11.549.771
10.603.918
9.611.275
8.467.078
JLR
2.691.596
2.334.768
1.992.349
1.830.406
1.661.035
1.468.931
Kia
13.620.640
11.641.709
9.915.831
9.091.180
8.185.434
7.219.243
Lucid
-
-
-
-
-
-
Ma/.da
7.436.302
6.311.034
5.400.534
4.979.559
4.486.060
3.971.235
McLaren
-
-
-
-
-
-
Mercedes Ben/.
8.962.63 1
7.714.081
6.582.712
6.055.126
5.503.215
4.839.005
Mitsubishi
2.445.355
2.096.377
1.793.052
1.655.038
1.492.114
1.320.328
Nissan
11.325.833
9.697.592
8.292.925
7.577.166
6.848.080
6.045.768
Rivian
1.250.064
1.077.654
918.242
807.311
727,744
638.008
Stellantis
62.588.968
53.980.289
46.143.424
41.917.961
37,857,253 ;
33.389.881
Subaru
17.043.371
14.599.846
12.455.656
11.477.299
10.358.520
9.165.862
Tcsla
7.805.005
6.470.334
5.483.939
5.000.067
4.506.321
3.949.565
Toyota
51.119.780
43.672.510
37.312.056
34.072.469
30,673,607 ;
27.055.432
Volvo
3.814.056
3.408.766
2.893.016
2.655.312
2.404.244
2.114.621
VW
14.958.968
12.789.892
10.865.312
9.911.830
9.008.122
7.969.711
TOTAL
389,145,827
334,603,349
285,802,315
261,141,132
235,898,531
208,235,441
Table
12-28: Projected GHG Targets (Mg),
Alternative
A - Combined
Manufacturer
2027
2028
2029
2030
203 1
2032
Aston Martin
35.100
30.153
25.474
23.388
21.200
18.768
BMW
14.088.903
12.020.703
10.195.537
9.336.730
8.430.738
7.441.949
Ferrari
117.462
99.953
83.891
76.418
68.648
60.646
Ford
71.694.937
62.003.243
53.082.894
48.590.966
43.806.358
38.697.734
General Motors
79.794.166
68.778.895
58.708.183
53.403.393
48.320.137
42.682.330
Honda
41.797.273
35.403.485
30.101.111
27.499.929
24.751.166
21.786.523
Hyundai
26.874.181
22.934.971
19.476.165
17.818.752
16.080.590
14.142.292
JLR
2.749.749
2.383.910
2.033.684
1.867.948
1.694.722
1.498.537
Kia
22.199.862
18.860.261
15.998.725
14.606.386
13.094.987
11.514.389
Lucid
78.004
64.234
54.044
48.714
43.533
37.941
Ma/.da
8.331.588
7.053.447
6.027.246
5.549.271
4.993.797
4.417.565
McLaren
25.997
22,077 ;
18.565
16.922
15.222
13.420
Mercedes Ben/
12.196.401
10.450.019
8.896.933
8.160.569
7.396.298
6.505.102
Mitsubishi
4.588.004
3.911.357
3.336.446
3.068.558
2.760.262
2.436.938
Nissan
22.117.594
18.747.709
15.914.226
14.463.412
13.017.809
11.461.216
Rivian
1.250.064
1.077.654
918.242
807.3 11
727,744
638.008
Stellantis
67.811.314
58.363.647
49.833.207
45.244.630
40.836.533
36.006.392
Subaru
19.364.669
16.544.362
14.089.585
12.962.814
11.684.275
10.329.517
Tcsla
16.885.501
14.139.748
11.951.794
10.856.733
9.760.503
8.553.598
Toyota
76.009.234
64.495.042
54.913.219
49.997.499
44.890.692
39.535.339
Volvo
4.657.560
4.117.610
3.492.630
3,202,205 ;
2.897.606
2.549.165
VW
20.109.213
17.120.748
14.510.130
13.201.830
11.954.430
10.552.095
TOTAL
512,776,775 :
438,623,227
373,661,931
340,804,379
307,247,248
270,879,465
12-17
-------
Table 12-29: Achieved GHG Levels (Mg), Alternative A - Cars
Manufacturer
2027
2028
2029
2030
203 1
2032
Aston Martin
24.1%
17.828
14.391
12.489
11.482
10.433
BMW
6.620.403
4.332.500
3.730.298
3.642.928
2.931.843
2.163.617
Ferrari
198.680
131.579
124.112
102.831
85.465
77,729
Ford
4.382.640
3.975.711
3.434.000
2.892.694
2,832,372 i
2.490.976
General Motors
13.622.587
10.992.458
9.267.250
8.539.349
8.383.835
7.311.169
Honda
13.544.529
10.386.138
8.548.891
7.468.971
5.789.957
4.991.378
Hyundai
10.140.495
8.469.385
7.519.471
6.531.045
5.762.619
4.527.623
JLR
64.066
43.150
49.069
43,722 :
37.441
31.604
Kia
7.217.169
5.205.112
4.218.848
3,550,087 ;
3.108.606
2.520.883
Lucid
(893)
(857)
(854)
(838)
(826)
(812)
Ma/.da
819.778
492.657
365.193
326.216
204.799
263.286
McLaren
48.252
31.848
29.950
23,887 :
19.089
16.955
Mercedes Ben/.
3.215.233
2.496.990
2,275,823 ;
2.110.879
2.044.083
1.577.148
Mitsubishi
1.870.259
1.501.960
1.334.380
1.223.947
1.090.162
935.236
Nissan
10.263.508
7.855.083
7.032.888
6.169.637
5.006.278
4.231.581
Rivian
-
-
-
-
-
-
Stellantis
6.434.195
4.995.574
4.228.153
4.376.071
3.768.652
3.573.156
Subaru
2.520.301
1.904.576
1.741.859
1.620.893
1.465.549
1.093.005
Tcsla
(105.527)
(103.695)
(103.570)
(102.178)
(101.043)
(99.83 1)
Toyota
21.042.279
15.325.199
13.392.983
13.468.484
10.734.926
8.721.522
Volvo
506.528
591.710
501.881
533.234
460.566
389.168
VW
5.209.920
3.929.395
3.238.889
2.711.836
2.592.686
2.600.071
TOTAL
107,638,598
82,574,302
70,943,905
65,246,182
56,228,541
47,425,898
Table 12-30: Achieved GHG Levels (Mg), Alternative
A - Trucks
Manufacturer
2027
2028
2029
2030
203 1
2032
Aston Martin
23,205 ;
20.393
20.012
16.524
13.043
11.166
BMW
11.034.411
8.172.295
7.809.117
6.407.002
5.877.902
5.646.276
Ferrari
-
-
-
-
-
-
Ford
75.394.625
61.959.107
52.419.717
49.187.915
43.863.165
39.825.691
General Motors
78.408.524
64.499.169
53.912.619
48.305.389
45.000.534
39.722.596
Honda
26.170.008
22.891.750
21.522.096
19.065.427
17,072,232
14.611.900
Hyundai
15.380.303
12.536.689
11.549.446
10.351.865
8.773.490
7.883.467
JLR
3.236.638
2.573.424
2,283,027
2,075,932 i
1.827.751
1.578.115
Kia
14.215.134
13.058.259
12.288.573
11.040.809
9.362.164
8.085.282
Lucid
-
-
-
-
-
-
Ma/.da
7.644.354
5.995.401
5.490.585
5.103.118
4.406.229
3.734.936
McLaren
-
-
-
-
-
-
Mercedes Ben/
9.571.172
8.493.769
7,595,787
6.741.393
5.743.452
5.030.786
Mitsubishi
2.594.701
2.074.235
1.865.032
1.716.803
1.503.557
1.290.796
Nissan
13.760.138
11.537.885
9.937.743
8.385.197
7.829.047
6.964.732
Rivian
(12.791)
(12.785)
(12.932)
(12.824)
(12.746)
(12.632)
Stellantis
74.896.622
59.753.621
50.528.462
44.151.946
41.315.274
36,374,037
Subaru
17.075.140
14.165.462
11.961.831
10.838.931
9.125.381
8.261.270
Tcsla
(94.891)
(94.513)
(95.097)
(94.268)
(93.673)
(92.808)
Toyota
56.201.589
47.611.412
43.179.671
35.515.971
32.006.898
28,757,320
Volvo
3.660.968
2.990.183
2.781.591
2.491.990
2.272.408
2,055,207
VW
15.051.351
12.522.240
11.777.292
10.765.750
8.549.697
7.312.638
TOTAL
424,211,202
350,747,997
306,814,573
272,054,870
244,435,803
217,040,776
12-18
-------
Table 12-31: Achieved GHG Levels (Mg), Alternative A - Combined
Manufacturer
2027
2028
2029
2030
203 1
2032
Aston Martin
47.401
38.220
34.403
29.013
24.525
21.599
BMW
17.654.814
12.504.795
11.539.415
10.049.930
8.809.745
7.809.893
Ferrari
198.680
131.579
124.112
102.831
85.465
77,729
Ford
79,777,265 :
65.934.819
55.853.717
52.080.609
46.695.536
42.3 16.666
General Motors
92.031.111
75.491.628
63.179.869
56.844.738
53.384.370
47.033.765
Honda
39.714.537
33,277,888
30.070.987
26.534.398
22.862.189
19.603.279
Hyundai
25.520.798
21.006.074
19.068.917
16.882.910
14.536.109
12.411.090
JLR
3.300.704
2.616.574
2.332.096
2.119.655
1.865.192
1.609.719
Kia
21.432.303
18.263.370
16.507.422
14.590.896
12.470.770
10.606.165
Lucid
(893)
(857)
(854)
(838)
(826)
(812)
Ma/.da
8.464.132
6.488.058
5.855.779
5.429.333
4.611.028
3.998.222
McLaren
48.252
31.848
29.950
23,887 :
19.089
16.955
Mercedes Ben/.
12.786.406
10.990.759
9.871.610
8.852.271
7,787,535 :
6.607.934
Mitsubishi
4.464.960
3.576.195
3.199.412
2.940.749
2.593.718
2,226,032
Nissan
24.023.646
19.392.968
16.970.63 1
14.554.834
12.835.325
11.196.314
Rivian
(12.791)
(12.785)
(12.932)
(12.824)
(12.746)
(12.632)
Stellantis
81.330.817
64.749.194
54.756.615
48.528.017
45.083.926
39.947.193
Subaru
19.595.441
16.070.038
13.703.690
12.459.824
10.590.930
9.354.275
Tcsla
(200.418)
(198.207)
(198.667)
(196.446)
(194.715)
(192.639)
Toyota
77.243.869
62.936.611
56,572,655 ;
48.984.455
42.741.824
37.478.842
Volvo
4.167.497
3.581.894
3.283.472
3,025,225 ;
2.732.974
2.444.376
VW
20.261.271
16.451.634
15.016.180
13.477.587
11.142.382
9.912.708
TOTAL
531,849,800
433,322,298
377,758,478 :
337,301,053
300,664,344
264,466,674
Table 12-32: GHG Credits/Debits Earned (Mg), Alternative A - Combined
Manufacturer
2027
2028
2029
2030
203 1
2032
Aston Martin
(12.301)
(8.067)
(8.930)
(5,625)
(3,325)
(2.830)
BMW
(3.565.911)
(484.092)
(1.343.878)
(713.199)
(379.007)
(367.943)
Ferrari
(81.218)
(31.627)
(40.221)
(26.413)
(16.817)
(17.083)
Ford
(8.082.328)
(3.931.576)
(2,770,822) ;
(3.489.643)
(2.889.178)
(3.618.932)
General Motors
(12.236.945)
(6.712.733)
(4.471.686)
(3.441.345)
(5.064.232)
(4.351.435)
Honda
2,082,735 :
2.125.597
30.124
965.53 1
1.888.977
2.183.244
Hyundai
1.353.383
1.928.896
407.248
935.842
1.544.481
1.731.202
JLR
(550.954)
(232.664)
(298.412)
(251.706)
(170.470)
(111.182)
Kia
767,559
596.891
(508.696)
15.489
624.217
908.224
Lucid
78.897
65.090
54.898
49.552
44.359
38,753
Ma/.da
(132.544)
565.389
171.467
119.938
382.768
419.342
McLaren
(22,255)
(9,771)
(11.385)
(6.965)
(3.867)
(3,535)
Mercedes Ben/
(590.004)
(540.740)
(974.677)
(691.702)
(391.237)
(102.832)
Mitsubishi
123.044
335.162
137.034
127.809
166.543
210.906
Nissan
(1.906.052)
(645.259)
(1.056.405)
(91.422)
182.484
264.902
Rivian
1.262.855
1.090.439
931.174
820.135
740.490
650.639
Stellantis
(13.519.503)
(6.385.547)
(4.923.408)
(3.283.386)
(4.247.394)
(3.940.801)
Subaru
(230,772)
474.324
385.895
502.991
1.093.345
975.242
Tcsla
17.085.919
14.337.955
12.150.461
11.053.179
9.955.218
8.746.237
Toyota
(1.234.635)
1.558.431
(1.659.436)
1.013.044
2.148.868
2.056.497
Volvo
490.063
535,717 :
209.158
176.981
164.633
104.789
VW
(152.058)
669.114
(506.050)
(275,757) :
812.047
639.387
TOTAL
(19,073,025)
5,300,929
(4,096,547)
3,503,326
6,582,904
6,412,791
12-19
-------
12.1.1.2.3 Alternative B
OEM-specific GHG emissions targets for Alternative B (in Mg) are shown in Table 12-33,
Table 12-34, and Table 12-35 for cars, trucks, and the combined fleet, respectively. Projected
achieved GHG emissions (in Mg) are given for cars, trucks, and the combined fleet in Table
12-36, Table 12-37 and Table 12-38. Overall credits or debits earned are provided for the
combined fleet on a manufacturer-specific basis, in Table 12-39.
Table 12-33: Projected GHG Targets (Mg), Alternative B - Cars
Manufacturer
2027
2028
2029
2030
203 1
2032
Aston Martin
18.364
16.355
14.527
12.693
11.519
10.407
BMW
5.014.624
4.431.554
3.940.841
3.439.193
3.105.050
2.788.030
Ferrari
121.781
108.197
96.097
83.439
76.060
68.811
Ford
5.3 10.198
4,722,537 :
4.224.772
3.693.627
3,363,287 :
3.036.754
General Motors
13.796.598
12.263.910
10.956.595
9.568.952
8.686.978
7.846.696
Honda
16.550.350
14.484.669
12.856.683
11.183.968
10.074.102
9.033.468
Hyundai
11.532.691
10.208.726
9.076.994
7.927.415
7.191.835
6.466.301
JLR
60.417
53.445
47.335
41.239
37,330 =
33,530
Kia
8.949.899
7,865,057 ;
6.978.863
6,078,557
5.454.675
4.875.507
Lucid
80.860
69.495
61.739
53,392 i
48.141
42.925
Ma/.da
937,477
817.567
723.943
630.664
566.814
509.384
McLaren
26.952
23.893
21.211
18.537
16.833
15.201
Mercedes Ben/.
3.355.463
2.974.796
2.649.233
2.309.635
2.095.615
1.891.147
Mitsubishi
2.243.890
1.993.394
1,779,737 i
1.563.291
1.414.825
1.274.392
Nissan
11.235.160
9.877.112
8,757,056 :
7.607.251
6.865.028
6.156.892
Rivian
-
-
-
-
-
-
Stellantis
5.426.308
4.779.135
4.244.444
3.671.607
3.3 14.232
2,977,737
Subaru
2.431.214
2.133.816
1.888.000
1.645.871
1.482.446
1.326.570
Tcsla
9.415.009
8.298.704
7.388.181
6.419.294
5.810.561
5.209.894
Toyota
25.923.626
22.773.448
20.202.984
17,552,287 i
15.783.267
14.200.732
Volvo
873.324
772.849
689.358
603.159
548.262
494.623
VW
5.388.004
4.736.067
4,177,375
3.646.041
3.279.444
2.936.364
TOTAL
128,692,206
113,404,725
100,775,968
87,750,113
79,226,304
71,195,364
12-20
-------
Table 12-34: Projected GHG Targets (Mg), Alternative B - Trucks
Manufacturer
2027
2028
2029
2030
203 1
2032
Aston Martin
19.485
17.384
15.444
13.472
12.098
10.660
BMW
10.444.228
9.169.951
8.151.405
7.107.061
6.320.162
5.543.206
Ferrari
-
-
-
-
-
-
Ford
74.792.135
66.638.053
59.544.850
51.774.739
46.060.313
40.355.63 1
General Motors
74.942.023
66.467.901
59.323.475
51.259.533
45.499.491
39.824.146
Honda
29.224.238
25.754.178
22.974.828
19.964.640
17,727,735 :
15.458.712
Hyundai
17.713.855
15.738.890
13.968.445
12.150.910
10.848.736
9.493.299
JLR
3.035.956
2.711.262
2.407.915
2.100.156
1.868.649
1.635.349
Kia
15.343.939
13.531.508
12.044.471
10.470.018
9.268.297
8.087.284
Lucid
-
-
-
-
-
-
Ma/.da
8.409.914
7.413.507
6.614.633
5.783.913
5.104.585
4.464.962
McLaren
-
-
-
-
-
-
Mercedes Ben/.
10.002.461
8.901.503
7.908.243
6.885.141
6.148.366
5.400.393
Mitsubishi
2,765,383 :
2.453.363
2.188.333
1.916.804
1.693.827
1.481.498
Nissan
12.782.525
11.288.527
10.059.416
8.704.434
7,708,427 i
6,732,780
Rivian
1.411.205
1.252.409
1.114.658
929.346
821.045
711.980
Stellantis
70.652.605
62.679.696
55.955.455
48.213.787
42.719.770
37.287.614
Subaru
19.290.132
17.053.794
15.232.449
13.309.221
11.766.366
10.283.145
Tcsla
8.811.119
7,519,577
6.656.974
5.755.891
5.084.054
4.407.486
Toyota
57.626.836
50.947.545
45.222.716
39.154.121
34.559.293
30.3 10.503
Volvo
4.291.509
3.811.360
3.399.896
2,972,733 ;
2.643.561
2.317.413
VW
16.859.642
14.877.053
13.250.299
11.514.106
10.155.419
8.896.346
TOTAL
438,419,187
388,227,463
346,033,905
299,980,026
266,010,193
232,702,406
Table 12-35: Projected GHG
Targets (Mg),
Alternative
B - Combined
Manufacturer
2027
2028
2029
2030
203 1
2032
Aston Martin
37.849
33.740
29.971
26.165
23.617
21.067
BMW
15.458.852
13.601.505
12.092.246
10.546.254
9.425.211
8.331.236
Ferrari
121.781
108.197
96.097
83.439
76.060
68.811
Ford
80.102.333
71.360.591
63.769.622
55.468.367
49.423.600
43.392.385
General Motors
88.738.621
78.731.810
70.280.070
60.828.485
54.186.469
47.670.842
Honda
45.774.588
40.238.847
35.831.511
31.148.608
27.801.837
24.492.179
Hyundai
29.246.546
25.947.616
23.045.439
20,078,325 :
18.040.571
15.959.600
JLR
3.096.372
2.764.708
2.455.250
2.141.395
1.905.979
1.668.879
Kia
24.293.838
21.396.565
19.023.334
16.548.575
14.722.972
12.962.791
Lucid
80.860
69.495
61.739
53.392
48.141
42.925
Ma/.da
9.347.391
8.231.074
7,338,576
6.414.577
5.671.399
4.974.346
McLaren
26.952
23.893
21.211
18.537
16.833
15.201
Mercedes Ben/
13.357.924
11.876.299
10.557.475
9.194.776
8.243.981
7.291.540
Mitsubishi
5.009.273
4.446.757
3.968.070
3.480.095
3.108.652
2.755.890
Nissan
24.017.685
21.165.639
18.816.471
16.311.685
14.573.455
12.889.672
Rivian
1.411.205
1.252.409
1.114.658
929.346
821.045
711.980
Stellantis
76.078.913
67.458.83 1
60.199.899
51.885.394
46.034.001
40.265.350
Subaru
21.721.346
19.187.610
17.120.449
14.955.091
13.248.812
11.609.715
Tcsla
18.226.128
15.818.281
14.045.155
12.175.185
10.894.615
9.617.380
Toyota
83.550.461
73.720.993
65.425.700
56.706.408
50.342.560
44.511.236
Volvo
5.164.833
4.584.208
4.089.254
3.575.892
3.191.823
2.812.036
VW
22.247.645
19.613.120
17.427.674
15.160.147
13.434.863
11.832.710
TOTAL
567,111,394
501,632,188
446,809,873
387,730,138
345,236,496
303,897,770
12-21
-------
Table 12-36: Achieved GHG Levels (Mg), Alternative B - Cars
Manufacturer
2027
2028
2029
2030
203 1
2032
Aston Martin
24.467
20.896
17.618
15.219
13.210
11.550
BMW
6.419.499
4.627.779
3.434.662
2.762.484
2.512.396
2,275,505
Ferrari
184.016
124.915
114.938
106.911
92.407
72.591
Ford
4.382.344
4.221.600
3.669.753
3.199.797
3.365.391
2.949.683
General Motors
13.708.894
9.261.390
7,774,573
7.318.029
6.864.338
6.120.908
Honda
13.674.168
11.817.877
9.968.93 1
7.781.356
7.987.912
5.901.858
Hyundai
10.161.457
8.870.371
7.256.526
6.167.604
6.337.888
4.972.211
JLR
72.928
62,733 i
61.737
60.825
58.824
55.583
Kia
7.058.204
5.938.179
5.036.444
4.245.130
4.374.595
3.494.760
Lucid
(9,100) ;
(7,282) :
(5,763)
(4.245)
(3,562)
(3.500)
Ma/.da
799.471
670.010
595.636
577.819
521.874
478.094
McLaren
46.569
31.859
28.962
26.852
21.192
16.171
Mercedes Ben/.
3,423,307 i
3.127.050
2.354.705
1.957.226
1.989.831
1.766.539
Mitsubishi
2.039.413
1.899.972
1.635.507
1.454.852
1.301.631
1.042.241
Nissan
10.158.213
8.908.545
7.505.541
6.498.846
5.026.855
3.731.504
Rivian
-
-
-
-
-
-
Stellantis
7.160.115
6.646.815
5.393.059
5.219.601
4.928.520
4.009.809
Subaru
2,577,236 ;
2.007.047
1.845.549
1.763.973
1.445.867
1.364.094
Tcsla
(1.075.056)
(881.406)
(699.101)
(517.278)
(435.746)
(430.521)
Toyota
19.883.802
18.310.279
14.294.900
14.065.893
10.837.081
9.852.150
Volvo
553.511
341.195
307.712
324.133
281.572
245.769
VW
4.509.599
4.660.966
4.311.701
4.141.986
3.192.102
2.638.747
TOTAL
105,753,057
90,660,789
74,903,592
67,167,014
60,714,180
50,565,745
Table 12-37: Achieved GHG Levels (Mg), Alternative B - Trucks
Manufacturer
2027
2028
2029
2030
203 1
2032
Aston Martin
25,477
21,752 i
20,322 :
19.056
14.853
11.698
BMW
11.622.737
9.483.470
9.137.410
8.593.927
7.437.152
5.716.997
Ferrari
-
-
-
-
-
-
Ford
77.471.438
70.368.301
60.530.643
54.442.788
48.151.794
39.678.509
General Motors
79.677.689
76.283.685
65.286.474
57.034.202
51.533.459
45.524.052
Honda
28.025.417
26.595.964
23,579,577 :
22,007,703 :
17.498.521
15.078.332
Hyundai
16.965.537
15.367.139
13.973.281
12.657.913
10.375.852
8.117.906
JLR
3.469.363
3.241.855
2.898.421
2.533.947
2,275,026 :
1.833.264
Kia
15.644.293
14.227.745
12.830.406
11.862.660
9.626.557
7.427.836
Lucid
-
-
-
-
-
-
Ma/.da
8.367.551
7.794.003
6.775.089
5.962.477
5.275.440
4.302.935
McLaren
-
-
-
-
-
-
Mercedes Ben/
10.144.382
9.405.740
8.629.871
7.804.485
6.655.466
5.192.953
Mitsubishi
2.890.651
2.511.051
2.204.934
1.950.928
1.735.615
1.395.353
Nissan
14.049.490
12.736.509
11.197.094
10.069.251
9.233.801
7.674.395
Rivian
(133.029)
(110.593)
(89.877)
(66.687)
(57,356) :
(56.842)
Stellantis
75.393.661
67.120.363
58.584.919
51.300.782
46.500.414
39.565.333
Subaru
18.797.153
17.808.717
15.159.556
13.423.750
12.096.503
8.964.482
Tcsla
(986.869)
(817.534)
(660.924)
(490.193)
(421.528)
(417.637)
Toyota
58.985.659
51.580.634
47.126.036
39.998.875
34.928.540
27.138.130
Volvo
3.444.548
3,575,255
3.171.665
2.817.705
2.768.713
2.314.770
VW
17.162.006
14.592.991
11.257.995
10.874.524
10.174.060
7,657,352
TOTAL
441,017,153
401,787,048
351,612,891
312,798,094
275,802,881
227,119,818
12-22
-------
Table 12-38: Achieved GHG Levels (Mg), Alternative B - Combined
Manufacturer
2027
2028
2029
2030
203 1
2032
Aston Martin
49.943
42.648
37.941
34,275 ;
28.063
23.249
BMW
18.042.236
14.111.250
12,572,072 ;
11.356.411
9.949.548
7.992.502
Ferrari
184.016
124.915
114.938
106.911
92.407
72.591
Ford
81.853.781
74.589.901
64.200.396
57.642.586
51.517.185
42.628.192
General Motors
93.386.583
85.545.076
73.061.047
64.352.23 1
58.397.797
51.644.960
Honda
41.699.585
38.413.841
33.548.509
29.789.059
25.486.433
20.980.189
Hyundai
27.126.994
24.237.510
21.229.807
18.825.518
16.713.740
13.090.117
JLR
3.542.291
3.304.588
2.960.158
2.594.772
2,333,850 i
1.888.848
Kia
22.702.497
20.165.924
17.866.850
16.107.790
14.001.152
10.922.596
Lucid
(9,100) ;
(7,282) :
(5,763) :
(4.245)
(3,562)
(3.500)
Ma/.da
9.167.022
8.464.013
7,370,725 !
6.540.296
5.797.314
4.781.028
McLaren
46.569
31.859
28.962
26.852
21.192
16.171
Mercedes Ben/.
13.567.689
12.532.790
10.984.577
9.761.712
8.645.297
6.959.492
Mitsubishi
4.930.064
4.411.023
3.840.441
3.405.780
3.037.246
2.437.594
Nissan
24,207,703
21.645.054
18.702.635
16.568.097
14.260.656
11.405.899
Rivian
(133.029)
(110.593)
(89.877)
(66.687)
(57,356) :
(56.842)
Stellantis
82,553,776 ;
73,767,178
63.977.978
56.520.383
51.428.934
43.575.142
Subaru
21.374.388
19.815.764
17.005.105
15.187.723
13.542.370
10.328.575
Tcsla
(2.061.924)
(1.698.940)
(1.360.025)
(1.007.471)
(857.274)
(848.157)
Toyota
78.869.461
69.890.913
61.420.936
54.064.768
45.765.621
36.990.280
Volvo
3.998.059
3.916.450
3,479,377
3.141.838
3.050.285
2.560.539
VW
21.671.605
19.253.957
15.569.695
15.016.510
13.366.162
10.296.099
TOTAL
546,770,210
492,447,837
426,516,483
379,965,107
336,517,061
277,685,563
Table 12-39: GHG Credits/Debits Earned (Mg), Alternative B - Combined
Manufacturer
2027
2028
2029
2030
203 1
2032
Aston Martin
(12.095)
(8.908)
(7.969)
(8.110)
(4.446)
(2.182)
BMW
(2.583.385)
(509.745)
(479.825)
(810.157)
(524.337)
338.734
Ferrari
(62,235) :
(16.718)
(18.840)
(23.472)
(16.347)
(3.780)
Ford
(1.751.449)
(3.229.310)
(430.774)
(2.174.219)
(2.093.585)
764.194
General Motors
(4.647.962)
(6.813.265)
(2,780,977) :
(3.523.746)
(4.211.328)
(3.974.118)
Honda
4.075.003
1.825.006
2,283,003
1.359.549
2.315.404
3.511.990
Hyundai
2.119.552
1.710.107
1.815.632
1.252.807
1.326.831
2.869.483
JLR
(445.919)
(539.881)
(504.908)
(453,377)
(427.871)
(219.969)
Kia
1.591.341
1.230.641
1.156.484
440.785
721.820
2.040.195
Lucid
89.960
76,777 :
67.502
57.636
51.703
46.425
Ma/.da
180.369
(232.939)
(32.149)
(125.719)
(125.915)
193.317
McLaren
(19.617)
(7.966)
(7,752)
(8,315)
(4,359)
(970)
Mercedes Ben/
(209.766)
(656.491)
(427.101)
(566.936)
(401.316)
332.048
Mitsubishi
79.209
35,734 :
127.628
74.315
71.406
318.296
Nissan
(190.018)
(479.415)
113.836
(256.411)
312.799
1.483.772
Rivian
1.544.234
1.363.002
1.204.535
996.033
878.401
768.822
Stellantis
(6.474.863)
(6.308.346)
(3,778,079) :
(4.634.988)
(5.394.933)
(3.309.792)
Subaru
346.957
(628.154)
115.344
(232.631)
(293.558)
1.281.140
Tcsla
20.288.053
17.517.221
15.405.180
13.182.656
11.751.888
10.465.538
Toyota
4.681.000
3.830.080
4.004.765
2.641.641
4.576.939
7.520.956
Volvo
1.166.774
667.759
609.877
434.054
141.538
251.498
VW
576.040
359.163
1.857.979
143.637
68.701
1.536.611
TOTAL
20,341,184
9,184,351
20,293,390
7,765,031
8,719,436
26,212,208
12-23
-------
12.1.2 Projected Manufacturing Costs per Vehicle
EPA has performed an assessment of the estimated per-vehicle production costs for
manufacturers to meet the final MY 2027-2032 standards, relative to the No Action case. The
fleet average costs per vehicle have been grouped by regulatory class. EPA's OMEGA model
also tracks vehicles by body style (sedans, crossovers/SUVs and pickups). We have included
summary tables in this format. The costs in this section represent compliance costs to the
industry and are not necessarily the same as the costs experienced by the consumer when
purchasing a new vehicle. For example, the costs presented here do not include any state and
Federal purchase incentives that are available to consumers. Also, the manufacturer decisions for
the pricing of individual vehicles may not align exactly with the production cost impacts for that
particular vehicle. EPA's OMEGA model assumes that manufacturers distribute compliance
costs through limited cross-subsidization of prices between vehicles in order to maintain an
appropriate mix of debit- and credit-generating vehicles that achieves compliance in a cost-
minimizing fashion.
12.1.2.1 Final GHG Standards
Incremental costs per vehicle for the final standards (compared to the No Action case) are
summarized by regulatory class in Table 12-40 and by body style in Table 12-41.
Table 12-40: Projected Manufacturing Costs Per Vehicle, Final Standards
2027
2028
2029
2030
203 1
2032
Cars
$135
$348
$552
$968
$849
$934
T nicks
$276
$642
$1,199
$1,703
$2,318
$2.561
Total
$232
$552
$1,002
$1,481
$1,875
S2,074
Table 12-41: Projected Manufacturing Costs Per Vehicle, Final Standards (by Body Style).
2027
2028
2029
2030
203 1
2032
Sedans
$115
$277
$555
$1,036
$666
$821
Crossovers/SUVs
$185
$694
$961
$1,443
$2,249
$2,558
Pickups
$528
$349
$1.611
$2,066
$1,816
$1,659
Total
$232
$552
$1,002
$1,481
$1,875
S2,074
Incremental costs per vehicle for the final standards, compared to the No Action case, are
shown for each OEM in Table 12-42, Table 12-43, and Table 12-44 for cars, trucks, and the
combined fleet, respectively.279
279 For some manufacturers in these tables, costs for the final standards are projected to be lower than in the No
Action case. This reflects the combined effects of cost learning with the higher accumulated battery production
under the final standards, and more ICE technology applied by some manufacturers in the No Action case.
12-24
-------
Table 12-42: Projected Manufacturing Costs Per Vehicle, Final Standards - Cars
Manufacturer
2027
2028
2029
2030
203 1
2032
Aston Martin
$474
$969
$1,069
$1,935
$1,634
$1,450
BMW
-$10
$238 :
$917
$2,052 i '
$843
$1,406
Ferrari
$422 Y
$2,039
$1,602
$2,169
$2,627
$2,598
Ford
$165
-$682
-$906
-$745 ]
$2
$114
General Motors
$129
$2,338 j
$2,429
$2,415
$2,587
$2,141
Honda
$43 :
$73 7
$556 ;
$1,307
$1,290
$1,070
Hyundai
$2%
$53 1
$566
$883
$862
$753
JLR
$523 1
$81
$76 =
-$388
$89
$70
Kia
-$80
$290
$510
$1,132
$737
$763
Lucid
So
$o ]
$o;
SO
$0
$0
Ma/.da
-$154
$1,426
$858
$1,027
$708
$1,469
McLaren
$382 !
$2,121
$1,772 ;
$2,370:
$2,876
$3,080
Mercedes Ben/.
-$217 1
-$109
-$353 1
$956
-$172
$642
Mitsubishi
$153
-$204
$105
$495
$711
$708
Nissan
$180
$409
$723 ]
$948
$813
$1,069
Rivian
-
-
-
-
-
-
Stellantis
-$768 ;
-$1,294
-$1,020
-$869
-$166
-$99
Subaru
$177 :
$2,490
$2,825 ;
$3,030
$2,888
$2,517
Tcsla
$0 ]
SO
$0 1
So
$0
$0
Toyota
$363 ;
$38 !
$327 :
$832 !
$486
$914
Volvo
$110
$895
$1,183
$412
$638
$805
VW
$5%
-$579 r
-$862
-$365 T
-$631
-$433
TOTAL
St 35
$348
$552 ;
$968
S849
$934
Table 12-43: Projected Manufacturing Costs Per Vehicle, Final
Standards
- Trucks
Manufacturer
2027
2028
2029
2030
203 1
2032
Aston Martin
$197
$414
$2,040
$2,151
$3,674
$4,450
BMW
$362 r
$1,489
$1,129
$1,295
$2,331
$3,160
Ferrari
-
-
-
-
-
-
Ford
$385
$744
$1,308
$1/773 T"
$1,485
$1,714
General Motors
$337 V"
$396
$1.118
$1,277 ;
$2,142
$2,801
Honda
$305 I
$63 1
$890
$1,201
$1,832
$2,289
Hyundai
$47 f
$449
$1,091
S 1.752
$2,490
$2,262
JLR
$185
$595
$1,597
$2,432 !
$2,973
$3,201
Kia
$413
HI $57° :
$1,182
$1,636
$2,685
$2,777
Lucid
-
-
-
-
-
-
Ma/.da
IZZZ$337 II
$353
$1,371
$2.116
$2,667
$2,599
McLaren
-
-
-
-
-
-
Mercedes Ben/
$548
$925 ;
$1,827 ;
$2,509
$3,821
$3,418
Mitsubishi
$169
$1,386
$1,961
$2,533 T
$2,981
$3,005
Nissan
$269
$534 I
$499
S 1.767
$2,638
$2,776
Rivian
SO
So
SO
So
$0
$0
Stellantis
$355 :
$679
$1,387
$1,960
$2,844
$3,045
Subaru
$190
$534 T"
$1,106
$1,886
$2,313
$2,375
Tcsla
SO '
So
$o
SO
$0
$0
Toyota
$87 i
$745 ;
$1,092
$1,746
$2,601
$2,747
Volvo
$288
$185
$786 i
$1,749
$2,015
$1,973
VW
-$21 :
$849
$1,717 :
$2,109
$2,895
$2,960
TOTAL
$2761
$642
SI, 199
$1,703
$2,318
$2,561
12-25
-------
Table 12-44: Projected Manufacturing Costs Per Vehicle, Final Standards - Combined
Manufacturer
2027
2028
2029
2030
203 1
2032
Aston Martin
$360
$740
$1,473
$2,026
$2,490
$2,712
BMW
$206
$971
$1,041
$1,605
$1,723 ]
$2,446
Ferrari
$422 1
$2,039
$1,602
$2,169
$2,627 1
$2,598
Ford
$362]
$600 ]
$1,086
$1,521
$1,337 ]
$1,555
General Motors
$288
$850
$1,423
$1,541
$2,245 ]
$2,648
Honda
$187
$382 ]'"
$742]
$1,248
$1,593
$1,754
Hyundai
$167
$489
$841
$1,338
$1,717 ]
$1,547
JLR
$195
$581
$1,554
$2,354 ]
$2,894
$3.116
Kia
$186
$442
$877 ]
$1,408
$1,805
$1,869
Lucid
So
$0 ] "
So
$0 ]
So
$0
Ma/.da
$273 ;
$492
$1,306
$1,977 :
$2,419
$2,456
McLaren
$382 |
$2,121
$1,772 ]
$2,370 ]
$2,876 ]
$3,080
Mercedes Ben/.
$294
$584
$1.112
$2,002
$2,522 ]
$2,517
Mitsubishi
$160
$561
$1,000 ]
$1,478
$1,807
$1,817
Nissan
$219
$464
$624
$1,310
$1,622
$1,827
Rivian
$o ]"
So
SO
SO
So
$0
Stellantis
$234 :
$469
$1,133
$1,663
$2,531 ;
$2,718
Subaru
$188
$816
$1,351
$2,048
$2,394
$2,395
Tcsla
$o
$o:
SO
SO
SO
$0
Toyota
$198
$462
$788 !
$1,384
$1,767 :
$2,025
Volvo
$247 r
$346
$876 :
$1,448
$1,706
$1,711
VW
$173 j
$405
$921
$1,349
$1,817
$1,927
TOTAL
$232]'
$552 ;
SI,002
SI, 481
SI,875
S2,074
12.1.2.2 Alternative A
Incremental costs per vehicle for Alternative A (compared to the No Action case) are
summarized by regulatory class in Table 12-45 and by body style in Table 12-46.
Table 12-45: Projected Manufacturing Costs Per Vehicle, Alternative A
2027
2028
2029
2030
203 1
2032
Cars
$597
$832
$932
$1,085
$1,022
$1,085
T nicks
$1,345
$2,218
$2,594
$2,958
$3,021
$2,999
Total
SI, 114
SI,794
S2,088
$2,390
S2,418
S2,425
Table 12-46: Projected Manufacturing Costs Per Vehicle, Alternative A (by Body Style
2027
2028
2029
2030
203 1
2032
Sedans
$337
$686
$799
$913
$808
$978
Crossovcrs/SUVs
$1,277
$1,937
$2,249
$2.681
$2,848
$2,824
Pickups
$1,385
$2,469
$2,874
$2,905
$2,571
$2,499
Total
S 1,114
SI,794
S2,088
$2,390
S2,418
S2,425
Incremental costs per vehicle for Alternative A, compared to the No Action case, are shown
for each OEM in Table 12-47, Table 12-48, and Table 12-49 for cars, trucks, and the combined
fleet, respectively.
12-26
-------
Table 12-47: Projected Manufacturing Costs Per Vehicle, Alternative A - Cars
Manufacturer
2027
2028
2029
2030
203 1
2032
Aston Martin
$1,014
$2,377
$2,326 f
$2,835 ;
$2,585
$2,583
BMW
$227 r
$1,144
$1,398
$1,435
$1,135
$1,854
Ferrari
$491
$2,677 T
$2,506
$2,944
$3,123
$2,983
Ford
$825 1
-$20 i
-$174
$168
$354
$339
General Motors
$682
$904
$1,334
$1,454
$1,442
$1,150
Honda
$644
$933 f
$1,107
$1,330
$1,498
$1,306
Hyundai
$882
$725 l
$841
$1,107
$888
$1,027
JLR
$1,986
$2,060
$1,437
S 1.323 ]
$2,003
$2,223
Kia
$369
$1,264
$1,513
$1,807
$1,397
$1,284
Lucid
SO
SO
$0 ]
$0 ;
$0
$0
Ma/.da
$255 !
$3,251 :
$2,394
$2,786 ;
$2,055
$2,588
McLaren
$553 f
$2,687 f
$2,326 r
$2,817 ]"
$2,970
$3,080
Mercedes Ben/.
$689
$1,447
$927 j
$1,567
$765
$1,061
Mitsubishi
$1,238
$1,017
$1,089
$1,243
$1,307
$1,275
Nissan
$664
$1.117
$1,044
$1,403
$1,399
$1,482
Rivian
-
-
-
-
-
-
Stellantis
$467
-$2,626
-$2,401
-$2,741
-$2,098
-$2,049
Subaru
$921
$1.117
$1,392
$1,625
$1,499
$1,735
Tcsla
SO
SO
$0 ]"
$0 ;
$0
$0
Toyota
$606 :
$1,410
$1,486
$1,341
$1,294
$1,635
Volvo
$1,090
$275;'
$516
$246
$50
$381
VW
$240
$451
$181
$1,064
$498
$146
TOTAL
$597 ;
S832
$932
SI,085
SI,022
SI,085
Table 12-48: Projected Manufacturing Costs Per
Vehicle, Alternative A -
Trucks
Manufacturer
2027
2028
2029
2030
203 1
2032
Aston Martin
$1,898
$1,930
$2,489
$3,456
$4,648
$4,609
BMW
$1,675
$2,939
$2.691
$3,466
$4,202
$3,647
Ferrari
-
-
-
-
-
-
Ford
$1,319
$2,365 :
$2,773 T'
$3,094
$1259
$2,138
General Motors
$1,176
$2,264
$2,896
$2,673 ]
$2,765
$2,890
Honda
$1,528
$1,925
$1,921
$2,408
$2,542
$2,896
Hyundai
$1,547
$2,214
$2,326 r
$2,756;
$3,358
$3,037
JLR
$1,696
$2,807 1
$3,198
$3,673 ]
$3,876
$4,038
Kia
$1,967
$1,855
$1,990
$2,447
$2,790
$3,038
Lucid
-
-
-
-
-
-
Ma/.da
$1,742
$2,209
$2,737 1
$2,988
$3,358
$3,272
McLaren
-
-
-
-
-
-
Mercedes Ben/
$2,087 '1"
$1352]
$2,970 ;
$3^22 V
$3,843
$3,359
Mitsubishi
$1,756
$2,504
$2,813
$3,148
$3,416
$3,532
Nissan
$1,124
$1,870
$2.119
$2,809
$2,859
$2,833
Rivian
So
SO
$0 T
So
$0
$0
Stellantis
$1,124
$2,360
$3.03 1
$3,571 !
$3,588
$3,476
Subaru
$1,604
$2,256 ]
$2,933 V
$3,316
$3,640
$3,378
Tcsla
$o;
$0 :"
So
SO
$0 '
$0
Toyota
$1,163
$1,821
$2,103
$2,888
$3,095
$3,052
Volvo
$942
$4,780
$4,669
$4.861
$4,844
$4,622
VW
$1,803
$2,752 ]
$2,376 "
S2.582
$3,519
$3,892
TOTAL
$1,345
$2,218
$2,594
$2,958
S3,021
$2,999
12-27
-------
Table 12-49: Projected Manufacturing Costs Per Vehicle, Alternative A - Combined
Manufacturer
2027
2028
2029
2030
203 1
2032
Aston Martin
$1,376
$2,192
$2,394
$3,095
$3,451
S3.435
BMW
$1,069
$2,196
$2,159
$2,634
$2,950
$2,917
Ferrari
$491
$2,677 ;"
$2,506
$2,944
S3.123
$2,983
Ford
$1,269
$2,125
$2,478 i"
$2,802 I
$2,068
$1,959
General Motors
$1,059
$1,945
$2,533 1"
$2,390
$2,459
$2,487
Honda
$1,130
$1,483
$1,560
$1,932
$2,083
$2,198
Hyundai
$1,227
$1,502
$1,618
$1,971
$2,186
$2,084
JLR
$1,705
$2,786 :
$3,149
$3,609
$3,825 ]
$3,989
Kia
$1,231
$1,585
$1,773 ;
$2,157 7
$2,161
$2,248
Lucid
So
SO
SO
$o:
So
$0
Ma/.da
$1,547
$2,344
$2,693
$2,962
$3,193
$3,186
McLaren
$553 1
$2,687 :"
$2,326 j
$2,817 7
$2,970 ]
$3,080
Mercedes Ben/.
$1,622
$2,054
$2,301
$2,614
$2,842
$2,613
Mitsubishi
$1,487
$1,733 T
$1,920
$2,162
$2,325 7
$2,365
Nissan
$863
$1,447
$1,517
$2,024
$2,046
$2,082
Rivian
So
$0 j
$o r
So
SO
$0
Stellantis
$1,053
$1,829
$2,458
$2,909
$2,995
$2,902
Subaru
$1,504
$2,092
$2,713 7
$3,076 1
$3,339 ;
$3,148
Tcsla
$o ]
So
So
$0 |
$o :
$0
Toyota
$938
$1,657
$1,858
$2,276 :
$2,384
$2,494
Volvo
$976
$3,757 I
$3,731 r
$3,821
S3.768
$3,671
VW
$1,311
$2,036
$1,698
$2.116
$2,595 7
$2,751
TOTAL
SI,114
SI,794
S2,088
$2,390
$2,418
S2,425
12.1.2.3 Alternative B
Incremental costs per vehicle for Alternative B (compared to the No Action case) are
summarized by regulatory class in Table 12-50 and by body style in Table 12-51.
Table 12-50: Projected Manufacturing Costs Per Vehicle,
Alternative B
2027
2028
2029
2030
203 1
2032
Cars
$114
$71 I
$373 j
$692
$607
$716
T nicks
$259
$598
$1,182
$1,671
$1,971
$2,358
Total
$214
S437
$936
SI,375
$1,561
SI,867
Table 12-51: Projected Manufacturing Costs Per Vehicle, Alternative B (by Body Style)
2027
2028
2029
2030
203 1
2032
Sedans
$91
-$63
$335
$756
$499
$572
Crossovers/SUVs
$198
$644
$1,008
$1,320
$1,775
$2,314
Pickups
$409
$251
$1,315
$2,202
$1,898
$1,617
Total
$214
S437
S936
$1,375
$1,561
$1,867
Incremental costs per vehicle for Alternative B, compared to the No Action case, are shown for
each OEM in 12-52, Table 12-53, and Table 12-54 for cars, trucks, and the combined fleet,
respectively.
12-28
-------
Table 12-52: Projected Manufacturing Costs Per Vehicle, Alternative B - Cars
Manufacturer
2027
2028
2029
2030
203 1
2032
Aston Martin
$394
$1,303
$1,293
$1,933
$2,051
$2,227
BMW
-$8 T"
$495
$1,306
$2,046
$1,324
$1,352
Ferrari
$422 :
$2,324i
$2,266
$2,356 i
$2,590
$2,880
Ford
$2171
-$748
-$975 r"
-$583 f
-$457
-$390
General Motors
$40
$2,276 1
$2,450
$2,476
$2,439
$1,975
Honda
$39
$0 ; "
$336 ]
$996 ;
$707
$897
Hyundai
$295 '!
$55 !
$507 1
$911
$359
$594
JLR
$265 ;
-$34
$55 ; "
-$384
$95
$117
Kia
-$82 V"
$165
$442
$873 7
$286
$387
Lucid
So
SO
SO
So
So
$0
Ma/.da
-$153
$235 :
-$364
-$3 !
-S3 97
$366
McLaren
$372T
$2,108
$1,948
$2,017
$2,335
$2,939
Mercedes Ben/.
-$171
S67
$499
$1,675
$785
$699
Mitsubishi
$153
-$343 |
$63 1
$395 :
$537
$701
Nissan
$162
$143
$430
$889
$1,081
$1,434
Rivian
-
-
-
-
-
-
Stellantis
-$797 ;
-$4,300
-$3,588
-$3,718
-$3,134
-$2,632
Subaru
$178
$363 7'
$679
$759 !
$981
$956
Tcsla
$o:
So
SO
$01
$0
$0
Toyota
$342 !
-$33 |
$443
$557 :
$692
$867
Volvo
$110
$1,038
$982
$856
$577
$679
VW
$426
-$719
-$1,263
-$863
-$427
-$336
TOTAL
SI 14
$71
$373]
$692
$607
$716
Table 12-53: Projected Manufacturing Costs Per Vehicle, Alternative B -
Trucks
Manufacturer
2027
2028
2029
2030
203 1
2032
Aston Martin
$117
$639
$1,671
$1,859
$3,585
$3,952
BMW
$361
$1,053
$966
$1,152
$2,532
$3,075
Ferrari
-
-
-
-
-
-
Ford
$338 "i
$701
$1,187
$1,829
$1,103
$1,623
General Motors
$296
$331 :
$1,069
$1,250
$1,671
$1,838
Honda
$300 i
$521 7
$1,033
$1,360
$2,156
$2,449
Hyundai
$47 ]""
$484
$895
$1,398
$2,371
$2,586
JLR
$418
$921
$1,459
$2,319
$2,661
$3,190
Kia
$429
$558:'
$1,089
$1,506
$2,398
S3.073
Lucid
-
-
-
-
-
-
Ma/.da
$3387.'.'.
$334
$1,410
$2,072 77
$2,470
$2,605
McLaren
-
-
-
-
-
-
Mercedes Ben/
$539
$727 !
$1,355
$1,554
$2,569
$2,844
Mitsubishi
$169
$698
$1,421
$2,121
" $2,452
$2,883
Nissan
$244
$394
$784
$1,180
$1,529
$1,908
Rivian
SO '
SO
$0:
SO
$0
$0
Stellantis
$341
$805
$1,488
$2,121
$2,401
$2,499
Subaru
$190
$471
$1,240
$1,903
$2,165
$2,789
Tcsla
$o ;
So
$0;
$0 :
$0
$0
Toyota
$78 i
$669
$1,028
$1,792
$2,288
$2,862
Volvo
$288
$98
$843
$1,556
$1,800
$2,226
VW
-$2i :
$891
$2,002
$2,000
$2,262
$3,300
TOTAL
$259
$598
SI, 182
SI, 671
$1,971
$2,358
12-29
-------
Table 12-54: Projected Manufacturing Costs Per Vehicle, Alternative B - Combined
Manufacturer
2027
2028
2029
2030
203 1
2032
Aston Martin
$281
$1,028
$1,451
$1,902
$2,695
S2.952
BMW
$207:'"
$822 :
$1,106
$1,519
$2,039
$2,374
Ferrari
$422 !
$2,324]
$2,266
$2,356 ]
$2,590
$2,880
Ford
S326
$555:
$970
$1,587 :
$947
$1,422
General Motors
$236!"
$786 :
$1,390
$1,535 !
$1,849
$1,870
Honda
$183
$289
$724'!""
$1,199
$1,519
$1,768
Hyundai
$166
$279;
$710
$1,166
$1,416
$1,642
JLR
$413
$894
$1,420
$2,245 !
$2,591
$3,107
Kia
$194
S3 78
$795 !
$1,219
$1,444
$1,863
Lucid
SO
So
$0 !
SO
SO
$0
Ma/.da
$273 i
$321 !
$1,183
$1,808
$2,107
S2.323
McLaren
$372 ]
$2,108
$1,948
$2,017 ':
S2.335
$2,939
Mercedes Ben/.
$303 1
$510
$1,074
$1,594
$1,989
$2,148
Mitsubishi
$161
$158
$717 :
SI.227
$1,461
$1,754
Nissan
$197
$253 r
$586
$1,018
$1,279
$1,644
Rivian
SO
So
So
So
So
$0
Stellantis
$218
$261
$952 i
$1,509
$1,824
$1,965
Subaru
$188
$455 ;
$1,160
$1,741
$1,998
S2.532
Tcsla
SO
$o ] "
SO
So
So
So
Toyota
$185
$389
$796
$1,303
$1,658
S2.076
Volvo
$247 V
$311
$874
$1,398
S 1.525
$1,879
VW
$120
$390 !
$993
$1,121
$1,440
$2,192
TOTAL
S214
S437
$936
St,375
SI, 561
SI,867
12.1.3 Technology Penetration Rates
Presented below are the projected technology penetration rates, by manufacturer, for cars,
trucks, and the combined fleet, for the No Action case, and the final standards and alternatives.
Tables are provided by manufacturer and regulatory class for BEV and PHEV penetrations.
Summary tables for strong HEV penetrations and a few key ICE technology groupings (TURB12
and Atkinson engines) are also provided.
12.1.3.1 No Action Case
Table 12-55 through Table 12-57 give BEV penetrations for the No Action case, by
manufacturer. Similarly, Table 12-58 through Table 12-60 provide PHEV penetrations.
12-30
-------
Table 12-55: Projected BEV Penetrations, No Action - Cars
Manufacturer
2027
2028
2029
2030
203 1
2032
Aston Martin
30%
27%
34%
34%
39%
41%
BMW
23%
30%
32%
32%
42%
43%
Ferrari
23%
22%
26%
26%
28%
30%
Ford
35%
34%
41%
39%
40%
43%
General Motors
27% i
25%
26%
28%
31%
35%
Honda
29%
29%
31%
32%
35%
36%
Hyundai
28%
29%
32%
33%
39%
38%
JLR
26%
31%
31%
36%
33%
34%
Kia
35%
33%
36%
37%
39%
41%
Lucid
100%
100%
100%
100%
100%
100%
Ma/.da
34%
32%
45%
43%
53%
46%
McLaren
26%
27%
30%
30%
32%
35%
Mercedes Ben/.
32%
31%
38%
31%
41%
43%
Mitsubishi
26%
27%
30%
30%
34%
36%
Nissan
26%
27%
3o%
31%
36" o
38%
Rivian
-
-
-
-
-
-
Stellantis
: 32%
32%
41%
46%
44%
40%
Subaru
25%
28%
29%
30%
34%
34%
Tcsla
100%
100%
100%
100%
100%
100%
Toyota
27%
28%
31%
31%
37%
36%
Volvo
51%
51%
54%
55%
61%
62%
VW
34%
31%
40%
37%
43%
46%
TOTAL
34%
34%
37%
38%
42%
43%
12-31
-------
Table 12-56: Projected BEV Penetrations, No Action - Trucks
Manufacturer
2027
2028
2029
2030
203 1
2032
Aston Martin
20%
24%
23%
24%
24%
25%
BMW
19%
25% I ... J
"7. 27% ^
28%
26%
1 27%
Ferrari
-
-
-
-
-
-
Ford
21%
23%
25%
26%
29%
30%
General Motors
19%
22%
26%
27%
29%
30%
Honda
21%
23%
26%
27%
29%
30%
Hyundai
23%
24%
26%
27%
27%
29%
JLR
23%
23%
27%
27%
30% :
31%
Kia
20%"""
23%
25% ¦ -¦¦ ¦¦
26%
28%
29%
Lucid
-
-
-
-
-
-
Ma/.da
22%
j"; 23%;
""25%
26%
28%
**** 30%
McLaren
-
-
-
-
-
-
Mercedes Ben/.
22% ;
24%
25%
28%
28%
28%
Mitsubishi
22%
23%
26%
26%
29%
29%
Nissan
22%
24%
27%
29%
29%
31%
Rivian
100%
100%
100% j
100%
100%
100%
Stellantis
18%
21%
23%
24%
27%
29%
Subaru
23%
24%
27%
27%
30%
31%
Tcsla
100%
100%
100%
100%
100%
100%
Toyota
20%
22%
25%
25%
27%
29%
Volvo
16%
16%
19%
20%
22%
25%
VW
26%
25%
30%
32%
33%
34%
TOTAL
22%
24%
27%
28%
30%
31%
Table 12-57: Projected BEV Penetrations,
No Action
- Combined
Manufacturer
2027
2028
2029
2030
203 1
2032
Aston Martin
26%
26%
30%
30%
110/
JJ/O
34%
BMW
21%
27%
29%
29%
33%
34%
Ferrari
23%
22%
26%
26%
28%
30%
Ford
22%
24%
27%
28%
30%
31%
General Motors
21%
23%
26%
27%
30%
31%
Honda
25%
26%
28%
29%
32%
33%
Hyundai
26%
27%
29%
30%
33%
33%
JLR
23%
23%
27%
28%
30%
31%
Kia
27%
28%
30%
31%
33%
35%
Lucid
100%
100%
100%
100%
100%
100%
Ma/.da
23%
24%
28%
28%
31%
32%
McLaren
26%
27%
30%
30%
32%
"3^0/
JJ/O
Mercedes Ben/
25%
26%
29%
29%
' 32%
110/
JJ/O
Mitsubishi
24%
25%
28%
28%
31%
33%
Nissan
24%
26%
29%
30%
33%
34%
Rivian
100%
100%
100%
100%
100%
100%
Stellantis
20%
22%
25%
26%
29%
30%
Subaru
23%
""25%
27%
28%
31%
31%
Tcsla
100%
100%
100%
100%
100%
100%
Toyota
23%
24%
27%
28%
31%
32%
Volvo
24%
24%
27%
28%
31%
33%
VW
29%
27%
33%
33%
36%
37%
TOTAL
26%
27%
30%
31%
34%
35%
12-32
-------
Table 12-58: Projected PHEV Penetrations, No Action - Cars
Manufacturer
2027
2028
2029
2030
203 1
2032
Aston Martin
4%
8%
8%
9%
9%
10%
BMW
5%
6%
7%
8%
8%
10%
Ferrari
20%
20%
20%
23%
25%
27%
Ford
10%
11%
12%
15%
12%
13%
General Motors
3%
5%
6%
6%
5%
9%
Honda
4%
5%
6%
7%
9%
14%
Hyundai
4%
5%
6%
7%
7%
13%
JLR
5%
7%
8%
9%
9%
11%
Kia
5%
5%
6%
7%
8%
12%
Lucid
0%
0%
0%
0%
0%
0%
Ma/.da
3%
6%
5%
7%
7%
8%
McLaren
8%
8%
10%
14%
18%
17%
Mercedes Ben/.
5%
7%
7%
10%
10%
12%
Mitsubishi
7%
8%
9%
10%
10%
14%
Nissan
4%
4%
6%
6% ^
7 %
10%
Rivian
-
-
-
-
-
-
Stellantis
4%
5%
6%
8%
8%
15%
Subaru
5%
6%
7% '
7%
8%
14%
Tcsla
" 0%
0%
0%
0%
0%
0%
Toyota
5%
5%
6%
6%
7%
9%
Volvo
7%
7%
7%
7%
7%
8%
VW
6%
; 6%
8%
9%
9%
11%
TOTAL
4%
5%
6%
7%
7%
10%
Table 12-59: Projected PHEV Penetrations, No Action - Trucks
Manufacturer
2027
2028
2029
2030
203 1
2032
Aston Martin
5%
8%
8%
10%
11%
19%
BMW
8" o
8%
10%
13%
13%
19%
Ferrari
-
-
-
-
-
-
Ford
4%
4%
8%
10%
7%
12%
General Motors
5%
4%
5%
8%
7%
12%
Honda
4%
6%
7%
9%
12%
14%
Hyundai
6%
7%
8%
11%
12%
19%
JLR
7%
9%
8%
9%
10%
14%
Kia
; 5%""
6%
7%'I ' ""'
1 9% 2
9% 1
13%
Lucid
-
-
-
-
-
-
Ma/.da
6% *"7.
9% ^
7% ^
8% '
' 9% '_'
15%
McLaren
-
-
-
-
-
-
Mercedes Ben/
5%
7%
8%
9%
11%
15%
Mitsubishi
3%
4%
5%
6%
7%
12%
Nissan
4%
4%
8%
6%
10%
13%
Rivian
0%
0%
0%
0%
0%
0%
Stellantis
9%
7%
9%
10%
10%
13%
Subaru
5%
7%
8%
9%
10%
16%
Tcsla
0%
0%
0%
0%
0%
0%
Toyota
7%
7%
8%
9%
10%
13%
Volvo
24%
24%
24%
24%
24%
26%
VW
6%
7%
8%
9%
10%
12%
TOTAL
6%
6%
8%
9%
9%
13%
12-33
-------
Table 12-60: Projected PHEV Penetrations, No Action - Combined
Manufacturer
2027
2028
2029
2030
203 1
2032
Aston Martin
5%
8%
8%
9%
10%
14%
BMW
7%
7%
9%
11%
11%
15%
Ferrari
20%
20%
20%
23%
25%
27%
Ford
5%
5%
9%
10%
7%
12%
General Motors
5%
4%
5%
7%
6%
11%
Honda
4%
5%
7%
8%
10%
14%
Hyundai
5%
6%
7%
9%
10%
16%
JLR
7%
9%
8%
9%
10%
14%
Kia
5%
6%
7%
8%
8%
13%
Lucid
0%
0%
0%
0%
0%
0%
Ma/.da
5%
9%
7%
8%
9%
14%
McLaren
8%
8%
10%
14%
18%
17%
Mercedes Ben/.
5%
j 7%
8%
9%
10%
14%
Mitsubishi
5%
6%
7%
8%
9%
14%
Nissan
4%
4%
7%
6%
8%
11%
Rivian
0%
0%
0%
0%
0%
0%
Stellantis
8%
1 7%
9%
10%
10%
13%
Subaru
5%
7%
8%
9%
10%
15%
Tcsla
0%
0%
0%
0%
0%
0%
Toyota
6%
6%
7%
8%
9%
11%
Volvo
20%
21%
21%
21%
21%
22%
VW
6%
7%
8%
9%
10%
12%
TOTAL
5%
6%
7%
8%
8%
12%
The tables below provide summary technology penetrations for the No Action case for strong
hybrids, TURB12 (only non-Miller engines) and MIL (Miller cycle engines). For these tables,
strong hybrids include all engine types, while the TURB12 and MIL penetrations shown are only
for non-hybrid versions of those vehicles.
Table 12-61: Projected Strong HEV Penetrations, No Action
2027 2028 2029 2030 2031 2032
Cars 3% 2% 2% 2% 1% 1%
Trucks 4% 4% 4% 4% 6% 6%
Total 4% 3% 3% 3% 5% 5%
Table 12-62: Projected TURB12 Penetrations, No Action
2027 2028 2029 2030 2031 2032
Cars 37% 28% 26% 25% 23% 21%
Trucks 53% 50% 46% 44% 40% 36%
Total 48% 43% 40% 39% 35% 32%
Table 12-63: Projected MIL Penetrations, No Action
2027 2028 2029 2030 2031 2032
Cars 5% 3% 2% 2% 2% 2%
Trucks 2% 1% 1% 1% 1% 1%
Total 3% 1% 1% 1% 1% 1%
12-34
-------
12.1.3.2 Final Standards
Table 12-64 through Table 12-66 give BEV penetrations for the Final Standards, by
manufacturer. Similarly,
Table 12-67 through Table 12-69 provide PHEV penetrations.
Table 12-64: Projected BEV Penetrations, Final Standards - Cars
Manufacturer
2027
2028
2029
2030
203 1
2032
Aston Martin
31%
29%
39%
48%
49%
50%
BMW
21%
35%
47%
60%
58%
63%
Ferrari
23%
31%
31%
39%
47%
53%
Ford
35%
37%
42%
45%
53%
56%
General Motors
27% 1
29%
37%
40%
49%
54%
Honda
28%
35%
46%
56%
60%
66%
Hyundai
30%
37%
41%
47%
51%
58%
JLR
29%
35%
35%
35%
37%
37%
Kia
32%
39%
44%
56%
55%
64%
Lucid
100%
100%
100%
100%
100%
100%
Ma/.da
31%
39%
45%
46%
49%
53%
McLaren
27%
33%
35%
46%
57%
61%
Mercedes Ben/.
29%
31%
36%
40%
42%
53%
Mitsubishi
26%
31%
38%
45%
49%
54%
Nissan
26%
34%
45%
47%
53%
59%
Rivian
-
-
-
-
-
-
Stellantis
23%
30%
46%
55%
66%
68%
Subaru
26%
32%
44%
46%
50%
62%
Tcsla
100%
100%
100%
100%
100%
100%
Toyota
31%
36%
46%
53%
53%
60%
Volvo
51%
57%
65%
56%
68%
71%
VW
40%
36% ;
44%
47%
50%
57%
TOTAL
35%
39%
48%
54%
57%
63%
12-35
-------
Table 12-65: Projected BEV Penetrations, Final Standards - Trucks
Manufacturer
2027
2028
2029
2030
203 1
2032
Aston Martin
20%
24%
34%
37%
49%
59%
BMW
20%
28%
29%
32%
46%
47%
Ferrari
-
-
-
-
-
-
Ford
21%
25%
35%
40%
48%
53%
General Motors
19%
24%
33%
37%
46%
54%
Honda
22%
27%
30%
'" 36%
45%
48%
Hyundai
23%
26%
35%
42%
49%
51%
JLR
24%
27%
36%
42%
47%
51%
Kia
26%
32%
48%
49%
Lucid
-
-
-
-
-
-
Ma/.da
23% ;
J 27%""'"
j" J'"''""35% ^
42%
49%
52%
McLaren
-
-
-
-
-
-
Mercedes Ben/.
24%
30%
37%
45%
""""52%
53%
Mitsubishi
23%
27%
35%
41%
47%
50%
Nissan
23%
27%
30%
41%
52%
56%
Rivian
100%
100%
100%
100%
100%
100%
Stellantis
20%
23%
32%
36%
43%
50%
Subaru
24%
27%
35%
42%
48%
50%
Tcsla
100%
100%
100%
100%
100%
100%
Toyota
19%
25%
30%
] 37%
47%
50%
Volvo
17%
16%
24%
33%
37%
40%
VW
25%
27%
40%
45%
52%
54%
TOTAL
23%
27%
34%
40%
48%
52%
Table
12-66: Projected BEV Penetrations,
Final Standards
- Combined
Manufacturer
2027
2028
2029
2030
203 1
2032
Aston Martin
26%
27%
37%
43%
49%
54%
BMW
21%
31%
36%
43%
51%
^"30/
JJ/O
Ferrari
23%
31%
31%
39%
47%
JJ/O
Ford
22%
26%
36%
40%
49%
53%
General Motors
21%
25%
34%
38%
47%
54%
Honda
25%
30%
37%
45%
51%
56%
Hyundai
26%
31%
38%
44%
50%
55%
JLR
24%
27%
36%
42%
47%
50%
Kia
27%
32%
38%
46%
51%
56%
Lucid
100%
100%
100%
100%
100%
100%
Ma/.da
24%
28%
36%
43%
49%
52%
McLaren
27%
33%
35%
46%
57%
61%
Mercedes Ben/
26%
30%
37%
43%
49%
53%
Mitsubishi
25%
29%
37%
43%
48%
52%
Nissan
25%
31%
38%
44%
52%
57%
Rivian
100%
100%
100%
100%
100%
100%
Stellantis
20%
24%
T 33%
38%
46%
52%
Subaru
24%
28%
36%
43%
48%
52%
Tcsla
100%
100%
100%
100%
100%
100%
Toyota
24%
29%
36%
43%
49%
54%
Volvo
25%
26%
"33%
38%
44%
47%
VW
30%
30%
41%
45%
52%
55%
TOTAL
26%
31%
39%
44%
51%
56%
12-36
-------
Table 12-67: Projected PHEV Penetrations, Final Standards - Cars
Manufacturer
2027
2028
2029
2030
203 1
2032
Aston Martin
6%
8%
9%
10%
11%
12%
BMW
5%
7%
7%
8%
8%
12%
Ferrari
20%
21%
23%
24%
25%
23%
Ford
11%
11%
13%
13%
14%
16%
General Motors
4%
5%
7%
8%
9%
12%
Honda
4%
5%
7%
10%
12%
9%
Hyundai
4%
4%
6%
7%
12%
10%
JLR
5%
7%
8%
9%
10%
13%
Kia
5%
6%
8%
7%
11%
10%
Lucid
0%
0%
0%
0%
0%
0%
Ma/.da
4%
5%
9%
10%
15%
16%
McLaren
8%
9%
10%
11%
12%
12%
Mercedes Ben/.
5%
6%
8%
16%
11%
15%
Mitsubishi
7%
7%
9%
10%
13%
15%
Nissan
5% ^
4%
5%
6%
6%
8%
Rivian
-
-
-
-
-
-
Stellantis
6%
5%
6%
7%
8%
10%
Subaru
5%
6%
5%
10%
16%
7%
Tcsla
0%
0%
0%
0%
0%
0%
Toyota
4%
5%
6%
7%
9%
10%
Volvo
7%
7%
7%
7%
7%
7%
VW
5%
7%
7%
9%
10%
10%
TOTAL
4%
5%
6%
8%
9%
10%
Table 12-68: Projected PHEV Penetrations, Final Standards - Trucks
Manufacturer
2027
2028
2029
2030
203 1
2032
Aston Martin
5%
10%
9%
10%
12%
12%
BMW
8" o
15%
17%
18%
9%
.11 22%
Ferrari
-
-
-
-
-
-
Ford
6%
5%
6%
8%
9%
11%
General Motors
6%
4%
6%
9%
10%
13%
Honda
4%
6%
9%
9%
10%
14%
Hyundai
6%
7%
8%
9%
10%
13%
JLR
6%
7%
9%
10%
14%
17%
Kia
11"5%i
6"o
8%
::io%;::::
9"o
14%
Lucid
-
-
-
-
-
-
Ma/.da
6%
7% ^
9%
9% ^
11%
14%
McLaren
-
-
-
-
-
-
Mercedes Ben/
5%
7%
9%
10%
15%
17%
Mitsubishi
3%
4%
5%
7%
9%
12%
Nissan
4%
5%
8%
8%
7%
9%
Rivian
0%
0%
0%
0%
0%
0%
Stellantis
8%
9%
10%
13%
16%
17%
Subaru
5%
7%
8%
9%
11%
15%
Tcsla
0%
0%
0%
0%
0%
0%
Toyota
7%
7%
9%
11%
12%
16%
Volvo
24%
24%
24%
24%
24%
26%
VW
6%
8%
8%
10%
11%
12%
TOTAL
6%
7%
8%
10%
11%
14%
12-37
-------
Table 12-69: Projected PHEV Penetrations, Final Standards - Combined
Manufacturer
2027
2028
2029
2030
203 1
2032
Aston Martin
5%
9%
9%
10%
11%
12%
BMW
6%
12%
13%
14%
9%
18%
Ferrari
20%
21%
23%
24%
25%
23%
Ford
6%
5%
7%
8%
10%
11%
General Motors
5%
4%
7%
9%
10%
13%
Honda
4%
6%
8%
9%
11%
12%
Hyundai
5%
5%
7%
8%
11%
12%
JLR
6%
7%
9%
10%
14%
16%
Kia
5%
6%
8%
9%
10%
12%
Lucid
0%
0%
0%
0%
0%
0%
Ma/.da
6%
7%
9%
9%
12%
14%
McLaren
8%
9%
10%
11%
12%
12%
Mercedes Ben/.
5%
| 7%
8%
12%
14%
16%
Mitsubishi
5%
6%
7%
8%
11%
14%
Nissan
4%
4%
6%
7%
7%
9%
Rivian
0%
0%
0%
0%
0% !
0%
Stellantis
8%
8%
9%
12%
15%
16%
Subaru
5%
7%
8%
10%
12%
14%
Tcsla
0%
0%
0%
0%
0%
0%
Toyota
6%
6%
8%
9%
11%
13%
Volvo
20%
21%
21%
21%
21%
22%
VW
6%
7%
8%
9%
10%
11%
TOTAL
6%
6%
8%
9%
11%
13%
The tables below provide summary technology penetrations for the Final Standards for strong
hybrids, TURB12 (only non-Miller engines) and MIL (Miller cycle engines). For these tables,
strong hybrids include all engine types, while the TURB12 and MIL penetrations shown are only
for non-hybrid versions of those vehicles.
Table 12-70: Projected Strong HEV Penetrations, Final Standards
2027 2028 2029 2030 2031 2032
Cars 3% 6% 5% 5% 4% 3%
Trucks 4% 3% 3% 3% 2% 2%
Total 4% 4% 4% 3% 3% 2%
Table 12-71: Projected TURB12 Penetrations, Final Standards
2027 2028 2029 2030 2031 2032
Cars 37% 23% 19% 16% 14% 11%
Trucks 52% 48% 41% 36% 29% 24%
Total 47% 40% 34% 30% 24% 20%
Table 12-72: Projected MIL Penetrations, Final Standards
2027 2028 2029 2030 2031 2032
Cars 5% 4% 4% 3% 3% 2%
Trucks 2% 1% 1% 1% 1% 1%
Total 3% 2% 2% 2% 1% 1%
12-38
-------
12.1.3.3 Alternative A
Table 12-73 through Table 12-75 give BEV penetrations for Alternative A, by manufacturer.
Similarly, Table 12-76 through Table 12-78 provide PHEV penetrations.
Table 12-73: Projected BEV Penetrations, Alternative A - Cars
Manufacturer
2027
2028
2029
2030
203 1
2032
Aston Martin
34%
43%
53%
57%
60%
63%
BMW
23%
36%
44%
45%
54%
62%
Ferrari
24%
30%
32%
41%
48%
51%
Ford
40%
42%
48%
56%
56%
59%
General Motors
30% 1
34%
42%
46%
45%
50%
Honda
35%
41%
51%
56%
65%
68%
Hyundai
36%
41%
46%
53%
57%
64%
JLR
33%
50%
43%
48%
55%
60%
Kia
38%
46%
55%
62%
65%
69%
Lucid
100%
100%
100%
100%
100%
100%
Ma/.da
34%
41%
56%
55%
72%
61%
McLaren
28%
35%
37% ;
49%
58%
61%
Mercedes Ben/.
40%
48%
52%
54%
55%
63%
Mitsubishi
37%
38%
45%
49%
53%
57%
Nissan
31%
40%
46%
^ 52% ^
60% ]
64%
Rivian
-
-
-
-
-
-
Stellantis
35%
36%
45%
42%
49%
50%
Subaru
31%
36%
40%
43%
46%
59%
Tcsla
100%
100%
100%
100%
100%
100%
Toyota
33%
42%
47%
48%
58%
64%
Volvo
60%
54%
60%
58%
64%
67%
VW
36%
40%
49%
* 7 52%I
57%
56%
TOTAL
39%
45%
51%
55%
60%
64%
12-39
-------
Table 12-74: Projected BEV Penetrations, Alternative A - Trucks
Manufacturer
2027
2028
2029
2030
203 1
2032
Aston Martin
29%
29%
27%
39%
52%
57%
BMW
26%
36%
38%
47%
51%
52%
Ferrari
-
-
-
-
-
-
Ford
27%
34%
43%
46%
48%
51%
General Motors
25%
33%
41%
46%
49%
53%
Honda
33% J
35%
37%
43%
47%
52%
Hyundai
35%
36%
40%
45%
52%
54%
JLR
36%
37%
43%
47%
52%
57%
Kia
""" 35% ;
34%
35%
41%
49%
52%
Lucid
-
-
-
-
-
-
Ma/.da
36% ^
35%
40%
44%
50%
56%
McLaren
-
-
-
-
-
-
Mercedes Ben/.
34%
34%
39%
45%
52%
56%
Mitsubishi
35%
36% j
42%
46%
52%
56%
Nissan
29%
31%
40%
45%
48%
51%
Rivian
100%
100%
100%
100%
100%
100%
Stellantis
25%
32%
41%
44%
46%
50%
Subaru
37%
36%
43%
47%
54%
57%
Tcsla
100%
100%
100%
100%
100%
100%
Toyota
30%
33% ;
39%
45%
49%
50%
Volvo
22%
25%
31%
37%
42%
45%
VW
40%
39%
42%
46%
56%
61%
TOTAL
31%
35%
42%
46%
50%
53%
Table 12-75: Projected BEV Penetrations,
Alternative A
- Combined
Manufacturer
2027
2028
2029
2030
203 1
2032
Aston Martin
32%
38%
42%
50%
56%
60%
BMW
25%
36%
41%
46%
52%
56%
Ferrari
24%
30%
32%
41%
48%
51%
Ford
29%
35%
43%
47%
48%
52%
General Motors
26%
34%
42%
46%
48%
52%
Honda
34%
37%
43%
49%
55%
59%
Hyundai
36%
39%
43%
49%
54%
59%
JLR
36%
38%
43%
47%
52%
57%
Kia
36%
39%
44%
50%
56%
60%
Lucid
100%
100%
100%
100%
100%
100%
Ma/.da
36%
36%
42%
45%
53%
57%
McLaren
28%
35%
37%
49%
58%
61%
Mercedes Ben/
36%
38%
43%
48%
53%
58%
Mitsubishi
36%
37%
44%
48%
52%
57%
Nissan
30%
36%
43%
49%
54%
59%
Rivian
100%
100%
100%
100%
100%
100%
Stellantis
26%
32%
41%
44%
46%
50%
Subaru
36%
36%
42%
46%
53%
57%
Tcsla
100%
100%
100%
100%
100%
100%
Toyota
31%
36%
42%
46%
52%
55%
Volvo
31%
32%
37%
42%
47%
50%
VW
39%
39%
44%
48%
56%
59%
TOTAL
33%
38%
45%
49%
53%
57%
12-40
-------
Table 12-76: Projected PHEV Penetrations, Alternative A - Cars
Manufacturer
2027
2028
2029
2030
203 1
2032
Aston Martin
7%
8%
9%
11%
12%
11%
BMW
5%
6%
8%
8%
9%
14%
Ferrari
20%
21%
23%
23%
24%
24%
Ford
12%
12%
15%
12%
14%
15%
General Motors
5%
6%
9%
10%
13%
15%
Honda
4%
7%
7%
8%
8%
9%
Hyundai
4%
5%
7%
7%
8%
9%
JLR
14%
7%
8%
9%
9%
10%
Kia
4%
; 6%
7%
7%
7%
8%
Lucid
0%
0%
0%
0%
0%
0%
Ma/.da
5%
6%
5%
15%
8%
17%
McLaren
9%
12%
14%
12%
12%
12%
Mercedes Ben/.
3%
7%
9%
9% r
10%
11%
Mitsubishi
7%
9%
10%
10%
12%
14%
Nissan
5% ^
5%
6%..^
6%
7% .131
7%
Rivian
-
-
-
-
-
-
Stellantis
4%
5%
6%
8% ;
8%
9%
Subaru
6%
6%
7%
10%
13%
9%
Tcsla
0%
0%
0%
0% ;
0%
0%
Toyota
4%
5%
8%
6%
7%
8%
Volvo
7%
7%
7%
7%
7%
9%
VW
5%
7%
8%
14%
10%
10%
TOTAL
4%
6%
7%
7%
8%
9%
Table 12-77: Projected PHEV Penetrations, Alternative A
- Trucks
Manufacturer
2027
2028
2029
2030
203 1
2032
Aston Martin
7%
9%
16%
15%
12%
13%
BMW
12%
14%
17%
18%
18%
18%
Ferrari
-
-
-
-
-
-
Ford
6%
7%
10%
10%
10%
11%
General Motors
6%
7%
11%
11%
12%
14%
Honda
3%
6%
9%
10% ;
12%
14%
Hyundai
5%
7%
9%
11%
13%
15%
JLR
3%
8%
10%
12%
13%
14%
Kia
4%
6%
11%
11%
9% 1
12%
Lucid
-
-
-
-
-
-
Ma/.da
10/
J /o
7% H
9%
9% """" f
11%
12%
McLaren
-
-
-
-
-
-
Mercedes Ben/
3%
8%
10%
10%
11%
12%
Mitsubishi
2%
5%
6%
7% 1
8%
10%
Nissan
5%
6%
7%
8%
9%
11%
Rivian
0%
0%
0%
0%
0%
0%
Stellantis
8%
10%
12%
13%
15%
17%
Subaru
3%
8%
9%
10%
11%
12%
Tcsla
0%
0%
0%
0%
0%
0%
Toyota
5%
8%
10%
10%
11%
16%
Volvo
24%
24%
24%
24%
24%
27%
VW
4%
8%
9%
10%
11%
11%
TOTAL
6%
8%
10%
11%
12%
14%
12-41
-------
Table 12-78: Projected PHEV Penetrations, Alternative A - Combined
Manufacturer
2027
2028
2029
2030
203 1
2032
Aston Martin
7%
8%
12%
12%
12%
12%
BMW
9%
11%
13%
14%
14%
17%
Ferrari
20%
21%
23%
23%
24%
24%
Ford
7%
7%
10%
10%
10%
11%
General Motors
6%
7%
11%
11%
12%
14%
Honda
3%
7%
8%
9%
10%
12%
Hyundai
4%
6%
8%
9%
11%
13%
JLR
3%
8%
10%
12%
13%
14%
Kia
4%
6%
9%
9%
8%
11%
Lucid
0%
0%
0%
0%
0%
0%
Ma/.da
3%
7%
8%
10%
10%
12%
McLaren
9%
12%
14%
12%
12%
12%
Mercedes Ben/.
3%
8%
10%
10%
11%
12%
Mitsubishi
5%
7%
8%
8%
10%
12%
Nissan
5%
5%
6%
7%
8%
9%
Rivian
0%
0%
0%
0%
0%
0%
Stellantis
8%
9%
11%
12%
14%
16%
Subaru
4%
7%
9%
10%
11%
12%
Tcsla
0%
0%
0%
0%
0%
0%
Toyota
5%
7%
9%
8%
10%
13%
Volvo
20%
21%
21%
21%
21%
23%
VW
4%
8%
9%
11%
11%
11%
TOTAL
5%
7%
9%
10%
11%
12%
The tables below provide summary technology penetrations for Alternative A for strong
hybrids, TURB12 (only non-Miller engines) and MIL (Miller cycle engines). For these tables,
strong hybrids include all engine types, while the TURB12 and MIL penetrations shown are only
for non-hybrid versions of those vehicles.
Table 12-79: Projected Strong HEV Penetrations, Alternative A
2027 2028 2029 2030 2031 2032
Cars 3% 4% 3% 3% 2% 2%
Trucks 3% 4% 3% 5% 7% 6%
Total 3% ; 4% ' 3% 5% 5% 5%
Table 12-80: Projected TURB12 Penetrations, Alternative A
2027 2028 2029 2030 2031 2032
Cars 34% 18% 15% 13% 11% 9%
Trucks 47% 35% 29% 24% 20% 17%
Total 43% 30% 25% 21% 17% 15%
Table 12-81: Projected MIL Penetrations, Alternative A
2027 2028 2029 2030 2031 2032
Cars 5% 2% 2% 1% 1% 1%
Trucks 1% 1% 1% 1% 1% 1%
Total 2% 1% 1% 1% 1% 1%
12-42
-------
12.1.3.4 Alternative B
Table 12-82 through Table 12-84 give BEV penetrations for Alternative B, by manufacturer.
Similarly, Table 12-85 through Table 12-87 provide PHEV penetrations.
Table 12-82: Projected BEV Penetrations, Alternative B - Cars
Manufacturer
2027
2028
2029
2030
203 1
2032
Aston Martin
31%
39%
46%
51%
56%
60%
BMW
21%
39%
53%
58%
62%
65%
Ferrari
23%
27%
32%
32%
44%
52%
Ford
36%
37%
44%
48%
49%
53%
General Motors
26%
28%
38%
41%
45%
47%
Honda
28%
33%
42%
52%
49%
59%
Hyundai
30%
34%
45%
51%
49%
56%
JLR
27%
34%
35%
35%
37%
38%
Kia
32%
40%
49%
55%
54%
58%
Lucid
100%
100%
100%
100%
100%
100%
Ma/.da
31%
39% * j
45%
42%
46%
48%
McLaren
27%
110/
JJ/O
38%
36%
52%
61%
Mercedes Ben/.
29%
33%
47%
52%
53%
57%
Mitsubishi
26%
30%
38%
43%
49%
56%
Nissan
26%
33% ;
42%
49%
59%
67%
Rivian
-
-
-
-
-
-
Stellantis
25%
28%
39%
41%
41%
52%
Subaru
26%
33%
37%
39%
49%
47%
Tcsla
100%
100%
100%
100%
100%
100%
Toyota
31%
35%
48%
48%
59%
60%
Volvo
51%
61%
65%
65%
70%
73%
VW
39%
35%
39%
41%
50%
58%
TOTAL
34%
39%
48%
52%
56%
60%
12-43
-------
Table 12-83: Projected BEV Penetrations, Alternative B - Trucks
Manufacturer
2027
2028
2029
2030
203 1
2032
Aston Martin
20%
22%
26%
26%
44%
56%
BMW
20% ]
25% J
28%
30%
40%
51%
Ferrari
-
-
-
-
-
-
Ford
21%
25%
34%
38%
45%
52%
General Motors
19%
24%
33%
39%
43%
48%
Honda
22%
26%
33%
35%
48%
53%
Hyundai
23%
26%
32%
36%
47%
56%
JLR
24%
25%
36%
43%
47%
56%
Kia
' 23%
24%
31%
"" 35%
46%
"" 55%
Lucid
-
-
-
-
-
-
Ma/.da
23% ;
J 27%""'"
j" J '""""36% ~
42%
48%
54%
McLaren
-
-
-
-
-
-
Mercedes Ben/.
24%
28%
JJ/O
37% *
46%
"55%
Mitsubishi
23%
27%
35%
41%
47%
55%
Nissan
23%
24%
1 32%
33%
38%
46%
Rivian
100%
100%
100%
100%
100%
100%
Stellantis
19%
23%
31%
34%
40%
47%
Subaru
24%
26%
36%
42%
47%
57%
Tcsla
100%
100%
100%
100%
100%
100%
Toyota
19%
24%
30%
34%
41%
52%
Volvo
17%
16%
24%
31%
36%
43%
VW
25%
27%
42%
44%
47%
57%
TOTAL
23%
26%
34%
38%
45%
52%
Table 12-84: Projected BEV Penetrations,
Alternative B
- Combined
Manufacturer
2027
2028
2029
2030
203 1
2032
Aston Martin
27%
32%
38%
41%
51%
58%
BMW
21%
31%
38%
41%
49%
57%
Ferrari
23%
27%
32%
32%
44%
52%
Ford
22%
26%
35%
39% I
45%
52%
General Motors
21%
25%
35%
40%
44%
48%
Honda
25%
29%
37%
42%
49%
56%
Hyundai
26%
30%
38%
43%
48%
56%
JLR
24%
25%
36%
42%
47%
55%
Kia
27%
31%
39%
44%
50%
56%
Lucid
100%
100%
100%
100%
100%
100%
Ma/.da
24%
28%
37%
42%
47%
54%
McLaren
27%
33%
38%
36%
52%
61%
Mercedes Ben/
26%
29%
38%
42%
48%
56%
Mitsubishi
25%
28%
36% ;
42%
48%
55%
Nissan
25%
29%
37%
42%
50%
58%
Rivian
100%
100%
100%
100%
100%
100%
Stellantis
20%
24%
32%
35%
41%
48%
Subaru
24%
"27%
36%
42%
47%
56%
Tcsla
100%
100%
100%
100%
100%
100%
Toyota
24%
28%
37%
40%
48%
55%
Volvo
25%
26%
34%
39%
43%
50%
VW
29%
30%
41%
43%
48%
57%
TOTAL
26%
30%
39%
43%
48%
55%
12-44
-------
Table 12-85: Projected PHEV Penetrations, Alternative B - Cars
Manufacturer
2027
2028
2029
2030
203 1
2032
Aston Martin
4%
6%
8%
10%
11%
12%
BMW
5%
5%
6%
10%
9%
9%
Ferrari
20%
20%
20%
25%
22%
22%
Ford
11%
11%
12%
13%
14%
15%
General Motors
4%
5%
6%
8%
8%
12%
Honda
4%
5%
6%
9%
12%
13%
Hyundai
4%
5%
6%
8%
10%
13%
JLR
5%
7%
8%
9%
10%
13%
Kia
5%
6%
6%
7%
8%
12%
Lucid
0%
0%
0%
0%
0%
0%
Ma/.da
4%
5%
6%
11%
14%
16%
McLaren
8%
11%
12%
18%
12%
12%
Mercedes Ben/.
5%
7%
8%
9%
10%
10%
Mitsubishi
7%
8%
10%
11%
12%
13%
Nissan
5% ^
4%
6% 'J
5% '
6%
7%
Rivian
-
-
-
-
-
-
Stellantis
3%
4%
6%
6%
10%
7%
Subaru
5%
6%
7%
8%
8%
15%
Tcsla
0%
0%
0%
0%
0%
0%
Toyota
4%
5%
6%
7%
7%
9%
Volvo
7%
7%
7%
7%
7%
7%
VW
5%
7%
8%
9%
12%
10%
TOTAL
4%
5%
6%
7%
9%
10%
Table 12-86: Projected PHEV Penetrations, Alternative B - Trucks
Manufacturer
2027
2028
2029
2030
203 1
2032
Aston Martin
5%
7%
9%
15%
12%
8%
BMW
8" o
8% ]
10%
13%
13%
12%
Ferrari
-
-
-
-
-
-
Ford
6%
4%
6%
8%
8%
10%
General Motors
6%
4%
6%
7%
10%
11%
Honda
4%
6%
8%
11%
9%
10%
Hyundai
6%
7%
9%
12%
11%
11%
JLR
8%
13%
8%
10%
12%
11%
Kia
' .'1. 5%''''
3...^.6%
7%
9%
9%
10%
Lucid
-
-
-
-
-
-
Ma/.da
6% *"""""
*7% ^
8%
9% ^
¦:::io%;
11%
McLaren
-
-
-
-
-
-
Mercedes Ben/
6%
7%
9%
10%
10%
10%
Mitsubishi
3%
4%
6%
7%
8%
9%
Nissan
4%
4%
7%
7%
9%
10%
Rivian
0%
0%
0%
0%
0%
0%
Stellantis
8%
8%
10%
10%
11%
11%
Subaru
5%
7%
8%
9%
11%
12%
Tcsla
0%
0%
0%
0%
0%
0%
Toyota
7%
8%
8%
10%
11%
12%
Volvo
24%
24%
24%
24%
24%
24%
VW
6%
8%
9%
10%
11%
12%
TOTAL
6%
6%
8%
9%
10%
11%
12-45
-------
Table 12-87: Projected PHEV Penetrations, Alternative B - Combined
Manufacturer
2027
2028
2029
2030
203 1
2032
Aston Martin
4%
6%
8%
12%
11%
10%
BMW
6%
7%
8%
12%
11%
11%
Ferrari
20%
20%
20%
25%
22%
22%
Ford
6%
5%
7%
9%
9%
11%
General Motors
5%
4%
6%
7%
10%
11%
Honda
4%
5%
7%
10%
11%
11%
Hyundai
5%
6%
8%
10%
10%
12%
JLR
8%
13%
8%
9%
12%
11%
Kia
5%
6%
7%
8%
9%
11%
Lucid
0%
0%
0%
0%
0%
0%
Ma/.da
6%
7%
8%
9%
11%
12%
McLaren
8%
11%
12%
18%
12%
12%
Mercedes Ben/.
5%
7%
9%
9%
10%
10%
Mitsubishi
5%
6%
8%
9%
10%
11%
Nissan
4%
4%
6%
6%
7%
8%
Rivian
0%
0%
0%
0%
0%
0%
Stellantis
8%
7%
9%
10%
11%
11%
Subaru
5%
7%
8%
9%
10%
12%
Tcsla
0%
0%
0%
0%
0%
0%
Toyota
6%
7%
7%
8%
10%
10%
Volvo
20%
21%
21%
21%
21%
21%
VW
6%
8%
8%
9%
11%
11%
TOTAL
6%
6%
7%
9%
10%
11%
The tables below provide summary technology penetrations for Alternative B for strong
hybrids, TURB12 (only non-Miller engines) and MIL (Miller cycle engines). For these tables,
strong hybrids include all engine types, while the TURB12 and MIL penetrations shown are only
for non-hybrid versions of those vehicles.
Table 12-88: Projected Strong HEV Penetrations, Alternative B
2027 2028 2029 2030 2031 2032
Cars 3% 8% 7% 6% 6% 5%
Trucks 4% 3% 3% 6% 5% 4%
Total 4% 5% 4% 6% 5% 4%
Table 12-89: Projected TURB12 Penetrations, Alternative B
2027 2028 2029 2030 2031 2032
Cars 37% 31% 25% 21% 19% 16%
Trucks 52% 48% 41% 34% 29% 24%
Total 47% 43% 36% 30% 26% 21%
Table 12-90: Projected MIL Penetrations, Alternative B
2027 2028 2029 2030 2031 2032
Cars 5% 5% 5% 4% 4% 3%
Trucks 2% 1% 1% 1% 1% 1%
Total 3% 2% 2% 2% 2% 1%
12-46
-------
12.1.4 Light-Duty Vehicle Sensitivities
Light-duty sensitivities are described in Section IV.F of the preamble. This section provides
the analytical results for the final standards across the various sensitivities. In addition, we
conducted additional sensitivity analysis regarding the IRA tax credit assumptions in a memo to
the docket (U.S. EPA 2024).
12.1.4.1 State-level ZEV Policies (ACC II)
Table 12-91: Projected targets with ACC II
CO2 grams/mile
2027
2028
2029
2030
2031
2032
No Action
169
170
171
172
171
172
Final Standards
171
153
136
119
102
85
- cars and trucks combined
Table 12-92: Projected achieved levels with ACC II (CO2 grams/mile) - cars and trucks
combined"
2027
2028
2029
2030
2031
2032
No Action
145
129
116
104
91
83
Final Standards
152
136
126
114
100
92
aThe No Action achieved levels for the State-level ZEV Policies sensitivity are
lower due to greater maximum off-cycle and A/C credits available in the No
Action case. We discuss this phenomenon in Section IV.D.3 of the preamble.
Table 12-93: BEV penetrations with ACC II - cars and trucks combined
2027
2028
2029
2030
2031
2032
No Action
31%
34%
39%
42%
42%
45%
Final Standards
31%
36%
40%
47%
52%
56%
Table 12-94: PHEV penetrations with ACC II - cars and trucks combined
2027
2028
2029
2030
2031
2032
No Action
5%
6%
6%
8%
14%
14%
Final Standards
5%
6%
7%
6%
8%
8%
Table 12-95: Average incremental vehicle cost vs. No Action case with ACC II - cars and
trucks combined
2027
2028
2029
2030
2031
2032
6-yr avg
Final Standards
$143
$82
$95
$227
$969
$1,003
$420
12-47
-------
12.1.4.2 Battery Costs
12.1.4.2.1 Low Battery Costs
Table 12-96: Projected targets for Low Battery Costs (CO2 grams/mile) - cars and trucks
combined
2027
2028
2029
2030
2031
2032
No Action
170
171
172
172
172
172
Final Standards
171
154
136
119
102
85
Table 12-97: Projected achieved levels for Low Battery Costs (CO2 grams/mile) - cars and
trucks combined
2027
2028
2029
2030
2031
2032
No Action
131
111
101
101
100
103
Final Standards
140
119
113
111
96
82
Table 12-98: BEV penetrations for Low Battery Costs - cars and trucks combined
2027
2028
2029
2030
2031
2032
No Action
37%
41%
44%
42%
43%
41%
Final Standards
37%
44%
47%
48%
54%
59%
Table 12-99: PHEV penetrations for Low Battery Costs - cars and trucks combined
2027
2028
2029
2030
2031
2032
No Action
5%
6%
7%
8%
8%
9%
Final Standards
5%
6%
7%
8%
9%
11%
Table 12-100: Average incremental vehicle cost vs. No Action case for Low Battery Costs -
cars and trucks combined
2027
2028
2029
2030
2031
2032
6-yr avg
Final Standards
$106
-$12
-$72
$25
$653
$1,416
$353
12.1.4.2.2 High Battery Costs
Table 12-101: Projected targets for High Battery Costs (CO2 grams/mile) - cars and trucks
combined
2027
2028
2029
2030
2031
2032
No Action
168
168
169
169
170
170
Final Standards
170
154
136
120
102
85
Table 12-102: Projected achieved levels for High Battery Costs (CO2 grams/mile) - cars and
trucks combined
2027
2028
2029
2030
2031
2032
No Action
163
149
148
144
134
128
Final Standards
168
137
126
108
95
83
12-48
-------
Table 12-103: BEV penetrations for High Battery Costs - cars and trucks combined
2027
2028
2029
2030
2031
2032
No Action
19%
20%
21%
22%
25%
26%
Final Standards
20%
25%
30%
38%
46%
50%
Table 12-104: PHEV penetrations for High Battery Costs - cars and trucks combined
2027
2028
2029
2030
2031
2032
No Action
10%
9%
8%
9%
11%
13%
Final Standards
10%
12%
12%
13%
15%
18%
Table 12-105: Average incremental vehicle cost vs. No Action case for High Battery Costs -
cars and trucks combined
2027
2028
2029
2030
2031
2032
6-yr avg
Final Standards
$230
$1,562
$2,300
$3,335
$3,818
$4,187
$2,572
12.1.4.3 Consumer Acceptance
12.1.4.3.1 Faster BEV Acceptance
Table 12-106: Projected targets for Faster BEV Acceptance (CO2 grams/mile) - cars and
trucks combined
2027
2028
2029
2030
2031
2032
No Action
170
171
172
173
173
174
Final Standards
171
154
136
120
102
85
Table 12-107: Projected achieved levels for Faster BEV Acceptance (CO2 grams/mile) -
cars and trucks combined
2027
2028
2029
2030
2031
2032
No Action
133
108
94
86
75
67
Final Standards
140
114
103
99
91
78
Table 12-108: BEV penetrations for Faster BEV Acceptance - cars and trucks combined
2027
2028
2029
2030
2031
2032
No Action
37%
43%
48%
51%
54%
56%
Final Standards
37%
46%
52%
54%
58%
62%
Table 12-109: PHEV penetrations for Faster BEV Acceptance - cars and trucks combined
2027
2028
2029
2030
2031
2032
No Action
4%
5%
6%
6%
8%
9%
Final Standards
5%
5%
5%
6%
6%
9%
12-49
-------
Table 12-110: Average incremental vehicle cost vs. No Action case for Faster BEV
Acceptance - cars and trucks combined
2027
2028
2029
2030
2031
2032
6-yr avg
Final Standards
$138
$193
$181
$40
-$19
$274
$134
12.1.4.3.2 Slower BEV Acceptance
Table 12-111: Projected targets for Slower BEV Acceptance (CO2 grams/mile) - cars and
trucks combined
2027
2028
2029
2030
2031
2032
No Action
168
170
170
170
171
171
Final Standards
170
153
136
119
102
85
Table 12-112: Projected achieved levels for Slower BEV Acceptance (CO2 grams/mile) -
cars and trucks combined
2027
2028
2029
2030
2031
2032
No Action
161
151
145
141
129
125
Final Standards
162
136
122
107
98
81
Table 12-113: BEV penetrations for Slower BEV Acceptance - cars and trucks combined
2027
2028
2029
2030
2031
2032
No Action
20%
17%
20%
21%
25%
27%
Final Standards
21%
25%
32%
38%
44%
52%
Table 12-114: PHEV penetrations for Slower BEV Acceptance - cars and trucks combined
2027
2028
2029
2030
2031
2032
No Action
9%
9%
10%
10%
11%
12%
Final Standards
10%
11%
13%
14%
15%
17%
Table 12-115: Average incremental vehicle cost vs. No Action case for Slower BEV
Acceptance - cars and trucks combined
2027
2028
2029
2030
2031
2032
6-yr avg
Final Standards
$426
$1,074
$1,512
$2,158
$2,291
$2,887
$1,725
12.1.4.4 No Credit Trading Case
Table 12-116: Projected targets for No Credit Trading (CO2 grams/mile) - cars and trucks
combined
2027
2028
2029
2030
2031
2032
No Action
170
169
169
170
170
170
Final Standards
171
153
136
119
102
85
12-50
-------
Table 12-117: Projected achieved levels for No Credit Trading (CO2 grams/mile) - cars and
trucks combined
2027
2028
2029
2030
2031
2032
No Action
142
141
133
129
121
117
Final Standards
146
132
116
103
89
77
Table 12-118: BEV penetrations for No Credit Trading - cars and trucks combined
2027
2028
2029
2030
2031
2032
No Action
27%
28%
30%
31%
34%
35%
Final Standards
28%
33%
41%
46%
52%
56%
Table 12-119: PHEV penetrations for No Credit Trading - cars and trucks combined
2027
2028
2029
2030
2031
2032
No Action
6%
6%
7%
8%
9%
10%
Final Standards
6%
7%
8%
9%
11%
13%
Table 12-120: Average incremental vehicle cost vs. No Action case for No Credit Trading -
cars and trucks combined
2027
2028
2029
2030
2031
2032
6-yr avg
Final Standards
$268
$1,055
$1,420
$1,983
$2,365
$2,807
$1,650
12.1.4.5 Alternative Manufacturer Pathways
12.1.4.5.1 Lower BEV Production
Table 12-121: Projected targets for Lower BEV Production (CO2 grams/mile) - cars and
trucks combined
2027
2028
2029
2030
2031
2032
No Action
168
169
169
170
171
171
Final Standards
170
153
136
119
102
85
Table 12-122: Projected achieved levels for Lower BEV Production (CO2 grams/mile) -
cars and trucks combined
2027
2028
2029
2030
2031
2032
No Action
160
153
142
137
128
118
Final Standards
160
146
133
117
102
88
Table 12-123: BEV penetrations for Lower BEV Production - cars and trucks combined
2027
2028
2029
2030
2031
2032
No Action
26%
27%
30%
31%
34%
35%
Final Standards
24%
29%
33%
37%
41%
43%
Table 12-124: PHEV penetrations for Lower BEV Production - cars and trucks combined
2027
2028
2029
2030
2031
2032
No Action
5%
6%
7%
8%
8%
12%
Final Standards
10%
12%
15%
18%
24%
29%
12-51
-------
Table 12-125: Average incremental vehicle cost vs. No Action case for Lower BEV
Production - cars and trucks combined
2027
2028
2029
2030
2031
2032
6-yr avg
Final Standards
$449
$788
$980
$1,639
$2,303
$2,575
$1,456
12.1.4.5.2 No Additional BEVs Beyond the No Action Case
Table 12-126: Projected targets for No Additional BEVs Beyond the No Action Case (CO2
grams/mile) - cars and trucks combined
2027
2028
2029
2030
2031
2032
No Action
168
169
169
170
171
171
Final Standards
170
155
137
121
103
86
Table 12-127: Projected achieved levels for No Additional BEVs Beyond the No Action
Case (CO2 grams/mile) - cars and trucks combined
2027
2028
2029
2030
2031
2032
No Action
160
153
142
137
128
118
Final Standards
159
124
112
100
95
90
Table 12-128: BEV penetrations for No Additional BEVs Beyond the No Action Case - cars
and trucks combined
2027
2028
2029
2030
2031
2032
No Action
26%
27%
30%
31%
34%
35%
Final Standards
24%
26%
30%
31%
34%
35%
Table 12-129: PHEV penetrations for No Additional BEVs Beyond the No Action Case -
cars and trucks combined
2027
2028
2029
2030
2031
2032
No Action
5%
6%
7%
8%
8%
12%
Final Standards
10%
17%
22%
27%
32%
36%
Table 12-130: Average incremental vehicle cost vs. No Action case for No Additional BEVs
Beyond the No Action Case - cars and trucks combined
2027
2028
2029
2030
2031
2032
6-yr avg
Final Standards
$536
$2,517
$2,630
$3,120
$3,334
$3,112
$2,542
12-52
-------
12.2 Medium-Duty Vehicles
12.2.1 GHG Targets and Compliance Levels
12.2.1.1 COi g/mi
Shown below are the projected average GHG targets for each manufacturer, as well as their
corresponding average achieved compliance, in g/mi, for vans and pickups.
12.2.1.1.1 Final Standards
OEM-specific GHG emissions targets for the final standards are shown in Table 12-131, Table
12-132, and Table 12-133 for vans, pickups, and the combined fleet, respectively. Similarly,
projected achieved GHG emissions levels are given in Table 12-134 through Table 12-136.
Table 12-131: Projected GHG Targets, Final Standards - Medium Duty Vans
Manufacturer
2027
2028
2029
2030
203 1
2032
Ford
384
383
348
311
275
240
General Motors
392
391
355
316
280
244
Mercedes Ben/.
426
424
386
345
306
267
Nissan
391
390
354
316
280
244
Slcllanlis
393
392
356
318
281
245
TOTAL
392
391
355
317
281
245
Table 12-132: Projected GHG
Targets, Final Standards - Medium Duty Pickups
Manufacturer
2027
2028
2029
2030
203 1
2032
Ford
486
472
426
368
329
288
General Motors
506
496
446
373'""
332 ""I
291
Mercedes Ben/.
-
-
-
-
-
-
Nissan
423
421
383
342
305
267
Slcllanlis
501
491
439
373 ;
'"332"'
291
TOTAL
497
486
437
371
331
290
Table
12-133: Projected GHG Targets, Final Standards -
Combined
Manufacturer
2027
2028
2029
2030
203 1
2032
Ford
446
438
396
346
309
270
General Motors
475
468
422
358
318
278
Mercedes Ben/.
426
424
386
345
306
267
Nissan
393
392
356
318
282
246
Slcllanlis
477
469
421
360
321
281
TOTAL
461
453
408
353
314
274
12-53
-------
Table 12-134: Achieved GHG Levels, Final Standards - Medium Duty Vans
Manufacturer
2027
2028
2029
2030
203 1
2032
Ford
424
418
330
243
142
99
General Motors
480
476
377
278
167
110
Mercedes Ben/.
375
371
294
207
133
92
Nissan
427
423
338
250
162
110
Slcllanlis
431
426
338 |
249
164
107
TOTAL
434
429
340
249
151
103
Table 12-135: Achieved GHG Levels, Final Standards -
Medium Duty Pickups
Manufacturer
2027
2028
2029
2030
203 1
2032
Ford
474
469
449
413
408
370
General Motors
469
465
444
403
386
353
Mercedes Ben/.
-
-
-
-
-
-
Nissan
396
391
382
344
349
360
Slcllanlis
456
451
433
395
394
360
TOTAL
468
463
443
405
396
361
Table 12-136:
Achieved GHG Levels, Final
Standards - Combined
Manufacturer
2027
2028
2029
2030
203 1
2032
Ford
455
450
404
348
307
267
General Motors
472
468
426
370
328
288
Mercedes Ben/.
375
371
294
207
133
92
Nissan
425
421
341
256
175
126
Slcllanlis
450
446
411
363
343
304
TOTAL
456
451
407
351
312
272
12.2.1.2 CO2 Mg
Shown below are the projected average GHG targets for each manufacturer, as well as their
corresponding average achieved compliance, in Mg, for cars, trucks, and the combined fleet.
Total emissions are calculated by multiplying the relevant CO2 emission rate, the production
volume of applicable vehicles, and the expected lifetime vehicle miles traveled (VMT) of those
vehicles. The equation to calculate total Mg (for either total emissions, or credits based on the
difference between target g/mi and achieved g/mi) is:
CO2 (Mg) = (CO2 (g/mi) x VMT x Production) / 1,000,000
In the above equation, "VMT" is in miles, and specified in the regulations as 150,000 miles.
When using these equations to calculate values for cars and trucks in aggregate, we use a
production weighted average of the car and truck VMT values.
12.2.1.2.1 Final Standards
OEM-specific GHG emissions targets for the final standards (in Mg) are shown in Table
12-137, Table 12-138, and Table 12-139 for vans, pickups, and the combined fleet, respectively.
Similarly, projected achieved GHG emissions (in Mg) are given for vans, pickups, and the
combined fleet in Table 12-140, Table 12-141, and Table 12-142. Finally, overall credits or
12-54
-------
debits earned are provided for the combined fleet on a manufacturer-specific basis, in Table
12-143.
Table 12-137: Projected GHG Targets (Mg), Final Standards - Medium Duty Vans
Manufacturer 2027 2028 2029 2030 2031 2032
Ford 6.527.530 6.510.205 5.927.714 5.320.365 4.753.308 4.189.625
General Motors 3.929.040 3.916.530 3.559.281 3.188.745 2.843.691 2.503.748
Mercedes Ben/. 1.862.709 1.853.565 1.688.648 1.516.903 1.356.793 1.196.749
Nissan 686.628 684.519 622.989 558.819 498.961 439.500
Slcllanlis 1.884.366 1.878.326 1.708.438 1.531.874 1.367.533 1.204.396
TOTAL 14,890,274 14,843,145 13,507,071 12,116,706 10,820,286 9,534,018
Table 12-138: Projected GHG Targets (Mg), Final Standards - Medium Duty Pickups
Manufacturer 2027 2028 2029 2030 2031 2032
Ford 13.210.563 12.910.330 11.705.812 10.201.534 9.239.318 8.190.773
General Motors 13.729.000 13.532.118 12.229.127 10.310.655 9.305.939 8.252.083
Mercedes Ben/. ......
Nissan 51.135 51.104 46.761 42.112 37.990 33.687
Slcllanlis 8.313.311 8.177.286 7.355.817 6.292.821 5.680.742 5.037.485
TOTAL 35,304,009 34,670,837 31,337,517 26,847,122 24,263,989 21,514,029
Table 12-139: Projected GHG Targets (Mg), Final Standards - Medium Duty Combined
Manufacturer 2027 2028 2029 2030 2031 2032
Ford 19.738.093 19.420.535 17.633.527 15.521.899 13.992.625 12.380.399
General Motors 17.658.040 17.448.647 15.788.408 13.499.400 12.149.630 10.755.832
Mercedes Ben/. 1.862.709 1.853.565 1.688.648 1.516.903 1.356.793 1.196.749
Nissan 737.763 735.623 669.750 600.931 536.951 473.187
Slcllanlis 10.197.677 10.055.612 9.064.255 7.824.694 7.048.275 6.241.881
TOTAL 50,194,283 49,513,982 44,844,587 38,963,827 35,084,275 31,048,047
Table 12-140: Achieved GHG Levels (Mg), Final Standards - Medium Duty Vans
Manufacturer 2027 2028 2029 2030 2031 2032
Ford 7.215.157 7.115.886 5.634.919 4.159.262 2.443.268 1.732.150
General Motors 4.808.881 4.762.100 3.780.310 2.799.449 1.699.714 1.129.733
Mercedes Ben/. 1.636.722 1.619.659 1.285.961 908.440 589.947 411.979
Nissan 749.064 741.627 594.912 442.075 288.983 197.713
Slcllanlis 2.064.545 2.043.632 1.621.964 1.200.857 796.160 526.696
TOTAL 16,474,368 16,282,904 12,918,066 9,510,083 5,818,073 3,998,272
Table 12-141: Achieved GHG Levels (Mg), Final Standards - Medium Duty Pickups
Manufacturer 2027 2028 2029 2030 2031 2032
Ford 12.882.609 12.827.681 12.332.351 11.439.735 11.459.668 10.526.250
General Motors 12.741.485 12.679.250 12.176.344 11.151.079 10.803.991 10.014.127
Mercedes Ben/. ......
Nissan 47.832 47.550 46.634 42.272 43.570 45.491
Slcllanlis 7.560.681 7.520.698 7.245.604 6.675.645 6.747.273 6.236.556
TOTAL 33,232,608 33,075,180 31,800,933 29,308,730 29,054,502 26,822,424
12-55
-------
Table 12-142: Achieved GHG Levels (Mg), Final Standards - Medium Duty Combined
Manufacturer 2027 2028 2029 2030 2031 2032
Ford 20.097.766 19.943.567 17.967.270 15.598.997 13.902.936 12.258.400
General Motors 17.550.366 17.441.351 15.956.654 13.950.527 12.503.705 11.143.860
Mercedes Ben/. 1.636.722 1.619.659 1.285.961 908.440 589.947 411.979
Nissan 796.896 789.177 641.546 484.347 332.553 243.204
Slcllanlis 9.625.226 9.564.330 8.867.567 7.876.502 7.543.433 6.763.252
TOTAL 49,706,976 49,358,084 44,718,999 38,818,813 34,872,574 30,820,696
Table 12-143: GHG Credits/Debits Earned (Mg), Final Standards - Medium Duty
Combined
Manufacturer 2027 2028 2029 2030 2031 2032
Ford (359.673) (523.032) (333.743) (77.098) 89.689 121.999
General Motors 107.674 7.297 (168.246) (451.128) (354.075) (388.029)
Mercedes Ben/. 225.987 233.906 402.686 608.463 766.846 784.770
Nissan (59.132) (53.555) 28.204 116.584 204.398 229.983
Slcllanlis 572.451 491.282 196.688 (51.807) (495.157) (521.372)
TOTAL 487,307 155,898 125,588 145,014 211,701 227,351
12.2.2 Projected Manufacturing Costs per Vehicle
EPA has performed an assessment of the estimated per-vehicle costs for manufacturers to
meet the MY 2027-2032 MDV GHG standards, relative to the No Action case. The fleet average
costs per vehicle are grouped by vans, MD pickups, and the fleet total. We have included
summary tables in this format. The costs in this section represent compliance costs to the
industry and are not necessarily the same as the costs experienced by the consumer when
purchasing a new vehicle. For example, the costs presented here do not include any state and
Federal purchase incentives that are available to consumers. Also, the manufacturer decisions for
the pricing of individual vehicles may not align exactly with the production cost impacts for that
particular vehicle. EPA's OMEGA model assumes that manufacturers distribute compliance
costs through limited cross-subsidization of prices between vehicles in order to maintain an
appropriate mix of debit- and credit-generating vehicles that achieves compliance in a cost-
minimizing fashion.
12.2.2.1 Final Standards
Incremental costs per vehicle for the final standards (compared to the No Action case) are
summarized by van and truck in Table 12-144.
Table 12-144: Projected Manufacturing Costs Per Vehicle, Final Standards - Medium Duty
Vehicles
2027
2028
2029
2030
203 1
2032
6-vcar avg
Vans
$178
$185
$1,443
$2,732
$4,128
$4,915
$2,264 "
Pickups
$97
$88
$53 1
$1,432
$1,516
$2,416
$1,013
Total
SI 25
SI 22
S847
SI,881
S2,416
$3,275
SI,444
12-56
-------
Incremental costs per vehicle for the final standards, compared to the No Action case, are
shown for each OEM in Table 12-145, Table 12-146, and Table 12-147 for vans, pickups, and
the medium duty combined fleet, respectively.
Table 12-145: Projected Manufacturing Costs Per Vehicle, Final Standards - Medium Duty
Vans
Manufacturer
2027
2028
2029
2030
203 1
2032
Ford
$212
$231
$1,594
$2,955
$4.581
$5,293
General Motors
$197
$194
$1,284
f $2,377
$3,639
$4,478
Mercedes Ben/.
$11
$10
$1,270
$2,771
$3,904
$4,649
Nissan
$179
$177
$1,312
$2,496
$3,641
$4,423
Slcllanlis
$167
$164
$1,447
$2,732
$3,925
$4,912
TOTAL
SI 78
SI 85
SI,443
$2,732
$4,128
$4,915
Table 12-146: Projected Manufacturing Costs Per Vehicle, Final Standards - Medium Duty
Pickups
Manufacturer
2027
2028
2029
2030
203 1
2032
Ford
$104
$84
$546
$1,410
$1,554
$2,536
General Motors
' $97 ^
$96
$539
$1,484
$1,587
$2,430
Mercedes Ben/.
-
-
-
-
-
-
Nissan
$781
$791
$954
$2,049
$1,761
$1,260
Slcllanlis
$78
$77
$493
$1,378
$1,337
$2,203
TOTAL
$97
$88
$531
$1,432
$1,516
$2,416
Table 12-147: Projected Manufacturing Costs Per Vehicle, Final Standards - Medium Duty
Combined
Manufacturer
2027
2028
2029
2030
203 1
2032
Ford
$146
$140
$947
$2,000
$2,707
$3,584
General Motors
$124
$122
$738
: $1,723
$2,134
$2,974
Mercedes Ben/.
$11
$10
$1,270
$2,771
$3,904
$4,649
Nissan
$217
$217
$1,289
$2,466
$3,518
$4,215
Slcllanlis
$98
$97
$705
$1,679
$1,910
$2,801
TOTAL
$125
$122
$847
$1,881
$2,416
$3,275
12.2.3 Technology Penetration Rates
Presented below are the projected technology penetration rates, by manufacturer, for vans and
pickups, for the No Action case and the Final Standards Tables are summarized by body style for
BEV penetrations, with the remainder of the fleet being ICE vehicles.
12.2.3.1 No Action Case
Table 12-148 through Table 12-150 give BEV penetrations for the No Action case, by
manufacturer. Similarly, Table 12-151 through Table 12-153 provide PHEV penetrations (no
PHEVs were projected for the No Action case).
12-57
-------
Table 12-148: Projected BEV Penetrations, No Action - Medium Duty Vans
Manufacturer
2027
2028
2029
2030
203 1
2032
Ford
3%
4%
5% "
6%
7%
8%
General Motors
3%
4%
5%
6%
7%
8%
Mercedes Ben/.
3% 1
4%
5%
6%
7%
8%
Nissan
3% r
4%
5%
6%
: 7%
8%
Slcllanlis
3%
4%
5%
6%
7%
8%
TOTAL
3%
4%
5%
6%
7%
8%
Table 12-149: Projected BEV Penetrations,
No Action -
Medium Duty Pickups
Manufacturer
2027
2028
2029
2030
203 1
2032
Ford
3"o *
4%
5%
6%
7%
8%
General Motors
10/
J /o
4%
5%
6%
rzi 7%
8%
Mercedes Ben/.
-
-
-
-
-
-
Nissan
10/
J /o
4%
5%
6%
7%
8%
Slcllanlis
3% i
4%
5%
6%
7%
8%
TOTAL
3%
4%
5%
6%
7%
8%
Table 12-150:
Projected BEV Penetrations, No Action - Medium Duty Combined
Manufacturer
2027
2028
2029
2030
203 1
2032
Ford
3%
4%
5%
6%
7%
8%
General Motors
3%
4%
5%
6%
7%
8%
Mercedes Ben/.
3%
4%
5%
6%
7%
8%
Nissan
3%
4%
5%
6%
7%
8%
Slcllanlis
3% i
4%
5%
6%
7%
8%
TOTAL
3%
4%
5%
6%
7%
8%
Table 12-151: Projected
PHEV Penetrations, No Action - Medium Duty Vans
Manufacturer
2027
2028
2029
2030
203 1
2032
Ford
0%
0%
0% "
0%
0%
0%
General Motors
0%
0%
0%
Mercedes Ben/.
0%
AO/
U /o
0%
0%
AO/
U /o
AO/
U /o
Nissan
0%
AO/
U /o
0%
0%
AO/
U /o
AO/
U /o
Slcllanlis
0%
0%
0%
0%
0%
0%
TOTAL
0%
0%
0%
0%
0%
0%
Table 12-152
Projected PHEV
Penetrations
No Action
- Medium Duty Pickups
Manufacturer
2027
2028
2029
2030
203 1
2032
Ford
0%
0%
0% "
0%
0%
0%
General Motors
0% 2
3 0%
Mercedes Ben/.
-
-
-
-
-
-
Nissan
0%
0%
0%
: o%
; o%
0%
Slcllanlis
0%
0%
0%
0%
0%
0%
TOTAL
0%
0%
0%
0%
0%
0%
12-58
-------
Table 12-153: Projected PHEV Penetrations, No Action - Medium Duty Combined
Manufacturer
2027
2028
2029
2030
203 1
2032
Ford
0%
0%
0%
0%
0%
0%
General Motors
0%
0%
0%
0%
0%
Mercedes Ben/.
0%
AO/
U /o
0%
0%
0%
0%
Nissan
0%
AO/
U /o
0%
0%
0%
0%
Slcllanlis
0%
0%
0%
0%
0%
0%
TOTAL
0%
0%
0%
0%
0%
0%
12.2.3.2 Final Standards
Table 12-154 through Table 12-156 give BEV penetrations for the final standards, by
manufacturer. Similarly, Table 12-157 through Table 12-159 provide PHEV penetrations.
Table 12-154: Projected BEV Penetrations, Final Standards - Medium Duty Vans
Manufacturer
2027
2028
2029
2030
203 1
2032
Ford
3%
5%
25%
45%
65%
75%
General Motors
3%
4%
24%
44%
64%
75%
Mercedes Ben/.
3%
4%
24%
44%
64%
75%
Nissan
3%
4%
23%
43%
63%
75%
Slcllanlis
3%
4%
24%
44%
63%
75%
TOTAL
3%
4%
24%
44%
64%
75%
Table 12-155: Projected BEV Penetrations, Final Standards - Medium Duty Pickups
Manufacturer 2027 2028 2029 2030 2031 2032
Ford 3% 4% 8% 10% 10% 10%
General Motors 3% 4% 8% 10% 10% 10%
Mercedes Ben/. ......
Nissan 2% 3% 6% 10% 10% 10%
Slcllanlis 3% 4% 8% 10% 9% 9%
TOTAL 3% 4% 8% 10% 10% 10%
Table 12-156: Projected BEV Penetrations, Final Standards - Medium Duty Combined
Manufacturer
2027
2028
2029
2030
203 1
2032
Ford
3%
4%
14%
23%
31%
35%
General Motors
3%
4%
12%
19%
24%
27%
Mercedes Ben/.
3% r
4%
24%
44%
64%
75%
Nissan
3% r
4%
22%
41%
60%
71%
Slcllanlis
3%
4%
11%
17%
21%
23%
TOTAL
3%
4%
13%
22%
28%
32%
12-59
-------
Table 12-157: Projected PHEV Penetrations, Final Standards - Medium Duty Vans
Manufacturer
2027
2028
2029
2030
203 1
2032
Ford
0%
0%
0%
0%
0%
0%
General Motors
0%
3 o%Z
0%
2%
Mercedes Ben/.
0%
AO/
U /o
AO/
U /o
AO/
U /o
0%
0%
Nissan
0%
AO/
U /o
AO/
U /o
AO/
U /o
0%
0%
Slcllanlis
0%
0%
0%
0%
0%
1%
TOTAL
0%
0%
0%
0%
0%
1%
Table 12-158: Projected PHEV Penetrations, Final Standards - Medium Duty Pickups
Manufacturer 2027 2028 2029 2030 2031 2032
Ford 0% 0% 0% 7% 0% 12%
General Motors 0% 0% 0% 9% 7% 18%
Mercedes Ben/. ......
Nissan 0% 0% 0% 6% 5% 2%
Slcllanlis 0% 0% 0% 7% 10% 20%
TOTAL 0% 0% 0% 8% 5% 16%
Table 12-159: Projected PHEV Penetrations, Final Standards - Medium Duty Combined
Manufacturer
2027
2028
2029
2030
203 1
2032
Ford
0%
0%
0%
4%
0%
8%
General Motors
0%
00,. 7^
6%
5%
13%
Mercedes Ben/.
0%
AO/
U /o
AO/
U /o
0%
0%
0%
Nissan
0%
AO/
U /o
AO/
U /o
0%
0%
0%
Slcllanlis
0%
0%
0%
6%
8%
16%
TOTAL
0%
0%
0%
5%
3%
11%
12.2.4 Medium-Duty Vehicle Sensitivities
The tables below summarize the projected average GHG targets and average achieved
compliance, in g/mi, BEV penetrations, and incremental vehicle cost vs the No Action case, for
medium duty vehicles. They are prepared for the Low Battery Cost, High Battery Cost, and No
Trading sensitivities.
12.2.4.1 Battery Costs
12.2.4.1.1 Low Battery Costs
Table 12-160. Projected targets with Low Battery Costs for No Action case and final
standards (CO2 grams/mile) - Medium Duty Combined
2027
2028
2029
2030
2031
2032
No Action
475
475
474
474
474
474
Final Standards
461
453
408
353
314
274
12-60
-------
Table 12-161. Projected achieved levels with Low Battery Costs for No Action case and
final standards (CO2 grams/mile) - Medium Duty Combined
2027
2028
2029
2030
2031
2032
No Action
456
452
447
443
438
434
Final Standards
456
452
407
351
311
272
Table 12-162. BEV penetrations with Low Battery Costs for No Action case and final
standards - Medium Duty Combined
2027
2028
2029
2030
2031
2032
No Action
3%
4%
5%
6%
7%
8%
Final Standards
3%
4%
13%
22%
28%
32%
Table 12-163. PHEV penetrations with Low Battery Costs for No Action case and final
standards - Medium Duty Combined
2027
2028
2029
2030
2031
2032
No Action
0%
0%
0%
0%
0%
0%
Final Standards
0%
0%
0%
5%
5%
12%
Table 12-164. Average incremental vehicle manufacturing cost vs. No Action case for Low
Battery Costs, final standards - Medium Duty Combined
2027
2028
2029
2030
2031
2032
6-yr avg
Final Standards
$125
$122
$553
$1,356
$1,863
$2,696
$1,119
12.2.4.1.2 High Battery Costs
Table 12-165. Projected targets with High Battery Costs for No Action case and final
standards (CO2 grams/mile) - Medium Duty Combined
2027
2028
2029
2030
2031
2032
No Action
475
475
474
474
474
474
Final Standards
461
453
409
353
315
275
Table 12-166. Projected achieved levels with High Battery Costs for No Action case and
final standards (CO2 grams/mile) - Medium Duty Combined
2027
2028
2029
2030
2031
2032
No Action
456
452
447
443
438
434
Final Standards
456
451
408
352
314
273
Table 12-167. BEV penetrations with High Battery Costs for No Action case and final
standards - Medium Duty Combined
2027
2028
2029
2030
2031
2032
No Action
3%
4%
5%
6%
7%
8%
Final Standards
3%
4%
10%
18%
26%
31%
12-61
-------
Table 12-168. PHEV penetrations with High Battery Costs for No Action case and final
standards - Medium Duty Combined
2027
2028
2029
2030
2031
2032
No Action
0%
0%
0%
0%
0%
0%
Final Standards
0%
0%
4%
9%
6%
11%
Table 12-169. Average incremental vehicle manufacturing cost vs. No Action case for High
Battery Costs, final standards - Medium Duty Combined
2027
2028
2029
2030
2031
2032
6-yr avg
Final Standards
$125
$121
$1,120
$2,493
$3,247
$4,206
$1,885
12.2.4.2 No Credit Trading Case
Table 12-170: Projected targets for No Credit Trading (CO2 grams/mile) - Medium Duty
Combined
2027
2028
2029
2030
2031
2032
No Action
473
473
473
473
474
473
Final Standards
460
452
408
352
313
274
Table 12-171: Projected achieved levels for No Credit Trading (CO2 grams/mile) - Medium
Duty Combined
2027
2028
2029
2030
2031
2032
No Action
426
425
424
423
422
420
Final Standards
413
406
366
317
282
247
Table 12-172: BEV penetrations for No Credit Trading - Medium Duty Combined
2027
2028
2029
2030
2031
2032
No Action
8%
8%
8%
8%
8%
9%
Final Standards
10%
11%
20%
27%
29%
30%
Table 12-173: PHEV penetrations for No Credit Trading - Medium Duty Combined
2027
2028
2029
2030
2031
2032
No Action
0%
0%
0%
0%
0%
0%
Final Standards
0%
0%
0%
5%
11%
20%
Table 12-174: Average incremental vehicle cost vs. No Action case for No Credit Trading -
Medium Duty Combined
2027
2028
2029
2030
2031
2032
6-yr avg
Final Standards
$326
$412
$1,086
$2,072
$2,846
$3,806
$1,758
12-62
-------
12.3 Additional Illustrative Scenarios
12.3.1 No New BEVs Above Base Year MY 2022 Fleet - Light-Duty Vehicles
Table 12-175: Projected targets for No New BEVs Above Base Year MY 2022 Fleet (CO2
grams/mile) - cars and trucks combined
2027
2028
2029
2030
2031
2032
No Action-No New BEVs
167
167
166
168
167
167
Final Standards-No New BEVs
169
152
134
118
101
84
Table 12-176: Projected achieved levels for No New BEVs Above Base Year MY 2022 Fleet
(CO2 grams/mile) - cars and trucks combined
2027
2028
2029
2030
2031
2032
No Action-No New BEVs
165
165
164
166
164
165
Final Standards-No New BEVs
167
150
133
117
102
84
Table 12-177: BEV penetrations for No New BEVs Above Base Year MY 2022 Fleet - cars
and trucks combined
2027
2028
2029
2030
2031
2032
No Action-No New BEVs
5%
5%
5%
5%
5%
5%
Final Standards-No New BEVs
5%
5%
5%
5%
5%
5%
Table 12-178: PHEV penetrations for No New BEVs Above Base Year MY 2022 Fleet -
cars and trucks combined
2027
2028
2029
2030
2031
2032
No Action-No New BEVs
9%
8%
9%
7%
7%
7%
Final Standards-No New BEVs
10%
19%
31%
43%
69%
86%
Table 12-179: Average incremental vehicle cost vs. No Action case for No New BEVs Above
Base Year MY 2022 Fleet, final - cars and trucks combined
2027
2028
2029
2030
2031
2032
6-yr avg
Final Standards
$205
$1,538
$2,536
$3,019
$4,722
$5,459
$2,913
12.3.2 No New BEVs Above Base Year MY 2022 Fleet - Medium-Duty Vehicles
Table 12-180: Projected targets for No New BEVs Above Base Year MY 2022 Fleet (CO2
grams/mile) - Medium Duty Combined.
2027
2028
2029
2030
2031
2032
No Action-No New BEVs
477
477
477
478
478
478
Final Standards-No New BEVs
461
454
411
355
318
278
12-63
-------
Table 12-181: Projected achieved levels for No New BEVs Above Base Year MY 2022 Fleet
(CO2 grams/mile) - Medium Duty Combined.
2027
2028
2029
2030
2031
2032
No Action-No New BEVs
459
455
452
448
445
441
Final Standards-No New BEVs
459
454
411
356
317
279
Table 12-182: BEV penetrations for No New BEVs Above Base Year MY 2022 Fleet -
Medium Duty Combined.
2027
2028
2029
2030
2031
2032
No Action-No New BEVs
0%
0%
0%
0%
0%
0%
Final Standards-No New BEVs
0%
0%
0%
0%
0%
0%
Table 12-183: PHEV penetrations for No New BEVs Above Base Year MY 2022 Fleet -
Medium Duty Combined.
2027
2028
2029
2030
2031
2032
No Action-No New BEVs
3%
4%
5%
6%
7%
8%
Final Standards-No New BEVs
3%
4%
16%
30%
39%
51%
Table 12-184: Average incremental vehicle cost vs. No Action case for No New BEVs Above
Base Year MY 2022 Fleet - Medium Duty Combined.
2027
2028
2029
2030
2031
2032
6-yr avg
Final Standards
$129
$181
$1,284
$2,850
$4,189
$5,360
$2,332
12-64
-------
Chapter 12 References
U.S. EPA. 2024. Sensitivity Analysis of IRA Tax Credit Assumptions, Memorandum to Docket
EPA-HQ-OAR-2022-0829, March 13, 2024.
12-65
-------