EMFAC2021

Volume III
Technical
Document
Version 1.0.1

April, 2021

Mobile Source Analysis Branch

Air Quality Planning and Science Division
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Table of Contents
1 Executive Summary ............................................................................................. 4
1.1 Structure of This Document ................................................................................ 4
1.2 New Features ...................................................................................................... 5
1.3 Overview of Major Changes ............................................................................... 6
1.3.1 Fleet Characterization ................................................................................... 6
1.3.2 In-Use Emissions ............................................................................................ 7
1.3.3 Activity ......................................................................................................... 10
1.3.4 New Sales and VMT Forecasting ................................................................. 11
1.3.5 Regulations and Policies .............................................................................. 12
2 Introduction ...................................................................................................... 14
2.1 EMFAC2021 ...................................................................................................... 14
2.2 Modeling Architecture ...................................................................................... 15
2.3 Web Interface ................................................................................................... 16
3 New Modules in EMFAC2021........................................................................... 19
3.1 PHEV Module .................................................................................................... 19
3.1.1 Background ................................................................................................. 19
3.1.2 Emission Rates Analysis ............................................................................... 20
3.1.3 Activity Analysis ........................................................................................... 24
3.2 Energy Module ................................................................................................. 28
3.2.1 Light-Duty .................................................................................................... 29
3.2.2 Heavy-Duty .................................................................................................. 31
3.3 Ammonia Module ............................................................................................. 34
3.3.1 Background ................................................................................................. 34
3.3.2 Data Sources ............................................................................................... 35
3.3.3 Results and Conclusions .............................................................................. 39
4 Methodology Updates ...................................................................................... 40
4.1 Introduction ...................................................................................................... 40
4.2 Fleet Characteristics.......................................................................................... 40

1
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

4.2.1 Light Duty Fleet Characterization ................................................................ 40

4.2.2 Heavy Duty Fleet Characterization .............................................................. 48
4.2.3 Transit Bus Population ................................................................................. 68
4.2.4 Natural Gas Population ............................................................................... 72
4.3 Emission Rates .................................................................................................. 74
4.3.1 Light Duty Emission Rates ........................................................................... 74
4.3.2 Evaporative Emissions Module .................................................................... 90
4.3.3 Light Duty CO2 Emission Rates ................................................................... 95
4.3.4 Motorcycle Emission Rates ........................................................................ 102
4.3.5 HD Emission Rates..................................................................................... 110
4.3.6 Heavy Duty NOx Deterioration Rates ....................................................... 138
4.3.7 Break Wear Emissions ............................................................................... 143
4.4 Activity Profiles ............................................................................................... 153
4.4.1 Light Duty Activity Profiles ........................................................................ 153
4.4.2 Heavy Duty Activity Profiles ...................................................................... 155
4.5 Forecasting ..................................................................................................... 170
4.5.1 Light Duty New Sales and VMT Forecasting ............................................. 170
4.5.2 Heavy Duty New Sales and VMT Forecasting ........................................... 174
4.5.3 Zero Emission Vehicle Forecasting ............................................................ 181
4.6 Regulations and Policies ................................................................................. 191
4.6.1 SAFE Rule .................................................................................................. 191
4.6.2 Innovative Clean Transit ............................................................................ 192
4.6.3 Zero-Emission Airport Shuttle Bus............................................................. 192
4.6.4 HD Omnibus .............................................................................................. 192
4.6.5 Warranty .................................................................................................... 194
4.6.6 Advanced Clean Trucks ............................................................................. 196
4.6.7 Opacity ...................................................................................................... 196
5 Overall Impacts ............................................................................................... 198
5.1 Vehicle Population .......................................................................................... 198

2
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

5.2 Vehicle Miles Traveled (VMT) .......................................................................... 199

5.3 Emissions ........................................................................................................ 200
6 Appendices ..................................................................................................... 209
6.1 Vehicle Class Categorization .......................................................................... 210
6.2 Exhaust Technology Groups ........................................................................... 216
6.3 Emissions Impact in Air Basins ........................................................................ 224
6.4 Heavy Duty Natural Gas Penetration .............................................................. 230
6.5 Dynamometer Test Data of Light Heavy-Duty Trucks from CARB Project
2R1702 ............................................................................................................ 239
6.6 Dynamometer Test Data of Heavy Heavy-Duty Diesel Trucks from CARB Truck
and Bus Surveillance Program ........................................................................ 242
6.7 Dynamometer Test Data of Medium Heavy-Duty Diesel Trucks from CARB
Surveillance Program for On-Road Class 4-6 Heavy-Duty Vehicles ................ 250
6.8 Mapping of CEC Vehicle Classes to EMFAC Vehicle Classes ......................... 252
6.9 Heavy Duty VMT Distribution by Hour Figures ............................................... 254
6.10 Heavy Duty VMT Distribution by Speed Figures ............................................ 258
6.11 Engine Starts Distribution by Hour Figure ...................................................... 263
6.12 Engine Starts Distribution by Soak Time Figure ............................................. 267

3
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

1 Executive Summary
The California Air Resources Board (CARB) officially released EMFAC2021 (v1.0.0)
January 15, 2021. Used to inform and support and upcoming planning and policy
development, EMFAC2021 is the latest emission inventory model that CARB developed
to assess emissions from on-road motor vehicles including cars, trucks, and buses in
California. This version of model reflects CARB’s latest understanding of statewide and
regional vehicle activities, emissions, and recently adopted regulations such as Advanced
Clean Trucks (ACT) and Heavy Duty Omnibus.
Staff have also released an interactive web platform that allows EMFAC2021 users to
download emissions data with default activity; produce emissions using custom activity
inputs, the same as EMFAC’s Custom Activity (SG 1) mode; and generate emission factors
using user-defined ambient temperature and relative humidity for project level (PL)
conformity assessment.
This Technical Document provides technical details of major changes and updates in
EMFAC2021 and provides an overview of differences between EMFAC2021 and the
prior version of the model, EMFAC2017. For more information on how to use
EMFAC2021, including how to install and navigate through its user interface, please refer
to the EMFAC2021 User’s Guide. 2 Please note that some of the legacy components,
methodologies, data, and logic are carried over into EMFAC2021 from prior versions of
the model and are not covered within this document.

1.1 Structure of This Document

The Introduction chapter (Chapter 2) provides an overview of EMission FACtors model,
EMFAC2021, architecture and newly developed web platform. Chapter 3 provides
details of new modules implemented in EMFAC2021 including plug-in hybrid electric
vehicles (PHEVs), energy consumption from zero-emission vehicles (ZEVs), and the
ammonia modules. Chapter 4 provides extensive information on updates to vehicle
emission rates and activities in the EMFAC2021 model. Chapter 5 presents overall
impacts of EMFAC2021 on total emissions, fuel consumption, energy use, and vehicle
activities and includes a comparison of model outputs between EMFAC2021 and
EMFAC2017.

1
Scenario Generation mode
2
https://ww2.arb.ca.gov/sites/default/files/2021-01/EMFAC202x_Users_Guide_01112021_final.pdf
4
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

1.2 New Features

In response to stakeholders’ comments on the model’s design and structure, the
EMFAC2021 model offers a variety of new features such as:
• Expansion of fuel technologies to include Plug-in Hybrid Electric Vehicles (PHEV)
and Natural Gas (NG) powered vehicles: EMFAC2021 separates PHEV from
conventional gasoline vehicles as a new fuel type for light-duty vehicles. PHEV-
specific emission rates and activity were developed based on results from CARB’s
real-world emission testing and vehicle activity data collections programs.
Detailed information can be found in Section 3.1. In addition, for the first time,
EMFAC2021 models emissions from natural gas trucks in the default emission
inventory and incorporates natural gas vehicle emission rates based on real-world
emissions testing. Similar to EMFAC2017, EMFAC2021 estimates natural gas
vehicle activities based on historical vehicle registration database. Detailed
information on natural gas population updates can be found in Section 4.3.5.4.
• Energy Consumption: EMFAC2021 now includes estimates of energy consumption
from light- and heavy-duty zero emission vehicles (ZEV). The significant
penetration of these vehicles in the fleet for light-duty and in the future for heavy-
duty vehicles has important implications for statewide energy demand. Staff used
the consumer-based trip data from ZEV manufacturers to estimate energy
consumption rates by speed for light-duty vehicles. For heavy-duty vehicles, staff
utilized the second-by-second data from transit fleets and estimated heavy-duty
energy consumption rates by speed. Detailed information can be found in Section
3.2.
• Ammonia Emissions (Section 3.3): For the first time, ammonia (NH3) emissions are
being included in the EMFAC model. A variety of historical and new NH3
emissions studies are used to develop running exhaust emission factors. The
emissions are determined by multiplying these emission rates by the vehicle miles
traveled (VMT). No adjustments for speed were made in this update. The results
indicate substantially higher emissions for Selective Catalytic Reduction (SCR)-
equipped diesels and natural gas vehicles. These two categories show significantly
higher emission rates than the historical diesel emission factors.
• Expansion of Heavy-Duty Truck Categories (Section 4.2.2.3): Heavy-duty truck
categories in EMFAC2021 were expanded to allow for more specificity in EMFAC
activity and emission rate updates and to better support future transportation
policies that CARB is pursuing. The new fleet categories provide higher resolution
of weight classes with additional vocation types.
• An Updated Light-Duty Activity Forecasting Approach (Section 4.5.1): New vehicle
sales and VMT forecasting models in EMFAC2021 were updated using a new
dynamic approach where a program creates an arbitrary list of economic
indicators, and then evaluates the accuracy and reasonableness of the generated
models. Unlike the static approach used in EMFAC2017 where a handful of

5
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

models were manually evaluated, the program goes through a large list of various
combinations of economic indicators to optimize the performance of the models
in predicting the historical data. These models were then evaluated by the staff for
further refinement based on reasonableness of assumption. In addition to the
present value of the economic indicators, EMFAC2021 considered the impacts of
the lagged indicators by one or two years.
• A New Heavy-Duty VMT Forecasting Framework (Section 4.5.2): EMFAC2017
projected diesel heavy-duty VMT at a statewide level based on a regression model
fitted to historical diesel fuel sales data. EMFAC2021 uses the California Statewide
Travel Demand Model (CSTDM) that forecasts VMT by county as the primary
source for the future VMT trends to better reflect the regional disparities in freight
VMT growth.
• A Novel Light-Duty ZEV Forecasting Framework (Section 4.5.3): Unlike
EMFAC2017 that projected ZEV market share based on the most likely
compliance scenario with California’s ZEV mandate, EMFAC2021 uses California
Energy Commission’s (CEC) vehicle choice model coupled with CARB’s updated
ZEV input attributes for short-term projections (2020-2030). For long-term
projections (2031-2050), the ZEV market share is assumed to plateau in California,
starting in 2030.

1.3 Overview of Major Changes

1.3.1 Fleet Characterization
Updates to EMFAC2021 vehicle population data using most recent Department of Motor
Vehicle (DMV) registration data and other sources are as follows.
• Multiple years of DMV data: EMFAC2021 uses DMV populations for years 2000
through 2019. The additional three years of vehicle registration data (2017-2019)
compared to EMFAC2017 reflect the most recent changes to California light- and
heavy-duty fleet characteristics.
• Expanded HD Vocational Categories (Section 4.2.2.3): Heavy-duty truck categories
in EMFAC2021 were expanded to allow for more specificity in EMFAC activity and
emission rate updates to better support CARB programs. Medium Heavy-Duty
Trucks (MHDT) were divided into four different weight classes (Class 4 through 7)
to provide higher resolution weight classes. New fleet groups such as delivery or
concrete mixers were added to provide the ability to separate trucks with specific
vocational types.
• International Registration Plan (IRP) data: IRP Clearinghouse data is another
primary source to estimate heavy-duty vehicle population. Vehicles already
registered in California can be identified as interstate trucks (CA IRP fleet) or
buses (motor coach fleet). For out-of-state vehicles in states and provinces that
report to the IRP Clearinghouse, updates were made using vehicle characteristics
for fleets that travel to California. As part of EMFAC2021 updates, the most recent
6
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

IRP data were used. This included updated VMT for T6 OOS, T7 NNOOS and T7
NOOS categories of EMFAC2021 and their populations were back-calculated with
the use of mileage accrual schedules.

• Vehicle Data from CHP and major Ports: List of Vehicle Identification Numbers
(VIN) from California Highway Patrol (CHP) School Bus Inspections are used to
identify school buses. Staff were also able to acquire list of VINs from the major
Ports which were used to identify drayage trucks operating at the port of Los
Angeles/Long Beach (POLA) and the port of Oakland (POAK).
• National Transit Database (NTD) data: Similar to EMFAC2017, NTD data was used
to characterize the transit fleet. An additional four years of transit bus reporting
(2016-2019) are included in this update.

1.3.2 In-Use Emissions

Light Duty Vehicles
EMFAC2021 has substantial updates regarding light-duty vehicle emission rates,
including running and start exhaust emission rates for Low Emission Vehicle 1 (LEV1)
through LEV3 vehicles, speed dependent PM brake wear emission rates as well as
updates to motorcycle emission rates. The PHEVs are also modeled using PHEV-specific
running and start exhaust emission rates that reflect the high power cold start emissions
identified by CARB staff as part of the 2017 Midterm Review to the Advanced Clean Cars
program 3.
In EMFAC2017, both running and start exhaust emission rates were updated for the first
time since EMFAC2000 using new Federal Test Procedure (FTP) data obtained from the
US EPA In-Use Verification Program (IUVP) coupled with emission test data from the
CARB’s Vehicle Surveillance Program (VSP). For EMFAC2021, running and start exhaust
emission rates derived from the Unified Cycle (UC) test data are further assessed and
updated with three years of new data from IUVP and CARB’s VSP programs. The method
used to determine the base emission rates are similar to EMFAC2017. However, a new
“low” emission regime has been added to model deterioration at the lower end of
emission levels. New test data were used to update the Ratio of Standards (ROS) and
emission rate models for LEV3 vehicles. When combined together, these updates have
resulted in higher NOx and ROG exhaust emissions for most of the light duty vehicles.
The PM brake-wear emission rates in previous versions of EMFAC were estimated using
limited available data from 2000-2003 4 and was assumed constant across various drive
cycles. For EMFAC2021, staff developed new emission factors and speed correction

3
https://ww2.arb.ca.gov/sites/default/files/2020-01/appendix_h_phev_testing_ac.pdf
4
See chapter 9 of EMFAC2011 Technical Support Document at: https://ww3.arb.ca.gov/msei/emfac2011-
technical-documentation-final-updated-0712-v03.pdf
7
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

factors for PM brake wear through an extramural contract. The new emission factors
incorporate the effect of modern material and composites used in disc brakes along with
impacts of real-world driving patterns on them. Compared to EMFAC2017,
EMFAC2021's new light-duty PM brake wear emission rates are significantly lower.
EMFAC2021 also includes updated carbon dioxide emission rates for all 2016 through
2020 model year light-duty vehicles. The updates use the latest national fuel efficiency
data from www.fueleconomy.gov, the official U.S. government source for fuel efficiency
information. Further, unlike 2-cycle fuel economies used in EMFAC2017, EMFAC2021
benefits from the more realistic 5-cycle fuel economies. Additionally, staff implemented
the Final Safer Affordable Fuel-Efficient (SAFE) Vehicles Rule phase-in schedule on GHG
emissions in EMFAC2021. The final SAFE rule applies to passenger cars and light-duty
trucks in California. While the previously established federal GHG emission standards and
related “augural” fuel economy standards for model years 2021-2025 would have
achieved yearly improvements through MY2025, the SAFE rule results in far less stringent
standards and consequently higher carbon dioxide emissions.
Another area of improvement in EMFAC2021 is the on-road motorcycle emission rates.
This emission category has not been updated since 2000 and staff has been conducting
extensive emission testing on motorcycles using both dynamometers and Portable
Emission Measurement Systems (PEMS) to better understand their emissions
characteristics for the past few years. Due to the availability of emissions data in this
round, EMFAC2021 is updated with new motorcycle exhaust emission rates.
Considering that PHEVs will increasingly constitute a higher fraction of California's
vehicle population, EMFAC2021 adds a new fuel type specifically for PHEV. To further
advance the modeling accuracy of EMFAC for this vehicle technology, staff used data
collected through CARB PEMS testing and an extramural contact with University of
California, Davis (UCD) to derive new emission factors and activity profiles for this vehicle
technology and incorporated them into the current version of the model.
Additionally, staff updated evaporative emissions in EMFAC2021 and replicated the
USEPA’s MOVES approach for modeling evaporative emissions. It brings USEPA’s
extensive research on modeling evaporative emissions into EMFAC and estimates
evaporative emissions with California-specific input data.
Heavy Duty Vehicles
For light heavy-duty (LHD) diesel and gasoline trucks (also referred to as medium duty
vehicles in CARB LEV regulations), the running exhaust emission rates have not been
revised since EMFAC2007 and the emission rates currently used in EMFAC2017 were
originally derived from the test data of heavier diesel and gasoline vehicles. To address
this issue, CARB has conducted an extensive emission testing program using both chassis
dynamometer as well as PEMS to update emission factors for this category of vehicles.

8
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

For heavy heavy-duty (HHD) diesel trucks, the running exhaust emission rates in
EMFAC2017 were based on the test data from CARB’s Truck and Bus Surveillance
Program (TBSP) and those from a project carried out by the Engine and Truck
Manufacturers Association (EMA) and University of California Riverside (UCR). For
EMFAC2017, the start emission rates for HHD diesel trucks were updated mainly based
on PEMS testing data collected through CARB’s TBSP and the idle emission rates for
HHD diesel trucks were updated based on the idle emissions test data from both CARB
and the Texas Transportation Institute (TTI) emission testing programs.
For medium heavy-duty (MHD) diesel trucks, due to a lack of test data the emission rates
in EMFAC2017 for trucks equipped with after-treatment systems were estimated by
applying scaling factors to the rates of HHD diesel trucks. For transit buses, the running
exhaust emission rates in EMFAC2017 were updated using the emissions test data from
CARB transit bus testing, the West Virginia University, and the Federal Transit
Administration.
For both MHD and HHD, staff updated NOx deterioration rates for newer vehicles using
data from an extramural contract. For EMFAC2021, staff has updated and developed
emission rates using emissions test data from the following sources:
1. Project 2R1702: Chassis Dynamometer and PEMS Testing of Class 2b-3 Vehicles-
This project was carried out by CARB with the objectives of understanding the in-
use emission performance of newer on-road LHD diesel engines; supporting
emissions inventory improvement; and addressing stakeholders’ concern
regarding a lack of emissions data from Class 2b-3 diesel trucks. The test data
were used to update the base emission rates (BER) and speed correction factors
(SCF) of LHD trucks.
2. Truck and Bus Surveillance Program (TBSP) – TBSP is an ongoing CARB program
that is designed to collect in-use emissions data for improving the emissions
inventory of heavy-duty vehicles, among other objectives. To date, the program
provided test data for more than 50 HHD diesel trucks (Class 8) and the test data
was used for updating the HHD diesel truck BERs and SCFs as well as start
emission rates (StER).
3. Surveillance Program for On-Road Class 4-6 Heavy-Duty Vehicles – This program
was initiated by CARB staff to provide emissions test data of in-use MHD trucks
for supporting emissions inventory development as well as other CARB programs.
So far, the program has tested four MHD diesel trucks and the test data from
these trucks were used for updating the MHD diesel truck BERs and SCFs.
4. In-Use Emissions Testing and Fuel Usage Profile of On-Road Heavy-Duty Vehicles
– This testing project, commonly referred to as “200-Vehicle Project”, sponsored
by multiple government agencies and private parties, has been carried out jointly
by the University of California at Riverside and West Virginia University and many
of the 200+ procured vehicles have been tested for their emissions. The emissions
data from this project has provided a unique opportunity for developing and

9
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

revising the emission rates of natural gas powered HD vehicles, including BERs,
SCFs, and idle emission rates (IdleER).
5. Contract 17AQP006 to Updates Heavy-Duty Emission Deterioration in EMFAC –
Through this project, a large volume of on-board diagnostics (OBD) data were
collected to improve heavy-duty vehicle deterioration assumptions in EMFAC
2021. Staff used the results of this study to update NOx deterioration rates for
newer OBD-equipped (engine model year 2013 and newer) vehicles.

1.3.3 Activity
Activity profile refers to the operational characteristics that influence vehicle emissions,
including mileage accrual rates, odometer schedule, speed profile, starts per day, soak
time distribution, and temporal distribution of VMT and trips. EMFAC2021 implemented
major updates on activity profile for both light-duty and heavy-duty vehicles using the
latest vehicle data. The data sources and updates include:
The 2017 National Household Travel Survey-California Add-on: This dataset provides the
most recent source for on-road motorcycle odometer mileage data in EMFAC2021. For
the purpose of EMFAC update, data from 1,923 surveyed motorcycles were used to
develop on-road motorcycle mileage accrual rates and the odometer schedule.
PHEV activity data from UC Davis: Through CARB project 17AQP005, “Cold Start
Emission Impacts of Blended Plug-In Hybrids”, UC Davis collected real-time data
from164 PHEVs including date and time, accelerator pedal position, engine RPM, vehicle
speed, catalyst temperature, and state of charge. The dataset was utilized to estimate
PHEV starts per day, start temporal distribution, and soak time distribution 5.
2018 California Vehicle Inventory and Use Survey and Geotab Telematics Data (Section
4.4.2): This data is used to update HD mileage accrual rate and odometer schedule. HDV
mileage and odometer data in EMFAC2017 were based on 2002 Vehicle In-Use Survey
(VIUS), now almost 20 years old and less likely to be representative of the current
trucking industry. New data has become available for updating the annual accrual rates
for most of the HDV Fleet categories to reflect more current activity trends. Fleet
Management Company Geotab provided aggregate data from GPS data logging for
over 1.3 million GPS tracking devices in operation on HDVs. In addition, Caltrans
completed the California Vehicle Inventory and Use Survey (CA-VIUS) in 2018 to update
the CA portion of the discontinued national VIUS.
Portable Activity Measurement System (PAMS) from 200-Vehicle Project: Real world
activity data from the 200-Vehicle Project were collected by PAMS by University of

5
URL: https://ww2.arb.ca.gov/sites/default/files/2021-03/erg_finalreport_hdv_accruals_20190614_ada.pdf

10
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

California, Riverside (UCR) and West Virginia University through the “200-Vehicle
Project” were analyzed to update activity profiles within the EMFAC2021 model. These
data were pooled with data collected through the “90-Vehicle Project” by UCR CE-CERT
in 2017 6 to update VMT distribution by hour and speed, engine starts per day, soak time
distributions, and idle hours. Compared to EMFAC2017, EMFAC2021 shows that the
majority of HD categories have higher percentage of VMT at higher speed, higher
number of starts per day with longer soak time, and less extended idling time.

1.3.4 New Sales and VMT Forecasting

As described in EMFAC2017 7technical support documentation (Sections 4.5.1 and
4.5.2), EMFAC2017 uses socio-econometric regression model forecasting methods to
predict new vehicle sales and VMT growth trends. These models connect the activity
estimates, VMT and new vehicle sales, to state and national economic indicators such as
fuel prices, disposable income, human populations, and federal interest rates.
For EMFAC2021, staff created new models for estimation of light-duty VMT and new
vehicle sales. Unlike EMFAC2017, EMFAC2021 benefits from models that are created
dynamically by searching through a large pool of various combinations of economic
indicators, and optimizing the parameter selection based on the performance of the
model in predicting the historical data. To do that, staff used the latest available socio-
economic data from UCLA Anderson Forecast (UCLA), California Department of Finance
(DOF), California Board of Equalization (BOE), California Energy Commission (CEC), and
Federal Reserve Bank of St. Louis.
As for the light-duty ZEV market share, EMFAC2021 projects the shares based on
consumer choice models rather than a most likely compliance scenario used in
EMFAC2017. As a result, the future ZEV market share projected by EMFAC2021 is
higher than that predicted by the EMFAC2017 model. Specifically, the ZEV market share
in light-duty vehicle new sales increase from approximately 8 percent in EMFAC2017 to
12 percent of EMFAC2021 for the model year 2030 passenger vehicles.
In the heavy-duty space, EMFAC2017 forecasted statewide diesel VMT using a
regression model fitted to historical diesel fuel sales. For EMFAC2021, the primary
source for the future VMT trends is the CSTDM 8 forecasted VMT per county. Staff
utilized county level medium and heavy duty VMT estimated by the CSTDM to determine
the county specific VMT growth rates. Refer to Section 4.5.2 for more details.

6
https://ww2.arb.ca.gov/sites/default/files/classic//research/apr/past/13-301.pdf
7
EMFAC2017 Technical Documentation, URL:
https://ww3.arb.ca.gov/msei/downloads/emfac2017-volume-iii-technical-documentation.pdf
8
https://tmip.org/content/california-statewide-freight-forecasting-model-CSTDM
11
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

1.3.5 Regulations and Policies

Policies and regulations covered in this version of EMFAC are listed below.
HD Warranty phase 1 (Section 4.6.5): In 2018, CARB adopted amendments to the heavy
duty engine (above 14,000 lbs.) warranty regulation and further updated those warranty
mileages through the CARB’s HD Omnibus regulation. NOx and PM emissions
reductions from these programs are incorporated into EMFAC2021.
Innovative Clean Transit (ICT) (Section 4.2.3): Adopted in 2018, ICT regulation is a
program that enhance public health by improving air quality and to mitigate climate
change by transforming California transit bus fleets to zero-emissions technologies. This
regulation applies to all transit agencies that own, operate, or lease buses with GVWR
above 14,000 lbs. It requires all public transit agencies to gradually transition to a 100%
zero-emission bus fleet and encourages them to provide innovative first and last-mile
connectivity and improved mobility for transit riders. The zero emissions buses
introduced through this program are reflected in EMFAC2021.
Amendments to Heavy-Duty Vehicle Inspection Program (HDVIP) and Periodic Smoke
Inspection Program (PSIP) (Section 4.6): CARB adopted amendments to the HDVIP and
PSIP programs in 2019. This program reduces particulate matter (PM) from diesel-
powered vehicles (GVWR above 14,000 lbs.) and it will be achieved through more
stringent opacity limit for non-DPF and DPF-equipped vehicles. PM reductions from this
program are reflected in EMFAC2021.
Zero Emission Airport Shuttle Bus: CARB adopted the zero-emission airport shuttle bus
regulation in 2019 and required airport shuttle fleets to fully transition to zero emission
by 2035. This regulation is not explicitly accounted for in EMFAC2021 because this
category represents a very small fraction of the fleet (less than 2,000 vehicles).
Advanced Clean Truck (ACT) (Section 4.5.3): Adopted in 2020, ACT requires
manufacturers of Class 2b-8 chassis or complete vehicles with combustion engines to sell
zero-emission trucks as an increasing percentage of their annual California sales from
2024 to 2035. By 2035, zero-emission truck/chassis sales would need to be 55% of Class
2b – 3 truck sales, 75% of Class 4 – 8 vocational truck sales, and 40% of Class 7-8 truck
tractor sales. EMFAC2021 reflects ACT by modelling heavy-duty zero emission vehicles
(ZEVs) based on the sales percentage requirements for each model year and those
percentages were applied to vehicles first sold or certified in California.
Heavy-Duty (HD) Omnibus (Section 4.6.4): EMFAC2021 reflects NOx emissions
reductions from Heavy Duty Omnibus regulation adopted in August 2020. HD Omnibus
represents a comprehensive update to heavy-duty NOx emissions standards and ensures
that heavy-duty engines will emit much lower NOx emissions throughout their lifetimes.
SAFE Vehicle Rules and Actions (Section 4.6.1): In September 2019, the U.S.
Environmental Protection Agency (U.S. EPA) and the National Highway Traffic Safety
Administration (NHTSA) issued the Safer Affordable Fuel-Efficient or SAFE Vehicles Rule
12
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Part One: One National Program (SAFE Part One) that revoked California’s authority to
set its own greenhouse gas emissions standards and ZEV mandates, 84 Fed. Reg. 51,310
(Sept. 27, 2019). In April 2020, the federal agencies issued the SAFE Vehicles Rule for
Model Years 2021-2026 Passenger Cars and Light Trucks (Final SAFE Rule) that relaxed
federal greenhouse gas emissions and fuel economy standards, 85 Fed. Reg. 24,174
(Apr. 30, 2020). The Final SAFE Rule relaxed federal greenhouse gas emissions and
Corporate Average Fuel Economy (CAFE) standards to increase in stringency at only
about 1.5% per year from model year (MY) 2020 levels over MYs 2021–2026. CARB has
updated the carbon dioxide emission factors for gasoline-fueled passenger cars and
light-duty trucks in EMFAC2021 to reflect the impacts of both the SAFE Part One and
the Final SAFE Rule. CARB has additionally updated EMFAC2021 model to reflect the
expected impact of California’s Framework Agreements on Clean Cars 9. Automakers
who voluntarily agreed to the framework agreements are BMW of North America
(including Rolls Royce for purposes of the agreement), Ford, Honda, Volkswagen Group
of America (including VW and Audi), and Volvo. The framework agreements are voluntary
commitments that support continued annual reductions of vehicle greenhouse gas
emissions through the 2026 model year, and encourage innovation to accelerate the
transition to electric vehicles.

9
https://ww2.arb.ca.gov/news/framework-agreements-clean-cars

13
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

2 Introduction
2.1 EMFAC2021
Today, the California Air Resources Board (CARB) is tasked with three key mandates:
achieving health-based air quality standards for ozone, particulate matter and other air
pollutants; reducing public exposure to toxic air contaminants; and leading California’s
efforts to fight climate change by promoting clean, energy-efficient fuels and
technology. Emissions inventories are essential tools that CARB utilize in developing
regulations and control strategies that helps California to achieve its air quality, climate,
and toxics reduction goals.
An emissions inventory (for any source category) can be calculated, at the most basic
level as the product of an emission rate, expressed in grams of a pollutant emitted per
some unit of source activity, and a measure of that source’s activity. The following
expression illustrates this basic relationship between the emissions rate and source
activity used to calculate emissions:
(Emission Factor) X (Source Activity) = Emissions
For on-road motor vehicles, emissions rates are typically expressed as mass of pollutant
emitted per mile driven, per vehicle per day, or per trip made, depending on the
emissions process being analyzed. An emission process for a motor vehicle is the physical
mechanism that results in the emissions of a pollutant (e.g., the combustion of fuel, the
evaporation of fuel, tire or brake wear, or the start of an engine).
CARB developed an EMFAC model to calculate statewide or regional emissions
inventories by multiplying emissions rates with vehicle activity data from all motor
vehicles, including passenger cars to heavy-duty trucks, operating on highways, freeways,
and local roads in California.
Over the years, tougher emissions standards have been met with technological solutions
of increasing complexity. As a result, the emissions estimation models have also grown in
size and complexity. EMFAC2021 is the latest emissions inventory model that calculates
emissions inventories for motor vehicles operating on California roadways. EMFAC2021
represents the next step forward in the ongoing improvement process for EMFAC, and
reflects the CARB’s current understanding of how vehicles travel and how much they
pollute. The EMFAC2021 model is needed to support CARB’s regulatory and air quality
planning efforts and to meet the Federal Highway Administration’s transportation
planning requirements.
The EMFAC2021 model can be used to show how California motor vehicle emissions
have changed over time and are projected to change in the future. This information
helps CARB evaluate prospective control programs and determine the most effective,
science-based proposals for protecting the environment. EMFAC2021 includes the latest
data on California’s car and truck fleets and travel activity. New forecasting methods
14
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

have been incorporated for developing vehicle age distributions and estimating vehicle
miles traveled. The model also reflects the emissions impact of Federal and California
recent rulemakings.

2.2 Modeling Architecture

Starting from EMFAC2014, CARB staff gradually departed from using Fortran to Python
and MySQL. EMFAC2014 and EMFAC2017 had a legacy Fortran module for certain
emission processes. EMFAC2021 is developed completely in Python and MySQL
database. While EMFAC development migrates to Python and MySQL, staff with the
goal to maximize the following aspects, including user friendliness, transparency, ease of
maintenance, flexibility for incorporating and supporting future regulations, improved
computational efficiency, and convenience to transfer of EMFAC outputs to other
services.
One major architectural change of EMFAC2021 is the ability to run the model in parallel
on multiple CPU cores. Although the number of computations required for model runs
has increased substantially due to many new features and updates in EMFAC2021, this
parallelization allows EMFAC2021 to run significantly faster than its predecessors when
deployed on high performance computing clusters. EMFAC2021 runs for multiple
combinations of sub-areas and calendar years simultaneously up to a level supported by
user’s computer. Since EMFAC performs most computations on its MySQL server, it is
recommended to run EMFAC's MySQL server with multiple CPU cores and a high-
performance storage system. Figure 2.2-1 displays a flow chart indicating the GUI
selections necessary to generate the various outputs of EMFAC2021.
Figure 2.2-1. EMFAC2021 Overall Flow

15
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Important! “Custom Activity” or Scenario Generation (SG) Mode

Similar to EMFAC2017, the SG mode in EMFAC2021 can also be used to produce
emissions inventories for both conformity assessments and SB375 assessments. For
conformity assessments, emissions are estimated with all current controls active. For
SB375 assessments, the Advanced Clean Cars (ACC) and SAFE vehicle rules are
deactivated. Because the ACC regulation has certain assumptions about vehicle usage
built into it, default data in custom activity templates produced for conformity
assessments will not match the default data in templates for SB375 assessments. For the
same reason, estimates of GHG will also differ.
Moreover, the SG mode has been extended to provide the end users more flexibility
with technology mix using the updated EMFAC2021 vehicle categories. Now users can
customize the VMT with electric, plug-in hybrid, and natural gas, besides gasoline and
diesel.
The SG mode uses total VMT as input and output cVMT and eVMT separately. Just as
the previous version, end users can also use their own speed fractions by selecting the
“Custom Hourly Speed Fractions” option. However, the user input speed fractions will
only be used for cVMT calculation and the eVMT fraction will stay the same as default
mode.
For PHEV brake wear emission calculations, the SG by hourly speed mode ignores the
varying speed weighting factors by model years. For this reason, the estimates of PHEV
brake wear emissions in SG mode with “Custom Hourly Speed Fractions” option will
differ slightly from the default mode using user’s input activity profiles.

2.3 Web Interface

Along with EMFAC2021, a new Web Platform https://arb.ca.gov/emfac/ was developed
to provide easy and quick access to all of the features provided by EMFAC2021 on the
web (Figure 2.3-1). The new web interface not only provides all features of the former
EMFAC Web Databases, it also provides the Project Analysis and Scenario Analysis
features, which correspond to Project-Level Analysis (PL) and Scenario Generation (SG)
modes in the EMFAC PC application, respectively. Compared to SG and PL modes of
EMFAC PC application, the equivalent modes on the web are much faster as they are
processed using preprocessed EMFAC outputs. As the new web platform provides all
the functionalities of EMFAC, staff expect most EMFAC users to use the Web Platform
rather than running EMFAC PC application.

16
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure 2.3-1. EMFAC web application and video tutorials

17
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

18
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

3 New Modules in EMFAC2021

3.1 PHEV Module
3.1.1 Background
Because of California and Federal electric vehicle incentives, Plug-in Hybrid Electric
Vehicle (PHEV) sales have increased significantly in recent years. According to California
DMV vehicle registration database, by the end of 2019, about 251 thousand PHEVs,
including 20 makes across 42 models, are registered in California. One of the
EMFAC2021’s new features is to separate PHEVs from conventional light-duty gasoline
vehicles and to model them in a new module, called PHEV module. As a new fuel type,
the PHEV module provides emission and activity estimates of PHEV for light-duty vehicle
categories.
PHEV vehicles use battery power to run an electric motor and use another fuel type,
typically gasoline, to power an internal combustion engine (ICE). These vehicles have
various engine operations. For example, their engine starts can occur at any time during
a trip based on the vehicle's battery level and energy demand. When running on battery
power alone, PHEVs do not produce tailpipe emissions; however, the engine may help
power the vehicle when the battery is mostly depleted during rapid acceleration or at
high speeds or when intensive heating or air conditioning is demanded. Under those
circumstances, the ICE is started under high power demand. Some of these high-power
demands start events may occur while the catalytic converter is cold, which could result
in significantly higher emissions than a warm start. In fact, CARB's in-house test results
show that the high-power cold starts from PHEV could have two to five times higher cold
start emissions than traditional cold starts from conventional vehicles.
The PHEV emission factors are updated in EMFAC2021, based on new emissions data
from real-world emission testing of several PHEV vehicles conducted by CARB and sec-
by-sec PHEV activity data collected through CARB project 17AQ005, “Cold Start
Emission Impacts of Blended Plug-In Hybrids”. Eleven PHEVs across different makes and
models were tested using On-road Portable Emissions Measurement System (PEMS) over
various routes and different soak times throughout Southern California.
Table 3.1.1-1 shows the list of eleven PHEVs included in CARB’s PEMS testing program.
After post-processing and quality assurance of the test data, eight vehicles are included
in the final emission analysis and these vehicles are certified under LEV2 SULEV or LEV3
SULEV30. Table 3.1.1-2 details the PEMS testing routes and characteristics. Each route
was tested twice for each vehicle. Besides, through an extramural contract, CARB
collaborated with UC Davis to collect activity data on 164 PHEVs over 200-300 days.

19
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Table 3.1.1-1. List of PHEV in CARB’s PEMS testing program

Model US06
No. Make Model Tech Group Type Included
Year Capable
1 2017 Toyota Prius L3 SULEV30 Blended Yes Yes
2 2017 Audi A3 E-Tron L3 SULEV30 Blended Yes Yes
3 2012 Chevy Volt L2 SULEV Non-blended Yes Yes
4 2014 Ford Fusion L2 SULEV Blended No Yes
5 2016 Ford C-Max L2 SULEV Blended No Yes
6 2016 Hyundai Sonata L3 SULEV30 Blended No Yes
7 2017 BMW 330e L3 ULEV125 Blended No No
8 2016 Porsche Cayenne L2 ULEV Blended No No
9 2016 Mercedes C350e L3 SULEV30 Blended No Yes
10 2014 Toyota Prius L2 SULEV Blended No Yes
11 2017 Chevy Volt L3 ULEV125 Blended No No

Table 3.1.1-2. CARB PEMS emission testing routes

Distance Travel Average Elevation
Route Route Characteristics
(mile) Time (min) Speed (mph) Gain (m)
Downtown Los
16 60 15 Typical city driving 75
Angeles (DTLA)
Freeway around Freeway with higher power
18 30 35-40 50
El Monte demand
Local roads in El
1.3 3-6 3-6 Various soak periods 10
Monte
El Monte to Longer arterial and freeway
65 95 45-50 800
Lake Pyramid with high road grade

3.1.2 Emission Rates Analysis

Like conventional vehicles, the PHEV start exhaust emission rates for HC, NOx, and CO
are derived from PEMS testing performed by CARB. The emission rates are a function of
soak time and are grouped for blended and non-blended PHEVs separately. For blended
PHEVs, the engine starts and provides propulsion power when the driver's power
demand is higher than what the electric powertrain and battery can provide. In contrast,
for non-blended PHEVs, the electric powertrain provides propulsion regardless of the
driver demand until the car switches to charge sustaining mode. At this point, the battery
is maintained at a minimal level, and the engine's use is prioritized. This happens when
the battery reaches a low level of charge.
Additionally, blended PHEVs typically have smaller-sized batteries and show a different
starts frequency behavior than non-blended PHEVs. The US06 cycle is an aggressive test
cycle with high speed or high acceleration driving behavior. Of the vehicles being tested,
three out of eight vehicles are US06 capable, including one non-blended and two
blended PHEVs. For the starts analysis, staff used US06 capable PHEVs for the non-
blended results.

20
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

For this analysis, staff defined PHEV start as when the soak time is greater than 5
minutes, and the engine RPM is above 100 rpm while this condition lasts for more than
five but less than 100 seconds. All the PHEV starts from the PEMS data were binned by
EMFAC soak time bins using a piecewise linear regression.
Figure 3.1.2-1 shows a comparison between PHEV and conventional ICE vehicle starts
emissions in units of grams per start for the same certification level as LEV2 SULEVs with
gasoline fuel type. As shown, blended PHEVs have significantly higher start emissions
when compared to conventional and non-blended PHEVs; especially with longer soaking
times this is more pronounced.

21
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure 3.1.2-1. Comparison of PHEV HC and NOx start emission rates in g/start by
soak time in minutes with gasoline LEV2 SULEV

0.40
0.35
HC start emission (g/start)

0.30 Non-Blended
0.25 Blended
0.20 LEV2 SULEV
0.15
0.10
0.05
0.00
0 100 200 300 400 500 600 700 800
Soak time (min)
0.40
NOx start emission (g/start)

0.35
0.30
Non-Blended
0.25
Blended
0.20
LEV2 SULEV
0.15
0.10
0.05
0.00
0 100 200 300 400 500 600 700 800
Soak time (min)
The PHEV running exhaust emissions data obtained through PEMS testing were analyzed
using the U.S. EPA’s MOVES 10 Vehicle Specific Power (VSP) modal model approach. VSP
is an estimate of the power demand on the engine during driving. MOVES emission rates
for running exhausts are based on 23 operating modes called OpModes, which are
combinations of speed and VSP that account for different patterns of acceleration,
cruising, deceleration, and average speed. Figure 3.1.2-2 presents the approach in a flow
chart. All the data points from non-first starts with less than 5 minutes soaking time and
beyond the 100 seconds starts duration threshold were binned into 23 MOVES

10
U.S. EPA MOtor Vehicle Emission Simulator (MOVES): https://www.epa.gov/moves
22
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

OpModes based on VSP and speed. As the next step, the model average running
emission rate in gram per second was calculated for all data points within an OpMode
bin. Following that, sixteen CARB test cycles representing various cycle-average speeds
from 2 to 71 mph were selected, and for each one, time distribution of OpModes and a
time-weighted cycle-average emission rate were developed. The PHEV running exhaust
emission rates were derived based on the relationship between these cycle-average
emission rates versus cycle-average speed in miles per hour.
Figure 3.1.2-2. EMFAC2021 approach for estimating PHEV running exhaust emissions
using MOVES modal model bins

Figure 3.1.2-3 compares CO2, CO, HC, and NOx running exhaust emissions of PHEVs
against those of conventional vehicles with gasoline LEV2 SULEVs. As shown, PHEV
vehicles have similar CO2 emission rates per mile, higher CO emission rates at lower
speeds, slightly higher HC emission rates, and lower NOx emission rates than
conventional ICE vehicles certified at the same level.

23
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure 3.1.2-3. Comparison of PHEV running exhaust emissions with LEV2 SULEV for
CO2, CO, HC, and NOx

1750 2.0
PHEV PHEV
1500

CO (g/mile)
1.5
CO2 (g/mile)

1250 LEV2 SULEV Gas LEV2 SULEV Gas

1000 1.0
750
500 0.5
250
0 0.0
0 10 20 30 40 50 60 70 80 0 10 20 30 40 50 60 70 80
Speed (mph) Speed (mph)

0.030 0.025
PHEV PHEV
0.025 0.020
NOx (g/mile)
HC (g/mile)

LEV2 SULEV Gas LEV2 SULEV Gas

0.020
0.015
0.015
0.010
0.010
0.005 0.005

0.000 0.000
0 10 20 30 40 50 60 70 80 0 10 20 30 40 50 60 70 80
Speed (mph) Speed (mph)

As a pilot study with limited emission testing data, PHEV starts and running exhaust
emission rates in EMFAC2021 are assumed to be the same for different vehicle classes.
For other pollutants or emission processes such as the evaporative processes where
PHEV specific data was not available, EMFAC2021 is programmed to use the emission
rates of conventional gasoline vehicles that are certified under the same emissions level;
or to use emission rates from electric vehicles of the same vehicle class for PM brakewear
emission estimation.

3.1.3 Activity Analysis

PHEV’s fleet utility factor (FUF) refers to a measure of fleet-average distance driven on
electricity. Specifically, it is a utility factor based on the total VMT and gives a probability
that an average vehicle will be driven less than or equal to its charge depleting range
during a particular driving day. Hence, FUF is particularly useful to calculate the expected
fuel and electric energy consumption for a fleet of vehicles.
Due to PHEV’s ability to partially substitute electricity for gasoline, FUF is particularly
useful to calculate the expected fuel, electric energy consumption, and emissions for a

24
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

fleet of vehicles. It is because of the effect of FUF that PHEV vehicles produce less
criteria and GHG emissions compared to conventional vehicles. Table 3.1.3-1 shows the
PHEV FUFs used in EMFAC2021. Pre-2018 model year FUF was derived from telematics
data 11, and the rest are forecasted. More details for the forecasted data were presented
in EMFAC2017 Technical Documentation Section 4.5.3.
Table 3.1.3-1. EMFAC2021 PHEV fleet utility factors
Model Year Fleet Utility Factor (FUF)
≤ 2018 0.46
2019 0.47
2020 0.48
2021 0.50
2022 0.55
2023 0.56
2024 0.58
≥ 2025 0.59

PHEV population in EMFAC2021 for historical years are based on California DMV vehicle
registration data. EMFAC2017 and earlier versions categorized PHEV as gasoline
vehicles. As a result, EMFAC2021 used fractions of PHEV by model year to re-distribute
PHEV population from gasoline population in historical years. The earliest available PHEV
model year from the DMV database is 2010.

11
https://ww2.arb.ca.gov/sites/default/files/2020-
01/appendix_g_pev_in_use_and_charging_data_analysis_ac.pdf
25
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Table 3.1.3-2 shows the fractions by vehicle class that EMFAC2021 used to re-distribute
PHEV population in historical years.
Table 3.1.3-2. PHEV population fractions by vehicle class in historical years

Vehicle Class
Model Year
LDA LDT1 LDT2 MDV

2010 0.00% 0.00% 0.00% 0.00%

2011 0.18% 0.00% 0.00% 0.00%

2012 1.17% 0.00% 0.00% 0.00%

2013 1.83% 0.00% 0.00% 0.00%

2014 2.63% 0.00% 0.00% 0.00%

2015 1.65% 0.00% 0.06% 0.00%

2016 1.74% 0.00% 0.15% 0.99%

2017 4.13% 0.04% 0.00% 0.80%

2018 4.08% 0.00% 0.35% 2.05%

2019 3.44% 0.00% 1.12% 0.60%

Moreover, all future year activity data is forecasted using a newly introduced ZEV
forecasting framework in EMFAC2021 (Section 4.5.3). The ZEV forecasting framework
uses CEC’s vehicle choice models with CARB’s updated ZEV input attributes for short-
term projections (2020-2030). For long-term projections (2031-2050), it is assumed that
California ZEV market shares will plateau starting with vehicle model year 2030.
Through an extramural contract, CARB analyzed the PHEV activity data collected by UC
Davis to generate engine start activity inputs for EMFAC2021, including a starts per day
distribution, start temporal distribution, and soak time distribution 12. Staff also compared
PHEV activity with conventional gasoline vehicle activity used in EMFAC.
As shown in Table 3.1.3-3, PHEVs have much higher starts frequency per day than
conventional vehicles, and blended PHEVs have more first starts than non-blended
PHEVs. In Figure 3.1.3-1, for all the first starts from blended PHEVs, more than 40% have
a soak time less than 60 minutes. In comparison, non-blended PHEVs have higher
fraction of cold starts. For the temporal distribution of engine starts, as shown in Figure

12
URL: https://ww2.arb.ca.gov/sites/default/files/2021-03/erg_finalreport_hdv_accruals_20190614_ada.pdf

26
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

3.1.3-2, non-blended PHEVs have a lower fraction of starts in the morning and higher
fraction in the afternoon.
Table 3.1.3-3. PHEV starts frequency per day from UC Davis Dataset
Non-First Starts
Category Starts First Starts
> 5 mins soak
Conventional ICE 2.7 – 5.1 - -
PHEV Non-blended 31.9 2.5 1.6
PHEV Blended 96.6 4.2 1.7

Figure 3.1.3-1. PHEV Soak time distribution for first starts from UC Davis Dataset

60
Conventional ICE
50 Blended
Non-Blended
Percent of Soaks

40

30

20

10

0
60 120 180 240 300 360 420 480 540 600 660 720 > 720
Soak Time Bin Maximum (Minutes)

27
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure 3.1.3-2. Temporal distribution of PHEV engine starts from UC Davis Dataset
12
Non Blended First
10 Blended First
Percent of Daily Starts

Conventional ICE
8

0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Hour of the Day

In EMFAC2021, all these starts activity have been processed into a similar table format as
conventional vehicles. EMFAC2021 assumes the PHEV fleet consists of 50% blended and
50% non-blended. PHEV trips are computed for blended and non-blended, separately,
accounting for each different starts frequency and hourly distribution. Starts and running
emissions are calculated in the same way as conventional vehicles by only accounting for
the gasoline portion in the default emission mode. In the emission rate (project level, PL)
mode, PHEV emission rates are calculated based on their total activity, including both
gasoline and electric portions.

3.2 Energy Module

Meeting California’s air quality and climate goals will require an ongoing transformation
to cleaner technologies and fuels. As of January 2021, there are more than 800,000 light-
duty zero emission and plug-in hybrid vehicles are sold in California 13. Additionally, the
adoption of Innovative Clean Transit (ICT) and Advanced Clean Truck (ACT) regulations
will significantly increase the number of heavy-duty zero-emission buses and trucks in
California. Therefore, it is critical to understand the energy consumption associated with
these vehicles for the purpose of infrastructure planning and development as well as
utility grid upgrades.

13
https://www.veloz.org/sales-dashboard/
28
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

3.2.1 Light-Duty
Plug-in hybrid and electric vehicles present much higher energy efficiency as compared
to conventional vehicles. However, similar to conventional vehicles, operational
efficiencies of zero emission vehicles vary under different driving conditions. EMFAC2017
lacked sufficient information to accurately estimate energy consumptions associated with
plug-in electric vehicles. Accurate characterization of energy consumption is particularly
important because of the large recent increase in battery zero emission and plug-in
hybrid vehicles in California. To collect more accurate information, real-world consumer
trip information that was provided by manufacturers was analyzed for approximately
50,000 battery electric (BEV) and plug-in hybrid electric vehicles (PHEV). Trip information
was collected through telematics, advanced on-board diagnostics (OBD) technologies,
and mobile applications via consumer cell phones. Figure 3.2.1-1 below summarizes the
sample size for each BEV and PHEV make and model analyzed, as well as the OEM
sampling period.
Figure 3.2.1-1. OEM Sampling Period for Trip-by-Trip Energy Data
CY2012 CY2013 CY2014 CY2015 CY2016 CY2017 CY2018
Toyota Prius PHEV 1,523 vehicles
Toyota Prius Prime 3,118 vehicles
Honda Accord PHEV 189 vehicles
Ford C-Max Energi 10,253 vehicles
Ford Fusion Energi 12,842 vehicles
Chevrolet Volt 2,154 vehicles
Ford Focus EV 4,218 vehicles
Honda Fit EV 645 vehicles
Nissan LEAF 12,215 vehicles

The electric energy consumption and average speed were estimated for each PHEV and
BEV trip. Ford, Toyota, and Honda provided a direct estimate of energy consumption
and, energy consumption for Chevrolet and Nissan was estimated using state of charge
(SOC) and battery capacity. Energy consumption per mile was obtained by taking energy
consumption and dividing it by the distance for a given trip. Average speeds were either
provided directly or calculated as distance divided by time. For PHEVs, only pure electric
trips during which the engine did not turn on were used. The average speeds were
mapped to EMFAC speed bins. For each make and model, the average energy
consumption was calculated for each speed bin. A California sales weighted average of
each make and model were used to calculate an overall energy consumption curve for
BEVs and PHEVs. To develop a more accurate total energy consumption, staff accounted
for the energy loss that occurs during charging when power is drawn from the electrical
grid from the on-board charge modules. This energy loss approximation of 20% was
found using Idaho National Laboratory charging test data 14 that measured the efficiency
and power quality of the on-board charge modules for 4 BEVs and 1 PHEV. To obtain

14
https://avt.inl.gov/content/charging-system-testing/vehicle-charging-system-testing
29
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

charging losses, staff took the average of the Level 1 (110V) and the Level 2 (208-240V)
charger efficiencies, which had charging efficiencies of 73% and 87%, respectively.
In addition to estimating energy consumption as a function of speed, the electric vehicle
miles travelled (eVMT) fraction at each speed bin were also estimated. Here, eVMT is
defined as a pure electricity powered trip. These fractions were used to allocate total
eVMT at each speed for BEVs and PHEVs. In EMFAC2021, vehicle miles traveled (VMT) is
split into two outputs, combustion VMT (cVMT), which corresponds to VMT that is
powered by the gasoline engine, and eVMT. Note that PHEVs have a mix of eVMT and
cVMT, while BEVs 15 just have eVMT.
The overall sales-weighted electric energy consumption from the grid as a function of
speed are shown in Figure 3.2.1-2 for BEVs and PHEVs. Energy consumptions shown
here are corrected for charging energy losses.
Figure 3.2.1-2. BEV and PHEV Electric Trip Energy Consumption-CA Sales Fleet
Distribution

0.8
0.7
kWh per mile per vehicle

0.6
0.5
(kWh/mile)

0.4 BEV
0.3 PHEV

0.2
0.1
0
0 10 20 30 40 50 60 70
Speed Bin (MPH)

In general, BEVs and PHEVs exhibit high-energy consumption at relative low and high
speed operations, while they demonstrate highest efficiency at speeds of 25-35 mph.
Energy consumption is largest during low speed operation due to periods of
acceleration. At faster speeds, energy consumption increases due to aerodynamic drag.
Energy consumption rates reach a minimum 0.28 kWh per mile at 35 mph for PHEVs and
0.37 kWh per mile at 25 mph for BEVs. Some of this difference between BEVs and PHEVs
may be due to differences in the method for estimating energy consumption. For

15
Due to lack of data on FCEVs, staff use similar assumptions as in BEVs to estimate energy and emissions
from FCEVs
30
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

example, estimating energy consumption from SOC and battery capacity may lead to
overestimation of energy consumption, since battery capacity decreases over time.
However, some makes and models, like Chevy Volt, will not have this issue because the
SOC is reported as a function of usable battery capacity. Additionally, staff did not have
access to trip-by-trip Tesla data, so a significant portion of the BEV market is missing
from this analysis. For the next version of EMFAC staff will explore improvements in
energy efficiency with model year, make and model, and refine estimates of charging
losses. The eVMT distribution for BEVs and PHEVs are shown in Figure 3.2.1-3.
Figure 3.2.1-3. EVMT Speed Distribution

25%
BEVs PHEVs
20%
EVMT Percentage

15%

10%

5%

0%
0 10 20 30 40 50 60 70
Speed Bin (MPH)

The eVMT peak for PHEVs occurs at 25-30 mph, while the peak for BEVs is around 25-30
mph. PHEV eVMT is skewed towards lower speeds as most of the PHEVs in this sample
were not US06 capable. As mentioned above, the PHEV results are only for electric trips
only, during which the engine did not turn on.

3.2.2 Heavy-Duty
CARB has conducted a study, 16 through an extramural contract, to quantify the emission
reduction benefits and operational performance of the zero-emission technology vehicles
compared to the current conventional vehicles, and to review commercialization

16
Data Monitoring, Collection, and Analysis for Projects Granted Under Fiscal Year 14-15 Air Quality
Improvements Program and Low Carbon Transportation Greenhouse Gas Reduction Fund Investments
(15MSC006)
31
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

readiness of zero-emission technologies for industry adoption. The vehicle activity

information was collected by HEM-supplied data loggers accessible through SAE J1962
OBD connectors. Staff analyzed real-world second-by-second vehicle activity data
generated through this contract to better understand the electricity usage of heavy-duty
vehicles under different driving conditions. Wheel-based vehicle speed was used to
group vehicle activity by speed bin and estimate VMT. Battery pack current (Amp) and
voltage (V) were used to estimate energy consumption in kWh. Subsequently, the VMT
and energy consumption are aggregated by speed bin. The energy consumption per
mile was obtained by taking total energy consumption and dividing it by the total
distance for a given speed bin.
Until the data cutoff date of EMFAC2021 development, staff utilized data from 23 40-
foot (Class 8) battery electric transit buses in five transit fleets, including San Joaquin
Regional Transit District (RTD), Fresno, Visalia, Modesto, and City of Porterville. Due to
the poor data quality in the initial days of data collection, data collected in 2019 was
used for the former four transit agencies (15 vehicles), and data collected from April 2020
to June 2020 was used for City of Porterville (8 vehicles). To obtain energy consumption
during bus operation, charging events were flagged in the dataset and removed during
the analysis. Charging events were identified as a continuous segment of vehicle activity
that meets three criteria: (1) opposite current to energy consumption, (2) vehicle speed
less than 1 mph, and (3) the total duration of more than 1 minute. Also, idling events at 0
mph, which showed high energy consumption, were excluded. Energy consumption was
averaged for 23 buses by speed bin for all transit fleets. The results are shown below in
Figure 3.2.2-1.
Figure 3.2.2-1. Fitted Curves of Heavy Heavy-Duty Vehicle Energy Consumption by
Speed Bins

6
5
4 y = 15.906x-0.763
kWh/mile

3
2 y = 0.0004x2 - 0.0342x + 2.1773

1
0
0 5 10 15 20 25 30 35 40 45 50 55 60 65
Speed bin

32
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

The aggregated results were used to develop a base energy consumption at 20 mph.
which was then scaled to other speed bins using speed correction factors relative to 20
mph. Considering similar weights, the energy consumption is assumed to be the same
for transit bus and heavy heavy-duty vehicles (i.e. Class 8). To scale the energy
consumption from transit bus to medium heavy-duty categories (e.g., medium heavy-
duty trucks, school buses), the ratio (0.59) of CO2 emission rates between diesel medium
heavy-duty and heavy heavy-duty in EMFAC was used. Staff also assumed an energy
consumption rate of 0.56 kWh/mi for Class 2b-3 vehicles based on information from San
Diego Airport Parking Company who used Class 3 Zenith Motors vans17 in their fleet.
Energy consumptions per mile were adjusted for 15% energy loss from grid to vehicle
battery 17. EMFAC2021 energy consumption inputs for heavy-duty categories are shown
in Table 3.2.2-1. The assumptions align with National Renewable Energy Laboratory
(NREL) measured energy use for Class 2b-8 vehicles. 18
Note that transit bus experiences more stop-and-go than other duty cycles, yet due to
lack of data, the same set of speed correction factors were used for all the heavy-duty
categories to show the energy consumption variation by speed, as shown in Table 3.2.2-
2, derived from fitted curve of energy consumption by speed in Figure 3.2.2-1. Also note
that the numbers listed in
Table 3.2.2-1 are corrected for 15% energy loss from grid to vehicle battery. Upon
availability of data, staff will continue to refine the energy consumption rates and speed
correction factors for various vehicle weight class and vocations.
Table 3.2.2-1. EMFAC2021 Input Energy Consumption Rates in Kilowatt Hours per
Mile at 20 mph (Including 15% Energy Loss from Grid to Vehicle Battery)
Energy Consumption (kWh/mile) @
Category
20 mph
Light Heavy-Duty Trucks 0.66
Medium Heavy-Duty Trucks 1.45
Heavy Heavy-Duty Trucks 2.07
School Bus/All Other Buses/Other Bus 1.45
Urban Buses 2.07

17
California Air Resources Board, Appendix G Battery Electric Truck and Bus Energy Efficiency Compared
to Conventional Diesel Vehicles, Staff Report of the Proposed Advanced Clean Trucks Regulation (web
link: https://ww3.arb.ca.gov/regact/2019/act2019/appg.pdf, posted October, 2019)
18
Oak Ridge National Laboratory (ORNL) and National Renewable Energy Laboratory (NREL), Medium-
and Heavy-Duty Vehicle Electrification: Assessment of Technology and Knowledge Gaps (web link:
https://info.ornl.gov/sites/publications/Files/Pub136575.pdf, December 2019)
33
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Table 3.2.2-2. EMFAC2021 Speed Correction Factors for Heavy-Duty Energy

Consumption Rates
Speed Speed
Speed Correction
Speed Bin Speed Bin Correction Speed Bin Correction
Factor
Factor Factor
5 2.85 35 0.75 65 0.91
10 1.35 40 0.79
15 1.13 45 0.84
20 1.00 50 0.84
25 0.92 55 0.87
30 0.74 60 0.86

3.3 Ammonia Module

3.3.1 Background
The EMFAC model currently does not estimate ammonia emissions from light and heavy-
duty vehicles in California. In the past, such an estimate was provided using Metropolitan
Planning Organization (MPO) vehicle miles travelled (VMT) combined with the latest
available emission factors for ammonia. The resulting emissions inventory was then
uploaded to the CEPAM database to support the State Implementation Plans (SIP). Table
3.3.1 1 and Table 3.3.1 2 detail the historical emission rates used in the 2016 State SIP
strategy.
Table 3.3.1-1. Emission factors used in the 2016 SIP inventory for TWC light duty
vehicles
Vehicle Category Technology Type EF (mg/mi)
Tier0 69
Tier1 69
TLEV 76
Light Duty Vehicles (TWC) LEV 50
ULEV 20
SULEV 20
LEV II 21

Table 3.3.1-2. Emission factors used in the 2016 SIP inventory for other vehicle
categories and technologies
Vehicle Category Fuel and Catalyst Type EF (mg/mi)
Gasoline non-catalyst 5
Light Duty Vehicle Gasoline oxidation catalyst 15
Diesel 3.1

34
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Vehicle Category Fuel and Catalyst Type EF (mg/mi)

Gasoline non-catalyst 5
Gasoline oxidation catalyst 15
Heavy Duty Vehicle
Gasoline three-way catalyst 45
Diesel 27
Motorcycles Gasoline non-catalyst 5
Motorcycles Gasoline Ox cat and three-way catalyst 11

3.3.2 Data Sources

In addition to the historical data sets, staff was able to utilize NH3 data from several
recent studies. These are summarized in Table 3.3.2-1 with the new data highlighted in
bold.
Table 3.3.2-1. Final Emission Factors for EMFAC2017 NH3 Estimates and Data
Sources

Fuel Vehicle Class Model Year EF (mg/mi) Data Source

1965-1975 5
1975-1979 15 Historical data
1980-1983 50
Dynamometer studies at UC Riverside and
Light and 1984-1997 70
UCLA
Medium Duty
Vehicles Caldecott tunnel study by UC Berkeley
1998-2003 45
published in 2009
Dynamometer studies at UC Riverside and
2004-2015 20
Gasoline UCLA
2016+ 42 CARB Light Duty Surveillance Program
pre-77 5
Historical data
Heavy Duty 1977-1983 15
Vehicles Caldecott tunnel study by UC Berkeley
1984+ 45
published in 2009
1965-1994 5
Motorcycles 1995-2007 6.4
2008+ 9.2 Historical
Light and
Medium Duty All 3.1
Vehicles
Diesel 2011+ 220 CARB Truck & Bus Surveillance Program
Heavy Duty 2007-2010 38 SCAQMD Heavy Duty Truck Emission Tests
Vehicles
1965-2006 27 Historical
CNG Refuse All 580 200-Vehicle Project

35
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Fuel Vehicle Class Model Year EF (mg/mi) Data Source

Transit All 970

Other All 1060

The major data sources that are used to update ammonia emission rates in EMFAC2021
include:
CARB Light Duty Surveillance Program: MY 2016+ light-duty gasoline updated to 42
mg/mile based upon emissions data from ten vehicles tested under CARB’s light duty
surveillance program. Table 3.3.2-2 summarizes these data.

Table 3.3.2-2. Updated Light Duty Emission Factors (CARB Light Duty Surveillance
Program)
Test Distance NH3 NH3
MY Vehicle Test ID
(phase) (mi) (mg) (mg/mi)
1061931 1 1.2 58.68 48.9
Chevrolet
2018 1061931 2 8.6 151.4 17.6
Tahoe
1061931 3 1.2 73.44 61.2
1061976 1 1.2 110.2 91.8
Toyota
2017 1061976 2 8.6 103.2 12.0
Camry
1061976 3 1.2 27.12 22.6
1061975 1 1.2 150.1 125.1
2016 Dodge Dart 1061975 2 8.6 556.4 64.7
1061975 3 1.2 76.68 63.9
1062115 1 1.2 36.6 30.5
Jeep
2018 1062115 2 8.6 243.4 28.3
Wrangler
1062115 3 1.2 42.12 35.1
1062116 1 1.2 61.44 51.2
Hyundai
2018 1062116 2 8.6 61.06 7.1
Elantra
1062116 3 1.2 0 0.0
1062226 1 1.2 134.9 112.4
2018 Nissan Altima 1062226 2 8.6 411.1 47.8
1062226 3 1.2 197.6 164.7
1062227 1 1.2 90.72 75.6
Mercedes
2017 1062227 2 8.6 319.9 37.2
C300
1062227 3 1.2 22.68 18.9
1062352 1 1.2 12.72 10.6
2017 Mazda 3 1062352 2 8.6 105.8 12.3
1062352 3 1.2 14.4 12.0
2018 Ford Fusion 1062353 1 1.2 287 239.2

36
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Test Distance NH3 NH3

MY Vehicle Test ID
(phase) (mi) (mg) (mg/mi)
1062353 2 8.6 756.8 88.0
1062353 3 1.2 139.7 116.4
1062354 1 1.2 122.5 102.1
2018 BMW 430I 1062354 2 8.6 200.4 23.3
1062354 3 1.2 45 37.5
Average 110 4613.0 41.9

SCAQMD Heavy Duty Truck Emission Tests: MY 2007-2010 heavy-duty diesel vehicles
are updated to 38 mg/mile (UDDS) based upon three vehicles from the South Coast Air
Quality Management District (SCAQMD) testing. These are noted in Table 3.3.2-3.
Table 3.3.2-3. Pre-SCR Heavy-Duty Diesel Ammonia Emission Rates (SCAQMD Heavy
Duty Truck Emission Tests)
Model
Vehicle ID Manufacturer Main Control Cycle NH3 (g/mi)
Year
N12.3 2009 Navistar Adv. EGR UDDS 0.013
D14a 2008 DDC DOC/DPF UDDS 0.036
D14b 2008 DDC DOC/DPF UDDS 0.065

CARB Truck and Bus Surveillance Program (TBSP): MY 2011+ technology heavy-duty
diesel vehicles are updated to 220 mg/mile (UDDS) based upon 21 vehicles from the
CARB Truck and Bus Surveillance Program (TBSP) testing program. A summary of these
data is given in Table 3.3.2-4.
Table 3.3.2-4. Updated Heavy Duty Emission Factors (MY 2011+) (CARB TBSP
Testing)
Vehicle MY Manufacturer Test Repeats NH3 (g/mi)
18-Veh2 2015 Cummins UDDS 3 0.28
18-Veh2 2015 Cummins UDDS 3 1.25
18-Veh3 2014 Detroit Diesel UDDS 3 0.13
18-Veh4 2015 Paccar UDDS 3 0.07
19-Veh10 2016 Detroit Diesel UDDS 3 0.00
19-Veh11 2017 Detroit Diesel UDDS 3 0.098
19-Veh15 2019 Cummins UDDS 3 0.980
19-Veh8 2016 Cummins UDDS 3 0.00
19-Veh8 2016 Cummins UDDS 3 0.024
19-Veh9 2018 Volvo UDDS 3 0.00
20-Veh16 2017 Cummins UDDS 3 0.008
2S19H01-4 2015 Cummins UDDS 3 0.007
TBSP A 2010 Cummins ISB6.7 UDDS 3 0.01
TBSP A 2010 Cummins ISB6.7 UDDS 3 0.06

37
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Vehicle MY Manufacturer Test Repeats NH3 (g/mi)

TBSP A 2010 Cummins ISB6.7 UDDS 3 0.00
TBSP B 2010 Cummins ISX15 UDDS 3 0.54
TBSP B 2010 Cummins ISX15 UDDS 3 0.48
TBSP C 2010 Detroit Diesel DD15 UDDS 3 0.40
TBSP C 2010 Detroit Diesel UDDS 3 0.14
TBSP F 2011 Cummins ISX15 UDDS 3 0.46
TBSP F 2011 Cummins ISX15 UDDS 3 0.30
TBSP I 2012 Cummins ISX15 UDDS 3 0.13
TBSP J 2013 Isuzu 4HKITC UDDS 3 0.06
TBSP L 2013 Cummins ISX15 UDDS 3 0.01
TBSP L 2013 Cummins ISX15 UDDS 3 0.04
TBSP M 2014 Volvo D13 UDDS 3 1.10
TBSP M 2014 Volvo UDDS 3 0.01
TBSP N 2014 Detroit Diesel DD15 UDDS 3 0.14
TBSP N 2014 Detroit Diesel DD15 UDDS 3 0.03
TBSP O 2014 Cummins ISX15 UDDS 3 0.13
TBSP P 2014 Navistar UDDS 3 0.31

200-Vehicle Project: Natural Gas (NG) data were obtained from an SCAQMD test
program. These results are summarized Table 3.3.2-5.
Table 3.3.2-5. CNG Ammonia Emission Rates (200-Vehicle Project)
Engine Vocation Fuel Cycle Distance NH3 (g) NH3 (g/mi)
8062 Other CNG UDDS 5.4 1.513 0.280
8062 Other CNG UDDS 5.4 2.575 0.477
8062 Other CNG UDDS 5.4 3.123 0.578
18082 Other CNG UDDS 4.8 3.101 0.646
18082 Other CNG UDDS 5 3.858 0.772
18082 Other CNG UDDS 5 3.429 0.686
18045 Other CNG UDDS 5.4 8.307 1.538
18045 Other CNG UDDS 5.4 8.435 1.562
18045 Other CNG UDDS 5.4 8.302 1.537
8044 Other CNG UDDS 5.668 4.722 0.833
8044 Other CNG UDDS 5.424 15.53 2.864
8044 Other CNG UDDS 5.488 5.231 0.953
18023 Refuse CNG UDDS 5.5 2.85 0.518
18023 Refuse CNG UDDS 5.5 2.597 0.472
18023 Refuse CNG UDDS 5.4 2.913 0.539
8013 Refuse CNG UDDS 5.5 3.053 0.555
8013 Refuse CNG UDDS 5.5 3.606 0.656
8013 Refuse CNG UDDS 5.5 4.013 0.730

38
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Engine Vocation Fuel Cycle Distance NH3 (g) NH3 (g/mi)

18112 Transit CNG UDDS 5.6 4.379 0.782
18112 Transit CNG UDDS 5.6 6.451 1.152
18112 Transit CNG UDDS 5.5 5.388 0.980

3.3.3 Results and Conclusions

Comparing Table 3.3.1-1 and Table 3.3.1-2, it is apparent that newer light-duty vehicles
have higher emissions than those of the late 1990s and 2000s. It is not obvious whether
this is a true increase from the catalyst configurations, or if it is simply differences in how
the test programs were conducted. It shows that the newest SCR equipped diesels have
significantly higher emissions than older diesels. Historically, emission rates from natural
gas were treated the same as diesel, due to lack of data. However, the new testing data
shows that they present substantially higher ammonia emissions than diesel.

39
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

4 Methodology Updates
4.1 Introduction
This chapter discusses the updates that have taken place in EMFAC2021. It can be
broken up into four broad categories, including: fleet characteristics (Section 4.2),
emission rate (Section 4.3), activity profile (Section 4.4), and forecasting frameworks
(Section 4.5). Update to fleet characteristics include the methodology used in developing
the LD and HD vehicle population and age distribution matrices used in EMFAC2021.
Emission rate updates not only include changes in base emission rates, but also changes
to any associated correction factors. For example, emission rates changes by speed and
deteriorate as vehicle aging. Emission rates updates have been made mainly for exhaust
emission process. In EMFAC2021, the particulate matter (PM) emission rates from brake
wear (non-exhaust emission process) are also updated.
Activity profile updates are made to mileage accrual rates, odometer schedule, speed
profile, starts per day, soak time distribution, temporal distribution of VMT and engine
starts. New forecasting frameworks are adapted in EMFAC2021 to project vehicle new
sales and vehicle miles traveled (VMT) for both LDV and HDV.

4.2 Fleet Characteristics

4.2.1 Light Duty Fleet Characterization
This section describes the major updates to EMFAC2021 fleet characterization and
describes changes to the methodology, tools, and data sources used to characterize the
vehicle population in California. It also compares the fleet vehicle counts as modeled by
EMFAC2017 to that in EMFAC2021.
Starting in January 2018, DMV shared quarterly cuts of vehicle registration data with
CARB. The data cuts, containing approximately 53 million records and 100 data fields,
are available in January, April, July and October of each calendar year. EMFAC2021 uses
the October data cut as the main source of data for fleet characterization, but
incorporates vehicle registration status from the following April data cut to update any
vehicles with pending registration. The fleet characterization begins with loading the raw
DMV data into a secured database, and then removing the duplicate records to only
keep the last record associated with each vehicle identification number (VIN). All newly
acquired VINs are run through VINtelligence 19 to obtain vehicle related information that
may be missing from DMV (i.e., gross vehicle weight code, model year, make name,
series name, model name, body style, motive power, fuel type, displacement, battery

19
A VIN decoder developed by IHS Markit. For more information please see:
https://ihsmarkit.com/products/automotive-vin-interpretation-decoding.html
40
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

size, etc.). Only on-road vehicles in the October database are analyzed and assigned a
vehicle classification. Vehicles are classified based on manufacturer certification EOs
issued for each vehicle make, model, and model year. Finally, each record is distributed
to a geographic area index (GAI) based on the registered owner address and used in the
population numbers for EMFAC.
Figure 4.2.1-1 compares EMFAC2021 and EMFAC2017 vehicle populations for gasoline
passenger cars (PC). Note that EMFAC2017 projected vehicle population in 2017 and
onwards. As shown, EMFAC2021 has slowly declining vehicle populations of less than 1%
from 2017 to 2019 compared to the previous model.
Figure 4.2.1-1. Comparison between EMFAC2021 and EMFAC2017 PC gasoline
population

16,000,000
14,000,000
12,000,000
Population

10,000,000
8,000,000
6,000,000
4,000,000 EMFAC2021
2,000,000 EMFAC2017
0
2012 2013 2014 2015 2016 2017 2018 2019
Calendar Year

Similarly, Figure 4.2.1-2 displays lower counts for gasoline light-duty trucks (LDT) than
predicted in EMFAC2017. EMFAC2021 does show continuous growth for the light-duty
truck population, and the difference between the populations in both models is less than
3% in 2019.

41
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure 4.2.1-2. Comparison between EMFAC2021 and EMFAC2017 LDT gasoline

population
12,000,000

10,000,000

8,000,000
Population

6,000,000

4,000,000
EMFAC2021
2,000,000 EMFAC2017
0
2012 2013 2014 2015 2016 2017 2018 2019
Calendar Year

Figure 4.2.1-3 illustrates gasoline light heavy-duty truck (LHDT) populations obtained for
EMFAC2021 and EMFAC2017. For this vehicle class, EMFAC2017 predicted a
continuous decline in the population while actual DMV counts used in EMFAC2021
indicate substantial growth. The difference in populations show that EMFAC2017 under
predicted the counts by nearly 18% for 2019.
Figure 4.2.1-3. Comparison between EMFAC2021 and EMFAC2017 LHDT gasoline
population
700,000
600,000
500,000
Population

400,000
300,000
EMFAC2021
200,000
100,000 EMFAC2017

0
2012 2013 2014 2015 2016 2017 2018 2019
Calendar Year

Figure 4.2.1-4 indicates a sharp drop in the number of diesel passenger cars.
EMFAC2017 predicted relatively moderate growth for this category, however,
EMFAC2021 counts have dropped to less than 50% of the predicted population.

42
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure 4.2.1-4. Comparison between EMFAC2021 and EMFAC2017 PC diesel

population

140,000
120,000
100,000
Population

80,000
60,000
40,000
EMFAC2021
20,000 EMFAC2017
0
2012 2013 2014 2015 2016 2017 2018 2019
Calendar Year

As can be seen in Figure 4.2.1-5, EMFAC2017 predicted steady growth for diesel light
duty trucks (LDT). EMFAC2021 shows the growth to be significantly slower, and the
difference between the two models ranges from about 10-15% lower for calendar years
2017 to 2019.
Figure 4.2.1-5. Comparison between EMFAC2021 and EMFAC2017 LDT diesel
population
120,000

100,000

80,000
Population

60,000

40,000 EMFAC2021

20,000 EMFAC2017

0
2012 2013 2014 2015 2016 2017 2018 2019
Calendar Year

For diesel LHDT vehicles in Figure 4.2.1-6, there is close agreement between the
estimated populations from EMFAC2017 and DMV registration counts reflected in
EMFAC2021. EMFAC2021 populations have surpassed forecasted EMFAC2017
populations for this category.

43
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure 4.2.1-6. Comparison between EMFAC2021 and EMFAC2017 LHDT diesel

population
540,000
520,000
500,000
Population

480,000
460,000
440,000
EMFAC2021
420,000
EMFAC2017
400,000
380,000
2012 2013 2014 2015 2016 2017 2018 2019
Calendar Year

Figure 4.2.1-7 illustrates that the EMFAC2021 electric passenger car (PC) population (or
electric equivalent vehicles that have the motive power of electric combined with the
electric fraction of PHEVs) is substantially higher than the forecasted population of
EMFAC2017. Since EMFAC2017 projected a slower growth rate for the population
compared to EMFAC2021, the difference between the two models grows to nearly
double the earlier number predicted for 2019. The electric PC population is expected to
reach 400,000 vehicles by calendar year 2020.
Figure 4.2.1-7. Comparison between EMFAC2021 and EMFAC2017 PC electric
vehicle population
400,000
350,000 EMFAC2021
300,000 EMFAC2017
Population

250,000
200,000
150,000
100,000
50,000
0
2012 2013 2014 2015 2016 2017 2018 2019
Calendar Year

According to Figure 4.2.1-8, electric LDTs have not experienced as rapid growth as their
PC counterparts. As in the previous figure, these counts are for electric equivalent

44
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

vehicles. While electric LDT populations grow in EMFAC2021, the increase for these
alternative fuel vehicles is flatter than the rate predicted by EMFAC2017.
Figure 4.2.1-8. Comparison between EMFAC2021 and EMFAC2017 LDT electric
vehicle population
35,000
30,000 EMFAC2021

25,000 EMFAC2017
Population

20,000
15,000
10,000
5,000
0
2012 2013 2014 2015 2016 2017 2018 2019
Calendar Year
Figure 4.2.1-9 illustrate sales trends for new gasoline PCs and LDTs sold in California.
While EMFAC2021 new vehicle sales projections for LDTs are consistent with those in
EMFAC2017, there is a continuous decline in new vehicle sales for PCs. For calendar year
2019, the new vehicle sales counts for PC have dropped by over 400,000 vehicles.
Figure 4.2.1-9. Comparison between EMFAC2021 and EMFAC2017 New Sales by
Model Year for Gasoline PC and LDT
1,200,000

1,000,000

800,000
Population

600,000

400,000
PC EMFAC2021 PC EMFAC2017
200,000
LDT EMFAC2021 LDT EMFAC2017
0
2012 2013 2014 2015 2016 2017 2018 2019
Calendar Year

Figure 4.2.1-10 displays the diesel light-duty vehicle market sales. Both diesel PCs and
LDTs have experienced major declines in recent years due in part to emissions
irregularities found by CARB and EPA. EMFAC2017 predicted modest sales increases for
45
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

years 2017 to 2019, but EMFAC2021 shows that new sales have fallen short for the same
period.
Figure 4.2.1-10. Comparison between EMFAC2021 and EMFAC2017 New Sales by
Model Year for Diesel PC and LDT
20,000 PC EMFAC2021 PC EMFAC2017
LDT EMFAC2021 LDT EMFAC2017
15,000
Population

10,000

5,000

0
2012 2013 2014 2015 2016 2017 2018 2019
Calendar Year

Figure 4.2.1-11 through Figure 4.2.1-13 show the distribution of vehicle population by
model year for PCs, LDTs, and LHDTs for EMFAC2021 and EMFAC2017. The PC vehicle
count has dropped more significantly than predicted by EMFAC2017. For LDTs and
LHDTs, the reverse is true as EMFAC2021 has higher vehicle population for model years
2017 and 2018 when compared to the projected EMFAC2017 vehicle counts. Both
models, however, are consistent for model year 2019 LDT and LHDT vehicle populations.
Figure 4.2.1-11. Comparison between EMFAC2021 and EMFAC2017 Model Year
Distribution for PC
1,400,000

1,200,000

1,000,000
EMFAC2021
Population

800,000
EMFAC2017
600,000

400,000

200,000

0
1975 1985 1995 2005 2015
Model Year

46
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure 4.2.1-12. Comparison between EMFAC2021 and EMFAC2017 Model Year

Distribution for LDT
1,000,000
900,000
800,000 EMFAC2021
700,000
EMFAC2017
Population

600,000
500,000
400,000
300,000
200,000
100,000
0
1970 1980 1990 2000 2010 2020
Model Year

Figure 4.2.1-13. Comparison between EMFAC2021 and EMFAC2017 Model Year

Distribution for LHDT

100,000
90,000
80,000
70,000 EMFAC2021
Population

60,000
EMFAC2017
50,000
40,000
30,000
20,000
10,000
0
1970 1980 1990 2000 2010 2020
Model Year

47
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

4.2.2 Heavy Duty Fleet Characterization

4.2.2.1 Introduction about HD fleet
This section focuses on the EMFAC2021 population trends for diesel and natural gas
fueled medium and heavy heavy-duty (MHD and HHD) trucks and buses operating in
California. Medium heavy-duty trucks have a gross vehicle weight rating (GVWR) of
14,001 to 33,000 pounds. Heavy heavy-duty trucks have a GVWR greater than 33,000
pounds. Bus fleet types include school buses, transit buses, motor coaches and other
buses. The following sections show comparisons of the population counts, new vehicle
sales and age distributions for EMFAC2021 and EMFAC2017. EMFAC2017 had a base-
year of 2016 while EMFAC2021 has a base-year of 2019.
4.2.2.2 Major Data Sources for Update
Major data sources used to process the heavy-duty vehicle inventory are as follows.
Processed DMV data. As discussed in the light duty vehicle section, the DMV data sets
for historical years through 2019 were processed using additional inputs from various
data sources to provide updated vehicle information for vehicles registered in California.
DMV data field values are used to designate utility and public fleet vehicles, and to
identify tractors and solid waste collection vehicles. After identifying all other fleet types
using all of the various data sources, the remaining trucks are designated as instate
single trucks and the remaining buses are designated as all other buses.
International Registration Plan (IRP) Data. IRP Clearinghouse data is another primary
data source for historical heavy-duty vehicle updates through 2019. Vehicles already
registered in California can be identified as interstate trucks (CA IRP fleet) or buses
(motor coach fleet). In addition, for out-of-state vehicles in states and provinces that
report to the IRP Clearinghouse, updates were made using vehicle characteristics for
fleets that travel to California. Out-of-state fleets report into IRP their annual mileage to
California at a fleet level, and not per individual vehicle. Since out-of-state fleets may
send many or none of their fleet’s individual trucks to travel into California, it is more
important to estimate their VMT travel in California than to estimate counts of unique
out-of-state vehicles, which cannot be determined accurately. Using calendar years 2008
through 2018 International Fuel Tax Agreement (IFTA) mileage data, the historical ratio
of VMT for out-of-state trucks as compared to VMT by CA IRP trucks was updated to
1.206 for T7 Non-Neighboring Out-of-state truck (NNOOS) and to 0.374 for T7
Neighboring Out-of-state truck (NOOS) for EMFAC2021. These updates were made
under the assumption that HHDT vehicles represented 95% of all the reported VMT and
MHDT vehicles represented 5% (based on past studies). Using these ratios, VMT was
calculated for T6 OOS, T7 NNOOS and T7 NOOS categories of EMFAC2021 and their
populations were back-calculated with the use of mileage accrual schedules.

48
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

TRUCRS 20 data for diesel Truck and Bus Rule. Data extracted from the TRUCRS
database were used to update the heavy-duty inventory as needed for fleets utilizing
flexible compliance options to meet Truck and Bus Rule requirements.
Drayage Trucks Operating at Major Ports. The Port of Los Angeles/Long Beach (POLA)
and the Port of Oakland (POAK) provided lists of VINs for vehicles that actually visited
the ports to directly flag Class 8 vehicles used as port trucks. POAK provided annual lists
with VIN and license plate data, which was used to update the CY2017 to CY2019
inventory. In 2019, POLA provided updated VIN lists for CY2010 to CY2019 that included
details on the monthly trips per VIN and the monthly average number of trucks to use for
EMFAC modeling to update the inventory for all of these years. The selected VINs to flag
as POLA each year were based on having annual trips to the port above a certain
threshold. Thresholds were back calculated for each year such that the level of annual
trips used to identify which VINs to flag as POLA resulted in achieving the monthly
average truck counts provided by POLA. Trucks not displayed as POAK or POLA in
EMFAC may be in lower weight classes, have fuel types other than diesel/electric/natural
gas, be considered inactive trucks as annual visits are too low, or have out-of-state
registration status. These trucks are included in other EMFAC categories for vehicles
activity and emissions calculation purposes.
California Highway Patrol (CHP) School Bus Inspections. 21 The CHP provided data on
School Buses that receive safety inspections that are required by law.
4.2.2.3 New HD Vocational Categories
The EMFAC2011 heavy-duty fleet categories used in EMFAC2014 and EMFAC2017
reflected MHD truck (T6) and HHD truck (T7) groupings developed to meet regulatory
needs such as the Diesel Truck and Bus rule. For EMFAC2021, new fleet category
updates were desirable to allow for more specificity in EMFAC activity and emission rates
updates to better support future policies.
New fleet groups will assist with more focused emissions reduction programs such as
freight hubs, last mile delivery, and local communities. Due to difficulties identifying
construction fleet categories, they are no longer displayed as separate fleet groups. For
example, ‘T6 construction heavy’ will now be included with ‘T6 heavy’, ‘T7 construction
tractor’ will now be included with ‘T7 tractor’, and so on. Similarly, agricultural fleet
categories which displayed vehicles claiming an approved TRUCRS exemption are now
absorbed back into appropriate fleet categories such as T6 or T7 tractors, and T6 instate
or T7 Single categories.

20
https://www.arb.ca.gov/msprog/onrdiesel/reportinginfo.htm
21
California Highway Patrol (CHP), School Bus Program (https://www.chp.ca.gov/Programs-
Services/Programs/School-Bus-Program)
49
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

To provide further details on various weight categories, instate T6 trucks groupings have
been disaggregated into weight classes 4 through 7. Exceptions include classes 4-5 for
tractors and class 4 for utility trucks, both of which did not have sufficient counts for
disaggregation so those small counts were rolled up into the next weight class above.
EMFAC2021 include new subgroup updates for Instate T6 trucks. The new subgroups
are based on DMV body type for tractors and for delivery trucks. The remaining instate
T6 trucks reside in the T6 other category. The following Table 4.2.2-1 displays the
updated EMFAC2021 T6 fleet categories along with the EMFAC2017 fleet categories
from which they originate.
Table 4.2.2-1. EMFAC2021 T6 (14,001-33,000 lbs.) Fleet Groups

Type EMFAC2017 EMFAC2021

T6 CAIRP small T6 CAIRP Class 4
T6 CAIRP small T6 CAIRP Class 5
T6 CAIRP small T6 CAIRP Class 6
T6 CAIRP heavy T6 CAIRP Class 7
T6 Instate Small T6 Instate Delivery Class 4
T6 Instate Small T6 Instate Delivery Class 5
T6 Instate Small T6 Instate Delivery Class 6
T6 Instate Heavy T6 Instate Delivery Class 7
T6 Instate Small T6 Instate Other Class 4
T6 Instate Small T6 Instate Other Class 5
T6 Instate Small T6 Instate Other Class 6
T6 Instate Heavy T6 Instate Other Class 7
T6 Truck
(14,001-33,000 lbs.) T6 Instate Small T6 Instate Tractor Class 6
T6 Instate Heavy T6 Instate Tractor Class 7
T6 OOS Small T6 OOS Class 4
T6 OOS Small T6 OOS Class 5
T6 OOS Small T6 OOS Class 6
T6 OOS Heavy T6 OOS Class 7
T6 Public T6 Public Class 4
T6 Public T6 Public Class 5
T6 Public T6 Public Class 6
T6 Public T6 Public Class 7
T6 utility T6 Utility Class 5
T6 utility T6 Utility Class 6
T6 utility T6 Utility Class 7

T7 single trucks have been split into new sub-groups based on DMV body type for
concrete/transit mix trucks and for dump trucks. The remaining T7 single trucks reside in
50
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

the T7 other category. The following Table 4.2.2-2 displays the updated EMFAC2021 T7
fleet categories along with the EMFAC2017 fleet categories from which they originated.
Table 4.2.2-2. EMFAC2021 T7 (>33,000 lbs.) Fleet Groups

Type EMFAC2017 EMFAC2021

T7 CAIRP T7 CAIRP Class 8
T7 NNOOS T7 NNOOS Class 8
T7 NOOS T7 NOOS Class 8
T7 other port T7 Other Port Class 8
T7 POAK T7 POAK Class 8
T7 POLA T7 POLA Class 8
T7 Truck (>33,000
lbs.) T7 Public T7 Public Class 8
T7 single T7 Single Concrete/Transit Mix Class 8
T7 single T7 Single Dump Class 8
T7 single T7 Single Other Class 8
T7 SWCV T7 SWCV Class 8
T7 Tractor T7 Tractor Class 8
T7 utility T7 Utility Class 8

4.2.2.4 In-State Population

The following Figure 4.2.2-1 compares EMFAC2021 and EMFAC2017 vehicle population
for heavy-duty instate trucks. Included trucks in the figure are trucks that operate within
California, and those that are diesel and natural gas. Please note that estimates from
EMFAC2021 are based on the DMV vehicle registration data while EMFAC2017
estimates for years 2017 and onward are projected using forecasting method described
in EMFAC2014 technical support documentation. As shown below, EMFAC2021 has a
higher actual vehicle population than was forecasted by EMFAC2017 for the calendar
years of 2017 and 2018 while it shows lower population in 2019 when compared to
EMFAC2017 projections.

51
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure 4.2.2-1. Comparison between EMFAC2021 and EMFAC2017 Instate Heavy-

Duty vehicle population

EMFAC2021 Dsl & NG - Actual

EMFAC2017 Dsl & NG - Forecasted
500,000

400,000

300,000
Population

200,000

100,000

0
2017 2018 2019
Calendar Year

The following Figure 4.2.2-2 compares EMFAC2021 and EMFAC2017 new sales for
heavy-duty instate trucks. New sales include all vehicles with chassis model years equal to
or greater than the calendar year. For calendar years 2017 to 2019, the new vehicles
sales exceeded the EMFAC2017 forecasts with an increase of 2.5% in 2017, 30.3% in
2018, and 44.5% in 2019.

52
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure 4.2.2-2. Comparison between EMFAC2021 and EMFAC2017 Instate Heavy-

Duty New Vehicle Sales

35,000
EMFAC2021 Dsl & NG - Actual
30,000 EMFAC2017 Dsl & NG - Forecasted

25,000
Population

20,000

15,000

10,000

5,000

0
2017 2018 2019
Calendar Year

The following Figure 4.2.2-3 compares EMFAC2021 and EMFAC2017 counts of vehicles
with a chassis model year of 2011 and greater which would be compliant with the Truck
and Bus Rule model year 2010 engine standard requirements. For the majority of heavy-
duty trucks, there is typically a one-year lag in the chassis model year from the engine
model year. These population counts would include both new and used vehicle sales.
The EMFAC2017 forecasted population exceeded the updated population for
EMFAC2021. In calendar year 2019, EMFAC2021 showed a 10.4% decrease from the
EMFAC2017 projection. EMFAC2017 Truck and Bus Rule assumptions anticipated a
higher rate of model year 2010 engine compliant used vehicle sales. However, fleets
have various options for achieving compliance. For example, purchasing a model year
2007 engine standard vehicle with an original equipment manufacturer (OEM) particulate
filter which does not need to be replaced with a 2010 engine standard vehicle until
calendar year 2023. The penetration rate for chassis model year 2011+ vehicles has been
increasing over time but with a slight offset compared to what was forecasted in
EMFAC2017.

53
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure 4.2.2-3. Comparison between EMFAC2021 and EMFAC2017 MY2011+ Instate

Heavy-Duty Vehicle Counts

300,000
EMFAC2021 Dsl & NG - Actual
EMFAC2017 Dsl & NG - Forecasted
250,000

200,000
Population

150,000

100,000

50,000

0
2017 2018 2019
Calendar Year

For the new base-year of 2019, EMFAC2021 reflects an average age of 10.9 for instate
heavy-duty (MHD and HHD) vehicles. EMFAC2017 used calendar year 2016 as the base
year with an average age of 11.2. The following Figure 4.2.2-4 provides a comparison of
the age distributions for the 2019 calendar year base year in EMFAC2021 with
EMFAC2017 values. As the chart indicates, the EMFAC2021 base year shows decreases
in the pre-2011 model years and increases in the 2011+ model years (from both new and
used vehicle purchases) as compared to the EMFAC2017 base year. EMFAC2017 shows
peaks of 2012 and 2016 model years due to the Truck and Bus vehicle replacement
assumptions used. EMFAC2021 modified replacement assumptions to distribute across
more model years to reduce such peaks.

54
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure 4.2.2-4. Comparison between EMFAC2021 and EMFAC2017 Instate Heavy-

Duty Age Distributions

EMFAC2021 DSL & NG CY2019 Base Year

60,000 EMFAC2017 DSL & NG CY2019 Projected

50,000

40,000
Population

30,000

20,000

10,000

0
1977
1979
1981
1983
1985
1987
1989
1991
1993
1995
1997
1999
2001
2003
2005
2007
2009
2011
2013
2015
2017
2019
Model Year

Port trucks had to meet the drayage rule that required MY2007 or newer engines by the
beginning of calendar year 2014, which lowered the average age of this fleet group. The
Port Truck population in the calendar year 2016 had an average age of 5.24. After all the
drayage rule requirements in the past, no further vehicle replacements have been
required so the average age increased to 7.51 in calendar year 2019 in EMFAC2021. This
is slightly higher the projected average age of 7.27 for calendar year 2019 in
EMFAC2017. The following Figure 4.2.2-5 provides a comparison of the age distributions
for the 2019 calendar year in EMFAC2021 with EMFAC2017 values. Older chassis model
years would reflect engine repowers. It should be noted that Port trucks need to meet
the 2010 engine standard requirement by January 1, 2023 as required by the Truck and
Bus rule.

55
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure 4.2.2-5. Comparison between EMFAC2021 and EMFAC2017 Heavy Heavy-

Duty Port Truck Age Distribution

4,500 EMFAC2021 DSL & NG CY2019 Base Year

4,000 EMFAC2017 DSL & NG CY2019 Projected
3,500
3,000
2,500
Population

2,000
1,500
1,000
500
0
2007

2008

2009

2010

2011

2012

2013

2014

2015

2016

2017

2018

2019

2020
Model Year
For HHD instate tractors, the EMFAC2017 base year of calendar year 2016 had an
average age of 9.4 which has decreased to 8.0 in the updated EMFAC2021 base year of
calendar year 2019. This is similar to the average age of 8.8 years that EMFAC2017
projected for year 2019. A comparison of the age distributions for the 2019 calendar
year in EMFAC2021 with EMFAC2017 values for heavy heavy-duty instate tractors is
shown in Figure 4.2.2-6.

56
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure 4.2.2-6. Comparison between EMFAC2021 and EMFAC2017 Base Year Heavy
Heavy-Duty Instate Tractor Age Distribution

10,000
9,000 EMFAC2021 DSL & NG CY2019 Base Year
8,000 EMFAC2017 DSL & NG CY2019 Projected
7,000
6,000
Population

5,000
4,000
3,000
2,000
1,000
0
1976
1978
1980
1982
1984
1986
1988
1990
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
2014
2016
2018
2020
Model Year

Figure 4.2.2-7 displays a comparison of the updated 2019 calendar year base year in
EMFAC2021 with EMFAC2017 values for heavy heavy-duty instate singles. For HHD
instate singles, the EMFAC2017 calendar year 2016 base year had an average age of
12.0, which has decreased to 10.6 in the updated EMFAC2021 calendar year 2019 base
year. This is slightly lower than the 11.2 average age that EMFAC2017 projected for year
2019.

57
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure 4.2.2-7. Comparison between EMFAC2021 and EMFAC2017 Base Year Heavy
Heavy-Duty Instate Single Age Distribution

5,000
EMFAC2021 DSL & NG CY2019 Base Year
4,500
EMFAC2017 DSL & NG CY2019 Projected
4,000
3,500
3,000
Population

2,500
2,000
1,500
1,000
500
0
1976
1978
1980
1982
1984
1986
1988
1990
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
2014
2016
2018
2020
Model Year

Figure 4.2.2-8 displays a comparison of the updated 2019 calendar year base year in
EMFAC2021 with EMFAC2017 values for medium heavy-duty instate vehicles. The
EMFAC2017 base year of calendar year 2016 had an average age of 11.3, which has
decreased slightly to 11.1 in the updated EMFAC2021 base year of calendar year 2019.
This is higher than the average age of 9.8 years that EMFAC2017 projected for calendar
year 2019.

58
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure 4.2.2-8. Comparison between EMFAC2017 and EMFAC2021 Medium Heavy-

Duty Instate Age Distribution
35,000

30,000 EMFAC2021 DSL & NG CY2019 Base Year

EMFAC2017 DSL & NG CY2019 Projected
25,000

20,000
Population

15,000

10,000

5,000

0
1976
1978
1980
1982
1984
1986
1988
1990
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
2014
2016
2018
2020
Model Year

4.2.2.5 California Interstate (CAIRP) Heavy-Duty Fleet Population

The following figure compares EMFAC2021 and EMFAC2017 vehicle population for
heavy-duty Interstate trucks that report into the International Registration Plan (CAIRP)
with fleet’s base jurisdiction designated as California. These trucks are authorized to
operate within California and within other states or provinces. As shown in ehicle
population by EMFAC2017.
Figure 4.2.2-9, EMFAC2021 shows similar counts in calendar year 2018 and a 3.8%
decrease in the counts for calendar year 2019 relative to the forecasted vehicle
population by EMFAC2017.

59
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure 4.2.2-9. Comparison between EMFAC2021 and EMFAC2017 CAIRP Heavy-

Duty vehicle population

100,000 EMFAC2020 Updates EMFAC2017 Forecasted

80,000
Population

60,000

40,000

20,000

-
2017 2018 2019
Calendar Year
For the heavy heavy-duty CAIRP, the calendar year 2016 EMFAC2017 base-year had an
average age of 5.8, which has decreased to 5.4 in 2019 updated base-year for
EMFAC2021. This is a lower than the 6.1 average age that EMFAC2017 projected for
calendar year 2019. A comparison is shown in Figure 4.2.2-10.
Figure 4.2.2-10. Comparison between EMFAC2021 and EMFAC2017 Heavy Heavy-
Duty CAIRP Age Distribution

25,000
EMFAC2021 DSL & NG CY2019 Base Year
20,000 EMFAC2017 DSL & NG CY2019 Projected
Population

15,000

10,000

5,000

0
1976
1978
1980
1982
1984
1986
1988
1990
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
2014
2016
2018
2020

Model Year

60
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

4.2.2.6 Bus Fleet Population

Bus fleet types presented in this section include school buses, transit buses, motor
coaches and other buses. Urban transit bus inventory has a separate module that began
with EMFAC2017 and is discussed separately. Figure 4.2.1-11 displays the updated
heavy-duty bus population for EMFAC2021 new base-year of calendar year 2019 with a
comparison to EMFAC2017 values.
Figure 4.2.2-11.Comparison between EMFAC2017 and EMFAC2021 Heavy-Duty Bus
vehicle population (excluding Urban Transit Buses)

40,000 EMFAC2021 Dsl & NG - Actual

EMFAC2017 Dsl & NG - Forecasted
35,000
30,000
25,000
Population

20,000
15,000
10,000
5,000
0
2017 2018 2019
Calendar Year
EMFAC2021 continues to make use of California Highway Patrol inspection reports to
flag vehicles in DMV as school buses. The flagged DMV school bus population is then
scaled up to reflect the CHP list of total vehicle counts. The scaling process might have
resulted in some buses that are less than 14,000 pounds being presented as heavy-duty
buses. Figure 4.2.2-12 reflects a positive trend in EMFAC2021 2019 with decreased
counts of pre-2011 model years and increased counts of MY2011+ buses.

61
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure 4.2.2-12. Comparison between EMFAC2021 and EMFAC2017 Heavy-Duty Bus

Age Distribution

2,500 EMFAC2020 DSL & NG CY2019 Base Year

EMFAC2017 DSL & NG CY2019 Projected
2,000

1,500
Population

1,000

500

0
1977
1979
1981
1983
1985
1987
1989
1991
1993
1995
1997
1999
2001
2003
2005
2007
2009
2011
2013
2015
2017
2019
Model Year

4.2.2.7 Updates for Diesel In-Use Fleet Rules

EMFAC2014 and EMFAC2017 incorporated regulatory changes for diesel In-Use Fleet
Rules 22 using assumptions regarding the selection of the most likely compliance path
options. For EMFAC2021, compliance assumptions had to be updated starting with the
new calendar year 2019 base-year inventory and applying updated compliance
assumptions as appropriate. Replacement model year (MY) selections for EMFAC2021
spread replaced vehicles across several model years to avoid large spikes in a single
model year grouping.
Compliance assumptions for EMFAC2021 reflect the changes in actual inventory with the
new calendar year 2019 base-year and updated compliance path options for fleets to
meet the diesel In-Use Fleet Rules. The following sections present the updated
Retrofit/Replacement assumptions. In 2018, finalized Court decisions voided 2014
amendments to the Truck and Bus Rule 23,24 as discussed in the relevant sections below.
As in EMFAC2017, EMFAC2021 assumes a delay in the engine technology standards as
compared to the chassis model year for modeling purposes. Thus, a chassis model year

22
https://ww3.arb.ca.gov/msprog/truckstop/truckstop.htm
23
https://ww3.arb.ca.gov/msprog/truckstop/tb/truckbus.htm
24
California Air Resources Board (CARB), CTA/Lawson Lawsuit: 2014 Truck & Bus Regulation Amendments
Affected (https://ww3.arb.ca.gov/msprog/truckstop/azregs/flexoptions.htm)

62
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

of 2008 (MY2008) and newer are assumed to have OEM DPFs and MY2012 and newer
vehicles are assumed to meet 2010 engine standards.
EMFAC2014 assumed 100% compliance with the Truck and Bus Rule. EMFAC2017 did
not assume 100% compliance each year but did assume achievement of full compliance
by calendar year 2023. On April 28, 2017, Governor Brown signed the Road Repair and
Accountability Act of 2017 (SB 1). The bill includes a provision that modified the Vehicle
Code to prohibit DMV from registering or renewing the registration of medium and
heavy-duty diesel trucks unless the truck owner can demonstrate full compliance with
applicable emission requirements. Beginning in 2020, DMV has started implementing
their vehicle registration hold system25 therefore EMFAC2021 assumes full compliance
beginning in 2020.
The assumption tables below list the “Action”, which is either retrofitting with diesel
particulate filters (DPF), or replacing an older vehicle with a newer vehicle (turnover). The
“DPF” in the “Action” column designates a retrofit requirement for a pre-2008 vehicle
not equipped with OEM filters. The numbers (such as 2008, 2012, 2013, etc.) in the
“Action” column designate the model year of the replacement vehicles. EMFAC2017
assumptions began for January 1, 2017. EMFAC2021 assumptions begin for January 1,
2020 as vehicle inventory has been updated through calendar year 2019.
Low Use Vehicle 26
A low-use vehicle is one that operates less than 1,000 miles per calendar year within
California’s borders. To qualify for this permanent exemption, vehicles must report
annual odometer readings into TRUCRS and maintain records that are subject to CARB
audits. In EMFAC2017, low-use vehicles also included vehicles that travel less than 5,000
total miles per calendar year and assumed that all pre-2012 low use vehicles (with less
than 5,000 miles per year) would be replaced with MY2012 vehicles by January 1, 2020.
However, court decisions voided the 5,000 mile per year 2014 amendment and thus
EMFAC2021 is no longer reflecting this provision.
Low-Mileage Construction Truck Extension 27
The Work Truck Phase-in Option is no longer available for EMFAC2021 as this provision
was also a voided 2014 amendment. The remaining provision is now the Low-Mileage

25
California Air Resources Board (CARB), DMV Compliance Verification
(https://ww3.arb.ca.gov/msprog/truckstop/azregs/dmvreg.htm)
26
California Air Resources Board (CARB), Truck and Bus Regulation Low-Use Vehicle Exempt
(https://www.arb.ca.gov/msprog/onrdiesel/documents/fsLowuse.pdf, last updated September 2019)
27
California Air Resource Board (CARB), Truck and Bus Regulation Low Mileage Construction
Truck Phase-in Option (https://www.arb.ca.gov/msprog/onrdiesel/documents/faqconstructiontrucks.pdf,
last updated November 2018)

63
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Construction Truck Extension, which required 100% DPF installations by 2018, and that
owners must meet eligibility requirements and complete annual reporting into TRUCRS.
Low-mileage construction trucks must operate less than 15,000 miles per year; however,
dump trucks can operate up to 20,000 miles per year. At the time of this modeling, the
counts for vehicles approved to use this provision in TRUCRS were too low to include in
the modeling compliance assumptions. For EMFAC2014 and EMFAC2017, construction
truck categories provided separated estimates for the sole purpose of applying Truck
and Bus provisions. EMFAC2021x no longer provides the construction categories as a
separate fleet category.
Specialty and Limited Mileage Agricultural Truck Provisions 28
Agricultural truck provisions provided extensions for vehicles that applied in TRUCRS as
having eligible specialty equipment or that operated within limited mileage thresholds.
EMFAC2017 assumed the limited mileage vehicles above 10,000 miles/year and less than
15,000 miles/year would have 25% of the vehicles replaced with MY2012 vehicles by
2017, 50% by 2020, and 100% by 2023. By 2023, EMFAC2017 assumed that all of the
specialty equipment and the limited mileage vehicles of less than 10,000 miles/year
would be replaced with 2012 MY trucks. However, the voided 2014 amendments
eliminated the mileage limit steps of 15,000 miles per year from January 1, 2017 to
January 1, 2020 and the 10,000 miles per year from January 1, 2020 to January 1, 2023.
For EMFAC2021, the specialty vehicle definition that had no mileage restrictions has
already expired as of January 1, 2017. Agricultural vehicles that have not exceeded
10,000 miles per year between January 1, 2011 and January 1, 2017 would continue to
be exempt until 2023 as long as they do not exceed the 10,000 miles per year and meet
annual reporting requirements. At the time of this modeling, the remaining counts for
vehicles approved to use this provision in TRUCRS were too low to include in the
modeling compliance assumptions. EMFAC2014 and EMFAC2017 provided separate
vehicle counts for agricultural fleet groups to apply the Truck and Bus provisions based
on reported TRUCR vehicle counts claiming the agricultural exemption compliance
options. At the time of this modeling, the counts for vehicles approved to continue to
use remaining agricultural provisions in TRUCRS were too low to include in the modeling
compliance assumptions and EMFAC2020 no longer displays agricultural fleet categories
separately.
Small Fleet Rule Compliance (>26,000 lbs. GVWR) 29
The Small Fleet Option allowed small fleets to delay vehicle replacements until January
1, 2020 or later for heavier trucks with a gross vehicle weight rating (GVWR) greater than

28
https://www.arb.ca.gov/regact/2014/truckbus14/tbfrooal.pdf
29
https://ww3.arb.ca.gov/msprog/onrdiesel/documents/tbfinalreg.pdf?_ga=2.99117590.966000508.15771
25908-182099193.1574357539
64
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

26,000 lbs. To use this option, owners must have reported their fleet information by
January 31, 2014, demonstrated they had at least one PM filter no later than July 1, 2014
and report fleet information each January. Table 4.2.2-3 below lists the EMFAC2020
modeling assumptions for small fleet trucks with GVWR above 26,000 lbs. (including
single-unit, tractor and interstate IRP trucks).
Table 4.2.2-3. Retrofit/Replacement Assumptions for >26,000 GVWR Trucks in Small
Fleets
Vehicle
By Jan 1 Model Year 1st Truck Action 2nd Truck Action 3rd Truck Action
2014 1996-2007 100% DPF
2017 1996-2007 100% DPF
2018 1996-2007 100% DPF
40% 2011, 40% 2012, 40% 2011, 40% 2012, 40% 2011, 40% 2012,
2020 Pre-2001 20% 2013 20% 2013 20% 2013
40% 2012, 40% 2013, 40% 2012, 40% 2013, 40% 2012, 40% 2013,
2021 2001-2005 20% 2014 20% 2014 20% 2014
40% 2013, 40% 2014, 40% 2013, 40% 2014, 40% 2013, 40% 2014,
2022 2006-2007 20% 2015 20% 2015 20% 2015
40% 2014, 40% 2015, 40% 2014, 40% 2015, 40% 2014, 40% 2015,
2023 2008-2010 20% 2016 20% 2016 20% 2016

Large Fleet Rule Compliance (>26,000 lbs. GVWR) 30

Heavier trucks and buses with a GVWR greater than 26,000 pounds must comply with a
schedule by engine model year. For EMFAC2017, staff assumed that by January 1, 2014,
only a portion of pre-2008 trucks within this category were retrofitted with DPFs based
on the number of pre-2008 model year trucks in the 2016 DMV registration data and the
number of retrofits that were sold in California (excluding those retrofits that were used
for fleets to meet PAU, Transit, and SWCV rules). EMFAC2021 continues to use similar
assumptions for the portion of pre-2008 trucks retrofitted with DPFs by January 1, 2014.
The EMFAC2021 modeling assumptions for large fleets are shown below in Table 4.2.2-4
4.2.2-5 through 4.2.2-9.
Table 4.2.2-4. Replacement Assumptions for >26,000 GVWR Out of State Trucks
(Large Fleets)
By Jan 1 Vehicle Model Year Fleet Action (Turnover to)
2012 1996-1999 15% DPF
2013 2000-2004 15% DPF
2014 2005-2007 15% DPF
2020 Pre-2001 2018

30
https://www.arb.ca.gov/msprog/onrdiesel/documents/FSRegSum.pdf
65
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

By Jan 1 Vehicle Model Year Fleet Action (Turnover to)

2021 2001-2005 2019
2022 2006-2007 2020
2023 2008-2010 2021

Table 4.2.2-5. Replacement Assumptions for >26,000 GVWR Tractors (Large Fleets)
By Jan 1 Vehicle Model Year Fleet Action (Turnover to)
2012 1996-1999 15% DPF
2013 2000-2004 15% DPF
2014 2005-2007 15% DPF
2020 Pre-1997 40% 2011, 40% 2012, 20% 2013
2020 1997-2000 40% 2014, 40% 2015, 20% 2016
2021 2001-2005 40% 2015, 40% 2016, 20% 2017
2022 2006-2007 40% 2016, 40% 2017, 20% 2018
2023 2008-2010 40% 2016, 40% 2017, 20% 2018

Table 4.2.2-6. Replacement Assumptions for >26,000 GVWR Single Unit Trucks
(Large Fleets)
By Jan 1 Vehicle Model Year Fleet Action (Turnover to)
2012 1996-1999 15% DPF
2013 2000-2004 15% DPF
2014 2005-2007 15% DPF
2020 Pre-1997 40% 2011, 40% 2012, 20% 2013
2020 1997-2000 40% 2012, 40% 2013, 20% 2014
2021 2001-2005 40% 2013, 40% 2014, 20% 2015
2022 2006-2007 40% 2014, 40% 2015, 20% 2016
2023 2008-2010 40% 2014, 40% 2015, 20% 2016

NOx Exempt Area Extensions 31

The NOx Exempt Area Extension only applies to vehicles that travel exclusively within
specified NOx exempt areas, and excludes school buses. Regions of the state added in
the 2014 amendments are no longer applicable for these NOx provisions due to court
decisions. Only vehicles in the remaining regions of the state are qualified to make use of
this provision. Qualified vehicles that met the delayed schedule PM filter requirements
do not need to be replaced.

31
https://www.arb.ca.gov/msprog/onrdiesel/documents/fsnoxexempt.pdf
66
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Table 4.2.2-7. Retrofit Assumptions for >26,000 GVWR Trucks in the NOx Exempt
Areas
Vehicle Large Fleets % of Small Fleet Trucks that must have DPF
By Jan 1
Model Year Fleets with>3 Trucks 1 Truck Fleet 2 Truck Fleet 3 Truck Fleet
2019 Pre-2008 85% must have DPF 100% 100% 100%
2020 Pre-2008 100% must have DPF 100% 100% 100%

Table 4.2.2-8. Retrofit Assumptions for <=26,000 GVWR Trucks in the NOx Exempt
Areas
All Fleets
By Jan 1 Vehicle Model Year (Both Large and Small Fleets have the same requirements)
2019 Pre-2001 100% of these must have DPF
2020 2001-2004 100% of these must have DPF
2021 2005-2007 100% of these must have DPF

Assumptions for Trucks <= 26,000 GVWR 32

This section discusses compliance requirements and options that are available to lighter
vehicles. Lighter vehicles are those with a gross vehicle weight rating (GVWR) of 14,001
to 26,000 lbs. These requirements and options do not apply to school buses. Table 4.2.2-
9 illustrated EMFAC2021 assumptions on most likely compliance path for these vehicles.
Table 4.2.2-9. Replacement Assumptions for <=26,000 GVWR Trucks
Fleet Action
By Jan 1 Vehicle Model Year (Turnover to)
2020 2000-2004 40% 2012, 40% 2013, 20% 2014
2021 2005-2007 40% 2015, 40% 2016, 20% 2017
2023 2008-2010 40% 2016, 40% 2017, 20% 2018

School Bus Provision

Diesel-fueled school buses with a gross vehicle weight rating (GVWR) over 14,000 lbs.
are subject to the Truck and Bus Regulation. Owners needed to retire school buses
manufactured before April 1, 1977 by calendar year 2012 and DPFs were required to be
installed according to a phase-in schedule that was to be completed by CY2014. All 2-
stroke engine buses were being replaced by January 1, 2018.

32
https://ww3.arb.ca.gov/msprog/onrdiesel/documents/fsregsum.pdf
67
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Public/Utility/Solid Waste Collection Vehicles 33,34

CARB approved a regulation in 2003 that required diesel-fueled solid waste collection
vehicles (SWCV) use CARB verified control technology according to a phase-in schedule
completed by 2010. EMFAC2021 assumes all pre-2008 diesel SWCV installed DPFs by
January 1, 2012. In 2005, CARB approved a regulation to reduce diesel particulate
matter (PM) emissions from fleets operated by public agencies and utilities (PAU).
EMFAC2021 assumes all pre-2008 diesel public vehicles operating in higher population
regions installed DPFs by January 1, 2013, and by January 1, 2018 for those operating in
lower population regions. For utility trucks, EMFAC2021 assumes DPF retrofits were
completed by January 1, 2011 and that all pre-2012 utility vehicles will be replaced with
40% MY2012, 40% MY2013 and 20% MY2014 vehicles by January 1, 2021.

4.2.3 Transit Bus Population

The National Transit Database (NTD) is used as the primary source of information to
obtain population and activity for urban transit buses (UBUS) in EMFAC. The NTD was
established as required by Congress to be the nation's primary source of information and
statistics on the transit systems. The State requires transit agencies to report their activity
to NTD annually if they receive or benefit from federal grants 35, 36. NTD reports provide
rich and detailed accounts of transit fleet activity data for emission modeling purposes.
Examples of data provided include vehicle make, model year, fuel type, capacity, number
of active vehicles, VMT for each transit agency, and service mode. The most recent NTD
report, 2019, is used as the base year to project the urban buses' activity and population
for EMFAC2021.
In EMFA2017, staff processed the 2000-2015 annual NTD revenue inventory data to
generate the most recent historical transit bus population and VMT data for urban buses.
In EMFAC2021, staff further processed the 2016-2019 annual NTD data. Depending on
the funding sources that urban and rural transit agencies use, they have different
reporting requirements and structure. Staff include both urban and rural reports in their
analyses in order to obtain the UBUS activities.
The future population and VMT are forecasted at the regional level using specific growth
rates estimated for each subarea or region. For areas governed by MPOs, transit growth
rates extracted from the Regional Transportation Plan/Sustainable Communities Strategy

33
https://www.arb.ca.gov/msprog/publicfleets/publicfleetsfactsheet.pdf
34
https://www.arb.ca.gov/msprog/swcv/trashtruck.pdf
35
https://www.transit.dot.gov/funding/grants/urbanized-area-formula-grants-5307
36
https://www.transit.dot.gov/rural-formula-grants-5311

68
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

(RTP/SCS) 37 were used. For areas that are not covered by an MPO, or where local MPOs
do not forecast transit growth, the county-level human population growth rate published
by the Department of Finance were used as a surrogate for the transit growth. 38 Where
human population growth was negative, EMFAC2021 assumed a zero growth rate for
transit VMT in that region (i.e., the VMT neither increases nor decreases).
Similar to EMFAC2017, EMFAC2021 assumes transit buses have a fixed life span and will
be removed from the service after their useful life. For lighter vehicles with GVWR less
than 14,000 lbs., the useful life is assumed to be 10 years, and for the rest of the vehicles,
the useful life is assumed to be 14 years. Figure 4.2.3-1 compares the projected
population from EMFAC2017 with the most recent processed NTD in EMFAC2021.
Figure 4.2.3-1. EMFAC 2017 Transit Bus population compared to processed NTD
2016-2019

EMFAC 2017 EMFAC 2021

16,000

12,000
Population

8,000

4,000

0
2016 2017 2018 2019
Calendar Year

CARB’s Board adopted the Innovative Clean Transit (ICT) regulation in December 2018.
This regulation intends to enhance public health by improving air quality and to mitigate
climate change by transforming California transit bus fleets to zero-emissions
technologies. The ICT regulation will achieve its electrification goal by gradually
increasing the fraction of zero-emission buses (ZEBs) purchased and the number of
engines with low NOx emissions purchased in early years, if available, to replace the

37
Regional Plans and Evaluations (https://ww2.arb.ca.gov/our-work/programs/sustainable-communities-
program/regional-plans-evaluations)
38
California Department of Finance (DOF), Population Projections (baseline 2019), P-2A Total
Population for California and Counties
(https://www.dof.ca.gov/Forecasting/Demographics/Projections/)
69
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

existing conventional internal combustion engine buses. This regulation applies to all
transit agencies that own, operate, or lease buses with GVWR larger than 14,000 lbs.
Starting 2023, large agencies are required to have a certain percentage of their new
purchases as ZEB. For small agencies, this requirement starts in 2025. By the year 2030,
all new purchases are required to be ZEB for both small and large agencies. Table
4.2.3-1 shows the general phase-in schedule of ZEBs for standard buses in large and
small agencies. The purchase requirements for cutaways and other types of buses will
start in 2026 or later. The detailed assumptions and methods for modeling the ICT
regulation can be found in the Appendix L of ICT staff report. 39, 40
Table 4.2.3-1. ZEB phase in based on ICT scheduling
ZEB % of Total New Bus Purchase
Year Large Transit Agency Small Transit Agency
2023 25% -
2024 25% -
2025 25% -
2026 50% 25%
2027 50% 25%
2028 50% 25%
2029 and after 100% 100%

Figure 4.2.3-2 illustrates the population of the urban buses by fuel type. As shown, ZEB
population gradually increases over time starting from 2023 according to the ICT
schedule.

39
CARB (2018). Appendix L: Emission Inventory Methods and Results for the Proposed Innovative Clean
Transit Regulation,
https://ww3.arb.ca.gov/regact/2018/ict2018/appl.pdf?_ga=2.215362455.1626164022.1612202484-
1307567751.1567730621)
40
CARB (2018). Attachment C: Updates to the Emission Inventory Methods and Results for the
Proposed Innovative Clean Transit Regulation (https://ww2.arb.ca.gov/rulemaking/2018/innovative-
clean-transit-2018)

70
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure 4.2.3-2. Technology mixture for urban buses based on ICT regulation.

EMFAC2021 LPG

25,000 LNG
GAS_HYB
20,000 GAS
FCEB
Population

15,000
DSL_HYB

10,000 DSL
CNG_Low
5,000 CNG
BEB
0
2000

2005

2010

2015

2020

2025

2030

2035

2040

2045

2050
Calendar Year

EMFAC2017
LPG
25,000
LNG
20,000 GAS_HYB
GAS
Population

15,000
FCEB
10,000 DSL_HYB
DSL
5,000
CNG
0 BEB
2000

2005

2010

2015

2020

2025

2030

2035

2040

2045

2050

Calendar Year

71
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

4.2.4 Natural Gas Population

In previous EMFAC models, only emissions from natural gas transit buses and refuse
trucks were modeled explicitly. Starting in EMFAC2017, a module was developed to
estimate natural gas truck population based on the fraction of natural gas vehicles
among heavy-duty truck population at the air district level. However, due to limited
emission test data for natural gas powered vehicles, this module treated emission rates
from natural gas trucks the same as their diesel counterpart. Besides, it was kept as an
optional selection for end users, and the estimated natural gas population or emissions
were not included as parts of the default emission inventory output. With the availability
of emission tests for a wide variety of natural gas vehicles (Section 4.3.5.4), EMFAC2021
updated the natural gas truck population and modeled emissions from these trucks
explicitly in the default emission inventory.
EMFAC2021 followed similar method to EMFAC2017 estimate natural gas truck
population, yet incorporated more recent vehicle registration data to expand the
coverage of vehicle categories and model years. EMFAC2017 used the model years of
2011 as the cut-off point for estimating the natural gas population, while EMFAC2021
tracked back to model year 2008. As compared to EMFAC2017, EMFAC2021 reduced
the natural gas penetration fraction prediction classes to a flat average and a linear
regression class. The flat average class assumes that a constant fraction of natural gas
penetration, while the linear regression class assumes that natural gas fraction increases
as a function of model year until it reaches 100%. Details of natural gas truck penetration
fraction at each air district is provided in Appendix 6.4.
Moreover, EMFAC2021 treated the historical natural gas vehicle fraction differently.
EMFAC2017 used the model predicted values even for the historical years, while
EMFAC2021 applied fractions obtained from the DMV directly. Figure 4.2.4-1 shows two
examples of natural gas truck prediction classes for heavy heavy-duty public trucks: linear
for South Coast AQMD and flat-averaged for San Joaquin Valley Unified APCD.

72
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure 4.2.4-1. Natural gas fractions for heavy heavy-duty public trucks based on two
different prediction classes

100%
South Coast AQMD (Linear)
San Joaquin Valley Unified APCD (Flat-Avg)
Natural Gas Fraction

80%

60%

40%

20%

0%
2005 2015 2025 2035 2045
Model Year

Figure 4.2.4-2 illustrates the statewide population of natural gas heavy-duty trucks from
2000-2050. Please note that this chart also includes population of natural gas powered
transit buses that were discussed in Section 4.2.3. It is worth mentioning that refuse
trucks are a major portion of the natural gas truck population. The decrease in number of
natural gas trucks is caused by the ACT regulation, which generally reduces the
population of trucks with internal combustion engines.
Figure 4.2.4-2. Statewide natural gas vehicle population based on EMFAC2021

50,000

40,000
Population

30,000

20,000

10,000

0
2000 2010 2020 2030 2040 2050
Calendar Year

73
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

In 2013, California established optional low-NOx standards with the most aggressive
being 0.02 g/bhp-hr, which is 90% below the current standard. As a result, since 2016, a
number of natural gas- and propane-fueled Otto-cycle engine families have been
certified to the optional low NOx standards of 0.02 g/bhp-hr. 41 The analysis of sales
volume of CARB certified heavy-duty natural gas engines shows that engines certified to
0.02 g/bhp-hr NOx are roughly 50% starting from model year 2017 and approximately
100% for the model year 2018 and newer. Therefore, the emission rates for natural gas
engines have been adjusted accordingly to these fractions. Detailed discussion about
natural gas truck and bus emission rates can be found in Section 4.3.5.4.

4.3 Emission Rates

4.3.1 Light Duty Emission Rates
Emissions that emanate from the vehicle’s tailpipe are called exhaust emissions.
Incomplete combustion of the fuel is the primary cause of hydrocarbon (HC), carbon
monoxide (CO), and particulate matter (PM) emissions. These emissions occur at all times
but are more intense when the air-to-fuel ratio is richer than stoichiometric (14.7-to-1)
conditions, such as during hard acceleration. Oxides of nitrogen (NOx) emissions are
produced during combustion at high temperatures and pressures and can be enhanced
under lean air-to-fuel ratio conditions. Properly working catalysts reduce tailpipe
emissions from gasoline vehicles by over 90 percent when combined with electronic
systems that monitor the air-to-fuel ratio. Due to higher combustion temperatures,
excess air, and high pressures, a diesel-fueled vehicle emits comparatively more NOx
than a gasoline-fueled vehicle. The lean overall air-to-fuel ratios used by diesel vehicles
precluded the use of conventional reduction catalysts for emissions control systems.
Combustion engine vehicles also emit carbon dioxide (CO2) and contribute to statewide
greenhouse gas (GHG) emissions. It should be noted that EMFAC uses measured CO2
emissions data to predict CO2 emissions and emission rates.
Two light-duty vehicle operational modes contribute to exhaust emissions: the stabilized
running mode and the start mode. This section provides a brief overview of the model’s
handling of basic tailpipe emission rates and start emission rates. Emission rates, also
referred to as emission factors, related to these sources are typically measured at
standard temperature and humidity using driving cycles mimicking typical vehicle driving
and operating patterns. Emission rates are ultimately combined with vehicle activity data
(such as vehicle population counts) to estimate vehicle emissions inventories.
In EMFAC2017, the base emission rates (BER) used to compute the running and the
starts emission rates were updated for the first time since EMFAC2000. In EMFAC2021,

41
CARB (2020). Optional Low NOx Certified Heavy-Duty Engines
(https://ww2.arb.ca.gov/sites/default/files/classic//msprog/onroad/optionnox/optional_low_nox_certified_
hd_engines.pdf)

74
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

those BERs are further evaluated and updated based on three years of new data from
the US EPA’s In-Use Vehicle Program (IUVP) and CARB’s Vehicle Surveillance Program
(VSP). The method to determine BERs are similar to EMFAC2017, but a new “low”
emission regime has been added to model deterioration at lower emission levels. The
new test data are also used to update the Ratio of Standards (ROS) for characterizing
technology groups (i.e., new certification levels) lacking sufficient test data.
As described in the EMFAC2017 technical documentation, see Section 4.3.1.1.3, the
light-duty vehicle fleet can be categorized into unique “technology groups.” 42 Each
technology group represents vehicles with distinct emission control technologies with
similar in-use deterioration rates and response to repairs. Further, vehicles in each
technology group can be sub-divided into “emission regimes” defined by certification
standards (i.e., Standards defined over the Federal Test Procedures –FTP). In order to
calculate the fleet average emission rates (i.e., Base Emission Rate – BER), emission
factors associated with each regime (i.e., Regime emission factor) are weighted using the
percent of vehicles within each regime (i.e., regime fractions).
In EMFAC2017, staff included three emission regimes, including normal, moderate, and
high. In EMFAC2021, a new "low" regime is separated from the normal regime to model
deterioration at lower emission levels (Table 4.3.1-1). The low regime's emission factor is
the average FTP emissions of vehicles that have emissions less than 0.5 times of the
standard. The emission factor for the revised normal regime is the average emissions of
vehicles that have emissions between 0.5 and 1 times the standard.
Table 4.3.1-1. Emission regime definitions in EMFAC2021 model
Emission Regime Emission Range
Low 0 to 0.5 x Standard
Normal 0.5 to 1.0 x Standard
Moderate 1.0 to 2.0 x Standard
High > 2.0 x Standard

Other than adding the new “low” emission regime, the method of BER determination in
EMFAC2021 is similar to EMFAC2017, including the test cycle (UC cycle), data sources
(IUVP and VSP), and the definition of technology groups. These are described in detail in
Section 4.3.1.1 of EMFAC2017 Technical Documentation.
For EMFAC2021, UC BERs are updated with the past three years of new data from the
IUVP and VSP programs, including running and start exhaust emission rates of HC, NOx,
and CO for LEV1 and LEV2 technology groups. The new data are also used to update the
Ratio of Standards (ROS) for the LEV3 BERs. In general, to update BERs, the following
steps were taken:

42
See appendix 6.2
75
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

1. Acquire the IUVP data (FTP results) for the selected tech group
2. Acquire the California sales data for each tech group from the NMOG report, if
available
3. Determine the sales-weighted regime fractions and average emission rates versus
odometer bins for the FTP data
4. Gather tests data for the selected tech group from VSP under the FTP and UC
tests
5. Classify FTP composite emission rate data by low, normal, moderate, and high
regimes
6. Select vehicles tested under both the FTP and UC
7. Determine the average value of UC results for each emission regime.
Table 4.3.1-2 shows a sample of results for HC emissions associated with LEV2 LEVs. The
U.S. EPA’s IUVP data were used to determine the fractions of low, normal, moderate and
high emitters versus odometer for LEV1 and LEV2 vehicles. The IUVP results were
weighted by the California sales of each test group. Sales data were obtained from the
manufacturers’ non-methane organic gas (NMOG) reports to CARB. The CARB’s VSP
program provides the UC cycle-based emission levels for the emission regimes. The
emission rates are computed in terms of fraction of vehicles in an emission regime (either
low-, normal-, moderate-, or high-emission, as a function of odometer), averaging the
corresponding UC emission rate over odometer bin for that regime. Finally, a curve is
fitted to average emission rates for each bin versus odometer.

76
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Table 4.3.1-2. Sample of Results for LEV2 LEVs Hydrocarbon (HC)

IUVP Data
Odometer Bin Regime Count Regime Fraction Sales Weighted Regime Fraction
Low 214 0.88 0.80
Normal 27 0.11 0.19
0-50 kmi
Moderate 2 0.01 0.01
High 0 0.00 0.00
Low 190 0.56 0.39
Normal 138 0.41 0.57
50-100 kmi
Moderate 8 0.02 0.03
High 1 0.00 0.00
Low 26 0.48 0.39
Normal 26 0.48 0.59
100-150 kmi
Moderate 2 0.04 0.02
High 0 0.00 0.00

VSP Data:
UC Bag Regime Count UC Phase Mean HC (g/mile)
Low 12 1 0.440
Normal 13 1 0.927
UCP1
Moderate 5 1 1.045
High 1 1 2.080
Low 12 2 0.014
Normal 13 2 0.014
UCP2
Moderate 5 2 0.037
High 1 2 0.037
Low 12 3 0.028
Normal 13 3 0.062
UCP3
Moderate 5 3 0.076
High 1 3 0.076

Emission rates are modeled in the form of regression equations with vehicle mileage
(odometer in units of 10,000 miles) as an input to the model. To do this, staff fitted a
curve to regime fractions as a function of odometer, and calculated the weighted
average emission rates for each phase of the UC cycle for different odometers. The
resulting emission rates were again fitted with regressions as a function of odometer, and
these estimated emission rates were then used in the model.

77
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure 4.3.1-1 to Figure 4.3.1-5 show comparison between EMFAC2021 and

EMFAC2017 emission rates. Both running and start exhaust emission rates for HC and
NOx for LEV1 LEV and LEV1 ULEV showed higher rates in EMFAC2021 when compared
to EMFAC2017, especially for high odometer regimes. For LEV2 technology groups,
EMFAC2021 shows lower HC emissions for LEV2 ULEV. However, HC emission rates are
slightly higher for LEV2 LEV in high odometer ranges and significantly higher for LEV2
SULEV in EMFAC2021 than EMFAC2017. The NOx running exhaust emission rates of
LEV2 LEV and LEV2 ULEV are higher in EMFAC2021 with an increased odometer
reading, although LEV2 SULEV is mostly closer to EMFAC2017 values. For HC and NOx
start exhaust emission rates, all LEV2 technology groups showed higher emissions in
EMFAC2021 than in EMFAC2017.
For LEV3 technology groups (MY from 2015), EMFAC assumes LEV160, ULEV125, and
SULEV30 share the same running and start exhaust emission regressions with LEV2 LEV,
LEV2 ULEV, and LEV2 SULEV, separately. For other LEV3 certification levels such as
ULEV70, ULEV50, and SULEV20, a Ratio of Standards (ROS) approach is used from
EMFAC2014 to estimate future technologies' emission rates since there are limited
available data to accurately develop their own emission regressions. Previously, the ROS
were derived from the ratio of emission standards between the LEV3 HC+NOx
certification and its LEV2 base. Despite that, staff concluded that based on the most
recent CARB VSP test results, applying the same ROS to all test phases might not be
appropriate. Therefore, in EMFAC2021, the ROS for LEV3 tech groups were further
assessed and updated based on the latest VSP or FTP test data, available EPA fuel
economy test data, and staff engineering judgment. Figure 4.3.1-6 to Figure 4.3.1-8
compare the UC bag emission rates for various technology groups in EMFAC2021.

78
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure 4.3.1-1. UC Bags 1-3 HC and NOx emission rates (g/mi) for LEV1 LEVs –
EMFAC2021 vs. EMFAC2017

LEV1 LEV UC P1 HC LEV1 LEV UC P1 NOx

2.5 2.5
y = 5.12E-06x + 8.61E-01 y = 5.43E-06x + 5.22E-01
2.0 R² = 9.84E-01 2.0 R² = 8.83E-01
ER (g/mi)

ER (g/mi)
1.5 1.5
1.0 1.0
0.5 EMFAC2021 0.5 EMFAC2021
EMFAC2017 EMFAC2017
0.0 0.0
0 100,000 200,000 0 100,000 200,000
Odometer (mi) Odometer (mi)

LEV1 LEV UC P2 HC LEV1 LEV UC P2 NOx

0.12 0.6
y = 3.83E-07x + 1.40E-02 y = 1.69E-06x + 2.53E-02
0.10 R² = 9.67E-01 0.5 R² = 8.91E-01
0.08 0.4
ER (g/mi)
ER (g/mi)

0.06 0.3
0.04 0.2
0.02 EMFAC2021 0.1 EMFAC2021
EMFAC2017 EMFAC2017
0.00 0.0
0 200,000 0 100,000 200,000
Odometer (mi) Odometer (mi)

LEV1 LEV UC P3 HC
0.4 LEV1 LEV UC P3 NOx
y = 1.25E-06x + 3.91E-02
0.4 R² = 9.73E-01 1.0
y = 3.13E-06x + 6.65E-02
0.3 R² = 8.71E-01
0.8
0.3
ER (g/mi)

ER (g/mi)

0.2 0.6
0.2 0.4
0.1
EMFAC2021 0.2
0.1 EMFAC2021
EMFAC2017
0.0 EMFAC2017
0.0
0 100,000 200,000 0 100,000 200,000
Odometer (mi) Odometer (mi)

79
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure 4.3.1-2. UC Bags 1-3 HC and NOx emission rates (g/mi) for LEV1 ULEVs –
EMFAC2021 vs. EMFAC2017
LEV1 ULEV UC P1 HC LEV1 ULEV UC P1 NOx
1.5 1.0
y = 1.21E-06x + 7.17E-01 y = 6.16E-07x + 5.47E-01
R² = 4.06E-01 0.8 R² = 1.07E-01
1.0
ER (g/mi)

ER (g/mi)
0.6

0.4
0.5
EMFAC2021 0.2 EMFAC2021
EMFAC2017 EMFAC2017
0.0 0.0
0 100,000 200,000 0 100,000 200,000
Odometer (mi) Odometer (mi)

LEV1 ULEV UC P2 HC LEV1 ULEV UC P2 NOx

0.04 y = 5.33E-08x + 2.37E-02 y = 1.66E-07x + 7.99E-02
0.15
R² = 5.55E-01 R² = 1.28E-01
0.03
ER (g/mi)

0.10
ER (g/mi)

0.02
0.05
0.01 EMFAC2021 EMFAC2021
EMFAC2017 EMFAC2017
0.00
0.00
0 100,000 200,000
0 100,000 200,000
Odometer (mi) Odometer (mi)

LEV1 ULEV UC P3 HC LEV1 ULEV UC P3 NOx

0.12 0.25
0.10 0.20
0.08 y = 2.58E-07x + 6.69E-02
ER (g/mi)

ER (g/mi)

R² = 5.76E-01 0.15 y = 2.66E-07x + 1.53E-01

0.06
0.10 R² = 1.05E-01
0.04
0.02 EMFAC2021 0.05 EMFAC2021
EMFAC2017 EMFAC2017
0.00 0.00
0 100,000 200,000 0 100,000 200,000
Odometer (mi) Odometer (mi)

80
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure 4.3.1-3. UC Bags 1-3 HC and NOx emission rates (g/mi) for LEV2 LEVs –
EMFAC2021 vs. EMFAC2017
LEV2 LEV UC P1 HC LEV2 LEV UC P1 NOx
1.0 0.3
y = 2.39E-06x + 5.15E-01 y = 4.48E-07x + 1.76E-01
0.8 R² = 8.07E-01 0.3 R² = 9.58E-01
0.2

ER (g/mi)
ER (g/mi)

0.6
0.2
0.4
0.1
0.2 EMFAC2021 0.1 EMFAC2021
EMFAC2017 EMFAC2017
0.0 0.0
0 100,000 200,000 0 100,000 200,000
Odometer (mi) Odometer (mi)

LEV2 LEV UC P2 HC LEV2 LEV UC P2 NOx

0.020
0.04
0.015 0.03
ER (g/mi)

y = 7.25E-09x + 1.41E-02 0.03 y = 9.23E-08x + 2.26E-02

ER (g/mi)

0.010 0.02 R² = 9.85E-01

R² = 5.74E-01
0.02
0.005 EMFAC2021 0.01 EMFAC2021
EMFAC2017 0.01 EMFAC2017
0.000 0.00
0 200,000 0 100,000 200,000
Odometer (mi) Odometer (mi)

LEV2 LEV UC P3 HC LEV2 LEV UC P3 NOx

0.06 y = 1.66E-07x + 3.32E-02 0.06 y = 3.24E-07x + 1.48E-02
R² = 8.15E-01 0.05 R² = 9.92E-01
0.05
0.04 0.04
ER (g/mi)
ER (g/mi)

0.03 0.03
0.02 0.02
EMFAC2021 EMFAC2021
0.01 0.01
EMFAC2017 EMFAC2017
0.00 0.00
0 100,000 200,000 0 100,000 200,000
Odometer (mi) Odometer (mi)

81
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure 4.3.1-4. UC Bags 1-3 HC and NOx emission rates (g/mi) for LEV2 ULEVs –
EMFAC2021 vs. EMFAC2017
LEV2 ULEV UC P1 HC LEV2 ULEV UC P1 NOX
0.5 0.25
0.4 0.20
y = 9.63E-07x + 3.62E-01
ER (g/mi)

y = 5.44E-07x + 1.51E-01

ER (g/mi)
0.3 R² = 8.72E-01 0.15
R² = 9.96E-01
0.2 0.10
0.1 EMFAC2021 0.05 EMFAC2021
EMFAC2017 EMFAC2017
0.0 0.00
0 100,000 200,000 0 100,000 200,000
Odometer (mi) Odometer (mi)

LEV2 ULEV UC P2 HC LEV2 ULEV UC P2 NOX

0.014 0.03
0.012
0.010
0.02
ER (g/mi)

ER (g/mi)

0.008 y = 1.35E-08x + 7.83E-03 y = 9.45E-08x + 1.28E-02

0.006 R² = 9.10E-01 R² = 9.96E-01
0.004 0.01
EMFAC2021 EMFAC2021
0.002 EMFAC2017
EMFAC2017
0.000 0.00
0 100,000 200,000 0 100,000 200,000
Odometer (mi) Odometer (mi)

LEV2 ULEV UC P3 HC LEV2 ULEV UC P3 NOX

0.04 0.06
0.03 0.05
0.03 0.04
ER (g/mi)
ER (g/mi)

y = 7.30E-08x + 2.00E-02 y = 2.61E-07x + 2.45E-02

0.02 0.03
R² = 8.75E-01 R² = 9.92E-01
0.02
0.02
0.01 EMFAC2021
EMFAC2021 0.01
0.01 EMFAC2017
EMFAC2017
0.00 0.00
0 200,000 0 100,000 200,000
Odometer (mi) Odometer (mi)

82
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure 4.3.1-5. UC Bags 1-3 HC and NOx emission rates (g/mi) for LEV2 SULEVs –
EMFAC2021 vs. EMFAC2017
LEV2 SULEV UC P1 HC LEV2 SULEV UC P1 NOx
0.20 0.30 y = 8.35E-07x + 7.42E-02
y = 3.32E-07x + 1.23E-01
R² = 9.86E-01 0.25 R² = 9.99E-01
0.15
0.20
ER (g/mi)

ER (g/mi)
0.10 0.15
0.10
0.05 EMFAC2021 EMFAC2021
0.05
EMFAC2017 EMFAC2017
0.00 0.00
0 100,000 200,000 0 100,000 200,000
Odometer (mi) Odometer (mi)
LEV2 SULEV UC P2 HC LEV2 SULEV UC P2 NOx
0.006 y = 2.04E-08x + 2.96E-03 0.035 y = 7.32E-08x + 1.06E-02
R² = 9.99E-01 0.030 R² = 9.31E-01
0.005
0.025
0.004
ER (g/mi)

ER (g/mi)

0.020
0.003
0.015
0.002
0.010 EMFAC2021
0.001 EMFAC2021
EMFAC2017 0.005 EMFAC2017
0.000 0.000
0 100,000 200,000 0 100,000 200,000
Odometer (mi) Odometer (mi)
LEV2 SULEV UC P3 HC LEV2 SULEV UC P3 NOx
0.014
0.05
0.012 y = 1.12E-07x + 1.54E-02
0.010 y = 4.10E-08x + 8.36E-03 0.04 R² = 9.51E-01
ER (g/mi)

ER (g/mi)

0.008 R² = 9.97E-01 0.03

0.006
0.02
0.004
0.01 EMFAC2021
0.002 EMFAC2021
EMFAC2017 EMFAC2017
0.000 0.00
0 100,000 200,000 0 100,000 200,000
Odometer (mi) Odometer (mi)

83
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure 4.3.1-6. EMFAC2021 HC and NOx Cold Start Exhaust (UCP1) Emission Rates by
Tech Group

1.2 HC UC Bag 1 LEV2 LEV/LEV3 LEV 160

1.0
LEV2 ULEV/LEV3
ER (g/mile)

0.8 ULEV125
LEV3 ULEV 70
0.6
0.4 LEV3 ULEV 50
0.2 LEV3 SULEV 30/LEV2
0.0 SULEV
0 5 10 15 20 LEV3 SULEV 20
Odometer (10k miles)

NOx UC Bag 1
0.35 LEV2 LEV/LEV3 LEV 160
0.30
LEV2 ULEV/LEV3
0.25
ER (g/mile)

ULEV125
0.20 LEV3 ULEV 70
0.15
LEV3 ULEV 50
0.10
0.05 LEV3 SULEV 30/LEV2
0.00 SULEV
LEV3 SULEV 20
0 5 10 15 20
Odometer (10k miles)

84
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure 4.3.1-7. EMFAC2021 HC and NOx Running Exhaust (UCP2) Emission Rates by
Tech Group

HC UC Bag 2
0.020
LEV2 LEV/LEV3 LEV 160

0.015 LEV2 ULEV/LEV3 ULEV125

ER (g/mile)

LEV3 ULEV 70
0.010
LEV3 ULEV 50
0.005
LEV3 SULEV 30/LEV2
SULEV
0.000
LEV3 SULEV 20
0 5 10 15 20
Odometer (10k miles)

NOX UC Bag 2
0.045
LEV2 LEV/LEV3 LEV 160
0.040
0.035
LEV2 ULEV/LEV3 ULEV125
0.030
ER (g/mile)

0.025 LEV3 ULEV 70

0.020
0.015 LEV3 ULEV 50
0.010
0.005 LEV3 SULEV 30/LEV2
0.000 SULEV
0 5 10 15 20 LEV3 SULEV 20
Odometer (10k miles)

85
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure 4.3.1-8. EMFAC2021 HC and NOx Running Exhaust (UCP3) Emission Rates by
Tech Group

HC UC Bag 3
0.070
0.060 LEV2 LEV/LEV3 LEV 160

0.050 LEV2 ULEV/LEV3 ULEV125

ER (g/mile)

0.040
LEV3 ULEV 70
0.030
0.020 LEV3 ULEV 50
0.010
LEV3 SULEV 30/LEV2 SULEV
0.000
0 5 10 15 20 LEV3 SULEV 20
Odometer (10k miles)

NOx UC Bag 3
0.09
0.08 LEV2 LEV/LEV3 LEV 160
0.07
0.06 LEV2 ULEV/LEV3 ULEV125
ER (g/mile)

0.05
LEV3 ULEV 70
0.04
0.03
LEV3 ULEV 50
0.02
0.01 LEV3 SULEV 30/LEV2 SULEV
0
0 5 10 15 20 LEV3 SULEV 20
Odometer (10k miles)
Table 4.3.1-3 to Table 4.3.1-13 provide the final regression equations utilized in
EMFAC2021 for LD Base Emission Rates (BERs). They are listed by Tech Group IDs –
Numerical identifiers for technology groups in the EMFAC model. Each technology
group represents vehicles with distinct emission control technologies with similar in-use
deterioration. The correlations are in units of g/mi for pollutants and 10 kmi for
odometer readings.

86
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Table 4.3.1-3. Tech Group 23 (LEV1 LEV) BER Correlation Coefficients

Process Pollutant A (X0) B (X1) Regression Type
HC 0.861 0.0512 Linear
NOx 0.522 0.0543 Linear
UC1 CO 10.7 0.262 Linear
HC 0.014 0.00383 Linear
NOx 0.0253 0.0169 Linear
UC2 CO 1.59 0.0832 Linear
HC 0.0391 0.0125 Linear
NOx 0.0665 0.0313 Linear
UC3 CO 1.39 0.127 Linear

Table 4.3.1-4. Tech Group 24 (LEV1 ULEV) BER Correlation Coefficients

Process Pollutant A (X0) B (X1) Regression Type
HC 0.717 0.0121 Linear
NOx 0.547 0.00616 Linear
UC1 CO 6.59 0.293 Linear
HC 0.0237 0.000533 Linear
NOx 0.0799 0.00166 Linear
UC2 CO 0.31 0.142 Linear
HC 0.0669 0.0026 Linear
NOx 0.1526 0.0027 Linear
UC3 CO 0.131 0.5255 Exponential

Table 4.3.1-5. Tech Group 28 (LEV2 LEV) BER Correlation Coefficients

Process Pollutant A (X0) B (X1) Regression Type
HC 0.515 0.0239 Linear
NOx 0.176 0.00448 Linear
UC1 CO 11.1 0.0569 Linear
HC 0.0141 0.0000725 Linear
NOx 0.0226 0.000923 Linear
UC2 CO 1.4 0.0303 Linear
HC 0.0332 0.00166 Linear
NOx 0.0148 0.00324 Linear
UC3 CO 1.49 0.0283 Linear

87
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Table 4.3.1-6. Tech Group 29 (LEV2 ULEV) BER Correlation Coefficients

Process Pollutant A (X0) B (X1) Regression Type
HC 0.362 0.00963 Linear
UC1 NOx 0.151 0.00544 Linear
CO 4.71 0.0449 Linear
HC 0.00885 0.000177 Linear
UC2 NOx 0.0128 0.000945 Linear
CO 0.6513 0.0115 Linear
HC 0.02 0.00073 Linear
UC3 NOx 0.0245 0.00261 Linear
CO 0.789 0.0187 Linear

Table 4.3.1-7. Tech Group 31 (LEV2 SULEV) BER Correlation Coefficients

Process Pollutant A (X0) B (X1) Regression Type
HC 0.123 0.00332 Linear
UC1 NOx 0.0742 0.00835 Linear
CO 1.3 0.0585 Linear
HC 0.00296 0.000204 Linear
UC2 NOx 0.0106 0.000732 Linear
CO 0.324 0.032 Linear
HC 0.00836 0.00041 Linear
UC3 NOx 0.0154 0.00112 Linear
CO 0.19 0.0214 Linear

Table 4.3.1-8. Tech Group 38 (LEV3 SULEV20) BER Correlation Coefficients

Process Pollutant A (X0) B (X1) Regression Type
HC 0.09963 0.002689 Linear
UC1 NOx 0.056392 0.006346 Linear
CO 1.868 0.0252 Linear
HC 0.002664 0.000184 Linear
UC2 NOx 0.00954 0.000659 Linear
CO 0.4134 0.003373 Linear
HC 0.007524 0.000369 Linear
UC3 NOx 0.01386 0.001008 Linear
CO 0.3601 0.01292 Linear

88
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Table 4.3.1-9. Tech Group 39 (LEV3 ULEV50) BER Correlation Coefficients

Process Pollutant A (X0) B (X1) Regression Type
HC 0.199875 0.005395 Linear
UC1 NOx 0.08904 0.00852 Linear
CO 3.812857 0.036348 Linear
HC 0.005017 0.000346 Linear
UC2 NOx 0.01166 0.000805 Linear
CO 0.527243 0.00931 Linear
HC 0.01254 0.000615 Linear
UC3 NOx 0.01848 0.001344 Linear
CO 0.638714 0.015138 Linear

Table 4.3.1-10. Tech Group 44 (LEV3 ULEV70) BER Correlation Coefficients

Process Pollutant A (X0) B (X1) Regression Type
HC 0.27675 0.00747 Linear
UC1 NOx 0.10388 0.007784 Linear
CO 3.812857 0.036348 Linear
HC 0.007074 0.000263 Linear
UC2 NOx 0.01272 0.000878 Linear
CO 0.527243 0.00931 Linear
HC 0.01672 0.00082 Linear
UC3 NOx 0.02156 0.001568 Linear
CO 0.638714 0.015138 Linear

Table 4.3.1-11. Tech Group 45 (LEV3 SULEV30) BER Correlation Coefficients

Process Pollutant A (X0) B (X1) Regression Type
HC 0.123 0.00332 Linear
UC1 NOx 0.0742 0.00835 Linear
CO 1.3 0.0585 Linear
HC 0.00296 0.000204 Linear
UC2 NOx 0.0106 0.000732 Linear
CO 0.324 0.032 Linear
HC 0.00836 0.00041 Linear
UC3 NOx 0.0154 0.00112 Linear
CO 0.19 0.0214 Linear

89
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Table 4.3.1-12. Tech Group 55 (LEV3 ULEV125) BER Correlation Coefficients

Process Pollutant A (X0) B (X1) Regression Type
HC 0.362 0.00963 Linear
UC1 NOx 0.151 0.00544 Linear
CO 4.71 0.0449 Linear
HC 0.00885 0.000177 Linear
UC2 NOx 0.0128 0.000945 Linear
CO 0.6513 0.0115 Linear
HC 0.02 0.00073 Linear
UC3 NOx 0.0245 0.00261 Linear
CO 0.789 0.0187 Linear

Table 4.3.1-13. Tech Group 56 (LEV3 LEV160) BER Correlation Coefficients

Process Pollutant A (X0) B (X1) Regression Type
HC 0.515 0.0239 Linear
UC1 NOx 0.176 0.00448 Linear
CO 11.1 0.0569 Linear
HC 0.0141 0.0000725 Linear
UC2 NOx 0.0226 0.000923 Linear
CO 1.4 0.0303 Linear
HC 0.0332 0.00166 Linear
UC3 NOx 0.0148 0.00324 Linear
CO 1.49 0.0283 Linear

4.3.2 Evaporative Emissions Module

EMFAC2021 has updated the evaporative module to estimate evaporative emissions
from gasoline vehicles, which are the major sources of gaseous hydrocarbon pollutants. 43
This updated module in EMFAC2021 models physical processes that are involved in
generation of evaporative emissions, while EMFAC2017 and earlier versions modeled the
certification processes of evaporative emissions. This update is implemented mainly by
adopting the same method as used by the U.S. EPA’s MOVES3 model to estimate

43
Evaporative emissions are generated as gasoline fuels evaporate and escape from the vehicle’s fuel
system.

90
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

evaporative emissions. 44 Additionally, CARB staff customized the new method for the
state of California by using California specific meteorological data, vehicle activity data,
and fuel information.
The evaporative emissions module in previous versions of EMFAC estimated emissions
using an empirical method based on certification processes, which are a) diurnal and
resting loss (DIURN and RESTLOSS), b) running loss (RUNLOSS), and c) hot soak
(HOTSOAK). 45 Following the MOVES3 model, EMFAC2021 takes a different approach of
modeling three physical evaporative emission processes, namely, tank vapor venting,
permeation, and liquid leak. Each physical process is modelled for three modes of
activity: cold soak (or diurnal), running, and hot soak modes. The three processes are
illustrated in Figure 4.3.2-1 and defined as follows:
1. Tank vapor venting is a process of fuel vapors escaping from the fuel tank when
the carbon canister is saturated (or the amount of fuel vapor is larger than the
capacity of the carbon canister).
2. Permeation happens when fuel escapes through materials in the system such as
tank walls, hoses, and seals.
3. Liquid leak indicates non-vapor form emissions of fuel escaping the fuel system
such as dripping fuel, which ultimately evaporates into the atmosphere.
Figure 4.3.2-1. An illustration of evaporative processes of tank vapor venting,
permeation, and liquid leaks as modeled in EMFAC2021

44
USEPA (2020). “Evaporative Emissions from Onroad Vehicles in MOVES3” Assessment and Standards
Division Office of Transportation and Air Quality U.S. Environmental Protection Agency. EPA-420-R-20-012
(web link: https://nepis.epa.gov/Exe/ZyPDF.cgi?Dockey=P1010M0C.pdf, November 2020)
45
California Air Resources Board (CARB), Chapter 5 of Technical Support Document for EMFAC2000 (web
link: https://ww3.arb.ca.gov/msei/onroad/downloads/tsd/emfac2000-emissions.pdf, May 2000)

91
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

The evaporative emission module in EMFAC2021 not only incorporate U.S.EPA’s latest
research findings, but also provides a platform for further development of EMFAC
through refinement of model parameters and updates of its input data based on results
from new emissions testing and data collection efforts. CARB staff is tentatively
considering to update the module in the following areas in the future: (1) designing and
conducting new emission test methods to better characterize emission rates for
permeation and tank vapor venting processes; (2) more accurate speciation profiles for
LEV 2 and LEV 3 technologies, and advanced vehicle technologies such as plug-in
hybrids; (3) better quantification method for liquid leaks; and (4) more refined
characterization of fuels used in California.
Compared to MOVES' implementation of evaporative emissions, EMFAC2021 is
implemented differently in several areas to improve the module's computational
efficiency and to customize the module for use in California. First of all, EMFAC2021 and
MOVES have different sets of vehicle classes. Staff matched vehicle classes used in
EMFAC2021 to the most equivalent vehicle classes used in the MOVES model. The
matched pairs are listed in Table 4.3.2-1.

Table 4.3.2-1 EMFAC vehicle classes and their equivalent MOVES regulatory classes
and source types used in the evaporative emission module

EMFAC2021 MOVES3
EMFAC202x Vehicle Category Regulatory Class Source Type
MCY MC Motorcycle
LDA LDV Passenger Car
LDT1 LDT Passenger Truck
LDT2 LDT Passenger Truck
MDV LDT Passenger Truck
LHDT1 LHD ≤ 10k Light Commercial Truck
LHDT2 LHD ≤ 14k Single Unit Short-haul Truck
MH LHD ≤ 14k Single Unit Short-haul Truck
OBUS LHD45 Transit Bus
UBUS LHD45 Transit Bus
SBUS LHD45 Transit Bus
T6TS MHD67 Single Unit Long-haul Truck
T7IS HHD8 Combination Short-haul Truck

92
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Second, EMFAC2021 uses California-specific meteorology inputs, vehicle activity, and

fuel information. Instead of using meteorological data used in the MOVES database, staff
used meteorological data that are specifically developed for use in EMFAC. 46,47
Additionally, as EMFAC’s spatial resolution of sub-areas is finer than MOVES’ county-
level resolution, this allows EMFAC to use more accurate meteorological input at the
finer sub-area level. Besides, staff used vehicle trip data including vehicle starts, trips,
and parking activities from the California Household Travel Survey (CHTS) 48 to estimate
vehicle activity values for evaporative emissions (e.g., vehicle operation hours, cold-soak
hours, and hot-soak hours). Rather than using MOVES default values, staff ran MOVES
with the CHTS data to pre-generate data as an input to the updated module to estimate
the activity values. In addition, rather than using those used by MOVES, staff used the
sub-area specific gasoline volatility (or Reid vapor pressure) values used in EMFAC2017
to better represent fuels used in California.
Third, EMFAC2021 does not generate temperature profiles of fuel tanks during model
runs but utilize tank temperature profiles preprocessed using MOVES. The fuel tank
temperature is an important parameter used in the calculation of evaporative emission
rates. MOVES has a module that estimates tank temperatures numerically based on
meteorological inputs and vehicle trip information. To reduce the computational cost of
generating the temperature profile in the module and simplify the module
implementation, staff pre-generated tank temperature values using MOVES with
aforementioned California-specific meteorological inputs and vehicle trip information
and used the generated values as input data to the new module.
Lastly, EMFAC2021 still outputs evaporative emissions by activity modes (i.e., DIURN,
HOTSOAK, and RUNLOSS). Note that unlike previous versions of EMFAC, RESTLOSS is
no longer produced as a separate process and that DIURN in EMFAC2021 includes
RESTLOSS. Like the MOVES model, EMFAC2021 models internally each of the three
physical evaporative emission processes (tank vapor venting, permeation, and liquid
leaks) for the three modes of vehicle activities (diurnal, running, and hot soak). While
MOVES aggregates resulting emissions in the three physical processes, EMFAC2021
does the same thing in the three modes of vehicle activity to provide more consistent
outputs with previous versions.
Besides these differences, staff incorporated the same method and data implemented in
the MOVES3 model. Please refer to U.S.EPA MOVES (2020) for detailed description of
the algorithms of estimating evaporative emissions.

46
California Air Resources Board (2000). On-Road Emission Model Methodology Documentation, Section
7.8 County-Specific Diurnal Temperature Profiles
47
California Air Resources Board (2006). An Update to Summer Temperature and Relative Humidity
Profiles for the EMFAC2007 On-road Emissions Model
48
California Department of Transportation. (2013) 2010-2012 California Household Travel Survey Final
Report Version 1.0
93
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

In order to illustrate the difference between EMFAC's new and old evaporative modules,
comparisons of statewide evaporative emissions for several calendar years estimated by
EMFAC2021 and EMFAC2017 are presented in Figure 4.3.2-2. Despite the substantial
change in the modeling method, EMFAC2021 shows a similar magnitude and trend of
evaporative emissions than EMFAC2017. EMFAC2021 estimates higher diurnal emissions
than EMFAC2017, while EMFAC2021’s running loss emissions are somewhat smaller than
EMFAC2017’s. As shown, EMFAC2021’s hot soak emissions are similar to EMFAC2017’s.
In addition, Figure 4.3.2-3 shows a comparison between EMFAC2021 and MOVES3 for
evaporative emissions from passenger cars (LDA) in Los Angeles county. The comparison
was made in terms of a per-vehicle emission basis to exclude the effects of different
future population forecasts of the two models. The emissions from EMFAC2021 are
aggregated by MOVES3 output categories for the comparison. The evaporative
emissions from both models show similar trends and magnitudes. As described earlier,
EMFAC2021 uses California-specific inputs (meteorology, vehicle activity, and fuel
information) and assumes model year distributions that are different from MOVES3.
These are some of the major reasons for the differences between EMFAC2021 and
MOVES3 emission outputs.
Figure 4.3.2-2. Comparison of statewide evaporative emissions between EMFAC2021
and EMFAC2017. EMFAC2017’s RESTLOSS is included in DIURN for the comparison.

94
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure 4.3.2-3. Comparison of evaporative emissions of Passenger Cars (LDA) in Los

Angeles county between EMFAC2021 and MOVES3.

4.3.3 Light Duty CO2 Emission Rates

The GHG module was first introduced in EMFAC2017 and staff have further updated fuel
efficiency assumptions in the model through the EMFAC2021 update. The current
update to GHG module improved the accuracy of CO2 emission rates and provided new
emission rates for CO2 for all model year vehicles as old as 2009. The new approach
entails using U.S. EPA 5-cycle fuel economy ratings for motor vehicles to estimate the
total CO2 emissions.
EMFAC2021 is developed assuming complete combustion of fuel, similar to
EMFAC2017, and consistent with the official CARB, U.S. EPA, and IPCC methodologies.
Complete combustion of fuel means that 100% of carbon in the fuel is converted to
CO2. Additionally, EMFAC2021 assumes that using EPA fuel economy ratings provides
an acceptable basis for estimation of amount of burned fuel.
Moreover, EMFAC2017 used EPA 2-cycle fuel economy ratings, while EMFAC2021 takes
advantage of 5-cycle ratings. 5-cycle fuel economy ratings are more representative of
real-world condition due to accounting for effects of air conditioning and more frequent

95
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

vehicle acceleration events. Hence, the resulting CO2 emissions are assumed to be more
representative of California driving style.
In order to obtain a fuel economy rating for each vehicle that operates in California, DMV
registration database was first queried and a list containing all vehicle VINs associated
with passenger and light duty trucks operating in California was obtained. The queries,
other than VIN, provided other vehicle characteristics such as model name, series name,
model year, and fuel type. More detailed vehicle characteristics such as engine size, drive
train type, and number of cylinders were obtained from Polk/IHS VINtelligence service
and were appended to the DMV vehicle records as shown in Figure 4.3.3-1.
Next, the latest copy of EPA fuel economy data was obtained from EPA
fueleconomy.gov, and was queried for each of the vehicle records from the previous
step. Using the collected vehicle characteristics for each vehicle in the DMV database
(including parameters extracted from the VINtelligence), EPA fueleconomy.gov data
base was then queried using an advanced string-matching algorithm to link the most
similar vehicles in terms of their vehicle characteristics. This algorithm is further described
in Chapter 2.4 of SB 1014 Clean Miles Standard, 2018 Base Year Emission Inventory
Report. 49 Subsequently, the 5-cycle city MPGs of the most similar matches/vehicles from
EPA fueleconomy.gov were assigned to the input VINs for all vehicles. The analysis was
also quality checked to ensure that correct 5-cycle fuel economy values are assigned to
each vehicle. Details of the data processing flow in illustrated in Figure 4.3.3 2. The 5-
cycle fuel economy values were then converted to CO2 emission rates (g CO2/mile)
using a fuel specific conversion factor of 8,887 grams CO2 per gallon of gasoline, and
10,180 grams CO2 per gallon of diesel. 50
Figure 4.3.3-1. Flow chart to show how each DMV vehicle record a fuel economy
rating was obtained

Vehicle Identification Number (VIN)

Series Name etc.

Make Model Year

Model

49
SB1014 Base Year Emissions Inventory https://ww2.arb.ca.gov/sites/default/files/2019-
12/SB%201014%20-%20Base%20year%20Emissions%20Inventory_December_2019.pdf
50
U.S. EPA website, https://www.epa.gov/energy/greenhouse-gases-equivalencies-calculatorcalculations-
and-references
96
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure 4.3.3-2. Data processing flow chart to determine fuel economy ratings for the
DMV vehicle fleet

VINtelligence

Vehicle Specs incl. transmission type, # of cylinders, fuel type,

advance vehicle type, drivetrain type
VIN

VINs
+
VINs
Fuel
Vehicle class: passenger

Economies
only if : MY 2012-2019,

car and light duty truck

Model Name, Series

DMV and VINtelligence

Name, Model Year

EPA Fuel Economy

VIN, Make Name,

All vehicle specs from

(FuelEconomy.GOV)
including VIN

VIN and fuel

Economies
Fuel economies
and vehicle specs

Advanced
String
DMV Matching
Algorithm

EMFAC2017 CO2 emissions were calculated using the 2-cycle (i.e., unadjusted) U.S. EPA
city fuel as opposed to the 5-cycle (i.e., adjusted) fuel economies. For 5-cycle fuel
economies, EPA added three additional cycles (cold FTP, US06, and SC03) to the 2-cycle
fuel economy tests to further reflect real world condition that impact fuel efficiency of
light duty vehicles. This was done by considering a wider range of driving conditions such
as more frequent acceleration events, impact of air conditioning and heater usage, and
hot and cold temperatures. The results from the five cycle tests are then combined to
derive the adjusted 5-cycle fuel economy rating.
Moreover, EMFAC models CO2 running exhaust emissions using Phase 2 of UC cycle.
Considering that the FTP composite CO2 emissions are highly correlated with the UC
Phase 2 CO2 emissions, 51 EMFAC2017 used a combination of 2-cycle city and highway
fuel economy ratings to calculate CO2 emission rates. In contrast, to account for more
city-like driving behavior in California, EMFAC2021 uses the adjusted 5-cycle city fuel
economies to derive CO2 emissions. This method now ensures that the EMFAC’s
estimate of CO2 emissions for different model year vehicles accounts for real world
conditions impacting them.

51
Use of 2-cycle fuel economies and correlation between FTP composite and UC cycle is thoroughly
discussed in SB 1014 Clean Miles Standard, 2018 Base Year Emission Inventory Report. See page 29.
Download https://ww2.arb.ca.gov/sites/default/files/2019-
12/SB%201014%20-%20Base%20year%20Emissions%20Inventory_December_2019.pdf
97
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Comparison of emissions between EMFAC2017 and EMFAC2021 CO2 emission rates for
years 2008 through 2026 for passenger cars (a) and light duty trucks (b) is shown in
Figure 4.3.3 3. Please note that CO2 emission data for model year 2020 through 2026
are projected using the GHG emission standards. As can be seen prior to 2019,
EMFAC2017 was on a different trajectory as compared to EMFAC2021 for both
passenger cars and light duty trucks. This is mainly due to the latest methodology
updates that included the use of 5-cycle fuel economy values versus previously using 2-
cycle data, and improvements made by using the advanced string matching algorithm,
and specially adjusting the CO2 rates for the differences between the certification fuel
(E0) versus commercially available fuel which is assumed to be E10. Although, the impact
of these updates and improvements extends into the future, the final SAFE rule and six
manufacturer Framework Agreement on Clean Cars account for the major difference in
CO2 emission rates of EMFAC2017 and EMFAC2021 post 2020.
For passenger cars, EMFAC2021 shows a more moderate reduction rates of
approximately 6.6 grams per model year in CO2 emissions as compared to 8.34 grams
per model year of EMFAC2017 for passenger cars. For light duty trucks, however, CO2
emission rates reduce at a rate of roughly 13.2 grams per model year, while EMFAC2021
reduction in CO2 emission rates is approximately 12.1 grams per model year.
In terms of fuel economy, as shown in Figure 4.3.3-3, for passenger cars, EMFAC2021
average fleet fuel economy by model year increases at a slower pace of approximately
0.8 mpg per model year as compared to roughly 1.1 mpg per model year for
EMFAC2017 for passenger cars. For light duty trucks, however, EMFAC2017 fuel
economies grow 0.89 mpg per model year, while EMFAC2021 growth rate of fuel
economy is approximately 0.84 mpg per model year.

98
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure 4.3.3-3. CO2 emission factors by vehicle model year for (a) passenger cars and
(b) light duty trucks for EMFAC2017 and EMFAC2021

a) Passenger Cars
400 EMFAC2017
EMFAC2021
300
CO2 (g/mile)

200

100

0
2000 2005 2010 2015 2020 2025
Model Year

b) Light Duty Trucks

600

500
CO2 (g/mile)

400

300

200
EMFAC2017
100
EMFAC2021
0
2000 2005 2010 2015 2020 2025
Model Year

99
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure 4.3.3-4. Fuel Economy by vehicle model year for (a) passenger cars and (b)
light duty trucks for EMFAC2017 and EMFAC2021

a) Passenger Cars
60
Fuel Economy (mile/gallon)

EMFAC2017
50 EMFAC2021
40
30
20
10
0
2000 2005 2010 2015 2020 2025
Model Year

b) Light Duty Trucks

45
Fuel Economy (mile/gallon)

40
35
30
25
20
15
EMFAC2017
10
5 EMFAC2021
0
2000 2005 2010 2015 2020 2025
Model Year

The Final 2020 SAFE Rule modified the tailpipe carbon dioxide (CO2) emission targets
for passenger cars (PC) and light trucks (LT) for the 2021 to 2026 model years. Actual
fleet CO2 targets are based on current fleet average footprints (FP) of 46.2 ft2 for PC and
53.8 ft2 for LT vehicles. These targets were subsequently used to derive year-over-year
(YoY) % reductions in CO2 targets for model years 2021-2026 in the PC and LT fleets.
Under the Final SAFE Rule, a 1.5% YoY improvement in fuel economy in the CAFE
standards was finalized which translates to a slightly different YoY reduction in CO2

100
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

space (e.g., grams of CO2 emitted/mile which is more akin to the inverse of fuel
economy (e.g., miles/gallon). Derived from the finalized CO2 standards rather than the
finalized CAFE standards, a 1.84% YoY reduction from 2020 to 2026 for the CO2
emission factor (EF) values of gasoline passenger cars were applied. For light trucks, a
1.75% YoY reduction was applied. EMFAC2021 also accounts for the six manufacturer
Framework Agreements on Clean Cars on reducing CO2 emissions more stringently than
those in the SAFE Rule. Specifically, these manufacturers have agreed to meet CO2
emission targets for model years 2022 through 2026 that collectively represent a 3.7%
YoY reduction between 2021 and 2026. In addition, these manufacturers are allowed to
meet up to 1.0% of those reductions using ZEV credits. As a result, a 2.7% YoY reduction
was applied from model years2022 to 2026 for the CO2 emission factor values of
gasoline passenger cars and light duty trucks subject to the framework. Based on the
CARB 2018 GHG compliance data, 38% and 62% of sales splits are assumed between
the six manufacturer framework and the non-framework gasoline vehicles in the
California fleet for light duty passenger cars, respectively; and 34% versus 66% for the
light trucks categories, are assumed to be the split between framework and non-
framework vehicles respectively.
As mentioned above, EMFAC2021 incorporated EPA rated 5-cylce fuel economy. The
accuracy of 5-cycle fuel economy is further evaluated against real-world fuel economy
data. Following the On Board Diagnostic (OBD) II regulation updates that were adopted
by the Board in 2016, the OBD systems on new vehicles (starting from model year 2019)
are expected to record fuel consumption and distance traveled, with a phase-in schedule
of 30% in MY2019, 60% in MY2020, and 100% in MY2021 52. Staff was able to extract the
OBD data for a total of 73,000 vehicles of model years 2019 and 2020 from California’s
Bureau of Automotive Repair (BAR) Smog Check database. Data with known issues,
duplicate records, and vehicles that travelled less than 6000 km were removed, resulting
in a dataset of 41,000 vehicles.
The fuel economy matching was done in two steps. First, for each vehicle, the OBD real-
world fuel economy was calculated as lifetime miles travelled divided by lifetime gallons
of fuel consumed. As part of the second step, staff matched EPA rated 5-cycle combined
fuel economy to vehicles in the OBD dataset using VINtelligence 53, a VIN decoder, and
did manual corrections based on vehicle specifications and EPA Fuel Economy Guide 54.
This step resulted in a sample size of approximately 34,000 gasoline vehicles. Finally, the
EPA rated fuel economy of 34,000 gasoline light-duty vehicles were evaluated against
OBD real-world fuel economy. .

52
https://ww3.arb.ca.gov/regact/2015/obdii2015/finalregorder2.pdf?_ga=2.6318005.628219904.16076136
41-982443962.1602979098
53
https://ihsmarkit.com/products/automotive-vin-interpretation-decoding.html
54
https://www.fueleconomy.gov/

101
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure 4.3.3-5 shows a bubble plot for the comparison of real-world fuel economy versus
EPA 5-cycle fuel economy of 10 major manufacturers with number of data points in the
OBD dataset greater than 1,000. Interestingly, the current data demonstrated that the
EPA 5-cycle fuel economy is close to real-world fuel economy. Furthermore, the analysis
showed significant correlation with an R-square value of 0.84. While there are over-
performing and underperforming models, the average difference between EPA rated
fuel economy (27.35 mpg) and OBD real-world fuel economy (26.90 mpg) is only 0.45
mpg for these datasets. Therefore, this analysis justifies that EPA 5-cycle fuel economy is
a better representation of real-world fuel efficiency as compared to the EPA 2-cycle fuel
economy.
Figure 4.3.3-5. A Comparison of real-world fuel economy versus EPA 5-cycle fuel
economy 55.

4.3.4 Motorcycle Emission Rates

The motorcycle emission rates currently in EMFAC2017 have not been updated since
2000. Since EMFAC2000, the model had used emission rates resulting from analysis of

55
Each bubble represents a model in the model year 2019 or 2020. Different colors denote manufacturers
A-J. The sizes of the bubbles represent the number of vehicles in the OBD dataset. The dash line is the line
for 1:1 ratio.
102
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

FTP and UC dynamometer test data from about 125 model year 1998 and older
motorcycles tested in calendar year 1999. Additionally, the evaporative emission factors
used in EMFAC2017 and previous models were based on light-duty automobiles. To
provide updated motorcycle emission rates for EMFAC2021, CARB conducted an
internal study in 2019 at the Haagen-Smit Laboratory (HSL) in El Monte to test current
model motorcycles. The test plan intended to procure and test up to 18 motorcycles
from the in-use fleet (model years 2010-2018) and test 7 state-owned bikes (3 MY2019, 3
MY2008, and 1 MY2006) for exhaust and evaporative emissions. The dynamometer
exhaust tests included the UC, FTP, and the World Motorcycle Test Cycle (WMTC). For
EMFAC, the UC test data is used to analyze and update the model for motorcycles of
MY2008 and newer (fuel injected with catalyst). The evaporative tests included a 1-hour
hot soak test and a 3-day diurnal test. Due to COVID-19, the laboratory was closed in
March 2020, resulting in the partial completion of CARB’s motorcycle test plan.
However, dynamometer test data from 13 motorcycles were collected and analyzed to
estimate exhaust emission rates used to update EMFAC2021. Of these motorcycles, one
was tested in stock condition and tampered configuration (without catalyst), one was
tested in tampered condition only, while the remaining motorcycles were all tested in
stock condition.
Table 4.3.4-1 lists the 13 motorcycles tested in non-tampered stock condition with
corresponding engine size and odometer at the time of testing.
Table 4.3.4-2 lists two motorcycles tested in the tampered configuration. Note that
MCY1 was tested in both stock and tampered configurations. The state-owned
motorcycles were of model year 2008, 2018, 2019, and 2020. The motorcycle of model
year 2013 and 2016 were procured from private owners from the in-use fleet.

103
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Table 4.3.4-1. CARB Motorcycle List (Non-Tempered Stock Condition)

MCY MY Make Model Displ cc Odo
1 2008 Honda CBR600RR 600 10,466
2 2019 Harley Davidson Street Bob 1746 632
3 2016 Kawasaki NINJA 650 650 27,302
4 2018 Suzuki GSX R1000 1000 1,013
5 2019 Yamaha MT09 847 1,180
6 2013 Honda CBR250R 250 29,300
7 2020 Triumph Street Triple 765 596
8 2014 Triumph Daytona 675 675 5,052
10 2014 Yamaha Bolt 950 16,787
11 2015 Honda VT750 750 7,486
12 2015 Yamaha FZ-07 690 8,117
13 2011 BMW S1000RR 1000 11,115

Table 4.3.4-2. CARB Motorcycle List (Tampered Configuration)

MCY MY Make Model Displ cc Odo
1 2008 Honda CBR600RR 600 10,178
9 2008 Harley Davidson FXDWG 1584 3,437

For EMFAC, the results of UC dynamometer tests were used for analysis. Each
motorcycle was tested twice over the UC, and the results were averaged by each test
phase (UC Bag 1, Bag 2, Bag 3). The non-tampered motorcycles from dynamometer tests
were binned by odometer reading. The averaged emission rates (grams per mile) were
then compiled by bag and pollutant, as shown in the following tables. Similarly, the
average emission rates from the tampered motorcycles from were compiled for each UC
bag.

104
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Table 4.3.4-3. Non-Tampered MCY UC Bag 1 Emission Rates by Bin

Average
Process Bin Pollutant ER (g/mile)
Odometer
HC 0.725
NOx 0.119
2500 855.25
CO 3.29
CO2 306.93
HC 1.74
NOx 0.329
7500 7780.25
CO 8.52
CO2 239.19
HC 2.08
NOx 0.145
UC Bag 1 12500 11115
CO 10.71
CO2 359.13
HC 2.07
NOx 0.327
17500 16787
CO 12.67
CO2 238.04
HC 1.83
NOx 0.253
22500 28301
CO 18.00
CO2 186.77

105
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Table 4.3.4-4. Average Tampered MCY UC Bag 1 Emission Rates

Process Pollutant ER (g/mile)
HC 5.32
NOx 0.533
UC Bag 1
CO 47.85
CO2 276.60

Table 4.3.4-5. Non-Tampered MCY UC Bag 2 Emission Rates by Bin

Average
Process Bin Pollutant ER (g/mile)
Odometer
HC 0.067
NOx 0.036
2500 855.25
CO 0.491
CO2 201.87
HC 0.432
NOx 0.367
7500 7780.25
CO 3.76
CO2 167.00
HC 0.188
NOx 0.118
UC Bag 2 12500 11115
CO 0.829
CO2 236.36
HC 0.351
NOx 0.108
17500 16787
CO 3.72
CO2 162.06
HC 0.609
NOx 0.299
22500 28301
CO 12.38
CO2 133.11

Table 4.3.4-6. Average Tampered MCY UC Bag 2 Emission Rates

Process Pollutant ER (g/mile)
HC 2.67
NOx 0.823
UC Bag 2
CO 27.27
CO2 171.81

106
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Table 4.3.4-7. Non-Tampered MCY UC Bag 3 Emission Rates by Bin

Process Bin Pollutant ER (g/mile) Average Odometer
HC 0.294
NOx 0.056
2500 855.25
CO 0.564
CO2 261.67
HC 0.789
NOx 0.260
7500 7780.25
CO 2.99
CO2 210.07
HC 0.486
NOx 0.166
UC Bag 3 12500 11115
CO 0.755
CO2 308.59
HC 0.952
NOx 0.193
17500 16787
CO 6.14
CO2 195.94
HC 0.832
NOx 0.201
22500 28301
CO 10.82
CO2 160.55

Table 4.3.4-8. Average Tampered MCY UC Bag 3 Emission Rates

Process Pollutant ER (g/mile)
HC 4.34
NOx 0.527
UC Bag 3
CO 34.96
CO2 235.35

For non-tampered motorcycles, by UC bag, each pollutant was plotted by odometer bin
to determine a best fit trend to represent non-tampered emission rates by mileage. The
analysis of the CO2 UC test results yielded no significant trend in emission rates, so the
average of emission rates between the binned non-tampered motorcycles and the
tampered motorcycles was calculated. Table 4.3.4-9 below shows the regression
equations that are used in EMFAC2021 to calculate emissions of HC, NOx, CO, and CO2
for motorcycles.

107
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Table 4.3.4-9. Regression equations of HC, NOx, CO and CO2 emissions for
motorcycle
Process Pollutant A (X0) B (X1) Regression Type
HC -3.5865 0.5741 Log Linear
NOx -0.3103 0.0593 Log Linear
UC Bag1
CO -43.841 5.9258 Log Linear
CO2 271.309 0 Flat
HC -1.3164 0.179 Log Linear
NOx -0.3883 0.0625 Log Linear
UC Bag 2
CO -28.869 3.6008 Log Linear
CO2 175.95 0 Flat
HC -1.5668 0.2434 Log Linear
NOx -0.3391 0.056 Log Linear
UC Bag 3
CO -29.44 3.6649 Log Linear
CO2 231.364 0 Flat

Two other data sources were examined to supplement the analysis for updated emission
rates-one to provide information on tampering rates and the other to provide an
updated motorcycle odometer schedule. The first source involves two recent CARB
online studies of 2000 sales advertisements of motorcycles and their tampered
components (August 2016 to February 2017 and September 2019 to January 2020). This
study provided information to determine tamper rates by motorcycle age. By combining
the two studies, staff determined the percentage of tampered motorcycles by age.
Figure 4.3.4-1 shows the two tamper studies through age 20, and the best fit trend line
to represent tamper percentages by age.
Figure 4.3.4-1 road Motorcycle Tamper Percentage by Age

0.6

0.5
Percent Tamper Rate

0.4

0.3

0.2 Percent Tamper

0.1 Tamper Rate Trend
0.0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Age

108
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

The second source of information is the 2017 National Household Travel Survey-
California Add-on (NHTS CA), which provides motorcycle odometer mileage data to
create an odometer schedule by age. This data source contains a vehicle subset of 1,923
motorcycles surveyed for odometer information by age. 56 Figure 4.3.4-2 illustrates staff
analysis of the 2017 NHTS data results as compared to the odometer schedule used in
EMFAC2017. The updated odometer schedule for EMFAC2021 continues at a slightly
higher trend after age 10 than estimated by EMFAC2017.
Figure 4.3.4-2 Updated Motorcycle Odometer Schedule

5
Odometer (x10k miles)

2
EMFAC2021 Odo Schedule
1 EMFAC2017 Odo Schedule

0
0 5 10 15 20 25 30 35 40 45
Age

In order to calculate the weighted emission rates by age, the tamper rates were used to
calculate weighted average emission rates using the non-tampered and tampered
emission factors presented earlier. The equation to calculate the weighted emission rate
by bag is shown below.
Wt. ER = (Bag Non-Tamp*(1-Tamp Rate)) + (Bag Tamp*Tamp Rate) (Eq. 4.3.4-1)
Furthermore, regression equations were developed to model the weighted average
emissions rates as a function of odometer. These regressions are provided in the
following tables.

56
https://www.nrel.gov/transportation/secure-transportation-data/tsdc-nhts-california.html
109
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Table 4.3.4-10. On-road Motorcycle Weighted Emission Rates (g/mile) vs odometer

(10k miles)
UC Bag 1 Regression Type Equation
HC Log Linear y = 0.6919ln(Odometer) + 2.588
NOx Log Linear y = 0.0646ln(Odometer) + 0.3085
CO Log Linear y = 7.1147ln(Odometer) + 19.817
CO2 Flat y = 271.3

UC Bag 2 Regression Type Equation

HC Log Linear y = 0.3247ln(Odometer) + 0.9094
NOx Log Linear y = 0.097ln(Odometer) + 0.3438
CO Log Linear y = 4.3618ln(Odometer) + 9.9189
CO2 Flat y=175.95

UC Bag 3 Regression Type Equation

HC Log Linear y = 0.4848ln(Odometer) + 1.5789
NOx Log Linear y = 0.0672ln(Odometer) + 0.2625
CO Log Linear y = 5.091ln(Odometer) + 11.835
CO2 Flat y=231.3640

Overall, the new motorcycle emission rates embedded in EMFAC2017 are lower for HC
and NOx, and higher for CO and CO2.

4.3.5 HD Emission Rates

4.3.5.1 Light Heavy-Duty Diesel and Gasoline Vehicle Running Exhaust
Emission Rates
In EMFAC, LHD trucks are defined as vehicles with a gross vehicle weight rating (GVWR)
of 8,501-14,000 lbs. and further divided into LHDT1 (LHD trucks with GVWR 8,501-
10,000 lbs.) and LHDT2 (LHD trucks with GVWR 10,001-14,000 lbs.). Unlike the HD
vehicles with GVWR greater than 14,000 lbs., these vehicles are unique in terms of their
emission compliance. Some of these vehicles comply with emission regulations through
certification of their engines but others through chassis certification. In either case, they
must meet the respective engine or chassis emission standards. When chassis certified,
they are also referred to as medium duty vehicles 4 and 5 (MDV4 and MDV5) under
California Low Emission Vehicle (LEV) regulations. Chassis standards are considered to be
comparable in stringency to the corresponding engine standards, and in EMFAC all
vehicles with GVWR of 8,501-14,000 lbs. are modeled under LHD trucks whether they are
engine or chassis certified.

110
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

The emission rates for LHD trucks have not been updated since EMFAC2007. In recent
years, there have been increased interest in the emissions of criteria pollutants as well as
greenhouse gases (GHG) from these vehicles; therefore, it is desirable to have a better
understanding of their emission levels and their emissions impact on the overall on-road
emissions inventory. To support CARB regulatory programs and better serve
stakeholders’ needs, staff carried out a project to test in-use LHD trucks and collect
emissions data from the test vehicles. This section discusses the analysis of the emission
test data from the project and present revised emission factors based on the results of
test data analysis.
4.3.5.1.1 Emissions Test Data
CARB staff has completed an emissions testing project (Project 2R1702) to obtain
emission data from in-use LHD trucks. A total of 10 diesel and 2 gasoline LHD trucks of
several model years (MY) were tested. Each test vehicle was tested on dynamometer
over eight test cycles (Table 4.3.5-1) and emissions of THC, CO, NOx, PM, and CO2
were measured following the standard light-duty vehicle testing procedures.
Table 4.3.5-1. Test cycles used for dynamometer testing in CARB Project 2R1702
Test Cycle/Mode Average Speed (mph) Duration (sec) Length (mi)
FTP-75 21.2 1,877 11.04
UC 22.92 1,735 11.04
MAC1 6.8 798 1.50
MAC3 39 823 8.92
MFC5 57 517 8.14
MFC6 65 502 9.12
MFC7 73 515 10.45
HWFET 48.3 765 10.26

111
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Table 4.3.5-2 lists the vehicles procured for emissions testing, and dynamometer test
results are listed in Appendix 6.5.
Table 4.3.5-2. Test Vehicles for Project 2R1702
Emissions
Manufacturer Model MY Fuel Odometer Weight Class
STD
LEV3
FCA U.S. Dodge Ram 2500 2015 Diesel 51,380 LHDT1 / MDV4
ULEV340
Daimler AG Sprinter 2500 2017 Diesel 22,860 LEV2 ULEV LHDT1 / MDV4
General Motors Silverado 2500 2015 Diesel 64,600 LEV2 ULEV LHDT1 / MDV4
Ford F250 2015 Diesel 30,460 LEV2 ULEV LHDT1 / MDV4
Ford F350 2015 Diesel 70,630 LEV2 ULEV LHDT2 / MDV5
Daimler AG Sprinter 3500 2017 Diesel 8,630 LEV2 ULEV LHDT2 / MDV5
LEV3
FCA U.S. Dodge Ram 3500 2015 Diesel 139,340 LHDT2 / MDV5
ULEV570
General Motors Silverado 2500 2015 Gasoline 42,400 LEV3 LEV395 LHDT1 / MDV4
General Motors Sierra 2500 2015 Gasoline 43,110 LEV3 LEV395 LHDT1 / MDV4
General Motors Silverado 3500 2015 Diesel 24,570 LEV2 ULEV LHDT2 / MDV5
Ford F350 2006 Diesel 73,800 ULEV LHDT1 / MDV4
General Motors Silverado 2500 2006 Diesel 120,810 ULEV LHDT1 / MDV4

4.3.5.1.2 Running Exhaust Emission Rate

The emission rates of 2007+ MY LHD diesel trucks (USEPA 2007 standards) were
updated using the test data of six 2015 MY and two 2017 MY diesel test vehicles (Table
4.3.5-2). Plots of emission rates versus odometer readings of these test vehicles show
that it is not possible to use a regression method to develop meaningful zero mile rates
(ZMR) and deterioration rates (DR), and therefore the average emission rates of all test
vehicles were used to scale the ZMRs and DRs of LHD diesel trucks in EMFAC2017 to
obtain the new ZMRs and DRs, as described below in detail.
Additionally, as the eight test vehicles include four tested LHDT1 trucks and four LHDT2
trucks, the ZMRs and DRs calculated from the test data of all eight vehicles were further
split into two separate sets of ZMRs and DRs using the average CO2 emission rates of
the four LHDT1 trucks and four LHDT2 trucks. This is based on an assumption that to
some degree the emissions of criteria pollutants from a fleet tend to be positively related
to the amount of fuel consumed, which in turn is correlated with the CO2 emissions.
The following discusses staff updates to ZMRs and DRs of UC Bag 1 and Bag 2 for LHD
trucks.
UC Bag 1. First, for each pollutant, the UC bag 1 (B1) emissions test data of the eight
vehicles were averaged to obtain the pollutant’s UC B1 emission rate (UCB1ER*). Next,
as the average of the odometer readings of the eight vehicles is 54,700 miles, the
pollutant’s UC B1 emission rate at 54,700 mi (UCB1ER) was calculated from the B1 ZMR

112
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

and DR of 2007+ MY LHD diesel trucks in EMFAC2017. Finally, the ratio between
UCB1ER* and UCB1ER was used to scale the pollutant’s ZMR and DR of 2007+ MY LHD
diesel trucks in EMFAC2017 to obtain a new ZMR and a new DR. Using this method, new
ZMRs and DRs for all pollutants were calculated.
In addition, two weighting factors were calculated from the UC B1 CO2 emission rates of
the four tested LHDT1 trucks and four LHDT2 trucks, respectively, and the two factors
were then applied to the new ZMRs and DRs for all pollutants to obtain separate sets of
ZMRs and DRs for LHDT1 and LHDT2. Table 4.3.5-3 and Table 4.3.5-4 show the revised
ZMRs and DRs for 2007+ MY diesel LHDT1 and LHDT2, respectively, as well as the
corresponding ZMRs and DRs for LHDT1 and LHDT2 in EMFAC2017.
Table 4.3.5-3. UC Bag1 ZMR and DR of 2007+ MY diesel LHDT1
EMFAC2017 EMFAC2021
Pollutant
ZMR (g/mi) DR (g/mi/10K mi) ZMR (g/mi) DR (g/mi/10K mi)
HC 0.038 0.0080 0.14 0.031
CO 0.28 0.055 0.86 0.18
NOx 4.43 0.011 1.30 0.033
PM 0.10 0.0023 0.0048 0.0012
CO2 745 0.0 1,085 0.0

Table 4.3.5-4. UC Bag1 ZMR & DR of 2007+ MY diesel LHDT2

EMFAC2017 EMFAC2021
Pollutant
ZMR (g/mi) DR (g/mi/10K mi) ZMR (g/mi) DR (g/mi/10K mi)
HC 0.038 0.0080 0.15 0.035
CO 0.28 0.055 0.97 0.21
NOx 4.43 0.011 1.46 0.038
PM 0.10 0.0023 0.0054 0.0013
CO2 745 0.0 1,222 0.0

UC Bag 2 The UC Bag 2 (B2) ZMRs and DRs of all pollutants for 2007+ MY were revised
in the same way as described above, and the results are shown in Table 4.3.5-5 and
Table 4.3.5-6.
Table 4.3.5-5. UC Bag2 ZMR and DR of 2007+ MY diesel LHDT1
EMFAC2017 EMFAC2021
Pollutant
ZMR (g/mi) DR (g/mi/10K mi) ZMR (g/mi) DR (g/mi/10K mi)
HC 0.11 0.0022 0.066 0.0014
CO 0.23 0.045 0.088 0.017
NOx 0.19 0.011 0.44 0.025
PM 0.071 0.0017 0.0069 0.0017
CO2 642 0.0 932 0.0

113
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Table 4.3.5-6. UC Bag2 ZMR and DR of 2007+ MY diesel LHDT2

EMFAC2017 EMFAC2021
Pollutant
ZMR (g/mi) DR (g/mi/10K mi) ZMR (g/mi) DR (g/mi/10K mi)
HC 0.10 0.0022 0.076 0.0016
CO 0.23 0.045 0.10 0.020
NOx 0.19 0.010 0.51 0.029
PM 0.071 0.0017 0.0080 0.0019
CO2 642 0.0 1,082 0.0
The emission rates of 2004-2009 MY LHD diesel trucks (or ULEV) were updated using the
test data of two 2006 MY diesel test vehicles (Table 4.3.5-2). Similar to the revision of
emission rates of 2007+ MY diesel LHDT1 and LHDT2, the ZMRs and DRs of 2004-2009
MY diesel LHDT1 and LHDT2 were revised by scaling the corresponding ZMRs and DRs
in EMFAC2017 using the test data of the two 2006 MY test vehicles. No attempt was
made to calculate separate sets of ZMRs and DRs for LHDT1 and LHDT2 and therefore
the obtained ZMRs and DRs apply to both these two LHD truck categories. Table 4.3.5-7
show the revised B1 ZMRs and DRs and Table 4.3.5-8 shows the revised B2 ZMRs and
DRs for the 2004-2009 MY diesel LHDT1 and LHDT2.
Table 4.3.5-7. UC Bag1 ZMR and DR of 2004-2009 MY diesel LHDT1 & LHDT2
EMFAC2017 EMFAC2021
Pollutant
ZMR (g/mi) DR (g/mi/10K mi) ZMR (g/mi) DR (g/mi/10K mi)
HC 0.14 0.0075 0.31 0.017
CO 1.00 0.055 1.84 0.10
NOx 4.43 0.011 7.85 0.019
PM 0.0999 0.0023 0.19 0.0043
CO2 745 0.0 1,137 0.0

Table 4.3.5-8. UC Bag2 ZMR and DR of 2004-2009 MY diesel LHDT1 & LHDT2
EMFAC2017 EMFAC2021
Pollutant
ZMR (g/mi) DR (g/mi/10K mi) ZMR (g/mi) DR (g/mi/10K mi)
HC 0.12 0.0066 0.23 0.012
CO 0.82 0.045 0.73 0.040
NOx 4.33 0.010 3.69 0.0089
PM 0.071 0.0017 0.15 0.0036
CO2 642 0.0 687 0.0

As shown in Table 4.3.5-5 to Table 4.3.5-8, compared to EMFAC2017, NOx emission

rates for MY2007+ have increased while those for MY2004-2009 LHD diesels have
decreased. In terms of diesel PM, the 2004-2009 model year LHD vehicles show
significantly higher PM emissions (~2x higher) than what was assumed in EMFAC2017
model.

114
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

The emission rates of 2008+ MY LHD gasoline trucks were updated using the test data of
two 2015 MY gasoline test vehicles (Table 4.3.5-2). Similar to the emission rate update
for 2007+ MY diesel LHDT1 and LHDT2, the emission rates for 2008+ MY gasoline
LHDT1 and LHDT2 were revised by scaling the coefficients of the equations between
emission rate and odometer in EMFAC2017. The same scaling factors were used for both
LHDT1 and LHDT2. Table 4.3.5-9, Table 4.3.5-10, and Table 4.3.5-11 show the revised
emission rates of 2008+ MY gasoline LHDT1 and LHDT2 at zero and 100,000 miles and
the corresponding rates used in EMFAC2017.
Table 4.3.5-9. UC Bag1 emission rates of 2008+ MY gasoline LHDT1 and LHDT2 at
zero and 100,000 miles
EMFAC2017 Emission Rate (g/mi) EMFAC2021 Emission Rate (g/mi)
Pollutant
at 0 miles at 100K miles at 0 miles at 100K miles
HC 0.30 0.44 0.51 0.76
CO 7.17 7.67 16.3 17.5
NOx 0.39 0.53 0.31 0.42
PM 0.023 0.023 0.003 0.003
CO2 810 810 1,128 1,128

Table 4.3.5-10. UC Bag2 emission rates of 2008+ MY gasoline LHDT1 and LHDT2 at
zero and 100,000 miles
EMFAC2017 Emission Rate (g/mi) EMFAC2021 Emission Rate (g/mi)
Pollutant
at 0 miles at 100K miles at 0 miles at 100K miles
HC 0.015 0.016 0.011 0.012
CO 0.15 0.16 0.64 0.68
NOx 0.034 0.050 0.011 0.016
PM 0.004 0.004 0.002 0.002
CO2 873 873 707 707

Table 4.3.5-11. UC Bag3 emission rates of 2008+ MY gasoline LHDT1 and LHDT2 at
zero and 100,000 miles
EMFAC2017 Emission Rate (g/mi) EMFAC2021 Emission Rate (g/mi)
Pollutant
at 0 miles at 100K miles at 0 miles at 100K miles
HC 0.027 0.038 0.018 0.025
CO 0.10 0.11 1.61 1.72
NOx 0.13 0.20 0.009 0.013
PM 0.020 0.020 0.002 0.002
CO2 799 799 1,020 1,020

As shown, EMFAC2021 is showing lower NOx emissions for gasoline LHD vehicles as
compared to EMFAC2017.

115
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

4.3.5.1.3 Speed Correction Factors

EMFAC calculates vehicle running exhaust emissions by multiplying emission rate in g/mi
by vehicle mile travelled (VMT). Since VMTs are distributed across the entire spectrum of
vehicle driving speeds, emission rates at different speeds are needed to match the VMTs
at different speeds. Emission rates of a pollutant at various speeds are calculated by
applying SCFs to the BERs of that pollutant. An SCF for a pollutant is developed from
the pollutant’s emission rates measured over several dynamometer test cycles with
different average speeds.
For LHD trucks, all vehicles were tested over the FTP cycle as well as several other test
cycles with average speeds either lower or higher than that of the FTP. A pollutant’s
emission rates from all the test cycles were first plotted as a function of speed, and
regression curves were then fitted to find equations best representing the data. Data
fitting shows that for all pollutants, quadratic polynomials could best represent the
variations of emission rate with speed. For each pollutant, the emission rates calculated
from the polynomial for various speeds were then normalized to 16.0 mph (the average
speed of FTP Bag 2) to yield SCFs for the pollutant. The following equation can be used
to calculate the SCFs for all the pollutants:
𝑺𝑺𝑺𝑺𝑺𝑺 = 𝑨𝑨 ∙ 𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝟐𝟐 + 𝑩𝑩 ∙ 𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔 + 𝑺𝑺 (Eq.
4.3.5-1)
where A, B, and C are coefficients of the polynomials. The numerical values of these
coefficients for all pollutants and respective MY groups are provided in Table 4.3.5-12.
Table 4.3.5-12. Coefficients speed correction factor equations for light heavy-duty
diesel trucks
Pollutant Model Year Fuel Type A B C
HC 2007+ Diesel 1.37x10-4 -3.12x10-2 1.71
HC 2004-2009 Diesel 5.76x10-4 -6.36x10-2 2.09
HC 2008+ Gasoline 2.69x10-3 -0.260 6.50
CO 2007+ Diesel 1.43x10-3 -0.168 5.26
CO 2004-2009 Diesel 6.96x10-4 -7.67x10-2 2.31
CO 2008+ Gasoline 4.64x10-2 -0.366 85.3
NOx 2007+ Diesel 1.73x10-3 -0.164 4.93
NOx 2004-2009 Diesel 3.52x10-4 -3.84x10-2 1.74
NOx 2008+ Gasoline 4.32x10-4 -2.73x10-2 1.27
PM 2007+ Diesel 4.04x10-4 -4.66x10-2 1.78
PM 2004-2009 Diesel 9.35x10-4 -9.79x10-2 3.12
PM 2008+ Gasoline 3.14x10-3 -0.265 8.56
CO2 2007+ Diesel 5.21x10-4 -5.22x10-2 1.97
CO2 2004-2009 Diesel 6.00x10-4 -6.08x10-2 2.13
CO2 2008+ Gasoline 5.55x10-4 -5.69x10-2 2.04

116
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

4.3.5.2 Heavy Heavy-Duty and Medium Heavy-Duty Diesel Truck Running

Exhaust Emission Rates and Speed Correction Factors
In EMFAC2017, staff updated the BERs of running exhaust emissions and SCFs for 2010+
MY HHD (T7) diesel trucks using emissions test data from CARB TBSP and emissions
testing of late model diesel trucks by CARB and EMA (Engine and Truck Manufacturer
Association) 57. However, due to a lack of test data, the BERs for 2010+ MY MHD (T6)
diesel trucks were estimated by scaling the rates of HHD diesel trucks and the SCFs of
HHD diesel trucks were assumed to be also applicable to MHD diesel trucks.
Since the release of EMFAC2017, additional late model HHD diesel trucks were tested in
TBSP. Following a pilot phase, TBSP has become a regular ongoing truck emissions
surveillance program. It primarily performs testing of Class 8 trucks (HHD trucks) and
large buses but also conducts some testing of Class 6 or Class 7 trucks as needed. To
better serve the needs of CARB programs and regulations, staff has recently developed a
new surveillance program, Class 4-6 SP, to focus on the testing of MHD diesel trucks.
The program has started producing valuable emissions data for modeling the emissions
of MHD trucks in EMFAC.
4.3.5.2.1 Emissions Test Data
All new emissions test data of HHD diesel trucks came from TBSP. To date, emissions
data from an additional twenty-six 2013+ MY HHD diesel trucks were obtained from the
program. All 26 trucks were tested on dynamometer over six test cycles for the emissions
of THC, CO, NOx, PM, and CO2 among some other species. The key parameters of the
six test cycles are shown in Table 4.3.5-13, and all test vehicle information and measured
emissions of five pollutants are listed in Appendix 6.6.
Table 4.3.5-13. Test cycles for dynamometer testing in CARB TBSP
Test Cycle/Mode Average Speed (mph) Duration (sec) Length (mi)
UDDS 18.8 1060 5.54
Creep 1.8 253 0.12
Near Dock Drayage 6.6 3,046 5.59
Local Drayage 9.3 3,362 8.70
HHDT Cruise 39.9 2,083 23.1
Modified HS Cruise 47.9 1,560 20.8

The new emissions test data of MHD diesel trucks came from both CARB’s Class 4-6 SP
and TBSP, with each contributing data for four of the eight tested trucks, all of which had
an engine of 2013 or later MY. The MHD diesel truck testing in Class 4-6 SP were carried

57
https://ww3.arb.ca.gov/msei/downloads/emfac2017-volume-iii-technical-documentation.pdf

117
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

out on a dynamometer over seven test cycles and emissions of THC, CO, NOx, PM, and
CO2, among other species, were measured. Key parameters for the seven test cycle are
shown in Table 4.3.5-14, and the test vehicle information and the measured emission
rates of the five pollutants are listed in Appendix 6.7.
Table 4.3.5-14. Test cycles for dynamometer testing in CARB Class 4-6 Program
Test Cycle/Mode Average Speed (mph) Duration (sec) Length (mi)
UDDS 18.8 1,060 5.54
Creep 1.8 253 0.124
Parcel Delivery 10.1 2,552 7.09
Transient 15.4 668 2.85
Local 32.6 1,690 15.3
HDDT Cruise 39.9 2,083 23.1
Modified HS Cruise 47.9 1,560 20.8

4.3.5.2.2 HHD and MHD Running Exhaust Emission Rates

In EMFAC2021, except for NOX, the BER of running exhaust emissions for a given model
year of heavy-duty trucks can be calculated using the following equation:
𝐵𝐵𝐵𝐵𝐵𝐵 = (𝑍𝑍𝑍𝑍𝐵𝐵 + 𝐷𝐷𝐵𝐵 × 𝑂𝑂𝑂𝑂𝑂𝑂) (Eq. 4.3.5-2)
where Odo is the average odometer of all trucks within that model year.
ZMR and DR are typically developed on a model year group basis, with each group
including several consecutive model years that usually share the same emission standards
and/or emission control technology. An average emission rate (ERavg) of all tested trucks
in a given MY group and an average odometer (Odoavg) of these trucks are first
calculated, and from these two averages the ZMR and DR for that MY group can be
calculated using an HD emissions deterioration model.
4.3.5.2.3 HHD Diesel Truck ZMR and DR
Considering that the additional 26 HHD trucks tested in TBSP since the release of
EMFAC2017 were 2013 MY or newer, staff decided to only revise the running exhaust
emission rates of 2013+ MY. For EMFAC2017, there were eighteen 2013+ MY HHD
trucks that provided test data for emission rate update. Therefore, dynamometer test
data from the twenty-six 2013+ MY newly tested HHD trucks in TBSP were merged with
the data from the eighteen 2013+ MY HHD trucks in EMFAC2017 updating to form a 44-
truck dataset for EMFAC2021 updating.
On the basis of engine model year, these 44 trucks were divided into two MY groups: a
2013-2015 MY group and a 2016+ MY group. This grouping is aligned with the HD OBD
requirements, with the phase-in period for 2013-2015 MY and full implementation for

118
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

2016 and later MY 58. For each group, the UDDS emission rates of all tested trucks as well
as their odometer readings were averaged to obtain the ERavg and Odoavg.
For NOx, staff revised the deterioration model used in EMFAC2017 based on the OBD
data obtained from a CARB sponsored study, and a non-linear model was developed
from the data to determine NOx ZMR and DR for NOx. Details are provided in Section
4.3.6 of this document.
For THC, CO, and PM, the HD diesel truck emissions deterioration model of EMFAC2017
was used to calculate the ZMRs and DRs for these pollutants using the equations below:
𝑍𝑍𝑍𝑍𝐵𝐵 = 𝐵𝐵𝐵𝐵𝑎𝑎𝑎𝑎𝑎𝑎 / (1 + 𝐵𝐵𝐸𝐸𝐵𝐵 ∙ 𝑂𝑂𝑂𝑂𝑂𝑂𝑎𝑎𝑎𝑎𝑎𝑎 ) (Eq. 4.3.5-3)
𝐷𝐷𝐵𝐵 = (𝐵𝐵𝐵𝐵𝑎𝑎𝑎𝑎𝑎𝑎 − 𝑍𝑍𝑍𝑍𝐵𝐵)/𝑂𝑂𝑂𝑂𝑂𝑂𝑎𝑎𝑎𝑎𝑎𝑎 (Eq. 4.3.5-4)
where EIR is emission impact rate, which is used in EMFAC to quantify the emission
deterioration of HD trucks. Unlike NOx, for THC, CO, and PM, separate sets of ZMRs
and DRs were calculated for the 2013-2015 MY group and 2016+ MY group. Table
4.3.5-15 shows the revised ZMR and DR for 2013+ MY HHD diesel trucks. For
comparison, the corresponding ZMRs and DRs are also shown in the table.
Table 4.3.5-15. Revised ZMRs (g/mi) and DRs (g/mi/10K mi) for heavy heavy-duty
diesel trucks
EMFAC Model HC CO NOx PM
Model Year ZMR DR ZMR DR ZMR DR ZMR DR
2013-15 0.013 0.00012 0.20 0.0017 0.0041 0.00015
EMFAC2021 0.6 n/a*
2016+ 0.011 0.00010 0.13 0.0011 0.0029 0.00011
2013-14 0.019 0.00025 0.28 0.003 2.67 0.050 0.0025 0.00011
EMFAC2017
2015+ 0.019 0.00021 0.28 0.003 2.68 0.046 0.0025 0.00010

*See Section 4.3.6 for details on NOx deterioration

4.3.5.2.4 MHD Diesel Truck ZMR and DR

As discussed earlier, the eight MHD diesel trucks had engines with model years ranging
from 2013 to 2017. In addition, an examination of the test data suggested that it was
more appropriate to analyze all these trucks as a single 2013+ MY group. Thus, for this
EMFAC update staff developed the ZMRs and DRs of MHD diesel trucks for a 2013+ MY
group only and left these rates for 2010-2012 unchanged from EMFAC2017.
Similar to the calculations for HHD diesel trucks, the UDDS emission rates of the eight
tested trucks and their odometer readings were averaged to obtain the ERavg and Odoavg.
The calculations of ZMRs and DRs of all gaseous pollutants and PM for MHD diesel trucks
are the same as those used for HHD trucks. Table 4.3.5-16 shows the revised ZMRs and

58
https://ww3.arb.ca.gov/regact/2012/hdobd12/hdobdiiisor.pdf
119
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

DRs for 2013+ MY MHD diesel trucks. For comparison, also include in the table are the
ZMRs and DRs of EMFAC2017.
Table 4.3.5-16. Revised ZMRs (g/mi) and DRs (g/mi/10K mi) for medium heavy-duty
diesel Trucks

EMFAC HC CO NOx PM
Model Year
Model ZMR DR ZMR DR ZMR DR ZMR DR

EMFAC2021 2013+ 0.0044 0.00017 0.039 0.0015 0.15 N/A* 0.0016 0.00010

2013-14 0.0088 0.00025 0.12 0.0025 1.52 0.070 0.0014 0.00015

EMFAC2017
2015+ 0.0088 0.00023 0.12 0.0020 1.48 0.065 0.0014 0.00012
*
See Section 4.3.6 for details on NOx deterioration
4.3.5.2.5 HHD and MHD Diesel Truck Speed Correction Factors
As described earlier, the running exhaust emissions is calculated by multiplying BER in
g/mi by VMT, and emission rate at a specific speed is obtained by applying an SCF to the
BER calculated from the emission rate at the UDDS speed (18.8 mph). HHD trucks
typically are tested over the UDDS as well as a several other cycles that have an average
speed either lower or higher than the UDDS speed. The emission rates of all the cycles
are normalized to the UDDS rate to yield SCF.
For this EMFAC update, the SCFs of HC, CO, NOx, PM, and CO2 for HHD diesel trucks
were developed using the emissions test data from 41 of the 44 trucks used for running
exhaust emission rate calculations. Three of the trucks were from the testing by CARB
and EMA (see earlier), which used a set of testing cycles different from that used in TBSP.
As discussed earlier, staff had to combine the 2013-2015 MY and 2016+ MY group into a
single 2013+ MY group in calculating NOx running exhaust emission rate but used the
two separate MY groups for the emission rates of other pollutants; therefore, the SCF for
NOx was also calculated for a single 2013+ MY group and the SCFs for the other
pollutants for a 2013-2015 MY group and a 2016+ MY group.
For a given MY group, a pollutant’s emission rates of all test cycles were first plotted
versus the cycles’ speeds. Curves were then fitted to find the equations best
representing the data. In finding the empirical curves that best relates the emission rates
to speeds, a two-segment curve was used in order to reasonably fit all the data points.
Based on data fitting, it was found that for all pollutants Eq. 4.3.4-5 should be used to
calculate SCFs for speeds below 18.8 mph and Eq. 4.3.4-6 for speed between 18.8 and
65 mph.
𝑺𝑺𝑺𝑺𝑺𝑺 = 𝑨𝑨 ∙ 𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝐁𝐁 (Eq. 4.3.5-5)

𝑆𝑆𝑆𝑆𝑆𝑆 = 𝑆𝑆 ∙ 𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑂𝑂 2 + 𝐷𝐷 ∙ 𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑂𝑂 + 𝐵𝐵 (Eq. 4.3.5-6)

120
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

where A, B, C, D, and E are coefficients for the respective equations, and Table 4.3.5-17
lists the numeric values of these coefficients for calculating the SCF of NOx for 2013+
MY group and the SCFs of other 4 pollutants for 2013-2015 and 2016+ MY groups.
Table 4.3.5-17. Coefficients of speed correction factor equations for heavy heavy-
duty diesel trucks
5-18.8 mph 18.8-65 mph
Pollutant Model Year
A B C D E
2013-15 60.3 -1.40 4.83x10-4 -5.02x10-2 1.77
THC
2016+ 89.1 -1.53 8.92x10-4 -7.70x10-2 2.13
2013-15 31.2 -1.17 5.06x10-4 -6.02x10-2 1.95
CO
2016+ 33.2 -1.19 5.07x10-4 -6.12x10-2 1.97
NOx 2013+ 31.4 -0.874 2.16x10-3 -0.210 5.59
2013-15 2.05 -0.244 2.27x10-3 -8.84x10-2 1.86
PM
2016+ 4.63 -0.525 3.70x10-3 -0.182 3.11
2013-15 3.05 -0.380 2.92x10-4 -2.81x10-2 1.43
CO2
2016+ 2.96 -0.370 2.85x10-4 -2.59x10-2 1.39

The data analysis for SCFs of MHD diesel trucks is the same as that for HHD diesel
trucks. As is the case with the BER, SCFs of MHD diesel trucks were developed for only a
single 2013+ MY group due to the small number of trucks tested. The two equations (Eq.
4.3.5-5 and Eq. 4.3.5-6) used for HHD diesel trucks are also applicable for MHD diesel
trucks. Table 4.3.5-18 lists the numeric values of these coefficients for calculating the
SCFs of THC, CO, NOx, PM, and CO2 for 2013+ MY group of MHD diesel trucks.
Table 4.3.5-18. Coefficients of speed correction factor equations for medium heavy-
duty diesel trucks
Model Year 5-18.8 mph 18.8-65 mph
Pollutant
Group A B C D E
THC 2013+ 172 -1.75 1.05x10-3 -8.62x10-2 2.25
CO 2013+ 19.3 -1.01 4.04x10-4 -4.58x10-2 1.72
NOx 2013+ 23.6 -1.08 8.78x10-4 -8.30x10-2 2.25
PM 2013+ 6.90 -0.659 2.59x10-3 -0.144 2.78
CO2 2013+ 4.02 -0.474 3.59x10-4 -3.09x10-2 1.46

4.3.5.3Heavy Heavy-Duty Diesel Truck Start Emission Rates

In EMFAC2017, StER of NOx for HD trucks were developed based on the emissions data
collected from CARB Project 2R1406 (In-Use Testing of Heavy-Duty Vehicles Certified to
Applicable 2010 Emission Standards). In the project, four trucks with 2011-2014 engine
MY were tested at CARB Depot Park facility using PEMS for the emissions of gaseous
pollutants. In this project staff performed testing on a route called DPTODP, which is an
uninterrupted round trip starting from Depot Park and covering a distance of about 15

121
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

miles before ending at Depot Park. For each test vehicle, start emission test was
conducted following soak times of 5, 10, 15, 30, 60, 120, 240, and 720 minutes.
In CARB’s TBSP, 11 HD diesel trucks of 2013+ engine model years were tested using
PEMS to collect start emissions data. Staff made two modifications to the testing
procedures of CARB Project 2R1406 based on the analysis of the project’s test data.
First, it was found that for all the test vehicles, the start emissions occurred within the
first 20 minutes of the runs, and thus a shortened version of the DPTODP route was used
for all the test runs in order to reduce testing time. Second, the 4 soak times between 5
and 30 minutes did not contribute a great deal of information in terms of establishing the
relationship between the soak time and start emissions (i.e., soak time curve); therefore,
for each vehicle only a 20-min soak time run was performed instead of the 4 runs done in
Project 2R1406 and a 480-min soak time run was added to better define the soak time
curve at that point.
Since the engines of all 11 trucks tested in CARB TBSP were 2013 or newer model year
engines, staff merged the start emission data of these 11 trucks with the data of three
trucks with 2013-2014 model year engines from Project 2R1406 to form a larger dataset
for updating the StERs of 2013+ engine model years. As mentioned earlier, the three
Project 2R1406 trucks were tested with a different set of soak times; as a result, a soak
time curve was first fit using each truck’s actual test data and then emissions at the re-
defined five soak times were then calculated from the curve for merging with the TBSP
data.
The method for analyzing start emission data is the same as that used in the EMFAC2014
and EMFAC2017 updates. Briefly, the NOx emissions during the start phase are
considered to include start emissions as well as running exhaust emissions, which are
emissions that would otherwise be emitted had the SCR reached operating
temperatures. Thus, for a test run, the StER is obtained by subtracting the NOx emission
rate of the running phase from the emission rate of the start phase. A detailed discussion
of the StER calculation method can be found in Section 3.2.3.6 of EMFAC2014 technical
documentation 59.
It should be mentioned that in previous EMFAC StER updates, SCR temperatures of a
test run was used to determine the duration of the start phase. However, for this update
SCR temperature data are available for only three of the 11 tested trucks. As a result,
staff had to use the exhaust temperature as a surrogate for SCR temperature using the
exhaust-SCR temperature relationship of the three trucks for which the SCR
temperatures were recorded.
As in EMFAC2017 update, start emissions were only found for NOx in the test data of
the 11 trucks with SCR engines but no incremental emissions increases were identified

59
https://ww3.arb.ca.gov/msei/downloads/emfac2014/emfac2014-vol3-technical-documentation-
052015.pdf
122
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

for THC, CO, and CO2. Thus, staff assumes no start emissions for THC, CO, and CO2.
PM emissions data were not measured in the PEMS testing of these 11 trucks, but the
dynamometer test data of these trucks show PM levels mostly near or at detection limit.
Since all these trucks were equipped with a DPF system, it is assumed that there were no
start emissions for PM.
The StERs of NOx by soak time for the four HD diesel trucks were determined for all test
runs and the results are provided in Table 4.3.5-19. In Figure 4.3.5-1 the calculated NOx
StERs are plotted as a function of soak time. EMFAC2017 StER data are also provided for
comparison.
Table 4.3.5-19. Heavy heavy-duty diesel truck NOx start emission rates by soak time
Start Emission Rate (g/hr)
EMFAC Model Model Year
20 min 120 min 240 min 480 min 720 min
2013-15 2.67 11.1 15.0 17.5 19.8
EMFAC20201
2016+ 4.74 11.2 14.2 17.8 19.6
EMFAC2017 2013+ 1.38 5.84 10.2 16.0 17.7

Figure 4.3.5-1. NOx start emissions as a function of soak time for heavy heavy-duty
diesel trucks: EMFAC2021 vs EMFAC2017

25

20
NOx Start Emissions (g)

15

10
EMFAC2021: 2013-15

EMFAC2021: 2016+
5
EMFAC2017: 2010+

0
0 100 200 300 400 500 600 700 800
Soak Time (min)

As can be seen in Figure 4.3.5-1, compared to the EMFAC2017 curve, the two revised
curves for EMFAC2021 both show higher NOx start emissions especially for soak times
less than 400 minutes. This indicates that NOx start emissions would be higher for
EMFAC2021 if the soak times of EMFAC2017 were unchanged.

123
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

4.3.5.4 Natural Gas Emission Rate Updates

In previous EMFAC models, only emissions from natural gas (NG) transit buses and NG
refuse trucks were modeled explicitly using emissions data compiled from CARB internal
testing, published papers, and testing project reports. The data used were all obtained
from dynamometer testing and covered engines of both pre-2007 MY non-TWC and
2007+ MY TWC control technologies. Emissions from other categories of NG vehicles
such as NG HD trucks were implicitly accounted for by treating them as diesel vehicles.
In EMFAC2014, staff developed the running exhaust emission rates for 2007+ MY NG
transit buses (engines certified to 0.2 g/bhp-hr NOx or 0.2-g engines) and 2010+ MY
refuse (also 0.2-g engines) 60. The analysis was based on dynamometer test data from one
2008 MY three-way catalyst (TWC) NG refuse truck one 2008 MY TWC engine, both of
which were tested in a project carried out by South Coast Air Quality Management
District (SCAQMD). In EMFAC2017, staff further updated running exhaust emission rates
of NG transit buses using emissions data from multiple data sources that included NG
transit buses ranging from 2008 to 2015 MY 61.
In the 200-Vehicle Project, nearly one hundred HD vehicles were tested using PEMS and
about half of these were NG vehicles. Each test vehicle was instrumented with a PEMS
unit, and the PEMS continuously measured the vehicle’s emissions of gaseous pollutants
for a typical day of operation. PEMS data from 46 NG HD vehicles were obtained, and
Table 4.3.5-20 shows the distribution of the data among four categories of vehicles.
Table 4.3.5-20. Natural gas vehicles tested in 200-Vehicle Project
Goods Delivery
Technology Transit Bus School Bus Refuse Truck
Movement Truck Truck
0.2g TWC 5 5 11 8 3
0.02g TWC 5 -- 1 9 --

It should be noted EMFAC does not have a goods movement truck category and a
delivery truck category, and these trucks were together treated as NG HD trucks in the
emission rate analysis.
As each PEMS testing run in the 200-Vehicle Project was basically the vehicle’s operation
of a typical working day, the test data consisted of emissions produced under various
driving conditions. A simple average of all the data points in a test run would not only
mask many unique emission features of the test vehicles but also make a meaningful
comparison difficult among vehicles in a vehicle category as well as between two
different vehicle categories. As a result, an emissions vs speed relationship needs to first
established for each of the NG vehicle categories from their PEMS data. This relationship

60
https://ww3.arb.ca.gov/msei/downloads/emfac2014/emfac2014-vol3-technical-documentation-
052015.pdf
61
https://ww3.arb.ca.gov/msei/downloads/emfac2017-volume-iii-technical-documentation.pdf
124
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

in essence is a speed correction curve, and from such a curve a BER can be calculated
using the cycle speed of a dynamometer test cycle commonly used for a vehicle
category. Table 4.3.5-21 lists the test cycles chosen for the four NG vehicle categories.
Table 4.3.5-21. Cycle speeds for estimating base emission rates of natural gas
vehicles
Goods Delivery
Transit Bus School Bus Refuse Truck
Movement Truck Truck
Standard Cycle OCBC AQMD-SB AQMD RTC UDDS UDDS
Cycle Speed (mph) 12.1 12.3 7.31 18.8 18.8

With the analysis approach outlined above, the PEMS data collected from a daily
operation of a test vehicle was first grouped into 10-mph speed bins based on the
speeds of the data points to form seven speed bins, with the middle point speed
representing the speed for each speed bin. A preliminary data analysis has showed that
there are very few data points for speeds higher than 70 mph, and therefore all such data
points are included in the 65-mph bin. Table 4.3.5-22 lists the speed ranges of the seven
speed bins.
Table 4.3.5-22. Middle Speed Point of the seven speed bin ranges
Speed Speed Bin
Middle Speed Point 5-mph 15-mph 25-mph 35-mph 45-mph 55-mph 65-mph
Speed Range (mph) <10 10-<20 20-<30 30-<40 40-<50 50-<60 ≥60

For each vehicle, all data points in a speed bin were averaged to yield an average
emission rate for that bin. The average emission rates of a given speed bin (e.g., 25-mph
bin) of all test vehicles in a vehicle category (e.g., transit bus) were then averaged to
obtain a final bin average emission rate.
Not all test runs have data in all seven bins. Among the tested transit buses, none has
data in the 65-mph bin and several even lack data in the 45-mph bin. The tested refuse
trucks have top speed bins ranging from 45-mph to 65-mph. Since individual test
vehicles show very different overall emission levels, if emission rates of all the speed bins
in a category of vehicles had been averaged, it would have resulted in a distorted
relationship between emissions and speed at the higher speed end. Therefore, in
constructing emission rate vs speed curves, an engineering judgement was made to
decide how many speed bins would be used.
PM emission rate. No PM emissions data were collected during the PEMS testing.
However, in the 200-Vehicle Project, selected test vehicles were also tested on
dynamometer and for many of these vehicles total PM emissions were quantified using
gravimetrical method. Using the dynamometer PM data, staff estimated PM running
exhaust emission rates for all NG vehicle categories. It should be noted that PM
emissions of all the NG vehicles tested on dynamometer were very low and thus these

125
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

estimated PM emission rates should be viewed as an indication of the overall levels of

PM emissions.
Emission deterioration. In previous EMFAC, no emission deterioration was applied to NG
transit buses and NG refuse trucks mainly based on the assumption that these vehicles,
being operated by public agencies or large companies, would undergo regular
maintenance, thus greatly reducing the occurrence of control component tampering
and/or failures. The dataset of NG vehicle PEMS testing from the 200-Vehicle Project,
although relatively large, does not seem to be sufficient for an analysis of emission
deterioration. Therefore, the average emission rates of all four categories of NG vehicles
are used for all odometers of the fleets (i.e., emission rates do not change with
mileages). While this does not explicitly provide an estimate of emission deterioration,
the effect of any emission deterioration that has occurred is implicitly reflected in the
average emission rates.
Idle emissions In EMFAC, continuous operation of a heavy-duty vehicles for >5 minutes
at speed <5 mph is defined as extended idle and emissions from such operation is
specifically modeled as idle emissions. As with HD diesel vehicles, IdleERs were
determined for NG refuse trucks and NG HD trucks. No separate idle emissions were
determined for NG transit buses and school buses; for these buses all idle operations
were considered to be part of the normal trips. To calculate the IdleER of a test run, all
PEMS data points were first flagged as either idle or non-idle based on the EMFAC
criteria for extended idle. All data points flagged as idle were then averaged to yield an
IdleER. All the non-idle data points were processed for the vehicle’s running exhaust
emission rate by following the same procedures as described above.
Results of the speed bin analysis of the PEMS data of urban transit buses (UBUS) are
displayed in Figure 4.3.5-2 for NOx and CO2, two key contributors from HD vehicles in
terms of emissions inventory.
Figure 4.3.5-2. NOx and CO2 running exhaust emission rates of natural gas urban
transit bus by speed bin

8 15000
0.2 g NOx NG UBUS 0.2 g NOx NG UBUS
6
CO2 (g/mi)

0.02 g NOx NG UBUS

NOx (g/mi)

10000 0.02 g NOx NG UBUS

4
5000
2

0 0
5 15 25 35 45 5 15 25 35 45
Speed Bin (mph) Speed Bin (mph)

126
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

NOx emission results show that compared to 0.2-g UBUS, 0.02-g UBUS can achieve
significant emissions reduction under all operating conditions. As will be seen later, this is
also the case for NG refuse trucks and NG HD trucks although the degree of NOx
reduction varies. Also note the notably high CO2 rates for the 5-mph speed bin relative
to other higher speed bins. As mentioned earlier, for UBUS the vehicle operations that
could be characterized as idle according to the EMFAC definition of extended idle were
considered to be part of the normal trips of buses, and thus all the data points belong to
such operations were counted in the 5-mph speed bin. The operations below 5 mph
contribute fair amount of emissions but little distance, resulting in much higher emission
rate in g/mi.
As discussed earlier, for the running exhaust emissions of NG vehicles, a BER can be
obtained from its emission rate vs speed curve established based on the PEMS test data.
For UBUS, the commonly used dynamometer test cycle is OCBC (Orange County Bus
Cycle), which has a cycle speed of 12.1 mph. From the emission rate vs speed curves
established for UBUS, the BERs for THC, CO, NOx, and CO2 were calculated at the
OCBC speed. The BER for PM was calculated based on the dynamometer test results of
one 0.2-g and two 0.02-g NG UBUS from the 200-Vehicle Project. Table 4.3.5-23 shows
the BER of THC, CO, NOx, PM, and CO2 for two MY groups of UBUS. Also included in
the table are the BERs for the corresponding MY groups of UBUS in EMFAC2017.
Table 4.3.5-23. Base emission rates of natural gas urban transit buses
Base Emission Rate (g/mi) @ OCBC Speed (12.1 mph)
EMFAC Model Model Year
THC CO NOx PM CO2
2007-2017* 2.95 50.9 1.44 0.00037 3,801
EMFAC2021
2018+** 1.70 22.1 0.23 0.00071 3,507
2007 21.0 0.833 17.1 0.015 2,048
EMFAC2017
2008+ 8.17 58.0 0.61 0.0050 2,237

*The vast majority of 2007-2017 MY engines were certified to 0.2 g/bhp-hr NOx with TWC as the primary
control but some 2007 engines were also certified to 1.8-1.0 g/bhp-hr NOx with OxyCat and some 2017
engines were certified to 0.1-0.02 g/bhp-hr NOx with TWC.
**Starting 2018 MY, engines have been certified to 0.02 g/bhp-hr NOx with TWC as the primary control
but some 2018-2019 MY engines were also certified to 0.1-0.05 g/bhp-hr NOx with TWC.

From the calculated BERs, emission rates at other speeds can be obtained by applying
SCFs. The SCFs for NG UBUS were calculated by normalizing the rates of all speed bins
to the BERs at the OCBC speed. Table 4.3.5-24 gives the equations for calculating SCFs
for the five pollutants.

127
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Table 4.3.5-24. Speed correction factor equations for natural gas urban transit buses
Pollutant Model Year Speed* SCF Equation A B C
𝐵𝐵
2007-2017 5-45 𝐴𝐴 (𝑆𝑆𝑠𝑠𝑠𝑠𝑠𝑠𝑂𝑂) 13.5 -1.04
THC
2018+ 5-45 𝐴𝐴 𝑠𝑠𝑒𝑒𝑠𝑠𝐵𝐵 (𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆) 3.95 -0.112
2007-2017 5-45 𝐴𝐴 (𝑆𝑆𝑠𝑠𝑠𝑠𝑠𝑠𝑂𝑂)𝐵𝐵 10.9 -0.953
CO
2018+ 5-45 𝐴𝐴 𝑠𝑠𝑒𝑒𝑠𝑠𝐵𝐵 (𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆) 3.41 -0.100
2007-2017 5-45 𝐴𝐴 (𝑆𝑆𝑠𝑠𝑠𝑠𝑠𝑠𝑂𝑂)𝐵𝐵 27.7 -1.32
NOx
2018+ 5-45 𝐴𝐴 (𝑆𝑆𝑠𝑠𝑠𝑠𝑠𝑠𝑂𝑂)𝐵𝐵 89.2 -1.79
PM 2007+ 5-65 𝐴𝐴 (𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑂𝑂)2 + 𝐵𝐵 (𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑂𝑂) + 𝑆𝑆 5.90x10-4 -5.73 x10-2 1.62
2007-2017 5-45 𝐴𝐴 (𝑆𝑆𝑠𝑠𝑠𝑠𝑠𝑠𝑂𝑂)𝐵𝐵 20.8 -1.21
CO2
2018+ 5-45 𝐴𝐴 (𝑆𝑆𝑠𝑠𝑠𝑠𝑠𝑠𝑂𝑂)𝐵𝐵 26.3 -1.30

* For speeds >45 mph, SCF equal to values at 45 mph

The above discussion is pertinent to the heavier urban transit buses or heavy heavy-duty
buses, as all NG transit buses tested in the 200-Vehicle Project had a GVWR > 33,000
lbs. Since EMFAC also includes many lighter transit buses or medium heavy-duty buses,
emission rates were estimated for these buses from T7 buses using scaling factors.
In estimating the emission rates of T6 buses, an assumption was made that, everything
else being equal, emission rates of buses would generally relate to the vehicle weight.
Thus, the ratios of BERs of medium heavy-duty and heavy heavy-duty diesel trucks were
used as scaling factors and were applied to the BERs of heavy heavy-duty NG buses to
obtain the BERs of medium heavy-duty NG buses. As the BERs of MHD and HHD diesel
trucks were calculated at the UDDS speed (18.8 mph), the scaling factors based on MHD
and HHD diesel trucks were further adjusted to the OCBC speed (12.1 mph) using SCFs
of MHD and HHD diesel trucks. Table 4.3.5-25 shows the estimated BERs for T6 NG
UBUS.
Table 4.3.5-25. Estimated base emission rates for T6 natural gas urban buses
Base Emission Rate (g/mi) @ OCBC Speed (12.1 mph)
EMFAC Model Model Year
THC CO NOx PM CO2
2007-2017 1.34 11.2 0.43 0.0001 2,335
EMFAC2021
2018+ 0.96 7.91 0.070 0.0004 2,232

As no medium heavy-duty transit bus test data is available from the 200-Vehicle Project
for determining speed-emissions relationship, the SCFs of heavy heavy-duty NG UBUS
(Table 4.3.5-24) were used for medium heavy-duty NG UBUS.
For this EMFAC update, staff also developed emission rates for NG school buses (SBUS)
using the PEMS data from the 200-Vehicle Project. Results of the SBUS PEMS analysis
are displayed in Figure 4.3.5-3 for NOx and CO2.

128
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure 4.3.5-3. NOx and CO2 running exhaust emission rates of natural gas school
bus by speed bin

12 12000
10 0.2 g NOx NG SBUS 10000
0.2 g NOx NG SBUS
NOx (g/mi)

8 8000

CO2 (g/mi)
6 6000
4 4000
2 2000
0 0
5 15 25 35 45 55 5 15 25 35 45 55
Speed Bin (mph) Speed Bin (mph)

A test cycle, SCAQMD-SB Cycle, was used in the 200-Vehicle Project for school buses
being tested on chassis dynamometer. This cycle was developed based on telemetry
activity of school buses operating in Southern California and has a cycle speed of 12.3
mph. From the rate-speed curves determined for SBUS, the BERs for THC, CO, NOx, and
CO2 were calculated at 12.3 mph, the average speed of SCAQMD-SB Cycle. The BER for
PM was calculated based on the dynamometer test results of one 0.2g NG school bus
from the 200-Vehicle Project. No emissions data for school buses with 0.02-g engines are
available from the 200-Vehicle Project, and therefore the BERs for 0.02-g SBUS were
estimated by scaling the BERs of 0.2-g SBUS using the ratios of BERs of 0.2-g UBUS and
0.02-g UBUS. Table 4.3.5-26 shows the BERs of THC, CO, NOx, PM, and CO2 for the
2007-2017 MY and 2018+ MY NG SBUS.
Table 4.3.5-26. Base emission rates of natural gas school buses
Rate @ School Bus Cycle Speed (12.3 mph)
Model Year
THC CO NOx PM* CO2
2007-2017* 11.0 39.8 2.23 0.0102 3,402
2018+** 6.36 17.3 0.363 0.0102 3,139

* Most of 2007-2017 MY engines were certified to 0.2 g/bhp-hr NOx with TWC as the primary control.
** Rates were estimated based on ratio of NG transit bus rates and applied to engines certified to 0.02
g/bhp-hr NOx.

Similar to UBUS, SCFs were developed for SBUS using the PEMS data. Equations for
calculating SCFs of the five pollutants are listed in Table 4.3.5-27. It is assumed that the
SCFs based on 0.2-g SBUS are also applicable to 0.02-g SBUS.

129
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Table 4.3.5-27. Speed correction factor equations for natural gas school buses
Pollutant MY SCF Equation Speed* A B C
𝐵𝐵
THC 2007+ 𝐴𝐴 (𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆) 0-55 20.4 -1.20
CO 2007+ 𝐴𝐴 (𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆)𝐵𝐵 0-55 24.7 -1.28
NOx 2007+ 𝐴𝐴 (𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆)𝐵𝐵 0-55 46.2 -1.53
PM 2007+ 𝐴𝐴 (𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠)2 + 𝐵𝐵 (𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠) + 𝐶𝐶 0-55 5.90x10-4 -5.73x10-2 1.62
CO2 2007+ 𝐴𝐴 (𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆)𝐵𝐵 0-55 16.2 -1.11

* For speeds >45 mph; SCF equals to values at 45 mph

For this version of EMFAC model, staff also updated emissions rates for solid waste
collection vehicles (SWCV). Results of the speed bin analysis of the PEMS data for SWCV
(refuse trucks) are displayed in Figure 4.3.5-4 for NOx and CO2.
Figure 4.3.5-4. NOx and CO2 running exhaust emission rates of natural gas solid
waste collection vehicles by speed bin

10 8000
0.2 g NOx NG SWCV
8 0.2 g NOx NG SWCV
6000 0.02 g NOx NG SWCV
CO2 (g/mi)

0.02 g NOx NG SWCV

NOx (g/mi)

6
4000
4
2000
2

0 0
5 15 25 35 45 55 5 15 25 35 45 55
Speed Bin (mph) Speed Bin (mph)

In recent years, dynamometer testing of refuse trucks have been using a cycle developed
by WVU for SCAQMD and consists of the refuse truck operation (SCAQMD-RTC) and
operation/compaction (SCAQMD-RCC) cycles to simulate typical transportation and
curbside pick-up mode operations. The emissions from the RCC are integrated into the
emissions from the RTC. The RTC has an average speed of 7.31 mph.
From the rate-speed curves determined for SWCV, the BERs for THC, CO, NOx, and
CO2 were calculated at the RTC speed of 7.31 mph. Again, the BER for PM was
calculated based on the dynamometer test results of one 0.2-g and one 0.02-g NG
refuse trucks from the 200-Vehicle Project. Table 4.3.5-28 shows the BERs of THC, CO,
NOx, PM, and CO2 for two MY groups of SWCV. Also included in the table are the BERs
for the corresponding MY groups of SWCV in EMFAC2017.

130
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Table 4.3.5-28. Base emission rates of natural gas solid waste collection vehicles
Base Emission Rate (g/mi) @ OCBC Speed (12.1 mph)
EMFAC Model Model Year
THC CO NOx PM CO2
2007-2017* 21.6 112 7.90 0.0017 4,559
EMFAC2021
2018+* 0.54 33.8 0.169 0.0034 4,197
2007-2009 22.8 18.5 18.8 0.0044 5,404
EMFAC2017
2010+ 10.1 36.6 0.879 0.0044 5,077

* 2007-2017 MY: mostly 0.2 g/bhp-hr NOx engines; 2018+ MY: mostly 0.02 g/bhp-hr NOx engines. See
Table 4.3.5-23 notes for more details.

From the calculated BER, emission rates at other speeds can be obtained by applying
the SCFs. The SCFs for NG SWCV were calculated by normalizing the rates of all speed
bins to the BER at the RTC speed. Table 4.3.5-29 lists the equations for calculating the
SCFs of NG SWCV for all the pollutants.
Table 4.3.5-29. Speed correction factor equations for natural gas school buses
Pollutant MY SCF Equation Speed* A B C
2007-2017 𝐵𝐵
𝐴𝐴 (𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆) 0-55 6.51 -0.942
THC
2018+ 𝐴𝐴 (𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆)𝐵𝐵 0-55 22.7 -1.57
2007-2017 𝐴𝐴 (𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆)𝐵𝐵 0-55 8.52 -1.08
CO
2018+ 𝐴𝐴 (𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆)𝐵𝐵 0-55 7.99 -1.05
2007-2017 𝐴𝐴 (𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆)𝐵𝐵 0-55 8.31 -1.06
NOx
2018+ 𝐴𝐴 (𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆)𝐵𝐵 0-55 2.92 -0.539
PM 2007-2017 𝐴𝐴 (𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠)2 + 𝐵𝐵 (𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠) + 𝐶𝐶 0-55 4.80x10-4 -4.67x10-2 1.32
2007-2017 𝐴𝐴 (𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆)𝐵𝐵 0-55 4.18 -0.719
CO2
2018+ 𝐴𝐴 (𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆)𝐵𝐵 0-55 4.14 -0.714

* SCF values for speeds >55 = SCF at 55 mph.

As discussed earlier, in addition to running exhaust emission rate a separate IdleER for
each pollutant was also determined for NG SWCV. IdleERs were calculated for SWCV
following the method described above, and the obtained rates are shown in Table
4.3.5-30.
Table 4.3.5-30. Idle emission rates of natural gas solid waste collection vehicles
Idle Emission Rate (g/hr)
Model Year
THC CO NOx PM CO2
2007-2017* 38.8 136 17.8 0.00090 12,350
2018+* 3.38 58.9 0.227 0.0029 13,991

* 2007-2017 MY: mostly 0.2 g/bhp-hr NOx engines; 2018+ MY: mostly 0.02 g/bhp-hr NOx engines. See
Table 4.3.5-23 notes for more details.

131
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

In previous EMFAC, heavy-duty NG trucks has been treated as heavy-duty diesel trucks
in emissions inventory calculations. In EMFAC2021 update, staff was able develop
emission rates for HHD (T7) NG trucks based on PEMS test data from the 200-Vehicle
Project. For HHD NG trucks, the method of data analysis has been described earlier and
results for NOx and CO2 are displayed in Figure 4.3.5-5.
Figure 4.3.5-5. NOx and CO2 running exhaust emission rates of heavy heavy-duty
diesel trucks by speed bin

8 6000
0.2 g NOx NG Truck
0.2 g NOx NG Truck 0.02 g NOx NG SWCV
6

CO2 (g/mi)
0.02 g NOx NG Truck
NOx (g/mi)

4000
4
2000
2

0 0
5 15 25 35 45 55 65 5 15 25 35 45 55 65
Speed Bin (mph) Speed Bin (mph)

Although two dynamometer test cycles – Goods Movement Truck Cycle and Delivery
Truck Cycle – were developed in the 200-Vehicle Project, staff decided to use the UDDS
as the cycle to determine the running exhaust emission rates of these trucks as they are
re-characterized as HHD NG trucks. Thus, BERs of HHD NG trucks were calculated at
18.8 mph from the rate-speed curve developed from the PEMS data. Table 4.3.5-31
shows the calculated BERs of THC, CO, NOx, PM, and CO2 of HHD NG trucks.
Table 4.3.5-31. Base emission rates of heavy heavy-duty natural gas trucks
Base Emission Rate (g/mi) @ UDDS Speed (18.8 mph)
Model Year
THC CO NOx PM CO2
2007-2017* 2.43 16.9 1.28 0.0033 2075
2018+* 2.18 14.1 0.24 0.0062 2193

* 2007-2017 MY: mostly 0.2 g/bhp-hr NOx engines; 2018+ MY: mostly 0.02 g/bhp-hr NOx engines. See
Table 4.3.5-23 notes for more details.

Emission rates at other speeds can be obtained using SCF. The SCFs for NG HHD trucks
were calculated by normalizing the rates of all speed bins to the BER at the UDDS speed
and regression equations were then obtained for all gaseous pollutants. The SCF for PM
was derived by re-normalizing the EMFAC2017 PM SCF for NG vehicles to the UDDS
speed. Table 4.3.5-32 presents the equations for calculating SCFs of NG HHD trucks.

132
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Table 4.3.5-32. Speed correction factor equations for heavy heavy-duty natural gas
trucks
Pollutant MY SCF Equation Speed A B C
𝐵𝐵
2007-2017 𝐴𝐴 (𝑆𝑆𝑠𝑠𝑠𝑠𝑠𝑠𝑂𝑂) 0-65 26.8 -1.12
THC
2018+ 𝐴𝐴 (𝑆𝑆𝑠𝑠𝑠𝑠𝑠𝑠𝑂𝑂)𝐵𝐵 0-65 14.7 -0.917
2007-2017 𝐴𝐴 (𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑂𝑂)2 + 𝐵𝐵 (𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑂𝑂) + 𝑆𝑆 0-65 4.37x10-4 -4.43x10-2 1.68
CO
2018+ 𝐴𝐴 (𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑂𝑂)2 + 𝐵𝐵 (𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑂𝑂) + 𝑆𝑆 0-65 4.13x10-4 -5.50x10-2 1.89
2007-2017 𝐴𝐴 (𝑆𝑆𝑠𝑠𝑠𝑠𝑠𝑠𝑂𝑂)𝐵𝐵 0-65 24.7 -1.09
NOx
2018+ 𝐴𝐴 (𝑆𝑆𝑠𝑠𝑠𝑠𝑠𝑠𝑂𝑂)𝐵𝐵 0-65 5.25 -0.565
PM 2007-2017 𝐴𝐴 (𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑂𝑂)2 + 𝐵𝐵 (𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑂𝑂) + 𝑆𝑆 0-65 7.94x10-4 -7.72x10-2 2.18
2007-2017 𝐴𝐴 (𝑆𝑆𝑠𝑠𝑠𝑠𝑠𝑠𝑂𝑂)𝐵𝐵 0-65 5.16 -0.559
CO2
2018+ 𝐴𝐴 (𝑆𝑆𝑠𝑠𝑠𝑠𝑠𝑠𝑂𝑂)𝐵𝐵 0-65 4.91 -0.542

Similar to NG refuse trucks, separate IdleERs were also developed for HHD NG trucks.
Same analysis method was used to calculate the IdleERs of gaseous pollutants and the
PM rate was estimated based on the dynamometer test data of one delivery truck (0.2-g
engine) and 3 goods movement trucks (one 0.2-g and two 0.02-g engines). Table
4.3.5-33 shows the IdleERs of THC, CO, NOx, PM, and CO2 of HHD NG trucks.
Table 4.3.5-33. Idle emission rates of natural gas heavy-duty trucks
Idle Emission Rate (g/hr)
Model Year
THC CO NOx PM CO2
2007-2017* 48.2 48.0 15.5 0.033 11,224
2018+* 33.9 109 14.4 0.066 14,164

* 2007-2017 MY: mostly 0.2 g/bhp-hr NOx engines; 2018+ MY: mostly 0.02 g/bhp-hr NOx engines. See
Table 4.3.5-23 notes for more details.

4.3.5.5CO2 Emission Rate

To align with the Federal Phase 1 and Phase 2 programs, CARB adopted California Phase
1 and Phase 2 GHG regulations in 2013 and 2018, respectively. These two regulations
allowed CARB to enforce its own GHG regulations. It also allowed CARB to certify the
heavy-duty manufacturers in California. Phase 1 covers engines and three vehicle
categories, including tractors (Class 7-8), vocational vehicles (Class 2b-8), and pickup
trucks and vans (Class 2b-3). Phase 2 covers trailers along with the engines and the three-
vehicle categories covered in Phase 1, and it is expected to lower CO2 emissions beyond
Phase 1 levels by an additional 13 percent in 2030, and by 2050, those reductions will
increase to roughly 24 percent.
To account for the impacts of CARB Phase 1 and Phase 2 GHG regulations, the
percentage reductions in CO2 emission rates with respect to 2010 are estimated to
adjust CO2 emissions by vehicle type, model year, and fuel type. The details of CO2
emission reduction percentage can be found under Section 4.3.3.1 in EMFAC2017

133
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

technical documentation and Appendix F 62 of Proposed California Greenhouse Gas

Emissions Standards for Medium- and Heavy-Duty Engines and Vehicles (i.e., CA Phase 2
staff report).
For EMFAC2021, staff further evaluated the impact of Phase 1 regulation
implementation using CARB certified subfamily CO2 Family Emission Level (FEL) along
with their production volume from 2016 through 2019. To evaluate the CO2 emission
reduction rate, composite baselines for model year 2010 in gram per mile are established
for medium heavy-duty and heavy heavy-duty vehicles based on baseline vehicle
performance 63 and California certified production volumes, as shown in Table 4.3.5-34
and Table 4.3.5-35.
Table 4.3.5-34. Composite CO2 baseline for Medium Heavy-Duty Vehicles
Vocational Tractor
Medium Heavy-Duty Vehicles
Class 4-5 Class 6-7 Class 7
CO2 baseline (gCO2/ton-mile) 408 247 236
Payload (tons) 2.85 5.6 12.5
CO2 baseline (gCO2/mile) 1163 1383 2950
Production Volume Share (%) 33% 63% 4%
Composite CO2 baseline (gCO2/mile) 1375

Table 4.3.5-35. Composite CO2 baseline for Heavy Heavy-Duty Vehicles

Vocational Tractor
Heavy Heavy-Duty Vehicles
Class 8 Class 8 Day-Cabs Class 8 Sleeper-Berth
CO2 baseline (gCO2/ton-mile) 236 95 93
Payload (tons) 7.5 19 19
CO2 baseline (gCO2/mile) 1770 1798 1758
Production Volume Share (%) 38.7% 28.3% 33.0%
Composite CO2 baseline (gCO2/mile) 1774

CO2 FEL and production volumes by subfamily, vehicle classification, model year are
collected from nine major manufacturers 64 for this analysis. The manufacturers account
for 76% of the total heavy-duty vehicle market from 2016 through 2019. The production
volume weighted average CO2 FEL in gram per mile is calculated for medium heavy-duty

62
California Air Resources Board, Appendix F-Emissions Inventory Analysis and Results, Staff Report of
Proposed California Greenhouse Gas Emissions Standards for Medium- and Heavy-Duty Engines and
Vehicles (web link:
https://ww3.arb.ca.gov/regact/2018/phase2/appf.pdf?_ga=2.133493582.1354806717.1613440332-
1307567751.1567730621, posted December, 2017)
63
U.S.EPA, Federal Register Volume 76 Number 179 57106-57513 (http://www.gpo.gov/fdsys/pkg/FR-
2011-09-15/pdf/2011- 20740.pdf, published September 15, 2011)
64
Daimler, Paccar, Ford, Isuzu, Volvo, FCA, Hino, Navistar, and Gillig
134
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

and heavy heavy-duty vehicles, respectively. Results from this analysis are presented in
Table 4.3.5-36 and Table 4.3.5-37. The analysis of California heavy-duty vehicle
certification data shows that the production volume weighted average CO2 FEL achieved
similar or even more CO2 emission reductions than the composite Phase 1 requirements
in EMFAC2017. The CO2 emission rates for heavy heavy-duty vehicles from the TBSP
also showed similar reduction trend. Therefore, the same Phase 1 CO2 emission
reduction factors from EMFAC2017 are used in EMFAC2021. To avoid double
accounting of Phase 1 impacts, the CO2 emission rates for model year 2014 and newer
are estimated as baseline emission rate for model year 2010 (2,350 g/mile for heavy
heavy-duty vehicles, and 1,413 g/mile for medium heavy-duty vehicles) multiplied by
CO2 reduction factors. According to emissions test data, the baseline emission rate is
2,350 g/mile for heavy heavy-duty vehicles and 1413 g/mile for medium heavy-duty
vehicles. The baseline CO2 emission rates for medium heavy-duty vehicles is scaled from
heavy heavy-duty ones, since there is no testing data available for model year 2010.
Table 4.3.5-36. Phase 1 CO2 Ratios to the Baseline for Medium Heavy-Duty Vehicle
Medium Heavy-Duty Vehicle
Model California Certificates
Year Phase 1 in
Production weighted average CO2 FEL Ratio to EMFAC2017
(gCO2/mile) Baseline
2010 1375 100% 100%
2014 No Data No Data 95%
2015 No Data No Data 95%
2016 1233 90% 95%
2017 1230 89% 91%
2018 1184 86% 91%
2019 1142 83% 91%
2020 No Data No Data 91%

135
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Table 4.3.5-37. Phase 1 CO2 Ratios to the Baseline for Heavy-Heavy Duty Vehicle
Heavy Heavy-Duty Vehicle
California Certificate TBSP Testing
Model Average
Year Production weighted average Phase 1 in
Ratio to Emission Ratio to EMFAC2017
CO2 FEL
Baseline Rate Baseline
(gCO2/mile)
(g/mil)
2010 1774 100% 2350 100% 100%
2014 No Data No Data 2215 88% 87%
2015 No Data No Data 2209 85% 87%
2016 1538 87% 2074 84% 87%
2017 1507 85% 2063 82% 85%
2018 1509 85% 1992 87% 85%
2019 1493 84% 1976 85% 85%
2020 No Data No Data No Data No Data 85%

ZEVs produced by manufacturers to meet the Advanced Clean Truck (ACT) regulation
can also be used to meet the Phase 2 GHG requirements. As such, corrections were
made to the Phase 2 CO2 reduction factors for vehicles originally sold in California to
ensure that emissions reductions are not double counted with ACT. The adjusted Phase
2 CO2 emission rate ratio to the baseline by model year for California certified trucks
after phase-in of ACT are estimated with the equation below and should not exceed
100%.
𝑂𝑂𝑂𝑂𝑅𝑅𝑂𝑂𝑅𝑅𝐵𝐵𝑎𝑎𝐵𝐵 𝑃𝑃ℎ𝑎𝑎𝑠𝑠𝑠𝑠2 𝑆𝑆𝑂𝑂2 𝐵𝐵𝑎𝑎𝐴𝐴𝑅𝑅𝑂𝑂 𝐴𝐴𝑂𝑂 𝐵𝐵𝑎𝑎𝑠𝑠𝑠𝑠𝐵𝐵𝑅𝑅𝐵𝐵𝑠𝑠
𝐴𝐴𝑆𝑆𝐴𝐴 𝐴𝐴𝑂𝑂𝐴𝐴𝐴𝐴𝑠𝑠𝐴𝐴𝑠𝑠𝑂𝑂 𝑃𝑃ℎ𝑎𝑎𝑠𝑠𝑠𝑠 2 𝑆𝑆𝑂𝑂2 𝐵𝐵𝑎𝑎𝐴𝐴𝑅𝑅𝑂𝑂 𝐴𝐴𝑂𝑂 𝐵𝐵𝑎𝑎𝑠𝑠𝑠𝑠𝐵𝐵𝑅𝑅𝐵𝐵𝑠𝑠 =
1 − 𝑍𝑍𝐵𝐵𝑍𝑍 % 𝑏𝑏𝑏𝑏 𝐴𝐴𝑆𝑆𝐴𝐴
However, once the Phase 2 CO2 ratios to baseline reaches a minimum, they were
assumed to stay constant for subsequent model years. In other words, staff assumed that
Phase 2 fuel efficiency gains would not be reversed through overlap with ACT. The
original and adjusted Phase 2 CO2 reduction percentage are shown in Table 4.3.5-38,
Table 4.3.5-39, Table 4.3.5-40. Note that the adjusted Phase 2 factors are used for
vehicles originally sold in California, while original Phase 2 reduction factors are applied
to vehicles originally sold out-of-state.

136
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Table 4.3.5-38. LHD1 and LHD2 Phase 1 and Phase 2 Original and Adjusted CO2
Reduction Percentage
Gasoline LHD1 and LHD2 Diesel LHD1 and LHD2
Model ACT Phase 1 and Phase 2 ACT Phase 1 and Phase 2
Year (% ZEV Phase 2 CO2 Adjusted (% ZEV Phase 2 CO2 Adjusted for
Sales) Ratio to Baseline for ACT Sales) Ratio to Baseline ACT
2014 0% 99% 99% 0% 98% 98%
2015 0% 98% 98% 0% 97% 97%
2016 0% 96% 96% 0% 94% 94%
2017 0% 94% 94% 0% 91% 91%
2018 0% 90% 90% 0% 85% 85%
2019 0% 90% 90% 0% 85% 85%
2020 0% 90% 90% 0% 85% 85%
2021 0% 88% 88% 0% 83% 83%
2022 0% 86% 86% 0% 81% 81%
2023 0% 83% 83% 0% 79% 79%
2024 5% 81% 83% 5% 77% 79%
2025 7% 79% 83% 7% 75% 79%
2026 10% 77% 83% 10% 73% 79%
2027 15% 75% 83% 15% 71% 79%
2028 20% 75% 83% 20% 71% 79%
2029 25% 75% 83% 25% 71% 79%
2030+ 30% 75% 83% 30% 71% 79%

Table 4.3.5-39. Medium Heavy-Duty Vocational Trucks and Buses, and Tractors Phase
1 and Phase 2 Original and Adjusted CO2 Reduction Percentage
MHD Vocational Trucks and Buses MHD Tractors
Mod ACT ACT
el Phase 1 and Phase 2 Phase 2 Phase 1 and Phase 2 Phase 2
(% (%
Year CO2 Ratio to Adjusted for CO2 Ratio to Adjusted for
ZEV ZEV
Baseline ACT Baseline ACT
Sales) Sales)
2014 0% 95% 95% 0% 95% 95%
2015 0% 95% 95% 0% 95% 95%
2016 0% 95% 95% 0% 95% 95%
2017 0% 91% 91% 0% 91% 91%
2018 0% 91% 91% 0% 91% 91%
2019 0% 91% 91% 0% 91% 91%
2020 0% 91% 91% 0% 91% 91%
2021 0% 82% 82% 0% 82% 82%
2022 0% 82% 82% 0% 82% 82%
2023 0% 82% 82% 0% 82% 82%
2024 9% 76% 82% 5% 76% 80%

137
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

2025 11% 76% 82% 7% 76% 80%

2026 13% 76% 82% 10% 76% 80%
2027 20% 73% 82% 15% 73% 80%
2028 30% 73% 82% 20% 73% 80%
2029 40% 73% 82% 25% 73% 80%
2030
50% 73% 82% 30% 73% 80%
+

Table 4.3.5-40. Heavy Heavy-Duty Vocational and Tractor Trucks Phase 1 and Phase 2
Original and Adjusted CO2 Reduction Percentage
HHD Vocational Trucks HHD Tractors
Mod ACT ACT
el Phase 1 and Phase 2 Phase 2 Phase 1 and Phase 2 Phase 2
(% (%
Year CO2 Ratio to Adjusted for CO2 Ratio to Adjusted for
ZEV ZEV
Baseline ACT Baseline ACT
Sales) Sales)
2014 0% 87% 87% 0% 87% 87%
2015 0% 87% 87% 0% 87% 87%
2016 0% 87% 87% 0% 87% 87%
2017 0% 85% 85% 0% 85% 85%
2018 0% 85% 85% 0% 85% 85%
2019 0% 85% 85% 0% 85% 85%
2020 0% 85% 85% 0% 85% 85%
2021 0% 74% 74% 0% 74% 74%
2022 0% 74% 74% 0% 74% 74%
2023 0% 74% 74% 0% 74% 74%
2024 9% 69% 74% 5% 69% 72%
2025 11% 69% 74% 7% 69% 72%
2026 13% 69% 74% 10% 69% 72%
2027 20% 65% 74% 15% 65% 72%
2028 30% 65% 74% 20% 65% 72%
2029 40% 65% 74% 25% 65% 72%
2030+ 50% 65% 74% 30% 65% 72%

4.3.6 Heavy Duty NOx Deterioration Rates

In EMFAC2017, heavy-duty (HD) vehicle base emission rates (BER) were calculated by
model year group using the following equation:
g
BER odo (mile) = (ZMR + DR × Odometer) × SCF (Eq. 4.3.6-1)

where ZMR is the zero-mile rate, DR is the deterioration rate, and SCF is the speed
correction factor. A basic assumption in assessing emission deterioration of HD trucks is
that emissions from engines remain stable in the absence of tampering, mal-

138
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

maintenance, and malfunction (TM&M). As the heavy-duty fleet ages and accrues
mileage, EMFAC models a greater fraction of the fleet having an engine or emissions
after-treatment malfunction. The DR represents the slope of this increase. ZMR and DR
are typically developed on a model year group basis. Each group includes several
consecutive model years that usually share the same emission standards and/or emission
control technology. DRs are calculated as follows.
DR (gmile-1 per 10,000 miles) = (ZMR×EIR)/100 (Eq. 4.3.6-2)
where EIR is the emission impact rate at 1,000,000 miles.
EMFAC2017 uses a framework for calculating EIR that was originally developed in
EMFAC 2000. To estimate the emission impact of TM&M, this method identifies several
specific types of TM&M affecting the average emissions of a truck fleet. The EIR is the
product of the frequency of occurrence of TM&M, and the emission increases over the
baseline level caused by the TM&M.
In EMFAC2017, there were some revisions to these TM&M frequencies for engine model
years 2010 and newer for vehicles with extended warranties. Furthermore, engine model
year 2013+ vehicles that are equipped with OBD systems were assumed to have 33%
reduction in TM&M frequencies for all categories (e.g., DPF filter issues). More details
can be found in section 4.3.2.1.2 of the EMFAC2017 technical documentation. When
EMFAC2017 was released, there was very limited information available for engine model
year 2013 and newer deterioration rates. With more and more engine model year 2013
and newer phase-in, there is a need to improve EMFAC modeling related to failure
frequencies associated with the engine and after-treatment components in the EMFAC
model for newer, OBD-equipped vehicles, especially to assess potential costs and
emission benefits of heavy-duty inspection and maintenance program. To better
understand heavy-duty in-use performance, EMFAC2021 utilized on-board diagnostics
(OBD), from which the malfunction indicator lamp (MIL) status can be determined, as well
as the fault codes that triggered the MIL.
Through Contract 17AQP006, CARB collected OBD data in the field, like the Port of LA
and Truck Stops, and through telematics companies allowing fleets to track data (e.g.,
GPS) on all of their vehicles. Overall, OBD data was collected from 457 vehicles from the
field locations, and 24,555 and 180,892 California-operating and US-operating vehicles,
respectively, from telematics data sources.
Figure 4.3.6-1 shows the malfunction indicator lamp (MIL) rates for field and telematics
data. Here “MIL On” represents the percentage of vehicles with an illuminated MIL and
thus are assumed to have an emissions-related engine or after-treatment malfunction.
Overall, US-operating vehicles have slightly higher MIL On, and there was no systematic
difference between field and telematics MIL rates. Notably, the observed MIL On
percentage was as large as 5% for low mileage vehicles (i.e., 0-50 kmiles), which indicates
that vehicles have emissions-related issues early in their life.

139
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure 4.3.6-1. Percentage of Vehicles with MIL On as a Function of Odometer

35.0%
Field MIL Telematics - US Telematics - CA
30.0%

25.0%
MIL On %

20.0%

15.0%

10.0%

5.0%

0.0%

Mileage Bins (kmiles)

For EMFAC2021, US-operating telematics dataset was selected to model the

deterioration rate due its high sample size. It is also noteworthy to mention that out-of-
state registered heavy-duty vehicles that travel to California are responsible for a large
fraction of heavy-duty vehicle miles traveled in the State.
A combination of MIL On frequency from this contract and in-use emission rate data
were used to develop new deterioration rates for engine model year 2013 and newer in
EMFAC2021. In-use test data was provided through CARB's Truck and Bus Surveillance
Program and the EMA/UCR Testing Projects. In a nutshell, in-use test data determines
the magnitude of deterioration, while the MIL On function determines the shape of
deterioration rate. First, the MIL On frequency (binned by 100 kMi) as a function of
odometer was fit to a power function, as shown in Figure 4.3.6-2.

140
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure 4.3.6-2. Power Function Fit to MIL On as a Function of Odometer

30%
Binned Data (100 kmiles)
25%
Best Fit Power Function

20%
Mil On %

15%

10%

5%

0%
0 200,000 400,000 600,000 800,000 1,000,000
Odometer Mileage

The fitted equation for MIL On frequency is

𝐌𝐌𝐌𝐌𝐌𝐌 𝐎𝐎𝐎𝐎 = 𝟎𝟎. 𝟎𝟎𝟎𝟎𝟎𝟎 × 𝐎𝐎𝐎𝐎𝐎𝐎𝐎𝐎𝐎𝐎𝐎𝐎𝐎𝐎𝐎𝐎 𝟎𝟎.𝟑𝟑𝟑𝟑 (Eq. 4.3.6-3)
Next, an iterative procedure was used to determine deterioration rates from the MIL On
function and in-use test data. Steps are described below.
• Step 1. Initiate an EIR and ZMR at 90,249 miles, the average odometer of the US-
wide data set
• Step 2. Scale EIR to other odometers using the MIL On function
• EIR (odometer) = MIL On (odometer)b/MIL On (90,249)b
• Step 3. Use EIRs to determine odometer-dependent emission rates
• ER (odometer) = ZMR + ZMR * EIR (odometer)
• Step 4. Calculate root mean square error (RMSE) between binned in-use NOx
emission rates and the modelled values
• Step 5. Update ZMR and EIR until the RMSE reaches a minimum
With the above method, the best fit EIR was 249% and a ZMR of 0.6 g/mile of NOx
emissions for vehicles with engine model year 2013 and newer; the resulting best fit
emission rate equation for heavy heavy-duty vehicles is shown in Figure 4.3.6-3. A ZMR

141
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

of 0.6 g/mile aligns well with the TBSP and EMA test data for vehicles with odometer
mileage less than 50,000 miles. Note that, unlike the linear model that was used
previously, the base EIR is at 90,249 miles instead of 1,000,000 miles.
Figure 4.3.6-3. Modelled and Observed NOx Emission Rates
12 Truck and Bus Surveillance Program Data
ER EMFAC 2021 - New Method
10 Test Data Binned
Standard
ARB/EMA
8
NOX ER (g/mi)

0
0 200,000 400,000 600,000 800,000 1,000,000
Odometer Mileage

For comparison, the linear method used for the previous versions of EMFAC is shown.
This linear method uses an average emission rate for all test data, as shown as the black
triangle in Figure 4.3.6-4. Then, the ZMR is back-calculated using an EIR of 170% at
1,000,000 miles. In general, the new method based on MIL On leads to higher emission
rates from 100,000 to 600,000 odometer mileage, but lower emission rates beyond
600,000 miles. Note that the OBD-based method is used for NOx, but not for PM due to
the lack of vehicles with high PM emissions in the in-use dataset. As described in Section
4.3.5.2, for PM, EMFAC2021 uses similar method as in EMFAC2017.

142
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure 4.3.6-4. Comparison of New Deterioration Model to that derived from a

Linear Method

6
ER EMFAC 2021 - New Method
ER EMFAC 2021 - Linear Method
5 Test Data Binned
Test Data Average

4
NOX ER (g/mi)

0
0 200,000 400,000 600,000 800,000 1,000,000
Odometer

4.3.7 Break Wear Emissions

The current EMFAC brake wear emission factors rely on data from 2000-2003. Both light
and heavy-duty brake wear emission rates have been updated with recently collected
data using dynamometer-based testing methods.
4.3.7.1 Light-Duty Brake wear
During EMFAC2021 development, CARB staff worked closely with U.S. EPA and Caltrans
to conduct a comprehensive brake wear testing using the European Commission Joint
Research Committee (JRC) protocol/procedure. Specifically, measuring emissions with a
brake dynamometer simulating real-world conditions. The testing would look at the most
popular brake configurations, and would address regenerative braking.
The testing was conducted under ARB contract 17RD016. This study utilized a LINK
Engineering (LINK) single wheel brake dynamometer for the measurement of PM
emissions over a prescribed driving cycle. The vehicles were driven on a representative
driving cycle (California Brake Driving Cycle or CBDC) developed from the 2010-2012
Caltrans Household Travel Survey. Brake temperatures were monitored, and temperature
profiles were controlled accordingly on the break dynamometer. Six vehicles were
tested, one of which used a regenerative braking system. PM was measured using both
particulate filters and in real time with a quartz crystal micro-balance (QCM). Figure
143
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

4.3.7-1 presents the vehicle-level results for each of the 6 tested models by three
different pad materials: Original Equipment Service non-asbestos organic (OES-NAO),
aftermarket NAO, and aftermarket Low-Metallic (LM). The F-150 and Sienna were also
tested with higher vehicle loads.
Figure 4.3.7-1. Whole Vehicle Braking Emissions by Model and Material
35
OES-NAO
Vehicle Level Emissions (mg/mi)

30
After-NAO
25 After-LM

20
HLW: Heavily Laden Weight
15

10

0
Camry Civic F-150 F-150 HLW Prius Rogue Sienna Sienna
HLW

Table 4.3.7-1 breaks out the emission rate results as a function of vehicle type. The light
trucks are slightly higher than the passenger vehicles because of a different distribution
of braking materials, especially in the rear axle. The Toyota Prius with the regenerative
braking have the lowest emission rates.
Table 4.3.7-1. Whole Vehicle Braking Emissions by Vehicle Class
Vehicle Type PM2.5 BER (mg/mi) PM10 BER (mg/mi)
Conventional Passenger 1.55 7.65
Light Truck 1.81 8.38
Regenerative-equipped 0.93 3.30

Figure 4.3.7-2 illustrates the relationship with respect to speed. The speed correction
factors are used in conjunction with the BERs to determine emissions at a given speed.

144
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure 4.3.7-2. Speed Correction Factors for PM2.5 and PM10

1.8
1.6
Speed Correction Factor

1.4 PM2.5 PM10

1.2
1
0.8
0.6
0.4
0.2
0
0-15 15-30 30-45 45-60 60+
Speed Bin (mph)

Interestingly, the study also found a form of deterioration to the basic emission rates.
This is not deterioration in the classical sense, but more of a representation of how there
is a switch from OEM equipment to aftermarket equipment as the vehicle ages.
Therefore, emission rates in the form of a linear regression were determined. The
adjusted emission rates and deterioration rates are given in Table 4.3.7-2
Table 4.3.7-2. Estimated PM10 Deterioration Rates and New-Vehicle Estimated
Emission Rates based on Friction Material Trend with Vehicle Age
New vehicle estimated
Vehicle Type Deterioration Rate (mg/10K mi)
emission rate (mg/mi)
Conventional Passenger Car 0.0492 7.65
Light Truck 0.1825 8.38
Regenerative-Equipped 0.0047 3.30

Figure 4.3.7-3 presents a comparison with past studies of brake wear emission rates of
PM2.5 and PM10 for light- and heavy-duty vehicles. The rectangles represent the range
of light duty emission rates determined in CARB’s recent study (i.e., contract 17RD016).
The squares represent heavy-duty vehicles and circles represent light-duty vehicles. The
findings from this light-duty study are lower in both PM2.5 and PM10 than values
currently in EMFAC2017, but reasonably close to most other studies.

145
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure 4.3.7-3. Various literature values for brake emissions, with the ranges from
this study overlaid for comparison

The study further revealed that:

• Front brakes emit more PM than rear brakes-since most of the braking occurs in
the front of the vehicle, the brakes seem to experience much higher temperatures
and emit significant higher PM.
• Non Asbestos Organic (NAO) friction material brakes emit less PM than Low
Metallic (LM) brakes-as vehicles age, the owners are more likely to replace the
brakes with low metallic materials resulting in higher emission rates as the vehicles
age.
• Speed effects are not monotonic as implied in previous versions of EMFAC-the
data suggest that brake wear emissions are highest at moderate speeds. At lower

146
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

speeds the braking events tend to be frequent but mild, and at high speeds
braking is relatively infrequent.
• There appears to be a correlation to weight as indicated in Figure 4.3.7-4.
Although beyond the scope of this update, it is possible that future updates may
want to address the vehicles load in use.
Figure 4.3.7-4. Total vehicle test cycle PM mass emissions vs simulated vehicle test
weight, categorized by pad material.
35
Total Vehicle PM Mass Emissions

30

25
F-150
20
(mg/mi)

Camry Sienna
15
HLW
Sienna
10
F-150
5 Rogue
Civic
0 Prius
1200 1700 2200 2700 3200
Vehicle Tested Weight (kg)
OES-NAO AfterMkt-NAO

The updated emission rates in EMFAC2021 are significantly lower than EMFAC2017. The
results indicate that emissions are approximately 75% lower than previous estimates.
Vehicles with regenerative braking may be 50% or less than those of conventional
vehicles. As illustrated in Figure 4.3.7-5, PM 2.5 shows a strong correlation with PM10.
Therefore, staff modeled PM10 and treat PM 2.5 as a fraction of the PM10. The ratio of
PM2.5/PM10 was found to be 0.35.

147
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure 4.3.7-5. PM2.5 mass emission rate vs. PM10 mass emission rate with linear
trend line as measured by 100S4

4 y = 0.309x + 0.2023
100S4 PM2.5 (mg/mi)

R² = 0.9513
3

0
0 2 4 6 8 10 12 14 16
100S4 PM10 (mg/mi)

4.3.7.2 Heavy-Duty Brake wear

Heavy-duty brake wear emission rates in EMFAC2021 were updated using dynamometer
test data from a Caltrans contract 65A0703. 65 Previous estimates relied on outdated data
and did not account for speed effects. Brake emissions tests were performed on LINK
Engineering (LINK)’s single wheel brake dynamometer for several heavy-duty brake wear
configurations, including Class 8 (GVWR > 33,000 lbs.) drum, Class 8 air disc, refuse truck
air disc, bus air disc, and hydraulic disc. Within these brake configurations, tests were
completed for different hardware configurations. steer, drive, and trailer axles (Class 8
only).
Table 4.3.7-3 shows the test matrix for HD brake wear tests for Class 8 (heavy heavy-
duty) vehicles with drum and disc brakes, medium heavy-duty (hydraulic disc; test weight
of 26,000 lbs.), as well as refuse, and urban bus with disc brakes. Various vocational
cycles with different average speeds, developed through the UC Riverside heavy-duty
activity study, 66 were used to characterize how brake wear emission rates vary by speed
heavy heavy-duty and medium heavy-duty truck tests. The UC Riverside study developed
vocational cycles using second-by-second activity data from 90 heavy-duty vehicles
operating in California, which were grouped into various vocations (e.g. line haul). Briefly,
“Drayage N” are drayage trucks operating in Northern California, “Cement” are cement

65
Link to final report once it’s released
66
Boriboonsomsin et al. 2017 Collection of Activity Data from On-Road Heavy Duty Vehicles, Final Report
for ARB Agreement 13-301, May 2017.
https://ww2.arb.ca.gov/sites/default/files/classic//research/apr/past/13-301.pdf
148
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

mixers, “LH OOS” are long-haul trucks that are registered out-of-state, “Refuse” are
garbage trucks, “Urban Buses” are urban or transit buses, “Beverage” are beverage
distribution trucks, and “Local Moving” are vehicles that do pick-up and delivery.
Table 4.3.7-3. Test Matrix for HD PM Brake Wear Emissions
Test Fixture Cycle 1 Cycle 2 Cycle 3
Drayage – Northern Long-haul out-of-state
Class 8 Drum Steer Loaded Cement
California (N) (LH OOS)
Class 8 Drum Drive Unloaded Drayage N Cement LH OOS
Class 8 Drum Drive Loaded Drayage N Cement LH OOS
Class 8 Drum Trailer Unloaded Drayage N Cement LH OOS
Class 8 Drum Trailer Loaded Drayage N Cement LH OOS
Class 8 Disc Steer Loaded Drayage N Cement LH OOS
Class 8 Disc Drive Unloaded Drayage N Cement LH OOS
Class 8 Disc Drive Loaded Drayage N Cement LH OOS
Refuse Truck ADisc Steer Refuse
Refuse Truck ADisc Drive Refuse
Urban Bus ADisc Steer Urban Bus
Urban Bus ADisc Drive Urban Bus
Hydraulic Disc Steer Beverage Local Moving
Hydraulic Disc Drive Beverage Local Moving

The test matrix was constructed with brake types, axle types, vocational cycle,
equipment, and loading:
• Urban Bus: one vehicle x 2 axle types (steer and drive) x 2 equipment types (OE
and AM), single test
• Refuse: one vehicle x 2 axle types x 2 equipment types, single test
• Medium Heavy-Duty: one vehicle x 2 axle types x 2 vocations x 2 equipment
types, single test
• Heavy Heavy-Duty: two vehicles (drum and disc brakes) x 3 axle types (including
trailer) x 3 vocations x 2 equipment types. Subset loaded and unloaded. Single
test with repeats on subset of trailer tests
Table 4.3.7-4 shows the average speeds of the vocational cycles listed above. Cycles with
smaller average speeds (e.g., drayage) generally have larger brake power densities.
Three vocational cycles were selected for heavy heavy-duty vehicles; the drayage cycle
represents lower speed operation (e.g., at a port) and therefore more frequent braking,
while the long-haul OOS cycle covers higher speeds (e.g., driving on the freeway) and
therefore less frequent braking. Finally, medium speed operation was best represented
by the cement cycle. Medium heavy-duty vehicles were tested on two vocational cycles,
one for lower speed operation (e.g., stop-and-go delivery) and one for medium speeds
(e.g., operating on city streets).

149
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Table 4.3.7-4. Average vocational cycle speeds for medium heavy-duty and heavy
heavy-duty brake wear tests
Class Vocation Average Speed (mph)
Drayage N 11.9
Heavy Heavy-Duty Cement 28.1
Long Haul OOS 48.6
Beverage 14.2
Medium Heavy-Duty
Local Moving 32.6
Refuse Refuse 11.1
Urban Bus Urban Bus 14.9

For HD trucks, raw filter results from single wheel dynamometer testing were
transformed to full vehicle EMFAC2021 emission rates. Brake wear ZMRs in EMFAC are
expressed as grams per vehicle-mile traveled (grams/mile). Transforming single wheel
dynamometer results to represent full-truck gram/mile rates required accounting for
load, axle, and speed factors designed into the emission test matrix, primarily for heavy
heavy-duty trucks. In rolling up single-wheel ZMR, the following factors were accounted
for:
• The mix of loaded and unloaded operation (heavy heavy-duty trucks only);
• The number of steers, drive, and (if applicable) trailer brakes per truck;
• The fraction of particles dispersing to the environment vs. residing within brake
housing (airborne fraction);
• The mix of drum and air disc brakes by model year range; (heavy heavy-duty
trucks only);
• Differences in PM emissions from original and aftermarket friction material, which
would form the basis of deterioration rates
• Vocation cycle results
Figure 4.3.7-6 shows a comparison between processed PM2.5 and PM10 brake wear
emission rates and the range of previous EMFAC assumptions, which represent PM10
emission rates for medium heavy-duty (133 mg/mile) and heavy heavy-duty (63 mg/mile).
These results are for full trucks and account for the list of factors above. Overall, the
updated results are similar to previous assumptions. Heavy heavy-duty vehicle emission
rates at lower speeds are higher than EMFAC2017 rates, especially for disc brake
systems. Refuse or solid waste collection vehicle emission rates were underestimated by
previous assumptions. T6 or medium-heavy trucks and bus emission rates are also
substantially smaller than the previously assumed emission rate of 133 mg/mile.

150
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure 4.3.7-6. Heavy-Duty Full Truck Brake Wear ZMRs

350
PM2.5-10
300 PM2.5
Brake Wear Emissions (mg/mile)

250 Range of Previous Assumptions

200

150

100

50

0
Urban Bus

Beverage
Refuse

Local Moving

3_LH OOS

3_LH OOS
1_Drayage

1_Drayage
2_Cement

2_Cement
ADisc ADisc HDisc ADisc Drum
Bus Refuse Medium Heavy Heavy-Duty
Heavy-Duty

Heavy heavy-duty disc and drum emission rates were combined into one factor by model
year groups. These groups reflect the shift from brakes to disc brakes due to NHTSA’s
Reduced Stopping Distance rules (NHTSA 2009 67). Market surveys indicate that this shift
has been slow, with currently 15% disc market penetration estimated for 2010-2025.
Trailer brakes will likely continue using drum brakes because tractors with discs can pull
older drum-equipped tractors. In the future, the percentage of disc brakes is expected to
increase. Table 4.3.7-5 lists the splits between drum and disc brakes used to develop
heavy heavy-duty brake wear emission rates by model year.

67
https://www.federalregister.gov/documents/2013/02/11/2013-02987/federal-motor-vehicle-safety-
standards-air-brake-systems
151
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Table 4.3.7-5. T7 Drum vs. Disc by Model Year Range

Pre 2010 2010-2025 2026+
Drum 100% 85% 50%
Disc 0% 15% 50%

Table 4.3.7-6 provides a mapping between EMFAC 2021 categories and the brake wear
categories and Table 4.3.7-7 provides ZMRs by model year. Unlike the light-duty brake
wear update, there was no significant deterioration (i.e., emission rate increases due to
aftermarket parts). Emission rate tests with aftermarket brake pads did not differ
significantly from OEM brake pads. Though this finding differed from LD vehicles, this is
consistent with market trends. In the U.S., the majority of commercial vehicle brake
components are supplied by only a few companies, which provide both original and
aftermarket parts. The friction material formulations do not vary significantly between
original and aftermarket, unlike the light vehicle market, where many aftermarket-only
suppliers produce many varieties of pads. The EMFAC 2021 brake wear emission rates
are assumed to be 50% lower for zero emission heavy-duty vehicles due to regenerative
braking.
Table 4.3.7-6. EMFAC2021 Brake Wear Category Mapping
Brake Wear Category EMFAC2021 Categories
Heavy Heavy-Duty All T7 truck categories (except T7 SWCV Class 8), Motor Coach
Medium Heavy-Duty All T6 truck categories, SBUS, OBUS, All Other Buses
Refuse T7 SWCV Class 8
Urban Bus UBUS

Table 4.3.7-7. EMFAC 2021 Input PM10 ZMRs by Vehicle Category

PM10 ZMR (g/mile)
Model Year Range
Refuse Medium Heavy-Duty Heavy Heavy-Duty UBUS
Pre2010 0.21 0.047 0.129 0.11
2010-2025 0.21 0.047 0.096 0.11
2026+ 0.21 0.047 0.106 0.11

To estimate the SCFs, full truck emission rates calculated as described for each vocation
cycle were mapped to speed bins based on closest cycle average speed. Emissions for
speed bins between the average speeds of two vocation cycles were interpolated based
on speed bin midpoint. Table 4.3.7-8 lists all the updated SCFs for heavy heavy-duty and
medium heavy-duty categories.

152
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Table 4.3.7-8. EMFAC 2021 Brake Wear PM Speed Correction Factors (SCFs)
PM SCF
Speed Bin
Medium Heavy-Duty Heavy Heavy-Duty
5 1.31 1.43
10 1.31 1.43
15 1.31 1.41
20 1.29 1.38
25 1.06 1.35
30 0.94 1.33
35 0.88 1.12
40 0.88 0.98
45 0.88 0.83
50+ 0.88 0.72

Aggregate HD PM10 and PM2.5 ZMRs were then calculated as a weighted average of
truck category ZMRs using EMFAC VMT fraction for heavy heavy-duty (0.42), medium
heavy-duty (0.27), refuse (0.18) and urban bus (0.13) categories. PM2.5 fraction was then
calculated as aggregate HD PM2.5 ZMR divided by aggregate HD PM10 ZMR. The result
was 0.35, which matches the light-duty PM2.5 fraction for brake wear. This is lower than
the previously assumed value in EMFAC2017 of 0.42.

4.4 Activity Profiles

4.4.1 Light Duty Activity Profiles
VMT distribution by speed. The VMT speed distributions by speed for LDV in
EMFAC2021 have been updated based on the most recent activity profiles from
Metropolitan planning organizations (MPOs) at the time of release and shown below. 68
Compared to the previous version, LDV activity profiles for more calendar years and sub-
areas are included in EMFAC2021. Figure 4.4.1-1 illustrates updates to light duty vehicle
speed distribution as compared to the one used by EMFAC2017.

68
Internal communication with CARB’s Sustainable Transportation and Communities Division. MPO009 is
used in EMFAC2021. This version of MPO data is reflective of latest data submittals to CARB as of
December 2020 and include SCAG’s 2020 RTP known as Connect SoCal.
153
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure 4.4.1-1 Light Duty Vehicle Speed Distribution Comparison between

EMFAC2017 and EMFAC2021

18%
EMFAC2017
16%
EMFAC2021
Fraction by Percentage, %

14%
12%
10%
8%
6%
4%
2%
0%
0 10 20 30 40 50 60 70 80
Speed (mph)

Updates to Motorcycles Accrual rates. For EMFAC2000, on-road motorcycle accrual

rates were estimated using data from the Motorcycle Industry Council’s (MIC) survey in
1990. For EMFAC2017, accrual rates were estimated for each GAI of the state. From the
EMFAC2017 model output, the statewide yearly mileage accrual rate can be determined
by model year using the VMT and population up to 45 years.
As a recent source of motorcycle mileage data, the 2017 National Household Travel
Survey-California Add-on (NHTS-CA) provided a vehicle subset of 1,923 motorcycles
surveyed for odometer information by age. 69 Figure 4.4.1-2 shows the comparison of the
mileage accrual rates by age from both EMFAC2017 and the 2017 NHTS-CA survey. As
shown in the figure, the accrual rates trend line from the 2017 NHTS-CA survey is lower,
which results in the reduction of total emissions in the EMFAC2021 model.

69
https://www.nrel.gov/transportation/secure-transportation-data/tsdc-nhts-california.html

154
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure 4.4.1-2 On-road Motorcycle Accrual Rates – EMFAC2017 vs. 2017 NHTS-CA
Add-on

6,000

5,000
EMFAC2017
Miles per year

4,000
2017 NHTS CA
3,000

2,000

1,000

0
0 5 10 15 20 25 30 35 40 45 50
Age (years)

4.4.2 Heavy Duty Activity Profiles

Similar to light duty vehicles, heavy-duty vehicle’s activity profiles have significant effects
on the emissions produced from these vehicles. Starting from EMFAC2017, utilization of
Portable Activity Measurement Systems (PAMS) data has enabled accurate
characterization of HD vehicle activity in the real world, which is critical to nowadays HDV
emission inventory development. In the development of EMFAC2021, more PAMS data
are collected and used for HD activity updates.
In total, PAMS data from 174 HD vehicles collected by UCR and West Virginia University
were utilized. Vehicle location and activity information were collected by Global
Positioning System (GPS) and electronic control unit (ECU) data loggers at 1 Hz
resolution. Activity information includes timestamp, vehicle instantaneous speed, engine
speed, etc. Data cleaning procedures were performed to remove any vehicle sample
with overly short sampling duration (i.e., less than 24 hours), low quality data, too many
missing data, abnormal behavior (e.g., too many engine starts, too slow speed). Finally,
168 vehicle samples are used for EMFAC2021 updates. These samples are pooled with
90 vehicle samples inherited from EMFAC2017 and together make a total sample size of
258 vehicles.

155
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

In EMFAC2017, the 90 sample vehicles from UCR CE-CERT study 70 were assigned vehicle
categories based on their vocations. EMFAC2021 adopts a new vehicle category schema:
EMFAC202x vehicle class. Hence the assignment of vehicle categories is re-done through
the following steps: 1) if the vehicle has VIN information, it is searched in DMV and IRP
dataset to find a matched record and corresponding EMFAC2021 vehicle class; 2) if the
vehicle does not have VIN, vocation and weight class are used to identify the vehicle’s
category as detailed in Table 4.4.2-1; 3) vocation and GPS records are used to validate
VIN identification results. For example, a truck is identified as T7 CAIRP based on VIN in
DMV but its vocation is assigned by the contractor as a line-haul in-state tractor. By
checking its GPS trajectory, it is known that the truck only has activities within a city
boundary in California, therefore this truck is more representative of an in-state tractors
and assigned as T7 Tractor in EMFAC.
Table 4.4.2-1. Sample sizes and vocation-category mapping used for EMFAC2021
HDV activity updates
Number of new
Number of
samples acquired Updated in
EMFAC2021 Vehicle Categories samples inherited
from 200-Vehicle EMFAC2021?
from EMFAC2017
Project
Do not exist in
Not Found 8 0
EMFAC2021
All Other Buses 10 0 No
T6 CAIRP, T6 OOS, T7 CAIRP, T7
5 18 Yes
NNOOS, T7 NOOS, Motor Coach
SBUS 0 27 Yes
T6 Instate Delivery 2 2 Yes
T6 Instate Tractor/Others 4 7 Yes
T7 POLA/POAK/Other Port 4 37 Yes
T6/T7 Public 22 1 No
T7 Single Concrete/Transit Mix, T7
11 4 Yes
Single Dump, T7 Single Other
T7 SWCV 6 27 Yes
T7 Tractor 8 35 Yes
T6/T7 Utility 2 0 No
UBUS 5 10 Yes

It should be noted that not all EMFAC202x HD vehicle categories’ activity profiles are
updated in EMFAC2021. For example, T6/T7 Public, T6/T7 Utility, All Other Buses are

70
Boriboonsomsin, K., Johnson, K., Scora, G., Sandez, D., Vu, A., Durbin, T., & Jiang, Y. (2017) Collection
of Activity Data from On-Road Heavy-Duty Diesel Vehicles. Available at
https://www.arb.ca.gov/research/apr/past/13-301.pdf

156
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

not updated in this version due to lack of samples in the new acquired PAMS dataset. In
addition, no significant difference was observed across fuel types. As a result, HD activity
updates are applied to EMFAC202x vehicle categories regardless of fuel types. Sample
size by region is too small to be representative, so HD activity updates are done to all
regions the same way, except for SCAG regions which have their own vehicle activity
profiles estimated by SCAG Truck model 71. Since EMFAC is designed to estimate
emissions on an average weekday, results were generated using weekday data, i.e.,
abandoning all activity data on weekends before analyses.
VMT distribution. Two dimensions of VMT distribution are modeled in EMFAC: by hour
and by speed. The hourly VMT distribution refers to when vehicles are active/inactive on
an average day, measured by VMT in each hour of the day. The speed distribution for
HD vehicles refer to the fractions of VMT in each speed bin. While EMFAC2017 updated
VMT distribution by speed but not by hour due to limited sample size, EMFAC2021
updated both. Similar to EMFAC2017, VMT distributions in EMFAC2021 are developed
for each vehicle category.
VMT distribution by hour. In general, VMT distribution by hour in EMFAC2021 shows
similar trends with EMFAC2017. Some vehicle categories’ by-hour distribution curve is
smoother and more continuous than EMFAC2017, for example out-of-state HD vehicles,
school bus, heavy-heavy duty port trucks, heavy-heavy duty tractors. School bus is
updated using real-world data for the first time. Out-of-state trucks, medium-duty in-
state tractor/other shows higher VMT early in the morning, while heavy heavy-duty port
trucks, heavy-heavy duty single trucks show higher VMT during work hours instead of late
night. Heavy-heavy duty solid waste collecting vehicle shows higher VMT in the
afternoon than EMFAC2017. Heavy heavy-duty tractors show higher VMT during night
time than day time compared to EMFAC2017, which can be contributed by the long-haul
travel patterns. Please see detailed VMT distribution by hour figures and comparison
with EMFAC2017 in Appendix 6.10.
VMT distribution by speed. In general, VMT distribution by speed trends are similar to
EMFAC2017. Some HD vehicle categories show higher VMT at higher speed (> 55mph),
for example out-of-state HD vehicles (including T6 CAIRP, T6 OOS, T7 CAIRP, T7 NOOS,
T7 NNOOS, Motor Coach due to their across-state-border traveling patterns), medium-
heavy duty in-state tractor/other, heavy-heavy duty port trucks, and heavy-heavy duty
single trucks. In EMFAC2021, school bus has lower VMT at high speed compared to
EMFAC2017, with majority of VMT around 40 mph, which is closer to real-world situation
since school buses operate a lot on local roads. Medium-heavy duty in-state delivery
truck has much lower activity at high speed compared to EMFAC2017, with majority of
the VMT at around 35-40 mph, possibly due to mostly traveling locally compared to
tractors. Urban bus has higher VMT around 35 mph, compared to EMFAC2017 at 20
mph. Prior to EMFAC2021, urban bus did not use real-world data to model its VMT

71
https://scag.ca.gov/heavy-duty-truck-model
157
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

distribution. Heavy-heavy duty solid waste collecting vehicle and heavy-heavy duty
tractor have very similar VMT distribution by speed to EMFAC2017. Please see detailed
VMT distribution by speed and comparison with EMFAC2017 in Appendix 6.10.
Starts and soak time distribution. Number of engine starts are directly related to start
emissions. On top of that, temperature of engine at the time of engine start, which is
often determined by time interval since last engine off (i.e. soak time), plays an important
role in affecting efficacy of pollutants removed through catalyst converters. Therefore,
both number of engine starts per weekday, and engine starts distribution by soak time
are critical information to estimate start emissions. As discussed earlier for VMT
distribution, selected EMFAC2021 vehicle category are mapped to available tested
groups, and then average starts per weekday and distributions by soak time were
applied to EMFAC2021 vehicle categories using the same mapping as in Table 4.4.2-1.
Similar to EMFAC2017, number of engine starts in EMFAC2021 are developed for each
vehicle category. All individual vehicle samples within each vehicle category are treated
the same.
Following the staff analysis, the average starts per weekday is provided in Table 4.4.2-2.
After including more vehicle samples, most vehicle categories have similar engine starts
per weekday compared to EMFAC2017, with an increasing trend in general. Out-of-state
(long-distance line hauls) trucks have the highest number of starts per day, while refuse
trucks and public fleets remain the lowest numbers. These align with those observed in
EMFAC2017.
Table 4.4.2-2. Engine starts per weekday of HDV in EMFAC2021 and EMFAC2017
EMFAC2021 Vehicle Number of engine starts Number of engine starts Updated in
Categories in EMFAC2021 in EMFAC2017 EMFAC2021?
All Other Buses 8.9 8.9 No
T6 Instate Delivery 14.27 11.54 Yes
T6 Instate Tractor 11.56 11.54 Yes
T7 POLA 16.36 7.6 Yes
T7 Single Other 9.42 11.54 Yes
T7 SWCV 4.60 3.9 Yes
T7 Tractor 14.53 12.7 Yes
OOS 22.98 14.6 Yes
SBUS 14.48 11.54 Yes
T6/T7 Public 3.0 3.0 No
T6/T7 Utility 11.5 11.5 No

The soak time distribution is defined as the fraction of starts with preceding soak time in
one of the 19 soak time bins at a specific hour of the day. The average starts with soak
time ≥ 2 hours per weekday is provided in Table 4.4.2-3. Similar to total number of starts
per weekday, starts with longer soak time (≥2hr) in EMFAC2021 have a slight increase
compared to EMFAC2017. For most vocations, the soak time distribution is dominated

158
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

by short soaking events of less than 5 minutes. The starts distributions by soak time and
by hour are provided in Appendices 6.11 and 6.12.
Table 4.4.2-3. Engine starts with soak time >= 2hr per weekday of HDV in
EMFAC2021 and EMFAC2017
Number of engine starts Number of engine starts in
EMFAC2021 Vehicle Categories
in EMFAC2021 EMFAC2017
All Other Buses 0.22 0.22
T6 Instate Delivery 3.10 2.78
T6 Instate Tractor 3.38 2.78
T7 POLA 1.41 0.56
T7 Single Other 1.05 2.78
T7 SWCV 1.35 1.19
T7 Tractor 2.15 1.11
OOS 1.59 1.47
SBUS 2.06 2.78
T6/T7 Public 0.30 0.30
T6/T7 Utility 0.20 0.20

Idling hours. The HD idling hours refers to time spent in extended idling activity that
usually occur at trip origins and destinations such as work site, or at rest stops. Duration
of idling events have a direct effect on idling emissions. In EMFAC2021, an HD Extended
Idling Event is defined as a continuous segment of vehicle activity that meets three
criteria: all instantaneous vehicle speeds being lower than 5 mph, the total distance of
less than 1 mile, and the total duration of more than 5 minutes. Adopting the same
method used in EMFAC2017, number of extended idling hours per weekday in
EMFAC2021 are developed for each vehicle category. All individual vehicle samples
within each vehicle category are treated the same. For all HDV except Heavy-heavy duty
CAIRP trucks, Heavy-heavy duty Out-of-state trucks, Heavy-heavy duty Single trucks,
Heavy-heavy duty Tractor, the extended idle hours per day is a set value depending only
on vehicle category, as presented in Table 4.4.2-4. For vehicle categories updated in
EMFAC2021, extended idling hours per weekday are longer compared to EMFAC2017.
School buses have the longest idling hours, while T6 Instate Delivery trucks and Tractors
have the shortest idling periods. These trends are observed in both EMFAC2021 and
EMFAC2017.
Table 4.4.2-4. Extended idling hours per weekday of HDV (except Heavy-heavy duty
Out-of-state trucks, Single trucks, and Tractors) in EMFAC2021 and EMFAC2017
EMFAC2021 Vehicle Extended idling hours in Extended idling hours Updated in
Categories EMFAC2021 in EMFAC2017 EMFAC2021
All Other Buses (OB) 0.098 0.098 No
T6 Instate Delivery 0.33 0.098 Yes
T6 Instate Tractor 0.35 0.098 Yes

159
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

EMFAC2021 Vehicle Extended idling hours in Extended idling hours Updated in

Categories EMFAC2021 in EMFAC2017 EMFAC2021
T7 POLA 1.44 1.38 Yes
T7 SWCV 0.58 0.63 Yes
SBUS (SB) 2.06 2.78 Yes
MCH 1.69 1.69 No
Public 0.51 0.51 No
Utility 0.27 0.27 No

Under the California truck idling regulation, heavy heavy-duty diesel trucks’ (HHDT) idling
requirements vary by both calendar year and model year, as presented in Table 4.4.2-5.
Newly acquired vehicle samples are all collected after year 2008 and for vehicle of model
year 2008 or newer. Therefore, only heavy heavy-duty single trucks and tractors in
calendar year later than 2008 and model year later than 2008 get updated in
EMFAC2021. For the rest of the HD fleets, or fleet of pre-2008 model year, EMFAC2017
assumptions were applied. Please note that for long-distance line hauls (T7 CAIRP and T7
OOS), the updated PAMS analysis results are significantly different than those historically
assumed in EMFAC. Due to lack of representative sample size and supporting evidences
from other empirical data, the idling hours for these two vehicle categories are not
updated in EMFAC2021 and remained the same as previous versions of the EMFAC
model.
Table 4.4.2-5. Extended idling hours per weekday of HHDT in EMFAC2021 and
EMFAC2017
EMFAC2021 Calendar Model Extended idling Extended idling
Updated in
Vehicle Year Year hours in hours in
EMFAC2021?
Categories Range Range EMFAC2021 EMFAC2017
T7 CAIRP 2005-2007 2008+ 4.28 4.28 No
T7 CAIRP 2005-2007 pre 2008 4.28 4.28 No
T7 CAIRP 2008+ 2008+ 0.97 4.41 No
T7 CAIRP 2008+ pre 2008 0.22 0.22 No
T7 CAIRP pre 2005 pre 2008 4.41 4.41 No
T7 OOS 2005-2007 2008+ 5.42 5.42 No
T7 OOS 2005-2007 pre 2008 5.42 5.42 No
T7 OOS 2008+ 2008+ 0.97 5.47 No
T7 OOS 2008+ pre 2008 0.21 0.21 No
T7 OOS pre 2005 pre 2008 5.47 5.47 No
T7 Single 2005-2007 2008+ 0.36 0.36 No
T7 Single 2005-2007 pre 2008 0.36 0.36 No
T7 Single 2008+ 2008+ 0.79 0.92 Yes
T7 Single 2008+ pre 2008 0.25 0.25 No
T7 Single pre 2005 pre 2008 0.79 0.79 No
T7 tractor 2005-2007 2008+ 0.36 0.36 No

160
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

EMFAC2021 Calendar Model Extended idling Extended idling

Updated in
Vehicle Year Year hours in hours in
EMFAC2021?
Categories Range Range EMFAC2021 EMFAC2017
T7 tractor 2005-2007 pre 2008 0.36 0.36 No
T7 tractor 2008+ 2008+ 1.46 0.79 Yes
T7 tractor 2008+ pre 2008 0.25 0.25 No
T7 tractor pre 2005 pre 2008 0.79 0.79 No

HD accrual rates
HDV mileage accrual rates in EMFAC2017 72 were similar to those used in EMFAC2014
and EMFAC2011. HDV mileage accrual rates in EMFAC2011 primarily relied on data
from the 2002 Vehicle Inventory and Use Survey (VIUS) 73 supplemented with CARB
survey data as documented in the 2008 Truck and Bus (T&B) Technical Appendix. 74
Updated EMFAC2021 HDV mileage accrual rates reflect new data that recently become
available for use. Caltrans completed a CalVIUS survey that had questions designed to
obtain annual freight truck activities specific to California as was discussed in the
Caltrans’ California Transportation Plan 2040 (CTP 2040) 75. The CalVIUS survey was
developed to fill the gap created by the discontinuance of the federal VIUS survey
process after 2002. Additionally, CARB’s contract to examine potential sources of HDV
accrual rate data for designated fleet types based on vocations and weight classes was
completed through an extramural contract 76. In addition to the CalVIUS survey data,
CARB utilized logged vehicle data provided by a telematics service provider with over
1.3 million tracking devices deployed on heavy-duty vehicles. The updated HDV mileage
accrual rates reflect the processed CalVIUS and telematics data results. The American
Transportation Research Institute 2019 Report on the Operational Costs of Trucking 77
indicated the average miles driven per year per truck was 91,506 in 2018 with an average
age of 4.4 for the trucks. The updated VMT in EMFAC2021 for long haul Class 8 trucks
agreed with these results, reflecting an average weighted annual mileage of 91,876 with
an average truck age between 4 to 5 years.
The updated annual accrual rates in EMFAC2021 will reflect more current activity trends.
EMFAC2017 underestimated annual accrual rates for some fleets, including Interstate
Tractors, Public/Utility Trucks and Solid Waste Collection Vehicles, while it overestimated

72
EMFAC2017 Technical Documentation at https://ww2.arb.ca.gov/our-work/programs/mobile-source-emissions-
inventory/road-documentation/msei-modeling-tools-emfac
73
https://www.census.gov/econ/overview/se0501.html
74
Table 1 in http://www.arb.ca.gov/regact/2008/truckbus08/appg.pdf cites sources used to derive mileage accrual
rates.
75
https://dot.ca.gov/-/media/dot-media/programs/legislative-affairs/documents/f0004899-ctp2040-
a11y.pdf
76
https://ww2.arb.ca.gov/sites/default/files/2021-01/erg_finalreport_hdv_accruals_20190614.pdf
77
Page 30 of https://truckingresearch.org/wp-content/uploads/2019/11/ATRI-Operational-Costs-of-Trucking-
2019-1.pdf
161
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

annual accrual rates for other fleets such as T7 Tractors, T6 Heavy Instate Other and T7
Single Other Vehicles and School Buses.
The following figures display the updated EMFAC2021 accrual rates based on the CA-
VIUS and Geotab data with the EMFAC2017 accrual for comparison. Figure 4.4.2-1
through Figure 4.4.2-3 display accrual for the Interstate long-haul tractors based in
California (CAIRP), based in near states/provinces (NOOS) and in not near
states/provinces (NNOOS). Figure 4.4.2-4 and Figure 4.4.2-5 display the accrual for Class
8 and Class 7 Instate tractors. Figure 4.4.2-6 and Figure 4.4.2-7 display accrual for
Delivery Vehicles which are a new vehicle category, and as such there is no EMFAC2017
accrual to display.
Figure 4.4.2-1. Interstate Accrual Rate for Class 7&8 CAIRP Tractors (N=1,382)

120,000

100,000
Annual Miles Accrued

80,000

60,000

40,000

20,000 EMFAC2021
EMFAC2017
-

Vehicle Ages

162
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure 4.4.2-2. Interstate Accrual Rate for Class 7&8 NOOS Tractors (N=6,874)

140,000

120,000

100,000
Annual Miles Accrued

80,000

60,000

40,000
EMFAC2021
20,000 EMFAC2017
-
10
11
12
13
14
15
16
17
18
19
0
1
2
3
4
5
6
7
8
9

20+
-1

Vehicle Ages
Figure 4.4.2-3. Interstate Accrual Rate for Class 7&8 NNOOS Tractors (N= 1,424)

140,000

120,000

100,000
Annual Miles Accrued

80,000

60,000

40,000 EMFAC2021
20,000 EMFAC2017

-
10
11
12
13
14
15
16
17
18
19
0
1
2
3
4
5
6
7
8
9

20+
-1

Vehicle Ages

163
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure 4.4.2-4. Instate Accrual Rate for Class 8 Tractors (N=5,551)

90,000
80,000
Annual Miles Accrued

EMFAC2021
70,000
EMFAC2017
60,000
50,000
40,000
30,000
20,000
10,000
-

10
11
12
13
14
15
16
17
18
19
0
1
2
3
4
5
6
7
8
9

20+
-1

Vehicle Ages
Figure 4.4.2-5. Instate Accrual Rate for Class 7 Tractors (N=456)

90,000
EMFAC2021
80,000
Annual Miles Accrued

EMFAC2017
70,000
60,000
50,000
40,000
30,000
20,000
10,000
-
10
11
12
13
14
15
16
17
18
19
0
1
2
3
4
5
6
7
8
9

20+
-1

Vehicle Ages

164
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure 4.4.2-6. Instate Accrual Rate for Class 7 Delivery Vehicles (N=392)

25,000
EMFAC2021
Annual Miles Accrued

20,000

15,000

10,000

5,000

-
10
11
12
13
14
15
16
17
18
19
0
1
2
3
4
5
6
7
8
9

20+
-1

Vehicle Ages

Figure 4.4.2-7. Instate Accrual Rate for Class 4-6 Delivery Vehicles (N=1,335)

18,000
EMFAC2021
16,000
Annual Miles Accrued

14,000
12,000
10,000
8,000
6,000
4,000
2,000
-
10
11
12
13
14
15
16
17
18
19
0
1
2
3
4
5
6
7
8
9

20+
-1

Vehicle Ages
HD odometer schedule
EMFAC uses odometer schedules to model increases in the emission rate due to
deterioration. In EMFAC2017, odometer schedules relied on the 2002 Federal Vehicle
Inventory and Use Survey (VIUS). In EMFAC2021, odometer schedules by age for heavy-

165
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

duty trucks were updated using the 2018 California Inventory and Use Survey (CAVIUS)
data. In EMFAC2021, odometer schedules were updated for heavy heavy-duty and
medium heavy-duty trucks, including T7 Interstate Registration Plan (IRP), T7 out-of-state
(OOS) registered, T7 in-state tractor T7 single-unit trucks, T6 classes 4-6 (GVWR 14,001 –
26,000), and T6 class 7 (GVWR 26,001 – 33,000 lbs.). Following a basics data cleaning,
(e.g., removing unrealistically large odometer mileages), average odometers by age were
calculated for each weight class and vehicle type. The average odometers as a function
of vehicle age were then fit to a polynomial function. The EMFAC 2021 odometer
schedules matches the polynomial fit at younger ages. At older ages, once the odometer
mileage hits a maximum value, the odometer mileage is assumed to be constant at the
maximum value of the fit.
Figures 4.4.2-8 through Figure 4.4.2-13 show the new odometer schedules, the
EMFAC2017 schedules, and the fitted CAVIUS data. Compared to EMFAC2017,
EMFAC2021 odometer schedules are larger for T7 OOS and IRP, leading to higher
deterioration-related emissions. On the other hand, odometer schedules are smaller for
T7 in-state tractors, T7 single unit trucks, and T6 trucks, leading to less deterioration-
related emissions.
Figure 4.4.2-8. T7 OOS Odometer Schedule

1,200,000
Average Odometer (miles)

1,000,000

800,000

600,000

400,000 CAVIUS

EMFAC 2017
200,000
EMFAC 2021
0
-1 9 19 29 39
Vehicle Age

166
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure 4.4.2-9. Updated T7 IRP Odometer Schedule

1,200,000

1,000,000
Average Odometer (miles)

800,000

600,000

400,000
CAVIUS
200,000 EMFAC 2017
EMFAC 2021
0
-1 9 19 29 39
Vehicle Age

Figure 4.4.2-10. Updated T7 Single Odometer Schedule

900,000
800,000 CAVIUS
700,000 EMFAC 2017
Average Odometer (miles)

600,000 EMFAC 2021

500,000
400,000
300,000
200,000
100,000
0
-1 9 19 29 39
Vehicle Age

167
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure 4.4.2-11. Updated T7 Tractor Odometer Schedule

900,000

800,000

700,000
Average Odometer (miles)

600,000

500,000

400,000

300,000
CAVIUS
200,000
EMFAC 2017
100,000 EMFAC 2021
0
-1 9 19 29 39
Vehicle Age (Years)

Figure 4.4.2-12. Updated T6 Class 4-6 Odometer Schedule

450,000

400,000
Average Odometer (miles)

350,000

300,000

250,000

200,000

150,000
CAVIUS
100,000
EMFAC 2017
50,000
EMFAC 2021
0
-1 9 19 29 39
Vehicle Age

168
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure 4.4.2-13. Updated T6 Class 7 Odometer Schedule

450,000

400,000
Average Odometer (miles)

350,000

300,000

250,000

200,000

150,000
CAVIUS
100,000
EMFAC 2017
50,000
EMFAC 2021
0
-1 9 19 29 39
Vehicle Age (Years)

169
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

4.5 Forecasting
4.5.1 Light Duty New Sales and VMT Forecasting
New Vehicle Sales
In EMFAC2021, the annual vehicle population is comprised of vehicles retained from the
prior calendar years, plus new vehicle sales. The count of retained vehicles is calculated
by counting how many vehicles existed in California for prior calendar years. To forecast
new vehicle sales, first the statewide new vehicle sales need to be estimated and then it
is disaggregated to regional level new sales. In EMFAC2017, the forecasting equation for
statewide new sales of LD vehicles, for all fuel types, was developed using a multivariate
regression analysis based on historical socio-econometric time-series data from 2000 –
2016. To do the econometric modeling, the selection of variables aimed to be consistent
with microeconomic theory which dictates that attention must be paid to the
reasonableness of coefficient magnitudes and signs. Hence, the goodness of fit and
significance criteria (such as t-statistic) from potential models, using different variable
combinations, was considered.
While CARB staff used the same socio-econometric modeling approach for development
of the new vehicle sales model for EMFAC2021, the current update included a dynamic
modeling approach where staff utilized an optimization methodology choose the best
socio-economic indicators that describe historical data and create a set of candidate
models for further assessment. As the next step, staff picked the best model that was
more consistent with past and future socioeconomic situation for use in EMFAC2021.
The same criteria for statistical modeling were applied to other regression analysis
efforts, including new vehicle sales, LDV VMT growth trends as discussed in subsequent
sections.
The primary data sources used for this analysis included UCLA Anderson Forecast
(UCLA), California Department of Finance (DOF), and DMV. Table 4.5.1-1 shows a more
detailed list for the sources used in this regression development, spanning the years
2003 – 2019, and in the forecasting equations, starting in 2020. All data variables used
were on a statewide, annual basis.
Table 4.5.1-1. Primary data sources used for EMFAC2021 new vehicle sales
forecasting.
Data Source
New vehicle sales Historical DMV data and UCLA Anderson Report, 2020
Human population and GDP DOF, Jan 2020
Gas price, unemployment rate, housing
UCLA Anderson Report, 2020
starts, disposable income
Gas price CEC Fuel Prices-2019 Forecasts
Federal Rates Federal Reserve Bank of St. Louis, 2019

170
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

To do the modeling, staff looked at seven different econometric indictors. To evaluate

whether any of the parameters listed have any delayed impact on new vehicle sales, all
the selected parameters were lagged by one year. The multivariate regression analyses
was then performed on all the socio-econometric parameters along with their lagged
counterparts.
For the modeling, a large number of models were created using ordinary least square
(OLS) method, including an array of two to seven variable models. Since it is impractical
to look at all available combination of the given parameters, a computer program was
developed to create random n-variable models using a random combination of the given
parameters such as housing starts and unemployment rate; the program then sieved
through the models to keep only the best ones.
The models were created based on historical socio-econometric data obtained for
calendar years 2003 through 2019; and projection or forecast for new vehicle sales was
developed for calendar years 2020 through 2050. Moreover, evaluation of each created
model based a set of automated reasonableness criteria – described below – allowed the
computer program to pick a handful of best models.
To further test the reasonableness of each model manually, a number of statistical
indicators including R2 value and t-statistic value or p-value were calculated and
evaluated for each of the computer-selected best models. R2 value measured the
correlation between the given data and the modeled ones, and p-value provided the
degree of significance of each parameter in the model. Other than these two measure,
the staff checked the sign validity for each model, and ensured that the coefficients of
the parameters of each model have the correct sign. Staff then evaluated parameter
coherency along with the overall impact of the model on forecasts up to 2050. This
exercise boiled down the available models to the best model for new vehicle sales as
shown below.

𝑁𝑁ew vehicle sales per capita Eq. 4.5.1-1

= 0.05744068
− 0.004672403 x UR
+ 0.00271036 x L1_UR
where UR is unemployment rate in percentage and L1_UR is 1-year lagged
unemployment rate. As shown, the count of new vehicles sold in California per capita is a
function of only unemployment rate and the 1-year lagged unemployment rate. The
select model has a high correlation coefficient of 0.89, and less than 0.05 p-values
illustrating the significance of the selected parameters. These values are shown in Table
4.5.1-2.

171
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Table 4.5.1-2. Statistics pertaining to the choice of parameters for the updated
vehicles sales per capita model for EMFAC2021
p-Value
1-year lagged Unemployment Rate R2
Intercept Unemployment Rate (UR)
(L1_ UR)
2.7x10-8 4.17x10-7 2.29x10-4 89%

A comparison between historical new vehicle vehicles sales and the new model forecasts
is provided in Figure 4.5.1-1. It is noteworthy to mention that an estimated new vehicles
for year 2020 based on the sales fata from California New Car Dealers Association 78
(CNCDA) was used to reflect the impact of COVID-19 pandemic on light duty vehicle
sales in year 2020. For years 2021 and onward, the model used the forecasted new
vehicles sales from the regression model described earlier.
Figure 4.5.1-1 Comparison between historical new vehicle sales and estimated new
sales by the regression model

2.5
New Vehicle Sales (million vehicles)

2.0

1.5
Historical
1.0 Forecast

0.5

0.0
2000 2010 2020 2030 2040 2050
Calendar Year

As described above, EMFAC2021 is updated using a new vehicle sales model. The new
model predicts that the count of new vehicles sold in California solely depends on
unemployment rate and 1-year lagged unemployment rate along with population. It
should be noted that the current model is only capable of representing business-as-usual
conditions and is made using the best available data, and factors such as COVID-19
introduce both short- and long-range uncertainties in the ability of the model to
accurately forecast future trends.

78
https://www.cncda.org/wp-content/uploads/Cal-Covering-4Q-20.pdf
172
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Vehicle Miles Travelled (VMT)

Similar to the new vehicles sales, VMT of light duty vehicles at the statewide level is
forecasted using multivariate regression analysis based on historical time-series data
from 2003-2019. Practices described above such as using a computer program to sieve
through automatically generated models and testing the reasonableness of the models
by first the computer program and then the staff were implemented to develop
forecasting models for light duty VMT. For this econometric modeling, it was assumed
that historical light duty vehicles VMT has followed similar trend as statewide gasoline
consumptions after discounting the impact of fuel efficiency improvements. Therefore, in
the absence of any future fuel efficiency improvement, the growth in the motor vehicle
gasoline consumptions should be directly correlated to the growth in the light duty
vehicle VMT. It needs to be noted that light duty vehicles in California are the major
consumers of motor vehicle gasoline.
The primary data sources used for this analysis included economic data from UCLA
Anderson forecasts along with those from Department of Finance (DOF), and fuel sales
data from the California Department of Tax and Fee Administration (CDTFA). Table
4.5.1-3 below shows a more detailed list for the sources used in this regression
development, spanning years 2003 through 2019, and in the forecasting equations,
starting in 2020. All data variables used were on a statewide, annual basis.
Table 4.5.1-3. Primary data sources used for EMFAC2021 new vehicle sales
forecasting.
Data Source
New vehicle sales UCLA Anderson Report, 2020
Human population and GDP DOF, Jan 2020
Gas price, unemployment rate, housing starts,
UCLA Anderson Report, 2020
disposable income
Gas price CEC Fuel Prices-2019 Forecasts
Federal Rates Federal Reserve Bank of St. Louis, 2019
California Department of Tax and Fee Administration
Fuel Usage
(CDTFA), 2019 Motor Vehicle Fuel Tax (MVF)

The chosen regression model for annual VMT of light duty vehicles at the statewide level
is:

𝑉𝑉𝑉𝑉ℎ𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀 𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇 (𝑉𝑉𝑉𝑉𝑉𝑉) Eq. 4.5.1-2.

= −381.4848092
− 13.74702754 × 𝐺𝐺𝐺𝐺𝐺𝐺𝐺𝐺
+ 18.91130716 × 𝑃𝑃𝑃𝑃𝑃𝑃
+ 0.024905874 × 𝐿𝐿1𝐻𝐻𝐻𝐻𝐻𝐻

173
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

where GASP is gas price in dollars, POP is human population in millions, and L1HST is
the 1-year lagged national housing starts in thousand units. The chosen model has a high
correlation coefficient of 0.98, and less than 0.05 p-values illustrating the significance of
the selected parameters. These values are shown in Table 4.5.1-4.
Table 4.5.1-4. Statistics pertaining to the choice of parameters for the updated VMT
model for EMFAC2021
p-Value
Gas price Human population 1-year lagged national housing starts R2
Intercept
(GASP) (POP) (L1HST)
2.7x10-8 1.1x10-5 ~0.0 9.3x10-7 98%

A comparison of historical VMT data and those estimated from the new model is shown
in Figure 4.5.1-2. It is noteworthy to mention that an estimated VMT for year 2020 based
on the CDTFA taxable fuel sales data was used to reflect the impact of COVID-19
pandemic on light duty vehicle VMT in year 2020. For years 2021 and onward, the model
used the forecasted VMT from the regression model described earlier.
Figure 4.5.1-2. Comparison between historical VMT data and estimates from the
VMT regression model

500
VMT (million miles per year)

400

300

200
Historical
100 Forecast

0
2000 2010 2020 2030 2040 2050
Calendar Year

4.5.2 Heavy Duty New Sales and VMT Forecasting

New Vehicle Sales
EMFAC2021 continues to use the heavy-duty new vehicle sales forecasting method used
in EMFAC2014 and EMFAC2017 (refer to Section 3.3.4.1.2.1 of the EMFAC2014

174
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Technical Documentation) 79. The forecasting method begins with the national HD new
vehicle sales growth trend, which is obtained from the Annual Energy Outlook (AEO)
report 80 of the U.S. Energy Information Administration (EIA). The national growth trend is
translated to California’s specific new HD vehicle sales growth trend by applying the ratio
between the national VMT growth based on AEO and California’s VMT growth. For more
details, please refer to Section 4.5.3 of the EMFAC2017 technical documentation 81.
While EMFAC2017used 2005 as the base year for new sales growth, EMFAC2021 uses
2019 as base year. The national VMT and growth trends are displayed in Figure 4.5.2-1 .
For more details, please refer to Section 4.5.3 of the EMFAC2017 technical
documentation 82.
Figure 4.5.2-1. National Heavy-Duty VMT and New Sales Growth Trend Reported by
AEO (Relative to 2019)

1.60
1.40
Growth Rates (Relative to CY2019)

1.20
1.00
0.80
0.60
0.40
AEO_VMT_Growth
0.20 AEO_NewSales_Growth
0.00

Calendar Year
In
EMFAC2021, except for CA IRP, out-of-state, and Class 8 in-state tractors , the rest of
83

HD categories needs a sales adjustment process to ensure that the forecasted new HD

79
https://ww3.arb.ca.gov/msei/downloads/emfac2014/emfac2014-vol3-technical-documentation-
052015.pdf
80
https://www.eia.gov/outlooks/aeo/
81
https://ww3.arb.ca.gov/msei/downloads/emfac2017-volume-iii-technical-documentation.pdf
82
https://ww3.arb.ca.gov/msei/downloads/emfac2017-volume-iii-technical-documentation.pdf
83
Including T6 CAIRP Class 4, T6 CAIRP Class 5, T6 CAIRP Class 6, T6 CAIRP Class 7, T6 OOS Class 4, T6
OOS Class 5, T6 OOS Class 6, T6 OOS Class 7, T7 CAIRP Class 8, 7 NNOOS Class 8, T7 NOOS Class 8, T7
Single Other Class 8, and T7 Tractor Class 8.
175
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

sales from the above process will meet the forecasted VMT target. The new vehicle sales
for these categories are estimated in such a way that VMT estimated by forecasted
population and accrual rate meets the target VMT determined by base year VMT and
growth rates. Therefore, the number of new vehicle sales for these categories was
determined based on how much VMT were needed to match the target VMT. Details can
be found in Section 3.3.4.1.2.1 of the EMFAC2014 technical documentation.
VMT forecasting
For years 2000-2016, EMFAC2017 used historical data on taxable diesel fuel sales to
normalize the statewide HD VMT rates, so that fuel usage results would match actual fuel
sales results. 84 For EMFAC2021, historical taxable diesel fuel sales continue to be used to
normalize the statewide VMT rates such that fuel usage results match the actual historical
fuel sales for 2000-2019. However, fuel sales data used to be obtained from the Board of
Equalization but now resides with the California Department of Tax and Fee
Administration. 85
EMFAC2017 forecasted future statewide level on-road diesel consumption using an
econometric regression model based on historical time-series data. This method was
based on the assumption that historical diesel consumption in California was directly
correlated to VMT associated with heavy-duty vehicle VMT (the major consumer of
taxable diesel in California), and that in the absence of any diesel fuel efficiency
improvement, the growth in future diesel fuel consumption should be directly correlated
with heavy-duty vehicle VMT. Statewide diesel fuel growth rates from the regression
model were used in EMFAC2017 to forecast the statewide diesel VMT for years 2017-
2050.
EMFAC2021 has improved HD VMT forecasting method. County level VMT growth rates
extracted from the California Statewide Travel Demand Model (CSTDM) 86 is used to
forecast future VMT from 2020 to 2050. CSTDM utilizes a commodity-based model which
forecasts future freight flows by mode on the transportation network under various policy
scenarios. The major benefits of using CSTDM freight projections as HD VMT surrogates
include:
• It helps create consistent forecasting methods across CARB’s freight inventory
sectors. Freight movement drives the growth in shipments by HD vehicles, ocean
going vessels (OGVs), locomotives, as well as activities in cargo handling
equipment (CHE) sectors. CARB’s OGV and CHE model primarily use Freight
Analysis Framework (FAF) projections, along with other freight forecasts. Using
freight flows as surrogates for forecasting HD VMT growth will ensure consistency
of CARB’s inventory projections in the main freight sectors.

84
refer to Section 4.5.3 in the EMFAC2017 Technical Documentation
85
https://www.cdtfa.ca.gov/taxes-and-fees/spftrpts.htm
86
https://dot.ca.gov/programs/transportation-planning/multi-modal-system-planning/statewide-modeling

176
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

• It allows VMT forecasting by truck class at regional level, in contrast to the

previous uniform statewide growth rate with the socio-economic regression
model. CSTDM forecasts freight activity by light, medium, and heavy duty trucks.
The CSTDM model projects freight flows among over 5000 zones in California.
• Finally, using CSTDM allows potential future explorations of HD VMT and
emissions under different policy scenarios, such as mode shifts, carbon tax,
alternative spatial development, change in infrastructure, etc.
The CSTDM county level VMT for the calendar years 2015, 2020, 2030 and 2040 were
used to predict the values between these years and 2030 – 2040 values were used to
predict VMT for 2041-2050. The CSTDM county level VMT growth rates were calculated
separately for the heavy heavy-duty vehicles (above33,000 lbs.) and the medium heavy-
duty vehicles (14,001-33,000 lbs.). Summary results are displayed in the following Figure
4.5.2-2 through Figure 4.5.2-4 for Statewide, South Coast, and San Joaquin Valley Air
Basins.
Figure 4.5.2-2. CSTDM HD VMT Growth Rates – Statewide

2.00
VMT Growth Rates (Relative to

1.80
1.60
1.40
CY2019)

1.20
1.00
0.80
0.60 Heavy VMT Growth Rates (CY2021/CY2019)

0.40 Medium VMT Growth Rates (CY2021/CY2019)

0.20
-
2010 2020 2030 2040 2050
Calendar Year

177
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure 4.5.2-3. CSTDM HD VMT Growth Rates-South Coast Air Basin

2.00
(Relative to CY2019)
VMT Growth Rates

1.80
1.60
1.40
1.20
1.00
0.80
Heavy VMT Growth Rates (CY2021/CY2019)
0.60
0.40 Medium VMT Growth Rates (CY2021/CY2019)
0.20
-
2010 2020 2030 2040 2050
Calendar Year

Figure 4.5.2-4. CSTDM HD VMT Growth Rates-San Joaquin Air Basin

2.50
(Relative to CY2019)
VMT Growth Rates

2.00

1.50

1.00

Heavy VMT Growth Rates (CY2021/CY2019)

0.50
Medium VMT Growth Rates (CY2021/CY2019)
-
2010 2020 2030 2040 2050
Calendar Year

178
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Drayage Trucks
As in EMFAC2014 and EMFAC2017, EMFAC2021 keeps using the activity growth rates
from CARB’s Ocean-Going Vessel (OGV) model as a surrogate for future drayage truck
VMT growth for the Port of Los Angeles (POLA) and the Port of Long Beach (POLB).
Mercator International Forecast from 2016 87 is used to project drayage truck VMT growth
operating at the POLA and POLB. It should be noted that POLA and POLB will reach
their combined capacity limit by 2035. At that point, the estimated growth rate is zero.
CARB staff applied this capacity limit. As a result, post-2035 growth rate is assumed to
be zero for POLA and POLB. For the Port of Oakland, the growth rates used for
EMFAC2021 are based on the moderate growth rate scenario described in the 2019-
2050 Bay Area Seaport Forecast report, which is prepared for the SF Bay Conservation
and Development Commission by the Tioga Group. 88 For the “Other Ports” drayage
truck category, EMFAC2021 assumes that VMT grows similar to the Port of Los
Angeles/Long Beach and Port of Oakland growth rates adjusted by Twenty-foot
Equivalent Unit (TEU) proportions. Figure 4.5.2-5 shows the VMT growth rates used in
EMFAC2021 for drayage trucks.

87
https://ww3.arb.ca.gov/msei/offroad/pubs/2019_OGV_Inv_Methodology.pdf
88
The Tioga Group, 2019-2050 Bay Area Seaport Forecast, Prepared for SF Bay Conservation and
Development Commission (https://www.bcdc.ca.gov/seaport/2019-2050-Bay-Area-Seaport-Forecast-
Draft.pdf, April 2020)

179
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure 4.5.2-5. VMT Growth Trend for Drayage Trucks

2.00
1.80
VMT Growth Rates (Relative to

1.60
1.40
1.20
CY2019)

1.00
0.80 EMFAC2021 VMT Growth Rate-POLA
0.60 EMFAC2021 VMT Growth Rate-POAK
0.40
EMFAC2021 VMT Growth Rate-Other Ports
0.20
-
2010 2020 2030 2040 2050
Calendar Year

Construction and Motor Coach Buses

The VMT growth rates for Construction and Motor Coach Buses in EMFAC2021 (Figure
4.5.2-6) are based on 10 years of projected employment from the June 2018 UCLA
Anderson Annual Economic Forecast 89. The T7 Single Concrete/Transit Mix and the T7
Single Dump Fleets computed VMT growth rates using the construction employment
data. The Motor Coach Bus Fleet computed VMT using the leisure/hospitality
employment data.

89
https://www.anderson.ucla.edu/centers/ucla-anderson-forecast
180
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure 4.5.2-6. VMT Growth Trend for Construction Trucks and Motor Coaches

1.20
VMT Growth Rates (Relative to

1.00

0.80
CY2019)

0.60
Construction Related Fleets VMT Growth Rates
Motor Coach VMT Growth Rates
0.40

0.20

-
2015 2025
2035 2045
Calendar Year
NOTE: CY2028 Annual Growth Rates were held constant from CY2029 through CY2050

Public, Utility and Solid Waste Collection Vehicles (SWCV), All Other Buses and School
Buses
Similar to EMFAC2017, the VMT of Public, Utility and Solid Waste Collection Vehicles
(SWCV) and the All Other Bus fleet in EMFAC2021 is assumed to follow the same activity
growth trend as the DOF based statewide human population 90 displays. The School Bus
growth rates to forecast VMT are set at 1.0 (as it was in EMFAC2017), reflecting no
growth.

4.5.3 Zero Emission Vehicle Forecasting

LDV ZEV sales
EMFAC2017 predicted the sales of new light-duty ZEV vehicles based on a most likely
compliance scenario that was dictated by ZEV mandates outlined in the Appendix A of
Advanced Clean Car Regulation (ACC) Mid-term Review 91. The comparison between ZEV
sales projected by EMFAC2017 and DMV data indicates that EMFAC2017
underestimated the new vehicle sales of light-duty ZEV vehicles in California in recent
years (shown in Figure 4.5.3-3). This shows the important role of consumer preference

90
http://www.dof.ca.gov/Forecasting/Demographics/Projections/
91
https://ww2.arb.ca.gov/sites/default/files/2020-
01/appendix_a_minimum_zev_regulation_compliance_scenarios_formatted_ac.pdf
181
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

that promotes ZEV sales beyond the existing ZEV mandate. Therefore, EMFAC2021
updates its methodology for projecting ZEV market shares using California Energy
Commission (CEC)’s light-duty vehicle choice models, coupled with CARB’s customized
input variable projections for the state incentives and vehicle attributes.
For short-term projection, California Energy Commission (CEC)’s consumer vehicle
choice models along with CARB’s updated input assumptions for projecting ZEV market
share from 2020 to 2030 are used. The models used for short-term projection are the
personal vehicle choice (PVC) model for vehicles owned by residents and the commercial
vehicle choice (CVC) model for vehicles owned by businesses in California. In addition,
light-duty vehicles and zero-emission vehicle sales in the rental and governmental sectors
follow CEC’s Integrated Energy Policy Report (IEPR) 2020 92 projections, which are made
based on economic growth and guidelines established by the current requirements for
California state government’s vehicle fleet. For long-term projections, ZEV market share
is assumed to plateau in California starting 2030.
CEC has developed and has been updating the PVC and CVC models since 1983. These
models are important components for policymaking in California, such as predicting
demand for alternative fuel vehicles, forecasting future transportation energy
consumption, and performing analysis under various scenarios. California specific data
from the California Vehicle Survey 93 provide the foundation of the PVC and CVC models
and are used to derive model coefficients. The survey represents households and
businesses' geographic distribution across California and collects thousands of responses
from residents and businesses, including hundreds of plug-in electric vehicle owners.
Based on CEC's 2017 Staff report, 94 PVC is a dynamic model that forecasts the mix of
light-duty vehicles. The model starts forecasting using base year data on the mix of
vehicles by fuel type, age, and vehicle class. Forecasts of vehicle mix are conducted year
by year, with the vehicle mix in the next year based on the current year's vehicle mix. For
each projected year, the models need the following inputs:
1. Vehicle attributes including price, fuel economy, range, the number of available
makes and models
2. Demographic and economic information such as household population and
income
3. Incentives from both the federal government and the California government (e.g.,
HOV lane access, federal tax credit)
4. Other variables such as fuel price

92
CEC IEPR website: https://www.energy.ca.gov/data-reports/reports/integrated-energy-policy-report
93
California Travel Survey: https://www.energy.ca.gov/data-reports/surveys/california-vehicle-survey
94
CEC Staff report (2017):
https://efiling.energy.ca.gov/GetDocument.aspx?tn=223241&DocumentContentId=28845

182
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

With these input projections, the PVC model predicts the changes in the number of
households with zero, one, two, and three or more vehicles. These changes determine
the number of households adding or dropping vehicles. Furthermore, the model
estimates the probability of a household replacing previously owned vehicles. The model
then calculates households' probability of purchasing and dropping vehicles of particular
fuel types, classes, and model years. To do so, the model considers the competitiveness
of the vehicle attributes and the degree to which certain vehicles appear attractive to the
consumers. Finally, the model projects statewide vehicle stock and new sales for different
fuel types, model years, and vehicle classes. Similarly, the CVC model can be used to
estimate the new sales of business-owned vehicles. CEC calibrated their vehicle choice
models to 2019 DMV data.
CARB conducted a sensitivity analysis of the PVC model to determine the input variables
that are most important to output results. These variables include vehicle price, fuel
economy, range, and CVRP rebate for BEV and PHEV. Staff further updated input data
used for these variables using the latest available information.
Projected battery electric vehicles' (BEV) and plug-in hybrid electric vehicles (PHEV) 's
attributes including fuel economy, price, and range are projected as described in the
following steps:
• Step 1 base-year (2019) vehicle attributes are calculated based on the best
available data, listed below.
1. EPA’s fueleconomy.gov 95 providing fuel economy, tank size, and range
2. WARDS INTELLIGENCE 96 providing vehicle price
3. IHS/POLK 97 providing the sales of different makes and models
Using the above data sources, sales-weighted average of vehicle attributes across all
models for each vehicle class is calculated.
• Step 2 Assumptions are developed as to how these attributes will change in the
future using projections obtained from a white paper published by ICCT 98 and by
consulting with CARB’s regulatory teams. Details are as follows.

95
EPA fuel economy data download page: https://www.fueleconomy.gov/feg/download.shtml
96
WARDS INTELLIGENCE website: https://wardsintelligence.informa.com/datacenter
97
https://ihsmarkit.com/btp/polk.html
98
The International Council on Clean Transportation (ICCT) (2019). Update on electric vehicle costs in the
United States through 2030. Available at:
https://theicct.org/sites/default/files/publications/EV_cost_2020_2030_20190401.pdf

183
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Fuel economy. The fuel economy of BEV is assumed to grow by 0.7% every year,
according to the ICCT report. For PHEV, the fuel economy of electric motors are
assumed to grow 0.7% every year; the fuel economy of gas engines follows the 2020
SAFE Vehicle rules, hence, increasing 1.5% per year from 2021 to 2026 and flattening in
2026. Then the composite fuel economy of PHEV is calculated using the weighted
average of fuel economy of electric motor and gas engine, based on utility factor, which
is a function of e-range defined in SAE paper. 99
Range. The range of BEV is assumed to reach 300 miles, and the e-range of PHEV is
assumed to reach 60 miles in 2030, according to the ICCT report. The ranges for years
between 2019 and 2030 are interpolated. The gasoline range of PHEV is calculated
based on the projection of the gasoline engines' fuel economy, assuming gasoline tank
size will stay the same over the years.
Vehicle price. Vehicle price is calculated as the sum of the base-year non-battery price,
the change in non-battery cost, and the future battery cost. Battery cost is calculated by
multiplying battery cost per kWh and battery size (kWh). The battery size is estimated
based on the range and the fuel economy. The battery cost per kWh is projected to
decrease 7% per year, according to the ICCT White Paper. Non-battery cost is assumed
to change following the trend of indirect costs (i.e., depreciation, amortization, research
and development, administration, and warranty) projected by ICCT: a downward trend
for BEV and a slight upward trend for PHEV.
For vehicle classes that do not have available PHEV and BEV models in the base year and
fuel cell electric vehicle (FCEV), CEC's IEPR2020 projections are used for their vehicle
attributes. 100
CVRP Rebate. Staff also updated the projection of California’s Clean Vehicle Rebate
Program (CVRP) rebate and HOV lane policy from CEC’s IEPR2019 projection. Staff’s
projections were adopted by CEC when IEPR2019 was updated to IEPR2020 projections.
California’s CVRP program promotes clean vehicle adoption by giving rebates for
purchasing new ZEV. The amount of CVRP rebate vary by income levels: high-income
individuals are not eligible for the rebate of BEV and PHEV, while low-income individuals
receive additional rebate. To project CVRP rebate averaged for all Californians of
different income levels, the following data sources are used: CVRP dashboard 101 provides
the number of applications and the amount of rebate given to each application. BEV and

99
SAE International (2009). Utility Factor Definitions for Plug-In Hybrid Electric Vehicles Using 2001 U.S.
DOT National Household Travel Survey Data J2841_201009. Available at:
https://www.sae.org/standards/content/j2841_201009/
100
CEC IEPR website: https://www.energy.ca.gov/data-reports/reports/integrated-energy-policy-report
101
CVRP dashboard: https://cleanvehiclerebate.org/eng/rebate-statistics

184
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

PHEV new sales are obtained from the California New Car Dealers Association
(CNCDA)’s website 102.
Using these two data sources, the percentage of BEV and PHEV purchasers who did not
apply for the CVRP rebate is estimated as the total number of applications divided by
total ZEV sales in 2019. All FCEVs are assumed to be funded by CVRP because there is
no restriction on income. It is assumed that the individuals who did not apply for CVRP
rebates (due to high income or other reasons) get zero CVRP rebates. Among the other
individuals, the percentage of CVRP rebates given to low-income individuals is calculated
based on the CVRP dashboard database. These individuals receive higher CVRP rebate.
Lastly, the weighted average CVRP rebate for all individuals in California are calculated.
The resulting CVRP rebate projection for BEV, PHEV, and FCEV are $1023, $481, and
$4823 from 2020 to 2025, respectively; CVRP rebate is assumed to end in 2025,
following suggestions from CARB’s CVRP team.
HOV lane policy Following the California Vehicle Code (CVC) §§5205.5 and 21655.9 103,
HOV lane access to ZEVs is projected to end in 2025.
CEC models output sales by CEC vehicle classes, so the data are post-processed to the
input information that EMFAC can utilize. Accordingly, CEC’s 18 vehicle classes (e.g.,
Car-Compact and Cross/Ut-midsize) are mapped to EMFAC vehicle classes (i.e., LDA,
LDT1, LDT2, and MDV), utilizing 2019 DMV registration data as well as EMFAC and
CEC’s classification guides. Appendix 6.8 summarizes the results of the mapping.
CEC models output sales for all fuel types, including BEV, PHEV, and FCEV. The sum of
BEV and FCEV in total LDV new sales is inputted as the market share of “electric” fuel
type to EMFAC. The fraction of PHEV in new sales of CEC output is inputted as the
market share of “PHEV” fuel type to EMFAC. EMFAC2021 uses the ZEV market shares
for different EMFAC light-duty vehicle classes and new vehicle sales forecasting to
calculate ZEV new sales.
The market shares of different EMFAC light-duty vehicle classes are summarized in Table
4.5.3-1 for electric vehicles and Table 4.5.3-2 for PHEV. Figure 4.5.3-1 and Figure 4.5.3-2
show the electric and PHEV market shares in total LDV sales (LDA, LDT1, LDT2, and
MDV) in EMFAC2021, respectively. EMFAC2021 projects that BEV and PHEV market
share in new sales to reach 9% and 3%, respectively. The comparison of light-duty ZEV
market share of EMFAC2021 and EMFAC2017 are shown in Figure 4.5.3-3. EMFAC2021
projects ZEV market share in LDV new sales in 2030 to be 12%, which is significantly
higher than that in EMFAC2017 (~6%). Both EMFAC2017 and EMFAC2021 project
higher ZEV market share in 2030 than the base year. Note that the decrease in ZEV

102
CNCDA California Auto Outlook: https://www.cncda.org/news/auto-outlook-2019-q4/
103
https://codes.findlaw.com/ca/vehicle-code/veh-sect-21655-9.html
185
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

market share from 2025 to 2026 in the EMFAC2021 projection can be attributed to the
projected end of the HOV lane policy and CVRP rebate in 2025.
Figure 4.5.3-1. The Market Share of Electric Vehicles (BEV+FCEV) in New Sales

10%

EMFAC2021 DMV
8%
ZEV Sales Percentage

6%

4%

2%

0%

Model Year

Figure 4.5.3-2. The Market Share of PHEV in New Sales

4%
EMFAC2021
DMV
ZEV Sales Percentage

3%

2%

1%

0%

Model Year

186
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure 4.5.3-3. The Market Share of ZEV in New Sales

14%
EMFAC2021
12% EMFAC2017
10% DMV
ZEV percentages

8%

6%

4%

2%

0%

Model year

187
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Table 4.5.3-1. The Market Share of Electric Vehicles in New LDV Sales
MY LDA LDT1 LDT2 MDV
2020 9.4% 0.1% 0.2% 1.0%
2021 10.2% 0.2% 1.2% 1.9%
2022 10.9% 0.6% 2.0% 3.5%
2023 11.4% 0.9% 2.3% 4.6%
2024 11.9% 1.2% 2.6% 4.9%
2025 12.3% 1.5% 2.9% 5.3%
2026 10.8% 1.8% 2.8% 5.0%
2027 11.0% 2.1% 3.1% 5.3%
2028 11.3% 2.5% 3.4% 5.6%
2029 11.6% 2.9% 3.7% 5.9%
2030+ 11.9% 3.3% 4.0% 6.2%

Table 4.5.3-2. The Market Share of PHEV in New LDV Sales

MY LDA LDT1 LDT2 MDV
2020 4.2% 0.2% 1.3% 0.9%
2021 4.5% 0.4% 2.0% 1.0%
2022 4.7% 0.7% 2.1% 1.7%
2023 4.7% 0.9% 2.3% 1.8%
2024 4.6% 1.2% 2.2% 2.5%
2025 4.6% 1.4% 2.3% 2.7%
2026 4.1% 1.5% 2.2% 2.5%
2027 4.1% 1.8% 2.3% 2.6%
2028 4.1% 2.0% 2.4% 2.7%
2029 4.0% 2.2% 2.6% 2.8%
2030+ 4.0% 2.5% 2.7% 2.9%

HDV ZEV sales

For the first time EMFAC2021 now includes heavy-duty zero-emission truck and bus
population projections. The phase-in of heavy-duty ZEVs is driven by CARB’s recently
adopted regulations, including Innovative Clean Transit (ICT), Zero-Emission Airport
Shuttle Bus, and Advanced Clean Truck (ACT). Zero-emission transit bus forecasting
based on ICT is provided in Section 4.2.3. Zero-emission airport shuttle bus was not
explicitly accounted for in EMFAC2021 since this category represents a very small
fraction of the fleet.

188
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

The ACT regulation approved on June 25, 2020 104 is a measure to improve air quality and
to mitigate climate change by transforming the California heavy-duty vehicle fleet to
zero-emissions technologies. The ACT regulation will achieve its electrification goal by
gradually increasing the fraction of zero-emission vehicles (ZEVs) sold in California
starting with model year (MY) 2024 vehicles. The rule applies different sales requirement
fractions to Class 2b-3 (8,501- 14,000 lbs. GVWR), Class 4-8 (>14,000 lbs. GVWR)
Vocational and Class 7-8 (>26,000 lbs. GVWR) Tractors. The Class 2b-3 vehicle inventory
is comprised of two EMFAC vehicle categories: LHDT1 (light heavy-duty trucks with
GVWR 8,501-10,000 lbs.) and LHDT2 (light heavy-duty trucks with GVWR 10.001-14,000
lbs.). The class 4-8 vocational vehicles included instate buses, instate non-tractor class 7
and 8 heavy trucks (>26,000 lbs.), and instate class 4 – 6 trucks (14,001-26,000 lbs.). The
bus inventory was adjusted to account for light duty vehicles that are included in this
EMFAC vehicle category and to exclude transit and shuttle buses, which already have to
meet the ICT 105 or Zero-Emission Airport Shuttle Bus 106 regulations. Instate vehicles
include vehicles that are registered by International Registration Plan (IRP) as well as
California Department of Motor Vehicles (DMV). The detailed assumptions and methods
of emission benefit analysis for the ACT regulation shown in the Appendix F of its
ISOR. 107,108
The ACT requirements only apply to vehicles that are originally sold in California
(hereinafter, we refer to them as First Sold in CA). To estimate the portion of new vehicle
sales specific to California, staff reviewed the First Sold data field values in the California
DMV vehicle registration data from 2014 through 2017. The First Sold data field
identifies the year for vehicles that were first sold in California. The International
Registration Program (IRP) vehicles are not required to have this field populated in the
DMV data sets and they were excluded from this analysis. The average percentages of
vehicles first sold in California from DMV data from 2014-2017 were used to develop
trends to estimate future new sales by vehicle type. The estimated first sold percentages
(as shown in Table 4.5.3-3) were then applied to the total new sales forecasted by
EMFAC2021. For instate buses and vehicles below 14,001 lbs., it was assumed that the
rate of first sold in California is 100%.

104
https://ww2.arb.ca.gov/our-work/programs/advanced-clean-trucks
105
California Air Resources Board, Innovative Clean Transit (web link: https://ww2.arb.ca.gov/our-
work/programs/innovative-clean-transit, Last accessed June, 2019)
106
California Air Resources Board, Zero-Emission Airport Shuttle (web link: https://ww2.arb.ca.gov/our-
work/programs/zero-emission-airport-shuttle, Last accessed June, 2019)
107
California Air Resources Board, Appendix F: Emissions Inventory Methods and Results for the Proposed
Advanced Clean Trucks (web link: Regulation https://ww3.arb.ca.gov/regact/2019/act2019/appf.pdf,
posted October, 2019)
108
California Air Resources Board, Attachment D Emissions Inventory Methods and Results for the
Proposed Advanced Clean Trucks Regulation Proposed Modifications (web link:
https://ww3.arb.ca.gov/regact/2019/act2019/30dayattd.pdf, May, 2020)

189
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Table 4.5.3-3. Average Percentages for First Sold in California by Vehicle Type
Class 8 Class 8 Class 7 Class 7 Class 4-6
Age
Tractor Vocational Tractor Vocational Vocational
-1 or 0 89.00% 89.78% 84.31% 85.01% 90.97%
1 86.61% 85.80% 82.10% 80.35% 88.38%
2 79.17% 81.86% 76.91% 76.22% 85.68%
3 68.61% 78.34% 69.92% 72.74% 83.07%
4 56.87% 75.59% 62.30% 70.02% 80.74%
5 45.87% 74.00% 55.25% 68.18% 78.90%
6 37.55% 73.92% 49.92% 67.35% 77.76%
7 33.85% 73.92% 47.51% 67.35% 77.50%
8 33.85% 73.92% 47.51% 67.35% 77.50%
9+ 33.85% 73.92% 47.51% 67.35% 77.50%

In EMFAC2021, different heavy-duty vehicle categories are mapped to tractor and

vocational groups, as shown in Table 4.5.3-4. As already noted, the ACT regulation
requires an increasing percentage of new vehicle sales in California to be ZEVs beginning
with model year 2024.
Table 4.5.3-5 provides a summary of the ZEV sales requirements for new vehicle sales in
California.
Table 4.5.3-4. Heavy Duty Vehicle Classes Mapped to ACT First Sold Classes
First Sold Class EMFAC2021 HD Category
Class 2b-3 LHD1; LHD2
T6TS; T6 Public Class 4; T6 Public Class 5; T6 Public Class 6; T6 Utility
Class 5; T6 Utility Class 6; T6 Utility Class 7; T6 Instate Tractor Class
6;T6 Instate Delivery Class 4; T6 Instate Delivery Class 5; T6 Instate
Class 4 -6 Vocational
Delivery Class 6; T6 Instate Other Class 4; T6 Instate Other Class 5; T6
Instate Other Class 6; T6 CAIRP Class 4; T6 CAIRP Class 5; T6 CAIRP
Class 6
Class 7 Vocational T6 Public Class 7; T6 Instate Delivery Class 7; T6 Instate Other Class 7
Class 7 Tractor T6 Instate Tractor Class 7; T6 CAIRP Class 7
T7 Public Class 8; PTO; T7 Single Concrete/Transit Mix Class 8; T7
Class 8 Vocational
Single Dump Class 8; T7 Single Other Class 8; T7 SWCV Class 8
T7 CAIRP Class 8; T7 Other Port Class 8; T7 POAK Class 8; T7 POLA
Class 8 Tractor
Class 8; T7 Tractor Class 8
SBUS SBUS
OBUS OBUS

190
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Table 4.5.3-5. ZEV Sales Requirements Under the ACT Regulation

Model Year Class 2b-3 Class 4-8* Vocational Class 7-8 Tractors
2024 5% 9% 5%
2025 7% 11% 7%
2026 10% 13% 10%
2027 15% 20% 15%
2028 20% 30% 20%
2029 25% 40% 25%
2030 30% 50% 30%
2031 35% 55% 35%
2032 40% 60% 40%
2033 45% 65% 40%
2034 50% 70% 40%
2035+ 55% 75% 40%

*Excludes Class 7-8 Tractors

4.6 Regulations and Policies

4.6.1 SAFE Rule
In September 2019, the U.S. Environmental Protection Agency (U.S. EPA) and the
National Highway Traffic Safety Administration (NHTSA) issued the Safer Affordable
Fuel-Efficient or SAFE Vehicles Rule Part One: One National Program (SAFE Part One)
that revoked California’s authority to set its own greenhouse gas emissions standards
and zero-emission vehicle (ZEV) mandates in California, 84 Fed. Reg. 51,310 (Sept. 27,
2019). In April 2020, the federal agencies issued the SAFE Vehicles Rule for Model Years
2021-2026 Passenger Cars and Light Trucks (Final SAFE Rule) that relaxed federal
greenhouse gas emissions and fuel economy standards, 85 Fed. Reg. 24,174 (Apr. 30,
2020). The Final SAFE Rule then relaxed federal greenhouse gas emissions and
Corporate Average Fuel Economy (CAFE) standards to increase in stringency at only
about 1.5% per year from model year (MY) 2020 levels over MYs 2021–2026. In 2020,
CARB finalized with six participating automakers to reach a Framework Agreements on
Clean Cars to meet a more stringent CO2 reduction targets than the SAFE Rule.
Automakers who voluntarily agreed to the framework agreements are BMW of North
America (including Rolls Royce for purposes of the agreement), Ford, Honda,
Volkswagen Group of America (including VW and Audi), and Volvo. The six manufacturer
Framework Agreements on Clean Cars will collectively represent a 3.7% YoY reduction in
CO2 emissions from MY 2022 to 2026. In addition, these manufacturers are allowed to
meet up to 1.0% of those reductions using ZEV credits. To comply with this rule, the
assumption in EMFAC2021 for CO2 emission factor were updated for gasoline
passenger cars and light trucks. This is further described in section 4.3.3.

191
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

4.6.2 Innovative Clean Transit

The Innovative Clean Transit (ICT) regulation was adopted by CARB in 2019 and targets
reductions in transit fleets by requiring transit agencies to gradually transition their buses
to zero-emission technologies. ICT has helped to advance heavy-duty ZEV deployment,
with buses acting as a beachhead in the heavy-duty sector. Based on the size of the
transit agencies, they are categorized as small and large agencies. Starting calendar year
2023, large agencies follow the phase-in schedule to have a certain percentage of their
new purchases as ZEB. For the small agencies, the start calendar year will be 2025. By
2030, all the agencies need to have 100% of their new purchases as ZEB 109. More details
can be found in section 4.2.3.

4.6.3 Zero-Emission Airport Shuttle Bus

CARB adopted the zero-emission airport shuttle bus regulation in 2019 and required
airport shuttle fleets to fully transition to zero emission by 2035. This regulation is not
explicitly accounted for in EMFAC2021 since this category represents a very small
fraction of California fleet (<2,000 vehicles).

4.6.4 HD Omnibus
CARB adopted the Heavy-Duty Omnibus regulation in August 2020, which applies to
engine-certified vehicles with GVWR > 10,000 lbs. that are first sold or certified in
California (see Section 4.5.3). Note that the majority (> 95%) of vehicles with GVWR
10,001-14,000 lbs. are chassis-certified and thus are excluded from HD Omnibus
standards in EMFAC 2021. This program represents a comprehensive update to heavy-
duty NOx emissions standards and ensures that heavy-duty engines will emit much lower
NOx emissions throughout their lifetimes. This regulation includes
• A tightened standard on the Federal Test Procedure (FTP),
• A new low-load certification cycle (LLC),
• Improvements to the existing heavy-duty in-use testing (HDIUT) program,
• Improvements to the durability demonstration program (DDP),
• Lengthened warranty and useful life (UL) mileages, and
• Amendments to the emission warranty information reporting (EWIR) program and
corrective action procedures.
More details on each program element listed above, as well as emissions benefits
methodology can be found in Appendix D of the HD Omnibus ISOR. 110 The full
requirements of the HD Omnibus program are shown in Table 4.6.4-1.

109
https://ww3.arb.ca.gov/regact/2018/ict2018/isor.pdf
110
https://ww3.arb.ca.gov/regact/2020/hdomnibuslownox/appd.pdf
192
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Table 4.6.4-1. HD Omnibus Requirements by Engine Model Year Groups

Standards, Test
Engine Engine Engine
Procedures and Units
MY2024-26 MY2027-30 MY2031 & Newer
Elements
FTP g/bhp-hr 0.05 0.02 0.02
LLC g/bhp-hr 0.2 0.04 0.04
Idling g/hr 10 5 5
Binned MAW w/ Binned MAW w/
Binned MAW 1.5x
HDIUT Method g/bhp-hr Cold Start 1.5x Cold Start 1.5x
Standards
Standards Standards
DDP NA 100% UL aging 100% UL aging 100% UL aging
UL3 10^3×miles 435/185/110/110* 600/270/190/155* 800/350/270/200*
Warranty See Section 4.6.5 See Section 4.6.5 See Section 4.6.5 See Section 4.6.5
EWIR* -- Modified EWIR Modified EWIR Modified EWIR

Note: DDP= Durability Demonstration Program, EWIR= Emission Warranty Information

Reporting and corrective actions, FTP= Federal Test Procedure, HDIUT= Heavy-Duty In-
Use Testing, LLC= Low Load Cycle, MAW= Moving Average Window, MY=Model Year,
NTE= Not-to-Exceed, UL= Useful Life
Units: g/bhp-hr= Grams Per Brake Horsepower-Hour, g/hr= Grams Per Hour.
* Diesel Class 8; GVWR >33,000 lbs. / Diesel Class 6-7; 19,500 < GVWR ≤ 33,000 lbs.
/Diesel Class 4-5; 14,000 lbs. < GVWR ≤ 19,500 lbs. / HD Otto (gasoline).
Staff modelled reductions in the zero-mile emission rate (ZMR) as part of the HD
Omnibus emissions inventory analysis. These reductions reflect the tightened Federal
Test Procedure (FTP) standard, which is 75% (engine model year 2024-2026) to 90%
(engine model year 2027 and newer) lower than the 0.2 g/bhp-hr baseline standard. They
also reflect improvements to heavy-duty in-use testing program to ensure that real-world
emission rates are closer to the standard. Staff developed scaling factors for HD
Omnibus ZMRs that were derived from the rulemaking analysis, specifically 0.44 g/mile
for engine model year 2024-26 and 0.12 g/mile for 2027 and newer, divided by the new
baseline ZMR of 0.6 g/mile (derived in Section 4.3.6). The same scaling factors were used
at all speeds. Tightening the FTP standard results in emissions reductions across the full
range of speeds. In the rulemaking analysis, tightening the LLC to less than 2X the FTP
standard resulted in additional benefits relative to just an FTP standard. However, since
the LLC standards in Table 4.6.4-1.
are at least 2X the FTP, no additional benefits for LLC were modelled and speed
correction factors (SCFs) were assumed to be the same as 0.2 g/bhp-hr engines.
However, in reality, an LLC standard will ensure that FTP standard results in significant
reductions at lower speeds.
Idle emissions for HD vehicles are also adjusted in EMFAC2021 using the ratio of the new
standards and the baseline standard of 30 g/hr for engine model year groups 2024-26
193
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

and 2027 and newer HD deterioration rates (DR) are adjusted to account for HD
Omnibus warranty and durability requirements. More details can be found in Section 4.6.
Please note that the HD Omnibus regulation only applies to vehicles that are first sold or
certified in California. However, a portion of the California-certified pool of MY 2024 and
newer heavy-duty vehicles are assumed to be converted to ZEVs as required by
Advanced Clean Trucks (see Section 4.6.2.5). Therefore, the first sold fractions were
adjusted, such that HD Omnibus only applied to combustion vehicles remaining after
ACT.

4.6.5 Warranty
The Heavy-Duty Engine Warranty Amendments were adopted in 2018. This regulation is
designed to reduce emissions of nitrogen oxides (NOx) and particulate matter (PM) from
heavy-duty vehicles by requiring manufacturers to lengthen the warranty mileages of
heavy-duty vehicles that are first sold or certified in California. In addition, the Heavy-
Duty Omnibus regulation (see Section 4.6.2.5) further lengthened warranty mileages in
two steps: for engine model years 2027-2030 and lengthened further for 2031 and
newer.
Lengthened warranty requirements in 2018 HD Warranty Amendments and 2020 HD
Omnibus are expected to decrease the PM and NOx deterioration rates of heavy-duty
vehicles by decreasing the frequency of malfunction of after-treatment system and
engine components. To model the impact of lengthened warranty on deterioration rates,
the emission impact rate (EIR) at the lengthened warranty mileage was assumed to be
the same as the previous warranty mileage. Figure 4.6.5-1 shows an example of NOx
emission rates with a lengthened warranty in black and the baseline warranty in blue. The
red squares mark the original warranty mileage for the baseline case and the new
warranty mileage for the lengthened warranty case. The emission rate for the lengthened
warranty case at 400,000 miles is the same as that for the baseline case at 200,000 miles
because the assumed EIRs are equivalent. For example, to adjust the NOx deterioration
rate of OBD-equipped, the following equation was used
𝐵𝐵𝐵𝐵𝐵𝐵𝐵𝐵𝐵𝐵𝐵𝐵𝐵𝐵𝐵𝐵 𝑊𝑊𝑊𝑊𝑊𝑊𝑊𝑊𝑊𝑊𝑊𝑊𝑊𝑊𝑊𝑊 𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑎𝑎𝑔𝑔𝑔𝑔 𝑏𝑏
𝐸𝐸𝐸𝐸𝐸𝐸@ 𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜 𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑚 × 𝐿𝐿𝐿𝐿𝐿𝐿𝐿𝐿𝐿𝐿ℎ𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 𝑊𝑊𝑊𝑊𝑊𝑊𝑊𝑊𝑊𝑊𝑊𝑊𝑊𝑊𝑊𝑊 𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀 𝑏𝑏 (Eq. 4.6.5-1)

where ZMR is the zero mile rate, EIR is the emission impact rate at a given odometer
mileage, and b is the fitted power function coefficient (see Section 4.3.6). Warranty
mileages used in EMFAC2021 are shown in Table 4.6.5-1.

194
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure 4.6.5-1. The Impact of Lengthened Warranty on HD NOx Emission Rates

4.5

4 Baseline Lengthened Warranty

3.5

3
NOx (g/mile)

2.5

1.5

0.5

0
0 200,000 400,000 600,000 800,000 1,000,000
Odometer Mileage

Table 4.6.5-1. Baseline and Warranty Mileages Adjusted for Year Limits
Regulation/Engine Model Year LHDD* MHDD* HHDD*
Baseline 100,000 100,000 100,000
June 2018 Step 1 Warranty 2022-2026 103,000 139,000 289,000
HD Omnibus 2027-2030 135,000 172,000 308,000
HD Omnibus 2031 and Newer 189,000 221,000 400,000

* HHDD: Diesel Class 8; GVWR >33,000 lbs. / MHDD: Diesel Class 6-7; 19,500 < GVWR ≤ 33,000 lbs.
/LHDD: Diesel Class 4-5.

Although the current emission warranty requirements for classes 4 to 8 trucks is 100,000
miles, a recent CARB sponsored survey indicates that 40% of these trucks carry extended
warranties to their respective useful lives (110,000 miles for classes 4 & 5, 185,000 miles
for classes 5 & 6, and 435,000 miles for class 8). 111 In addition, according to data and
information from manufacturers as well as other sources, ~75% of the remaining 60% of
class 8 trucks have an extended warranty of 250,000 miles. As such, the baseline
mileages were adjusted for vehicle classes with warranties larger than the baseline. The
warranty mileages shown in Table 4.6.5-1 represent average miles covered when
considering the miles, years, and hours provisions within the proposed requirements. For
example, using the 2022 Step 1 warranty requirements for HHD, manufacturers must
cover emissions warranty to 350,000 miles or 5 years, whichever comes first. On average,

111
https://ww3.arb.ca.gov/regact/2018/hdwarranty18/appf.pdf
195
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

HHD vehicles reach 289,000 miles before they reach 5 years of age, so the table shows
“289,000 miles” for HHD.

4.6.6 Advanced Clean Trucks

CARB’s Advanced Clean Trucks (ACT) regulation was adopted by the Board in June
2020. EMFAC2021 reflects ACT by modelling heavy-duty zero emission vehicles (ZEVs)
based on the sales percentage requirements for each model year and those percentages
were applied to vehicles first sold or certified in California. More details can be found in
Section4.5.3. ZEVs produced by truck manufacturers to meet the ACT regulation can also
be used to meet the Phase 2 GHG regulation. For heavy-duty vehicles first sold or
certified in California, the Phase 2 reduction factors had to be adjusted to account for
overlap with ACT. More details can be found in Section 4.3.5.5

4.6.7 Opacity
The Amendments to the Heavy-Duty Vehicle Inspection Program (HDVIP) and Periodic
Smoke Inspection Program (PSIP) are designed to reduce emissions of particulate matter
(PM) from diesel powered heavy-duty vehicles (HDV) powered by diesel engines.
Emission reductions would primarily result from heavy-duty diesel (HDD) trucks, which
include heavy heavy-duty diesel trucks (above 33,000 lbs. GVWR) and medium heavy-
duty diesel trucks (14,001-33,000 lbs. GVWR).
The amendments to the HDVIP and PSIP regulation require diesel powered trucks and
buses that are found to exceed i) the 20-40% opacity limit for non-DPF vehicles to repair
their engines and ii) 5% opacity limit for DPF-equipped engines to either repair or
replace their DPFs in order to reduce the opacity below the respective limits. The PM
emissions benefits are the result of such repairs or replacement. EMFAC2021 assumed
that repaired engines have equivalent PM emissions to a new vehicle, i.e., the emission
rate would equal the zero-mile rate, ZMR.
These assumptions were incorporated into Appendix C of the Proposed Amendments of
the Heavy-Duty Vehicle Inspection Program and Periodic Smoke Inspection Program
Staff Report. 112 Briefly, repair effectiveness values were calculated for calendar years
2019 and 2025. This repair effectiveness was translated to reductions in TM&M
frequency, i.e., the frequency of malfunction, which reduces the deterioration rate, DR.
To implement these changes into the model, scalars were developed for calendar years
2019 through 2025 for different model year groups (e.g., engine model years 2007-
2009). These scalars are shown in Table 4.6.7-1. Below is an example of how PM
deterioration rates are applied in EMFAC 2021.

112
https://ww2.arb.ca.gov/rulemaking/2018/heavy-duty-vehicle-inspection-program-and-periodic-smoke-
inspection-program
196
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Odometer
PM Emission Rate = ZMR + PM Deterioration Scalar × DR × 10,000
(Eq. 4.6.7-1)

Table 4.6.7-1 PM Deterioration Rate Scalars developed for EMFAC2021

PM Deterioration Scalar by Calendar Year
Engine Model Year Range
2019 2020 2021 2022 2023 2024 2025
Pre1993 1 1 1 1 1 1 1
1993-02 0.92 0.88 0.87 0.87 0.86 0.86 0.85
2003-06 0.88 0.8 0.79 0.78 0.78 0.77 0.76
2007-09 0.82 0.71 0.7 0.69 0.67 0.66 0.65
2010-12 0.86 0.78 0.76 0.75 0.74 0.73 0.72
2013+ 0.85 0.77 0.76 0.75 0.73 0.72 0.71

197
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

5 Overall Impacts
As described earlier in this document, EMFAC2021 retains some of the EMFAC2017
updates but also has some unique additions and new features. Some of the noteworthy
updates to the EMFAC2021 include:
• Extended fuel technologies to PHEV and natural gas vehicles
• Estimation of energy consumption from BEV and PHEV vehicles as well as heavy-
duty vehicles
• Development of an ammonia module for running exhaust emissions based on
various historical and new studies
• Expansion of heavy-duty truck categories to include vocational categories for
more emission characterization specificity
• Adaption of new light-duty ZEV market share and heavy-duty VMT forecasting
frameworks
To examine the impact of these updates, this section presents charts of vehicle
populations, VMT, emissions, fuel and energy consumptions. A comparison is made
between emission and activity estimates from EMFAC2017 and those estimated using
EMFAC2021, at the statewide level. In order to better explain the differences, separate
comparisons are made for LD (GVWR below 8,500 lbs. and including motorcycles) and
HD (GVWR above 8,500 lbs.) vehicles. The EMFAC2021 results presented in this section
were generated using default VMT data. Please note that CARB’s SIP inventory is based
on VMT and speed profiles provided by Metropolitan Planning Organizations (MPOs),
which might be different from EMFAC default VMT.
Similar comparisons have been performed for the South Coast and San Joaquin Valley
sub areas; and, the explanations provided for the statewide results also apply to these
regions. The charts for these regional comparisons are not presented in this section, but
are provided in Appendix 6.3.
This section compares the statewide results of EMFAC2021 with EMFAC2017 for vehicle
populations (in millions), VMT (in million miles per day), emissions (in tons per day), fuel
consumption (million gallons), and energy consumption (giga watts-hour). Differences in
the results between the two model versions are discussed below.

5.1 Vehicle Population

The panels of figure below compare EMFAC2021 and EMFAC2017 total HD and LD
vehicle populations. The EMFAC2021 total vehicle population, which is dominated by LD
vehicles, is about 10 million lower than in EMFAC2017 by the year of 2050. EMFAC2021
accounts for the COVID-19 related restrictions and incorporated a new LD new vehicle
sales forecasting for future years. The statewide HD populations are slightly higher than
EMFAC 2017 since staff accounted for light-heavy categories, i.e., LHD1 and LHD2 with
registration addresses outside of California in EMFAC2021, as shown in Figure 5.1-1.
198
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure 5.1-1. Comparison of Vehicle Population between EMFAC2021 and

EMFAC2017

50
45
Population (million vehicles)

40
35
30
25
20
15
10
EMFAC2021
5 EMFAC2017
0
2000 2010 2020 2030 2040 2050
Calendar Year

50 Population-Statewide-Light Duty 3.0 Population-Statewide-Heavy Duty

Population (million vehicles)

Vehicles (incl. MCY) Vehicles (above 8,500 lbs. GVWR)

Population (million vehicles)

2.5
40

2.0
30
1.5
20
1.0
EMFAC2021 EMFAC202x
10
0.5
EMFAC2017 EMFAC2017
0 0.0
2000 2010 2020 2030 2040 2050 2000 2010 2020 2030 2040 2050
Calendar Year Calendar Year

5.2 Vehicle Miles Traveled (VMT)

Figure 5.2-1 shows a comparison of statewide VMT from EMFAC2021 and EMFAC2017
in million miles per day. EMFAC2021 shows a lower VMT as compared to EMFAC2017
for all vehicles statewide. The lower LD VMT in historical years is because EMFAC2021
further subtracted off-road gas usage from the total fuel consumption. The COVID-19
pandemic also restricted LD VMT to some extent. For future years, EMFAC2021

199
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

reflected the adjusted human population forecast from DOF in its LD VMT estimates. HD
VMT in EMFAC2021 is higher than EMFAC 2017 in the long-term because of higher
assumed VMT growth rates from the Caltrans CSTDM model.
Figure 5.2-1. Comparison of VMT between EMFAC2021 and EMFAC2017

1,600 VMT-Statewide-All Vehicles

1,400
VMT (million miles per day)

1,200

1,000

800

600

400
EMFAC 2021
200 EMFAC2017
0
2000 2010 2020 2030 2040 2050
Calendar Year

VMT-Statewide-Light Duty 200

VMT-Statewide-Heavy Duty
1,500 Vehicles (above 8,500 lbs.
Vehicles (incl. MCY)
VMT (million miles per day)

GVWR)
VMT (million miles per day)

150
1,000

100

500
EMFAC 2021 50 EMFAC 2021
EMFAC2017 EMFAC2017

0 0
2000 2010 2020 2030 2040 2050 2000 2010 2020 2030 2040 2050
Calendar Year Calendar Year

5.3 Emissions
Oxides of Nitrogen (NOx)
The following figure shows the comparison of estimates of statewide NOx emissions
between EMFAC2021 and EMFAC2017, in tons per day. EMFAC2021 results shows
slightly higher NOx emissions in historical years but lower emissions in future years. This
is because of LD updated NOx emission rates for LEV 1, LEV 2, and LEV 3 vehicles, which

200
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

resulted higher emission rates in historical years. On the other hand, higher compliance
due to CARB’s enforcement actions and SB 1 DMV registration holds was modelled in
EMFAC2021 for HD, which resulted in lower NOx emissions between 2020 and 2023
compared to EMFAC2017. In the long-term, ACT, HD Warranty, and HD Omnibus
resulted in lower NOx emissions in EMFAC 2021 relative to EMFAC 2017, but these
reductions are somewhat offset by the higher VMT growth rate.
Figure 5.3-1. Comparison of NOx emissions between EMFAC2021 and EMFAC2017
2,500
NOx-Statewide-All Vehicles
2,000 EMFAC 2021
EMFAC2017
NOx Emissions (tpd)

1,500

1,000

500

0
2000 2010 2020 2030 2040 2050
Calendar Year

1,200 NOx-Statewide-Light Duty 1,600 NOx-Statewide-Heavy Duty

NOx Emissions (tpd)

NOx Emissions (tpd)

1,000 Vehicles (incl. MCY) 1,400 Vehicles (above 8,500 lbs. GVWR)
1,200
800
1,000
EMFAC 2021 EMFAC2021
600 800
EMFAC2017 EMFAC2017
400 600
400
200
200
0 0
2000 2010 2020 2030 2040 2050 2000 2010 2020 2030 2040 2050
Calendar Year Calendar Year

Reactive Organic Gases (ROG)

The following figure shows a comparison of EMFAC2021 and EMFAC2017 estimates of
statewide ROG emissions. ROG is dominated by LD vehicles. EMFAC2021 assumes
higher emission rates for LEV 1, LEV 2, and LEV 3 LD vehicles, and higher LD evaporative
emissions. HD ROG emissions decreased across all years for heavy-duty due to mileage
accrual updates for certain EMFAC 2021 categories, notably the T7 Tractor Class 8
category.

201
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure 5.3-2. Comparison of ROG emissions between EMFAC2021 and EMFAC2017

1,200 ROG-Statewide-All Vehicles
ROG Emissions (tpd)

1,000
800 EMFAC2021
600 EMFAC2017
400
200
0
2000 2010 2020 2030 2040 2050
Calendar Year
1,000 ROG-Statewide-Light Duty 150 ROG-Statewide-Heavy Duty
Vehicles (incl. MCY) Vehicles (above 8,500 lbs.

ROG Emissions (tpd)

800 GVWR)
ROG Emissions (tpd)

100
600 EMFAC2021 EMFAC2021
EMFAC2017 EMFAC2017
400
50
200

0 0
2000 2010 2020 2030 2040 2050 2000 2010 2020 2030 2040 2050
Calendar Year Calendar Year

Particulate Matter (PM2.5)

PM2.5 emissions include both tailpipe exhaust emissions and brake wear emissions.
Figure 5.3-3 shows the comparison of estimates of statewide total tailpipe PM2.5
exhaust emissions between EMFAC2021 and EMFAC2017. There is not much difference
in LD PM2.5 exhaust emissions, but EMFAC2021 HD PM2.5 exhaust emissions are lower
in the historical years compared to EMFAC2017. Mileage accrual for certain EMFAC
categories, like T7 Tractor Class 8, was lower than previously assumed in EMFAC2017
(see Section 4.4.2). This decrease in mileage accrual for these categories led to a
decrease in emissions. In the longer-term, emissions are larger because of updated
emission rates for model years 2014 and newer. In addition, the larger assumed VMT
growth rates in EMFAC2021 also led to an increase in PM emissions over time, but these
increases were somewhat offset by PSIP amendments, HD Warranty, ICT, and ACT
regulations.

202
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure 5.3-3. Comparison of tailpipe PM2.5 emissions between EMFAC2021 and

EMFAC2017
60 PM2.5 Exhaust-Statewide-All Vehicles
50
PM2.5 Emissions (tpd)

40

30
EMFAC2021
20 EMFAC2017

10

0
2000 2010 2020 2030 2040 2050
Calendar Year

15 PM2.5 Exhaust-Statewide-Light 60 PM2.5 exhaust-Statewide-Heavy

Duty Vehicles (incl. MCY) Duty Vehicles (above 8,500 lbs.
PM2.5 Emissions (tpd)

PM2.5 Emissions (tpd)

GVWR)
10 40

EMFAC2021 EMFAC2021
EMFAC2017 EMFAC2017
5 20

0 0
2000 2010 2020 2030 2040 2050 2000 2010 2020 2030 2040 2050
Calendar Year Calendar Year

The following figure shows a comparison of statewide PM2.5 brake wear emissions
between EMFAC2021 and EMFAC2017. As described in in Section 4.3.7, both LD and
HD PM brake wear emission rates were updated based on new data and resulted in
substantially lower PM brake wear emissions in EMFAC2021. The smaller PM2.5/Total
PM fraction of 35% in EMFAC2021 compared to 42% in EMFAC2017 also resulted in
substantially lower PM2.5 emissions overall. In addition, zero emission heavy-duty
vehicles required by ICT and ACT were assumed to have 50% lower brake wear
emissions through regenerative braking.

203
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure 5.3-4. Comparison of PM2.5 brake wear emissions between EMFAC2021 and
EMFAC2017
30 PM2.5 Brakewear-Statewide-All Vehicles
25
PM2.5 Emissions (tpd)

20
EMFAC2021
15
EMFAC2017
10

0
2000 2010 2020 2030 2040 2050
Calendar Year

30 PM2.5 Brakewear-Statewide- 8 PM2.5 Brakewear-Statewide-

PM2.5 Emissions (tpd)

Light Duty Vehicles (incl. MCY) Heavy Duty Vehicles (above 8,500
PM2.5 Emissions (tpd)

6 lbs. GVWR)
20

4
10 EMFAC2021
EMFAC2017 2
EMFAC2021
EMFAC2017
0 0
2000 2010 2020 2030 2040 2050 2000 2010 2020 2030 2040 2050
Calendar Year Calendar Year

Carbon Dioxide (CO2)

Figure 5.3-5 shows a comparison of CO2 emissions between EMFAC2021 and
EMFAC2017. The lower CO2 emissions in historical years is mainly due to the subtraction
of taxable gasoline being used in the off-road section for LD. Higher LD CO2 emissions
for future years attributes to the updated fuel economy analysis and the incorporation of
the U.S. EPA’s Final SAFE Rule in EMFAC2021. For HD, the emission rates in
EMFAC2021 are calculated so that once they are corrected for Phase 1 GHG standards,
they match the in-use emission data from TBSP. Compared to EMFAC2017, HD emission
rates in EMFAC2021 are slightly higher for MY2014 and newer. In the long-term,
emissions begin to increase due a higher VMT growth rate, but this increase is somewhat
offset by the incorporation of the ACT regulation. As described earlier, staff accounted
for the potential overlap between the ACT and Phase 2 regulations when determining
the fleet average CO2 emissions for heavy duty vehicles.

204
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure 5.3-5. Comparison of CO2 emissions between EMFAC2021 and EMFAC2017

(MMTCO2e per year) 200 CO2-Statewide-All Vehicles

150
CO2 Emissions

100

50
EMFAC2021
EMFAC2017
0
2000 2010 2020 2030 2040 2050
Calendar Year

150 CO2-Statewide-Light Duty 60 CO2-Statewide-Heavy Duty

Vehicles (incl. MCY)
(MMTCO2e per year)

Vehicles (above 8,500 lbs. GVWR)

(MMTCO2e per year)
CO2 Emissions

100 40
CO2 Emissions

50 20
EMFAC2021
EMFAC2021
EMFAC2017
EMFAC2017
0 0
2000 2010 2020 2030 2040 2050 2000 2010 2020 2030 2040 2050
Calendar Year Calendar Year

Ammonia
Ammonia (NH3) is modeled for the first time in EMFAC2021 and the following figure
shows the trend of estimated NH3 emissions. The increase of NH3 emissions in future
years is because EMFAC2021 assumes newer LD vehicles have higher NH3 emissions.
HD NH3 emissions increase due to the increasing fraction of SCR-equipped vehicles (i.e.
2010-certified). These vehicles produce NH3 emissions via NH3 slip from their SCR
catalysts.

205
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure 5.3-6. NH3 emissions in EMFAC2021

60 NH3-Statewide-Light Duty Vehicles 30 NH3-Statewide-Heavy Duty Vehicles
(incl. MCY) (above 8,500 lbs. GVWR)

NH3 Emissions (tpd)

NH3 Emissions (tpd)

40 20

20 10

EMFAC2021 EMFAC2021
0 0
2000 2020 2040 2000 2020 2040
Calendar Year Calendar Year

Fuel Consumptions
Figure 5.3-7 shows a comparison of gasoline, diesel, and natural gas fuel consumptions
between EMFAC2021 and EMFAC2017. The higher estimates of gasoline consumption
in future years is mainly due to higher LD CO2 emission rates from the impact of the
Final SAFE Rule (section 4.3.3). Diesel fuel consumption in EMFAC2021 is larger in the
near-term starting in 2019 due to updated fuel consumption data from CDTFA. In the
longer-term, fuel consumption increases due to larger projected HD VMT growth rates
from CSTDM. To some extent, these increases in fuel consumption are mitigated by the
ACT regulation, however, the regulatory impact of ACT is somewhat limited by its
overlap with Phase 2 (see more details in Section 4.3.5.5). Natural gas (NG) consumption
in EMFAC2021 is larger than EMFAC2017 between 2012 and 2030 because the
separation of NG trucks from diesel ones. The decline in NG consumption starting in
2030 is driven by the ICT and ACT regulations. Transit buses comprises a significant
portion of NG vehicles, and ICT directly affects the population of combustion buses by
turning them into zero-emission ones. Therefore, starting 2030, which is the first year of
100% ICT implementation, the NG population and fuel consumption starts decreasing.

206
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure 5.3-7. Comparison of annual fuel consumption (in billion gallons)

between EMFAC2021 and EMFAC2017

18
4.0

Annual Diesel Consumption

16
Annual Gasoline Consumption

3.5
14

(Bil. Gallons DGE)

3.0
12
(Bil. Gallons GGE)

10 2.5

8 2.0
6 1.5
4 1.0
EMFAC 2021
2 EMFAC 2021
EMFAC2017 0.5
EMFAC2017
0 0.0
2000 2010 2020 2030 2040 2050 2000 2010 2020 2030 2040 2050
Calendar Year Calendar Year

0.18
0.16
Annual Natural Gas Consumption

0.14
0.12
(Bil. Gallons DGE)

0.10
0.08
0.06
0.04 EMFAC 2021
0.02 EMFAC2017
0.00
2000 2010 2020 2030 2040 2050
Calendar Year

Energy Consumption
Figure 5.3-8 shows EMFAC2021 estimated California statewide energy consumption
from on-road sections. LD BEV and PHEV energy consumption increases over time, as
they make up a greater fraction of the fleet. Energy consumption for MD and HD starts
when ACT is implemented in 2024 and continues to grow as MD and HD electric vehicles
further penetrate the fleet. In 2050, MD and HD are responsible for gigawatt hours
(GWhr) per day, which is larger than the LD energy consumption of 40.9 GWhr per day.
Although there are a larger number of LD BEVs and PHEVs of 2.4 million compared to 1
million MD and HD ZEVs, energy consumption per mile for HD and MD electric vehicles
are significantly higher than LD ones. Overall, the total energy consumption in 2050 is
94.4 GWhr per day.
207
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure 5.3-8. Statewide Energy Consumption (Gigawatt Hours)

100
LD MD/HD
90
Daily Energy Consumption (Gwhr)

80
70
60
50
40
30
20
10
0
2000 2005 2010 2015 2020 2025 2030 2035 2040 2045 2050
Calendar Year

208
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

6 Appendices

209
EMFAC2021 Volume III Technical Document, Version 1.0.0
March 12, 2021

6.1 Vehicle Class Categorization

EMFAC202x Vehicle EMFAC2011 EMFAC2007
EMFAC202x Vehicle & Fuel Description
Class Vehicle Class Vehicle Class
LDA-Dsl
LDA-Gas
LDA Passenger Cars LDA LDA
LDA-Elec
LDA-Phe
LDT1-Dsl Light-Duty Trucks
LDT1- Gas (GVWR* <6000 lbs.
LDT1 LDT1 LDT1
LDT1-Elec and ETW** <= 3750
LDT1-Phe lbs.)
LDT2-Dsl
LDT2-Gas Light-Duty Trucks
LDT2 (GVWR <6000 lbs. and LDT2 LDT2
LDT2-Elec ETW 3751-5750 lbs.)
LDT2-Phe
MDV-Dsl
MDV-Gas Medium-Duty Trucks
MDV (GVWR 5751-8500 MDV MDV
MDV-Elec lbs.)
MDV-Phe
MH-Dsl
MH Motor Homes MH MH
MH-Gas
MCY-Gas MCY Motorcycles MCY MCY
LHD1 – Dsl Light-Heavy-Duty
LHD1-Gas LHD1 Trucks (GVWR 8501- LHDT1 LHDT1
LHD1-Elec 10000 lbs.)
LHD2 – Dsl Light-Heavy-Duty
LHD2-Gas LHD2 Trucks (GVWR 10001- LHDT2 LHDT2
LHD2-Elec 14000 lbs.)
T6 Public Class 4-Dsl T6 Public Class 4 T6 Public MHDT

210
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

EMFAC202x Vehicle EMFAC2011 EMFAC2007

EMFAC202x Vehicle & Fuel Description
Class Vehicle Class Vehicle Class
T6 Public Class 4-Elec Medium-Heavy Duty
Public Fleet Truck
T6 Public Class 4-NG (GVWR 14001-16000
lbs.)
T6 Public Class 5-Dsl Medium-Heavy Duty
T6 Public Class 5-Elec Public Fleet Truck
T6 Public Class 5 T6 Public MHDT
(GVWR 16001-19500
T6 Public Class 5-NG lbs.)
T6 Public Class 6-Dsl Medium-Heavy Duty
Public Fleet Truck
T6 Public Class 6 T6 Public MHDT
T6 Public Class 6-Elec (GVWR 19501-26000
lbs.)
T6 Public Class 6-NG
T6 Public Class 7-Dsl Medium-Heavy Duty
T6 Public Class 7-Elec Public Fleet Truck
T6 Public Class 7 T6 Public MHDT
(GVWR 26001-33000
T6 Public Class 7-NG lbs.)
T6 Utility Class 5-Dsl Medium-Heavy Duty
T6 Utility Class 5-Elec Utility Fleet Truck
T6 Utility Class 5 T6 Utility MHDT
(GVWR 16001-19500
T6 Utility Class 5-NG lbs.)
T6 Utility Class 6-Dsl Medium-Heavy Duty
T6 Utility Class 6-Elec Utility Fleet Truck
T6 Utility Class 6 T6 Utility MHDT
(GVWR 19501-26000
T6 Utility Class 6-NG lbs.)
T6 Utility Class 7-Dsl Medium-Heavy Duty
T6 Utility Class 7-Elec Utility Fleet Truck
T6 Utility Class 7 T6 Utility MHDT
(GVWR 26001-33000
T6 Utility Class 7-NG lbs.)
T6 Instate Tractor Class 6-Dsl Medium-Heavy Duty
T6 Instate Tractor Class
T6 Instate Tractor Class 6-Elec Tractor Truck (GVWR T6 Instate small MHDT
6
T6 Instate Tractor Class 6-NG 19501-26000 lbs.)
T6 Instate Delivery Class 4-Dsl T6 Instate Delivery Class
T6 Instate small MHDT
T6 Instate Delivery Class 4-Elec 4

211
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

EMFAC202x Vehicle EMFAC2011 EMFAC2007

EMFAC202x Vehicle & Fuel Description
Class Vehicle Class Vehicle Class
Medium-Heavy Duty
T6 Instate Delivery Class 4-NG Delivery Truck (GVWR
14001-16000 lbs.)
T6 Instate Delivery Class 5-Dsl Medium-Heavy Duty
T6 Instate Delivery Class
T6 Instate Delivery Class 5-Elec Delivery Truck (GVWR T6 Instate small MHDT
5
T6 Instate Delivery Class 5-NG 16001-19500 lbs.)
T6 Instate Delivery Class 6-Dsl Medium-Heavy Duty
T6 Instate Delivery Class
T6 Instate Delivery Class 6-Elec Delivery Truck (GVWR T6 Instate small MHDT
6
T6 Instate Delivery Class 6-NG 19501-26000 lbs.)
T6 Instate Other Class 4-Dsl Medium-Heavy Duty
T6 Instate Other Class 4-Elec T6 Instate Other Class 4 Other Truck (GVWR T6 Instate small MHDT
T6 Instate Other Class 4-NG 14001-16000 lbs.)
T6 Instate Other Class 5 -Dsl Medium-Heavy Duty
T6 Instate Other Class 5-Elec T6 Instate Other Class 5 Other Truck (GVWR T6 Instate small MHDT
T6 Instate Other Class 5-NG 16001-19500 lbs.)
T6 Instate Other Class 6 – Dsl Medium-Heavy Duty
T6 Instate Other Class 6-Elec T6 Instate Other Class 6 Other Truck (GVWR T6 Instate small MHDT
T6 Instate Other Class 6-NG 19501-26000 lbs.)
T6 Instate Tractor Class 7-Dsl Medium-Heavy Duty
T6 Instate Tractor Class
T6 Instate Tractor Class 7 –Elec Tractor Truck (GVWR T6 Instate heavy MHDT
7
T6 Instate Tractor Class 7-NG 26001-33000 lbs.)
T6 Instate Delivery Class 7 -Dsl Medium-Heavy Duty
T6 Instate Delivery Class
T6 Instate Delivery Class 7 -Elec Delivery Truck (GVWR T6 Instate heavy MHDT
7
T6 Instate Delivery Class 7 -NG 26001-33000 lbs.)
T6 Instate Other Class 7-Dsl Medium-Heavy Duty
T6 Instate Other Class 7-Elec T6 Instate Other Class 7 Other Truck (GVWR T6 Instate heavy MHDT
T6 Instate Other Class 7-NG 26001-33000 lbs.)
T6 CAIRP Class 4-Dsl T6 CAIRP Class 4 T6 CAIRP small MHDT

212
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

EMFAC202x Vehicle EMFAC2011 EMFAC2007

EMFAC202x Vehicle & Fuel Description
Class Vehicle Class Vehicle Class
Medium-Heavy Duty
CA International
T6 CAIRP Class 4-Elec Registration Plan
Truck (GVWR 14001-
16000 lbs.)
T6 CAIRP Class 5-Dsl Medium-Heavy Duty
CA International
T6 CAIRP Class 5 Registration Plan T6 CAIRP small MHDT
T6 CAIRP Class 5-Elec Truck (GVWR 16001-
19500 lbs.)
T6 CAIRP Class 6-Dsl Medium-Heavy Duty
CA International
T6 CAIRP Class 6 Registration Plan T6 CAIRP small MHDT
T6 CAIRP Class 6-Elec Truck (GVWR 19501-
26000 lbs.)
T6 CAIRP Class 7- Dsl Medium-Heavy Duty
T6 CAIRP Class 7-Elec CA International
T6 CAIRP Class 7 Registration Plan T6 CAIRP heavy MHDT
T6 CAIRP Class 7-NG Truck (GVWR 26001-
33000 lbs.)
Medium-Heavy Duty
Out-of-state Truck
T6 OOS Class 4-Dsl T6 OOS Class 4 MHDT
(GVWR 14001-16000
lbs.)
Medium-Heavy Duty
Out-of-state Truck
T6 OOS Class 5-Dsl T6 OOS Class 5 T6 OOS small MHDT
(GVWR 16001-19500
lbs.)
Medium-Heavy Duty
Out-of-state Truck
T6 OOS Class 6-Dsl T6 OOS Class 6 MHDT
(GVWR 19501-26000
lbs.)
Medium-Heavy Duty
Out-of-state Truck
T6 OOS Class 7-Dsl T6 OOS Class 7 T6 OOS heavy MHDT
(GVWR 26001-33000
lbs.)

213
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

EMFAC202x Vehicle EMFAC2011 EMFAC2007

EMFAC202x Vehicle & Fuel Description
Class Vehicle Class Vehicle Class
T6TS-Gas Medium-Heavy Duty
T6TS T6TS MHDT
T6TS-Elec Truck
T7 Public Class 8-Dsl Heavy-Heavy Duty
T7 Public Class 8-Elec Public Fleet Truck
T7 Public Class 8 T7 Public HHDT
(GVWR 33001 lbs. and
T7 Public Class 8-NG over)
T7 CAIRP Class 8-Dsl Heavy-Heavy Duty CA
International
T7 CAIRP Class 8 Registration Plan T7 CAIRP HHDT
T7 CAIRP Class 8-Elec Truck (GVWR 33001
T7 CAIRP Class 8-NG lbs. and over)
T7 Utility Class 8-Dsl Heavy-Heavy Duty
Utility Fleet Truck
T7 Utility Class 8 T7 Utility HHDT
T7 Utility Class 8-Elec (GVWR 33001 lbs. and
over)
Heavy-Heavy Duty
Non-Neighboring
T7 NNOOS Class 8-Dsl T7 NNOOS Class 8 Out-of-state Truck T7 NNOOS HHDT
(GVWR 33001 lbs. and
over)
Heavy-Heavy Duty
Neighboring Out-of-
T7 NOOS Class 8-Dsl T7 NOOS Class 8 T7 NOOS HHDT
state Truck (GVWR
33001 lbs. and over)
T7 Other Port Class 8-Dsl Heavy-Heavy Duty
Drayage Truck at
T7 Other Port Class 8 Other Facilities T7 Other Port HHDT
T7 Other Port Class 8-Elec (GVWR 33001 lbs. and
over)
T7 POAK Class 8-Dsl Heavy-Heavy Duty
T7 POAK Class 8-Elec Drayage Truck in Bay
T7 POAK Class 8 T7 POAK HHDT
Area (GVWR 33001
T7 POAK Class 8-NG lbs. and over)
T7 POLA Class 8-Dsl Heavy-Heavy Duty
T7 POLA Class 8 T7 POLA HHDT
T7 POLA Class 8-Elec Drayage Truck near

214
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

EMFAC202x Vehicle EMFAC2011 EMFAC2007

EMFAC202x Vehicle & Fuel Description
Class Vehicle Class Vehicle Class
South Coast (GVWR
T7 POLA Class 8-NG
33001 lbs. and over)
T7 Single Concrete/Transit Mix Class
Heavy-Heavy Duty
8-Dsl
T7 Single Single Unit
T7 Single Concrete/Transit Mix Class
Concrete/Transit Mix Concrete/Transit Mix T7 Single HHDT
8-Elec
Class 8 Truck (GVWR 33001
T7 Single Concrete/Transit Mix Class
lbs and over)
8-NG
T7 Single Dump Class 8-Dsl Heavy-Heavy Duty
T7 Single Dump Class 8-Elec Single Unit Dump
T7 Single Dump Class 8 T7 Single HHDT
Truck (GVWR 33001
T7 Single Dump Class 8-NG lbs. and over)
T7 Single Other Class 8-Dsl Heavy-Heavy Duty
T7 Single Other Class 8-Elec Single Unit Other
T7 Single Other Class 8 T7 Single HHDT
Truck (GVWR 33001
T7 Single Other Class 8-NG lbs. and over)
T7 Tractor Class 8-Dsl Heavy-Heavy Duty
T7 Tractor Class 8-Elec T7 Tractor Class 8 Tractor Truck (GVWR T7 Tractor HHDT
T7 Tractor Class 8-NG 33001 lbs. and over)
T7 SWCV Class 8-Dsl Heavy-Heavy Duty
T7 SWCV Class 8-Elec Solid Waste Collection
T7 SWCV Class 8 T7 SWCV HHDT
Truck (GVWR 33001
T7 SWCV Class 8-NG lbs. and over)
T7IS-Gas Heavy-Heavy Duty
T7IS T7IS HHDT
T7IS-Elec Truck
PTO-Dsl
PTO Power Take Off PTO HHDT
PTO-Elec
SBUS-Gas
SBUS-Dsl
SBUS School Buses SBUS SBUS
SBUS-Elec
SBUS-NG
UBUS-Dsl
UBUS Urban Buses UBUS UBUS
UBUS-Gas

215
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

EMFAC202x Vehicle EMFAC2011 EMFAC2007

EMFAC202x Vehicle & Fuel Description
Class Vehicle Class Vehicle Class
UBUS-Elec
UBUS-NG
Motor Coach-Dsl
Motor Coach Motor Coach Motor Coach OBUS
Motor Coach-Elec
OBUS-Gas
OBUS Other Buses OBUS OBUS
OBUS-Elec
All Other Buses-NG
All Other Buses All Other Buses All Other Buses OBUS
All Other Buses-Dsl

6.2 Exhaust Technology Groups

Model Years Vehicle Type
Tech. Fuel
Using TG in Using TG in Catalyst Fuel Delivery Description
Group Type
FRAC arrays FRAC arrays
1 1965-1974 LDA-LDT-MDV GAS NonCat Carb Pre-1975, no secondary air
2 1966-1974 LDA-LDT-MDV GAS NonCat Carb Pre-1975, with secondary air
3 1975-1979 LDA-LDT-MDV GAS NonCat Carb 1975+
4 1975-1976 LDA-LDT-MDV GAS OxCat Carb 1975-76, with secondary air
5 1975-1979 LDA-LDT-MDV GAS OxCat Carb 1975-79, no secondary air
6 1980-1981 LDA-LDT-MDV GAS OxCat Carb 1980+, no secondary air
7 1977-1984 LDA-LDT-MDV GAS OxCat Carb 1977+, with secondary air
8 1978-1979 LDA-LDT-MDV GAS TWC TBI / Carb 1978-79
9 1981-1984 LDA GAS TWC TBI / Carb 1981-84, 0.7 NOx std.
10 1985-1993 LDA-LDT-MDV GAS TWC TBI / Carb 1985+, 0.7 NOx std.
11 1977-1980 LDA GAS TWC MPFI 1977-80,
12 1981-1985 LDA-LDT-MDV GAS TWC MPFI 1981-85, 0.7 NOx std.
13 1986-1993 LDA-LDT-MDV GAS TWC MPFI 1986+, 0.7 NOx std.
14 1989-1994 LDA-LDT-MDV GAS TWC TBI / Carb 1989+, 0.4 NOx std.
15 1989-1994 LDA-LDT GAS TWC MPFI 1989+, 0.4 NOx std.

216
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Model Years Vehicle Type

Tech. Fuel
Using TG in Using TG in Catalyst Fuel Delivery Description
Group Type
FRAC arrays FRAC arrays
16 1980 LDA-LDT-MDV GAS TWC TBI / Carb 1980,
17 1993-1995 LDA-LDT GAS TWC TBI / Carb 1993+, 0.25 HC std.
18 1993-1995 LDA-LDT GAS TWC MPFI 1993+, 0.25 HC std.
19 1996-1999 LDA-LDT GAS TWC TBI / Carb 1996+, 0.25 HC std, OBD2
20 1996-1999 LDA-LDT GAS TWC MPFI 1996+, 0.25 HC std, OBD2
21 1994-1995 LDA-LDT GAS Adv.TWC MPFI 1994-95, TLEV, AFC
22 1996 LDA-LDT GAS Adv.TWC MPFI 1996+, TLEV, OBD2, AFC
23 1997-2003 LDA-LDT-MDV GAS Adv.TWC MPFI 1996+,LEV, OBD2, GCL, CBC, AFC
24 1997-2003 LDA-LDT-MDV GAS Adv.TWC MPFI 1996+,ULEV, OBD2, GCL, CBC, AFC
25 2000-2040 LDA-LDT1 ELE na na ZEV-Pure Electric
26 1996-2000 LDT-MDV GAS TWC MPFI 1996+, 0.7 NOx std., OBD2
27 1996-2000 LDT-MDV GAS TWC TBI / Carb 1996+, OBD2
28 2004-2025 LDA-LDT-MDV GAS Adv.TWC MPFI 2004+, LEV2, OBD2
29 2004-2020 LDA-LDT GAS Adv.TWC MPFI 2004+, ULEV2, OBD2
30 2004-2014 LDA-LDT GAS Adv.TWC MPFI 2004+, SULEV, OBD2
31 2003-2040 LDA-LDT1 GAS Adv.TWC MPFI 2004+, PZEV, OBD2
32 2009-2040 LDA-LDT1 GAS Adv.TWC MPFI 2004+, Tier2-3 120K //0.055/2.1/0.03, OBD2
33 2007-2040 LDA-LDT GAS Adv.TWC MPFI 2004+, Tier2-4 120K //0.07/2.1/0.04, OBD2
34 2004-2006 MDV GAS Adv.TWC MPFI 2004+, Tier2-8 120K //0.156/4.2/0.2, OBD2
35 2004-2006 LDT2 GAS Adv.TWC MPFI 2004+, Tier2-9 120K //0.09/4.2/0.3, OBD2
36 2004-2006 MDV GAS Adv.TWC MPFI 2004+, Tier2-10 120K //0.23/6.4/0.6, OBD2
37 2003-2040 LDA-LDT1 GAS Adv.TWC MPFI 2003+, AT PZEV, OBD2
38 2020-2040 LDA-LDT GAS Adv.TWC MPFI 2015+, SULEV 20, OBD2
39 2020-2040 LDA-LDT GAS Adv.TWC MPFI 2015+, ULEV 50, OBD2
40 1965-1979 LDA GAS NonCat Carb Pre-1980, Mexican veh no secondary air
41 1975-1986 LDA GAS OxCat Carb 1975-76, Mexican veh with secondary air
42 1980-1987 LDA GAS TWC TBI/Carb 1980-87, Mexican veh, 0.7 NOx std.
43 1981-2040 LDA GAS TWC MPFI 1981-2040, Mexican veh, 0.7 NOx std.

217
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Model Years Vehicle Type

Tech. Fuel
Using TG in Using TG in Catalyst Fuel Delivery Description
Group Type
FRAC arrays FRAC arrays
44 2015-2025 LDA-LDT GAS Adv.TWC MPFI 2015+, ULEV 70, OBD2
46 1965-1976 LHDT1 GAS NonCat Carb Pre-1977
47 1977-1983 LHDT1 GAS OxCat Carb 1977-83
48 1984-1987 LHDT1 GAS TWC Carb 1984-87
49 1988-1990 LHDT1 GAS TWC FI 1988-90
50 1991-1995 LHDT1 GAS TWC FI 1991-94
51 1995-2001 LHDT1 GAS TWC MPFI 1995-01, MDV
52 2002-2003 LHDT1 GAS TWC MPFI 2002-03, LEV
53 2004-2008 LHDT1 GAS Adv.TWC MPFI 2004-08, ULEV
54 2008-2021 LHDT1 GAS Adv.TWC MPFI 2008+, USEPA 2008 stds.
58 2016-2040 LHDT1 GAS Adv.TWC MPFI 2016+ LEV 3 ULEV 250
59 2018-2040 LHDT1 GAS Adv.TWC MPFI 2018+ LEV 3 SULEV 170
60 1965-1974 LHDT1 DSL Pre-1975
61 1975-1976 LHDT1 DSL 1975-76
62 1977-1979 LHDT1 DSL 1977-79
63 1980-1983 LHDT1 DSL 1980-83
64 1984-1986 LHDT1 DSL 1984-86
65 1987-1990 LHDT1 DSL 1987-90
66 1991-1993 LHDT1 DSL 1991-93
67 1994-1995 LHDT1 DSL 1994
68 1995-2001 LHDT1 DSL 1995-01, MDV?
69 2002-2003 LHDT1 DSL 2002-03, LEV
70 2004-2009 LHDT1 DSL 2004-09, ULEV
71 2007-2021 LHDT1 DSL 2007+, USEPA 2007 stds.
73 2016-2040 LHDT1 DSL 2016+ LEV 3 ULEV 250
74 2018-2040 LHDT1 DSL 2018+ LEV 3 SULEV 170
76 1965-1976 LHDT2 GAS NonCat Carb Pre-1977,
77 1977-1983 LHDT2 GAS OxCat Carb 1977-83,

218
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Model Years Vehicle Type

Tech. Fuel
Using TG in Using TG in Catalyst Fuel Delivery Description
Group Type
FRAC arrays FRAC arrays
78 1984-1987 LHDT2 GAS TWC Carb 1984-87,
79 1988-1990 LHDT2 GAS TWC FI 1988-90,
80 1991-1995 LHDT2 GAS TWC FI 1991-94,
81 1995-2001 LHDT2 GAS TWC MPFI 1995-01, MDV
82 2002-2003 LHDT2 GAS TWC MPFI 2002-03, LEV
83 2004-2008 LHDT2 GAS Adv.TWC MPFI 2004-08, ULEV
84 2008-2040 LHDT2 GAS Adv.TWC MPFI 2008+, USEPA 2008 stds.
86 2016-2040 LHDT2 GAS Adv.TWC MPFI 2016+ LEV 3 ULEV 400
87 2018-2040 LHDT2 GAS Adv.TWC MPFI 2018+ LEV 3 SULEV 230
90 1965-1974 LHDT2 DSL Pre-1975,
91 1975-1976 LHDT2 DSL 1975-76,
92 1977-1979 LHDT2 DSL 1977-79,
93 1980-1983 LHDT2 DSL 1980-83,
94 1984-1986 LHDT2 DSL 1984-86,
95 1987-1990 LHDT2 DSL 1987-90,
96 1991-1993 LHDT2 DSL 1991-93,
97 1994-1995 LHDT2 DSL 1994
98 1995-2001 LHDT2 DSL 1995-01, MDV
99 2002-2003 LHDT2 DSL 2002-03, LEV
100 2004-2009 LHDT2 DSL 2004-09, ULEV
101 2007-2021 LHDT2 DSL 2007+, USEPA 2007 stds.
104 2016-2040 LHDT2 DSL 2016+ LEV 3 ULEV 400
105 2018-2040 LHDT2 DSL 2018+ LEV 3 SULEV 230
106 1965-1976 MHDV GAS NonCat Carb Pre-1977,
107 1977-1984 MHDV GAS OxCat Carb 1977-83,
108 1984-1987 MHDV GAS TWC Carb 1984-87,
109 1986-1990 MHDV GAS TWC FI 1988-90,
110 1987-1997 MHDV GAS TWC FI 1991-97,

219
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Model Years Vehicle Type

Tech. Fuel
Using TG in Using TG in Catalyst Fuel Delivery Description
Group Type
FRAC arrays FRAC arrays
111 1994-2003 MHDV GAS TWC MPFI 1998-03,
112 1998-2004 MHDV GAS TWC MPFI 2004,
113 2004-2040 MHDV GAS TWC MPFI 2005, 1g HC + NOx std.
114 2008-2040 MHDT GAS TWC MPFI 2008+, USEPA 2008 stds.
120 1965-1974 MHDT-MH DSL Pre-1975,
121 1975-1976 MHDT-MH DSL 1975-76,
122 1977-1979 MHDT-MH DSL 1977-79,
123 1980-1983 MHDT-MH DSL 1980-83,
124 1984-1986 MHDT-MH DSL 1984-86,
125 1987-1990 MHDT-MH DSL 1987-90,
126 1991-1993 MHDT-MH DSL 1991-93,
127 1994-1997 MHDT-MH DSL 1994-97,
128 1998-1998 MHDT-MH DSL 1998,
129 1999-2002 MHDT-MH DSL 1999-02,
130 2003-2006 MHDT-MH DSL 2003-06, 2g NOx std.
131 2007-2009 MHDT-MH DSL 2007-09, Transition 2010 stds.
132 2010-2040 MHDT-MH DSL 2010+, US EPA 2010 stds.
133 2010-2040 MHDT-MH DSL 2010+, US EPA 2010 stds/OBD
136 1965-1976 HHDV-LHV GAS NonCat Carb Pre-1977,
137 1977-1984 HHDV-LHV GAS OxCat Carb 1977-84,
138 1985-1985 HHDV-LHV GAS TWC Carb 1985
139 1986-1986 HHDV-LHV GAS TWC FI 1986
140 1987-1993 HHDV-LHV GAS TWC FI 1987-93,
141 1994-1997 HHDV-LHV GAS TWC MPFI 1994-97,
142 1998-2003 HHDV-LHV GAS TWC MPFI 1998-03,
143 2004-2040 HHDV-LHV GAS TWC MPFI 2004-06,
150 1965-1974 HHDV-LHV DSL Pre-1975, CA stds.
151 1975-1976 HHDV-LHV DSL 1975-76, CA Stds.

220
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Model Years Vehicle Type

Tech. Fuel
Using TG in Using TG in Catalyst Fuel Delivery Description
Group Type
FRAC arrays FRAC arrays
152 1977-1979 HHDV-LHV DSL 1977-79, CA Stds.
153 1980-1983 HHDV-LHV DSL 1980-83, CA Stds.
154 1984-1986 HHDV-LHV DSL 1984-86, CA Stds.
155 1987-1990 HHDV-LHV DSL 1987-90, CA Stds.
156 1991-1993 HHDV-LHV DSL 1991-93, CA Stds.
157 1994-1997 HHDV-LHV DSL 1994-97, CA Stds.
158 1998-1998 HHDV-LHV DSL 1998, CA Stds.
159 1999-2002 HHDV-LHV DSL 1999-02, CA Stds.
160 2003-2006 HHDV-LHV DSL 2003-06, CA 2g NOx Stds.
161 2007-2009 HHDV-LHV DSL 2007-2009, USEPA 2007 stds.
162 2010+ HHDV-LHV DSL 2010+, USEPA 2007 stds.
163 2010+ HHDV-LHV DSL 2010+ , USEPA 2007 stds. W/OBD2
170 1965-1974 LDA-LDT-MDV DSL Pre-1975,
171 1975-1979 LDA-LDT-MDV DSL 1975-79,
172 1980-1980 LDA-LDT-MDV DSL 1980,
173 1981-1981 LDA-LDT-MDV DSL 1981,
174 1982-1982 LDA-LDT-MDV DSL 1982,
175 1983-1983 LDA-LDT-MDV DSL 1983,
176 1984-1992 LDA-LDT-MDV DSL 1984-92,
177 1993-2003 LDA-LDT-MDV DSL 1993+,
178 2007-2025 LDA LT3 DSL DPF SCR 2008+, LEV 160 DSL, OBD2
179 2007-2025 LDA LT3 DSL DPF SCR 2008+, ULEV 125 DSL, OBD2
180 2020-2040 LDA LT3 DSL DPF SCR 2020+, SULEV 30 DSL, OBD2
200 1965-1973 HHDV-LHV DSL Pre-1974, Federal Stds.
201 1974-1978 HHDV-LHV DSL 1974-78, Federal Stds.
202 1979-1983 HHDV-LHV DSL 1979-83, Federal Stds.
203 1984-1987 HHDV-LHV DSL 1984-87, Federal Stds.
204 1988-1990 HHDV-LHV DSL 1988-90, Federal Stds.

221
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Model Years Vehicle Type

Tech. Fuel
Using TG in Using TG in Catalyst Fuel Delivery Description
Group Type
FRAC arrays FRAC arrays
205 1991-1993 HHDV-LHV DSL 1991-93, Federal Stds.
206 1994-1997 HHDV-LHV DSL 1994-97, Federal Stds.
207 1998-1998 HHDV-LHV DSL 1998, Federal Stds.
208 1999-2003 HHDV-LHV DSL 1999-02, Federal Stds.
209 2003-2009 HHDV-LHV DSL 2003-06, Federal Stds.
210 2007-2009 HHDV-LHV DSL 2007-2009, USEPA 2007 stds.
211 2010+ HHDV-LHV DSL 2010+, USEPA 2007 stds.
216 1965-1986 UB DSL Pre-87,
217 1987-1990 UB DSL 1987-90,
218 1991-1993 UB DSL 1991-93,
219 1994-1995 UB DSL 1994-95,
220 1996-1998 UB DSL 1996-98,
221 1999-2002 UB DSL 1999-02,
222 2003-2003 UB DSL 2003,
223 2004-2006 UB DSL 2004-06,
224 2007-2040 UB DSL 2007,
225 2008-2040 UB DSL 2008+, ZEV or ZEBS
228 1965-1976 SBUS GAS NonCat Carb Pre-77,
229 1977-1983 SBUS GAS OxCat TBI / Carb 1977-83,
230 1984-1987 SBUS GAS TWC FI 1984-87,
231 1988-1990 SBUS GAS TWC FI 1988-90,
232 1991-1997 SBUS GAS TWC FI 1991-97,
233 1998-2003 SBUS GAS TWC MPFI 1998-03,
234 2004-2004 SBUS GAS TWC MPFI 2004,
235 2005-2008 SBUS GAS TWC MPFI 2005,1g HC+NOx Stds.
236 2008-2040 SBUS GAS TWC MPFI 2008+, USEPAs Stds.
240 1965-1974 SBUS DSL Pre-75,
241 1975-1976 SBUS DSL 1975-76,

222
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Model Years Vehicle Type

Tech. Fuel
Using TG in Using TG in Catalyst Fuel Delivery Description
Group Type
FRAC arrays FRAC arrays
242 1977-1979 SBUS DSL 1977-79,
243 1980-1983 SBUS DSL 1980-83,
244 1984-1986 SBUS DSL 1984-86,
245 1987-1990 SBUS DSL 1987-90,
246 1991-1993 SBUS DSL 1991-93,
247 1994-1997 SBUS DSL 1994-97,
248 1998-1998 SBUS DSL 1998,
249 1999-2003 SBUS DSL 1999-02,
250 2003-2009 SBUS DSL 2003-06, 2g NOx Std
251 2007-2040 SBUS DSL 2007+, USEPA Std.
260 1965-1977 MCY GAS NonCat 2-Stroke All, 6g evap Std.
261 1965-1977 MCY GAS NonCat Carb Pre-1978, 6g evap Std.
262 1978-1979 MCY GAS NonCat Carb 1978-79, 6g evap Std.
263 1980-1981 MCY GAS NonCat Carb 1980-81, 6g evap Std.
264 1982-1984 MCY GAS NonCat Carb 1982-84, 6g evap Std.
265 1985-1987 MCY GAS NonCat Carb 1985-87, 2g evap Std.
266 1988-2003 MCY GAS NonCat Carb 1988-03, 2g evap Std.
267 1994-2003 MCY GAS NonCat FI 1988-03, 2g evap Std.
268 1995-2003 MCY GAS OxCat Carb 1988-03, 2g evap Std.
269 1994-2003 MCY GAS TWC FI 1988-03, 2g evap Std.
270 2004-2007 MCY GAS NonCat Carb 2003-08, 2g evap Std.
271 2004-2007 MCY GAS NonCat FI 2003-08, 2g evap Std.
272 2004-2007 MCY GAS OxCat Carb 2003-08, 2g evap Std.
273 2004-2007 MCY GAS TWC FI 03-08 MCY FI/cat/2g evap
274 2008-2040 MCY GAS NonCat Carb 2008+, 2 evap Std.
275 2008-2040 MCY GAS NonCat FI 2008+, 2 evap Std.
276 2008-2040 MCY GAS OxCat Carb 2008+, 2 evap Std.
277 2008-2040 MCY GAS TWC FI 2008+, 2 evap Std.

223
EMFAC2021 Volume III Technical Document, Version 1.0.0
March 12, 2021

6.3 Emissions Impact in Air Basins

Comparison of Vehicle Emissions in the South Coast Air Basin between EMFAC2017
and EMFAC2021
NOx
800 NOx-South Coast-All Vehicles
NOx Emissions (tpd)

600
EMFAC 2021
EMFAC2017
400

200

0
2000 2010 2020 2030 2040 2050
Calendar Year

500 NOx-South Coast-Light Duty Vehicles 400 NOx-South Coast-Heavy Duty

(incl. MCY) Vehicles (above 8,500 lbs. GVWR)
400
NOx Emissions (tpd)

NOx Emissions (tpd)

300
300 EMFAC 2021 EMFAC 2021
EMFAC2017 200 EMFAC2017
200

100
100

0 0
2000 2020 2040 2000 2020 2040
Calendar Year Calendar Year

224
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

ROG
400 ROG-South Coast-All Vehicles
350
ROG Emissions (tpd)

300 EMFAC2021
EMFAC2017
250
200
150
100
50
0
2000 2010 2020 2030 2040 2050
Calendar Year

400 ROG-South Coast-Light Duty 40 ROG-South Coast-Heavy Duty

Vehicles (incl. MCY) Vehicles (above 8,500 lbs. GVWR)
ROG Emissions (tpd)
ROG Emissions (tpd)

300 30

200 EMFAC2021 EMFAC2021

20
EMFAC2017 EMFAC2017
100 10

0 0
2000 2010 2020 2030 2040 2050 2000 2010 2020 2030 2040 2050
Calendar Year Calendar Year

225
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

PM2.5
30 PM2.5-South Coast-All Vehicles

25
PM2.5 Emissions (tpd)

EMFAC2021
20 EMFAC2017

15

10

0
2000 2010 2020 2030 2040 2050
Calendar Year

15 PM2.5-South Coast-Light Duty 15 PM2.5-South Coast-Heavy Duty

PM2.5 Emissions (tpd)

Vehicles (incl. MCY) Vehicles (above 8,500 lbs. GVWR)

PM2.5 Emissions (tpd)

10 10
EMFAC2021
EMFAC2017
EMFAC2021
5 5
EMFAC2017

0 0
2000 2010 2020 2030 2040 2050 2000 2010 2020 2030 2040 2050
Calendar Year Calendar Year

226
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Comparison of Vehicle Emissions in the San Joaquin Valley (SJV) between EMFAC2017
and EMFAC2021
NOx

450 NOx-SJV-All Vehicles

400

350 EMFAC2021
NOx Emissions (tpd)

EMFAC2017
300

250

200

150

100

50

0
2000 2010 2020 2030 2040 2050
Calendar Year

150 NOx-SJV-Light Duty Vehicles 300 NOx- SJV-Heavy Duty Vehicles

(incl. MCY) (above 8,500 lbs. GVWR)
NOx Emissions (tpd)

NOx Emissions (tpd)

100 EMFAC2021 200 EMFAC2021

EMFAC2017 EMFAC2017

50 100

0 0
2000 2010 2020 2030 2040 2050 2000 2010 2020 2030 2040 2050
Calendar Year Calendar Year

227
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

ROG

140 ROG-SJV-All Vehicles

120
EMFAC2021
100
ROG Emissions (tpd)

EMFAC2017
80

60

40

20

0
2000 2010 2020 2030 2040 2050
Calendar Year

100 ROG-SJV-Light Duty Vehicles 30 ROG-SJV-Heavy Duty Vehicles

(incl. MCY) (above 8,500 lbs. GVWR)
ROG Emissions (tpd)

ROG Emissions (tpd)

EMFAC2021 20
EMFAC2017 EMFAC2021
50 EMFAC2017
10

0 0
2000 2010 2020 2030 2040 2050 2000 2010 2020 2030 2040 2050
Calendar Year Calendar Year

228
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

PM2.5
15 PM2.5-SJV-All Vehicles
PM2.5 Emissions (tpd)

EMFAC2021
10 EMFAC2017

0
2000 2010 2020 2030 2040 2050
Calendar Year
3.5 PM2.5-SJV-Light Duty Vehicles (incl. 15 PM2.5-SJV-Heavy Duty Vehicles
PM2.5 Emissions (tpd)

3.0 MCY) (above 8,500 lbs. GVWR)

PM2.5 Emissions (tpd)

2.5
10
2.0 EMFAC2021
EMFAC2021 EMFAC2017
1.5
EMFAC2017
5
1.0
0.5
0.0 0
2000 2020 2040 2000 2010 2020 2030 2040 2050
Calendar Year Calendar Year

229
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

6.4 Heavy Duty Natural Gas Penetration

EMFAC 2021 Vehicle Prediction
Air District Name Intercept Slope
Category Class
T6 Utility Class 5 Amador County APCD 0.083333 0 Flat12Yr
T6 Utility Class 6 Amador County APCD 0.083333 0 Flat12Yr
T6 Utility Class 7 Amador County APCD 0.083333 0 Flat12Yr
T6 Instate Tractor Class 7 Amador County APCD 0.013889 0 Flat12Yr
T6 Instate Delivery Class 7 Amador County APCD 0.013889 0 Flat12Yr
T6 Instate Other Class 7 Amador County APCD 0.013889 0 Flat12Yr
T6 Instate Tractor Class 7 Antelope Valley AQMD 0.027778 0 Flat12Yr
T6 Instate Delivery Class 7 Antelope Valley AQMD 0.027778 0 Flat12Yr
T6 Instate Other Class 7 Antelope Valley AQMD 0.027778 0 Flat12Yr
T7 POLA Class 8 Antelope Valley AQMD 0.099306 0 Flat12Yr
T7 Single Concrete/Transit
Antelope Valley AQMD 0.006944 0 Flat12Yr
Mix Class 8
T7 Single Dump Class 8 Antelope Valley AQMD 0.006944 0 Flat12Yr
T7 Single Other Class 8 Antelope Valley AQMD 0.006944 0 Flat12Yr
SBUS Antelope Valley AQMD 0.130057 0 Flat12Yr
T6 Public Class 4 Bay Area AQMD 0.066853 0 Flat12Yr
T6 Public Class 5 Bay Area AQMD 0.066853 0 Flat12Yr
T6 Public Class 6 Bay Area AQMD 0.066853 0 Flat12Yr
T6 Public Class 7 Bay Area AQMD 0.066853 0 Flat12Yr
T6 Utility Class 5 Bay Area AQMD 0.008433 0 Flat12Yr
T6 Utility Class 6 Bay Area AQMD 0.008433 0 Flat12Yr
T6 Utility Class 7 Bay Area AQMD 0.008433 0 Flat12Yr
T6 Instate Tractor Class 6 Bay Area AQMD 0.014131 0 Flat12Yr
T6 Instate Delivery Class 4 Bay Area AQMD 0.014131 0 Flat12Yr
T6 Instate Delivery Class 5 Bay Area AQMD 0.014131 0 Flat12Yr
T6 Instate Delivery Class 6 Bay Area AQMD 0.014131 0 Flat12Yr
T6 Instate Other Class 4 Bay Area AQMD 0.014131 0 Flat12Yr
T6 Instate Other Class 5 Bay Area AQMD 0.014131 0 Flat12Yr
T6 Instate Other Class 6 Bay Area AQMD 0.014131 0 Flat12Yr
T6 Instate Tractor Class 7 Bay Area AQMD 0.018565 0 Flat12Yr
T6 Instate Delivery Class 7 Bay Area AQMD 0.018565 0 Flat12Yr
T6 Instate Other Class 7 Bay Area AQMD 0.018565 0 Flat12Yr
T7 Public Class 8 Bay Area AQMD 0.008157 0 Flat12Yr
T7 CAIRP Class 8 Bay Area AQMD 0.001827 0 Flat12Yr
T7 POAK Class 8 Bay Area AQMD 0.002252 0 Flat12Yr
T7 POLA Class 8 Bay Area AQMD 0.076389 0 Flat12Yr
T7 Single Concrete/Transit
Bay Area AQMD 0.057956 0 Flat12Yr
Mix Class 8

230
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

EMFAC 2021 Vehicle Prediction

Air District Name Intercept Slope
Category Class
T7 Single Dump Class 8 Bay Area AQMD 0.057956 0 Flat12Yr
T7 Single Other Class 8 Bay Area AQMD 0.057956 0 Flat12Yr
T7 Tractor Class 8 Bay Area AQMD 0.078304 0 Flat12Yr
SBUS Bay Area AQMD 0.05757 0 Flat12Yr
All Other Buses Bay Area AQMD 0.023012 0 Flat12Yr
T7 Single Concrete/Transit
Butte County AQMD 0.006944 0 Flat12Yr
Mix Class 8
T7 Single Dump Class 8 Butte County AQMD 0.006944 0 Flat12Yr
T7 Single Other Class 8 Butte County AQMD 0.006944 0 Flat12Yr
T7 Tractor Class 8 Butte County AQMD 0.002976 0 Flat12Yr
T6 Instate Tractor Class 6 Calaveras County APCD 0.015152 0 Flat12Yr
T6 Instate Delivery Class 4 Calaveras County APCD 0.015152 0 Flat12Yr
T6 Instate Delivery Class 5 Calaveras County APCD 0.015152 0 Flat12Yr
T6 Instate Delivery Class 6 Calaveras County APCD 0.015152 0 Flat12Yr
T6 Instate Other Class 4 Calaveras County APCD 0.015152 0 Flat12Yr
T6 Instate Other Class 5 Calaveras County APCD 0.015152 0 Flat12Yr
T6 Instate Other Class 6 Calaveras County APCD 0.015152 0 Flat12Yr
T7 Single Concrete/Transit
El Dorado County APCD 0.010417 0 Flat12Yr
Mix Class 8
T7 Single Dump Class 8 El Dorado County APCD 0.010417 0 Flat12Yr
T7 Single Other Class 8 El Dorado County APCD 0.010417 0 Flat12Yr
All Other Buses El Dorado County APCD 0.041667 0 Flat12Yr
T7 Tractor Class 8 Feather River AQMD 0.053407 0 Flat12Yr
T6 Instate Tractor Class 6 Great Basin Unified APCD 0.0625 0 Flat12Yr
T6 Instate Delivery Class 4 Great Basin Unified APCD 0.0625 0 Flat12Yr
T6 Instate Delivery Class 5 Great Basin Unified APCD 0.0625 0 Flat12Yr
T6 Instate Delivery Class 6 Great Basin Unified APCD 0.0625 0 Flat12Yr
T6 Instate Other Class 4 Great Basin Unified APCD 0.0625 0 Flat12Yr
T6 Instate Other Class 5 Great Basin Unified APCD 0.0625 0 Flat12Yr
T6 Instate Other Class 6 Great Basin Unified APCD 0.0625 0 Flat12Yr
T6 Public Class 4 Imperial County APCD 0.09035 0 Flat12Yr
T6 Public Class 5 Imperial County APCD 0.09035 0 Flat12Yr
T6 Public Class 6 Imperial County APCD 0.09035 0 Flat12Yr
T6 Public Class 7 Imperial County APCD 0.09035 0 Flat12Yr
T7 Public Class 8 Imperial County APCD 0.010417 0 Flat12Yr
T7 POLA Class 8 Imperial County APCD 0.171721 0 Flat12Yr
T7 Tractor Class 8 Imperial County APCD 0.014385 0 Flat12Yr
T7 POLA Class 8 Kern County APCD 0.041667 0 Flat12Yr
T7 Tractor Class 8 Kern County APCD 0.002604 0 Flat12Yr
T6 Instate Tractor Class 6 Lassen County APCD 0.027778 0 Flat12Yr

231
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

EMFAC 2021 Vehicle Prediction

Air District Name Intercept Slope
Category Class
T6 Instate Delivery Class 4 Lassen County APCD 0.027778 0 Flat12Yr
T6 Instate Delivery Class 5 Lassen County APCD 0.027778 0 Flat12Yr
T6 Instate Delivery Class 6 Lassen County APCD 0.027778 0 Flat12Yr
T6 Instate Other Class 4 Lassen County APCD 0.027778 0 Flat12Yr
T6 Instate Other Class 5 Lassen County APCD 0.027778 0 Flat12Yr
T6 Instate Other Class 6 Lassen County APCD 0.027778 0 Flat12Yr
T7 Single Concrete/Transit Mendocino County
0.024515 0 Flat12Yr
Mix Class 8 AQMD
Mendocino County
T7 Single Dump Class 8 0.024515 0 Flat12Yr
AQMD
Mendocino County
T7 Single Other Class 8 0.024515 0 Flat12Yr
AQMD
Mendocino County
T7 Tractor Class 8 0.005556 0 Flat12Yr
AQMD
T6 Public Class 4 Mojave Desert AQMD 0.070833 0 Flat12Yr
T6 Public Class 5 Mojave Desert AQMD 0.070833 0 Flat12Yr
T6 Public Class 6 Mojave Desert AQMD 0.070833 0 Flat12Yr
T6 Public Class 7 Mojave Desert AQMD 0.070833 0 Flat12Yr
T7 Public Class 8 Mojave Desert AQMD 0.055556 0 Flat12Yr
T7 CAIRP Class 8 Mojave Desert AQMD 0.000223 0 Flat12Yr
T7 POLA Class 8 Mojave Desert AQMD 0.046296 0 Flat12Yr
T7 Single Concrete/Transit
Mojave Desert AQMD 0.022751 0 Flat12Yr
Mix Class 8
T7 Single Dump Class 8 Mojave Desert AQMD 0.022751 0 Flat12Yr
T7 Single Other Class 8 Mojave Desert AQMD 0.022751 0 Flat12Yr
T7 Tractor Class 8 Mojave Desert AQMD 0.00743 0 Flat12Yr
Monterey Bay Unified
T6 Public Class 4 0.041667 0 Flat12Yr
APCD
Monterey Bay Unified
T6 Public Class 5 0.041667 0 Flat12Yr
APCD
Monterey Bay Unified
T6 Public Class 6 0.041667 0 Flat12Yr
APCD
Monterey Bay Unified
T6 Public Class 7 0.041667 0 Flat12Yr
APCD
Monterey Bay Unified
T6 Instate Tractor Class 7 0.017062 0 Flat12Yr
APCD
Monterey Bay Unified
T6 Instate Delivery Class 7 0.017062 0 Flat12Yr
APCD
Monterey Bay Unified
T6 Instate Other Class 7 0.017062 0 Flat12Yr
APCD
Monterey Bay Unified
T7 Public Class 8 0.01513 0 Flat12Yr
APCD
T7 Single Concrete/Transit Monterey Bay Unified
0.032989 0 Flat12Yr
Mix Class 8 APCD

232
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

EMFAC 2021 Vehicle Prediction

Air District Name Intercept Slope
Category Class
Monterey Bay Unified
T7 Single Dump Class 8 0.032989 0 Flat12Yr
APCD
Monterey Bay Unified
T7 Single Other Class 8 0.032989 0 Flat12Yr
APCD
Monterey Bay Unified
T7 Tractor Class 8 0.002924 0 Flat12Yr
APCD
Monterey Bay Unified
SBUS 0.002874 0 Flat12Yr
APCD
T7 Single Concrete/Transit North Coast Unified
0.022488 0 Flat12Yr
Mix Class 8 AQMD
North Coast Unified
T7 Single Dump Class 8 0.022488 0 Flat12Yr
AQMD
North Coast Unified
T7 Single Other Class 8 0.022488 0 Flat12Yr
AQMD
T6 Instate Tractor Class 7 Northern Sierra AQMD 0.070648 0 Flat12Yr
T6 Instate Delivery Class 7 Northern Sierra AQMD 0.070648 0 Flat12Yr
T6 Instate Other Class 7 Northern Sierra AQMD 0.070648 0 Flat12Yr
T7 Public Class 8 Placer County APCD 0.007564 0 Flat12Yr
T7 Single Concrete/Transit
Placer County APCD 0.018315 0 Flat12Yr
Mix Class 8
T7 Single Dump Class 8 Placer County APCD 0.018315 0 Flat12Yr
T7 Single Other Class 8 Placer County APCD 0.018315 0 Flat12Yr
T7 Tractor Class 8 Placer County APCD 0.003333 0 Flat12Yr
Sacramento Metropolitan
T6 Public Class 4 0.110258 0 Flat12Yr
AQMD
Sacramento Metropolitan
T6 Public Class 5 0.110258 0 Flat12Yr
AQMD
Sacramento Metropolitan
T6 Public Class 6 0.110258 0 Flat12Yr
AQMD
Sacramento Metropolitan
T6 Public Class 7 0.110258 0 Flat12Yr
AQMD
Sacramento Metropolitan
T6 Utility Class 5 0.002604 0 Flat12Yr
AQMD
Sacramento Metropolitan
T6 Utility Class 6 0.002604 0 Flat12Yr
AQMD
Sacramento Metropolitan
T6 Utility Class 7 0.002604 0 Flat12Yr
AQMD
Sacramento Metropolitan
T6 Instate Tractor Class 7 0.009707 0 Flat12Yr
AQMD
Sacramento Metropolitan
T6 Instate Delivery Class 7 0.009707 0 Flat12Yr
AQMD
Sacramento Metropolitan
T6 Instate Other Class 7 0.009707 0 Flat12Yr
AQMD
Sacramento Metropolitan
T7 Public Class 8 0.027083 0 Flat12Yr
AQMD
Sacramento Metropolitan
T7 CAIRP Class 8 0.010142 0 Flat12Yr
AQMD

233
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

EMFAC 2021 Vehicle Prediction

Air District Name Intercept Slope
Category Class
T7 Single Concrete/Transit Sacramento Metropolitan
0.030047 0 Flat12Yr
Mix Class 8 AQMD
Sacramento Metropolitan
T7 Single Dump Class 8 0.030047 0 Flat12Yr
AQMD
Sacramento Metropolitan
T7 Single Other Class 8 0.030047 0 Flat12Yr
AQMD
Sacramento Metropolitan
T7 Tractor Class 8 0.009123 0 Flat12Yr
AQMD
Sacramento Metropolitan
SBUS 0.040438 0 Flat12Yr
AQMD
Sacramento Metropolitan
All Other Buses 0.130316 0 Flat12Yr
AQMD
T6 Public Class 4 San Diego County APCD 0.012872 0 Flat12Yr
T6 Public Class 5 San Diego County APCD 0.012872 0 Flat12Yr
T6 Public Class 6 San Diego County APCD 0.012872 0 Flat12Yr
T6 Public Class 7 San Diego County APCD 0.012872 0 Flat12Yr
T6 Instate Tractor Class 6 San Diego County APCD 0.017149 0 Flat12Yr
T6 Instate Delivery Class 4 San Diego County APCD 0.017149 0 Flat12Yr
T6 Instate Delivery Class 5 San Diego County APCD 0.017149 0 Flat12Yr
T6 Instate Delivery Class 6 San Diego County APCD 0.017149 0 Flat12Yr
T6 Instate Other Class 4 San Diego County APCD 0.017149 0 Flat12Yr
T6 Instate Other Class 5 San Diego County APCD 0.017149 0 Flat12Yr
T6 Instate Other Class 6 San Diego County APCD 0.017149 0 Flat12Yr
T6 Instate Tractor Class 7 San Diego County APCD 0.007641 0 Flat12Yr
T6 Instate Delivery Class 7 San Diego County APCD 0.007641 0 Flat12Yr
T6 Instate Other Class 7 San Diego County APCD 0.007641 0 Flat12Yr
T7 Public Class 8 San Diego County APCD 0.009151 0 Flat12Yr
T7 CAIRP Class 8 San Diego County APCD 0.000848 0 Flat12Yr
T7 Single Concrete/Transit
San Diego County APCD 0.070608 0 Flat12Yr
Mix Class 8
T7 Single Dump Class 8 San Diego County APCD 0.070608 0 Flat12Yr
T7 Single Other Class 8 San Diego County APCD 0.070608 0 Flat12Yr
T7 Tractor Class 8 San Diego County APCD 0.013263 0 Flat12Yr
SBUS San Diego County APCD 0.020093 0 Flat12Yr
All Other Buses San Diego County APCD 0.14921 0 Flat12Yr
San Joaquin Valley Unified
T6 Public Class 4 0.123221 0 Flat12Yr
APCD
San Joaquin Valley Unified
T6 Public Class 5 0.123221 0 Flat12Yr
APCD
San Joaquin Valley Unified
T6 Public Class 6 0.123221 0 Flat12Yr
APCD
San Joaquin Valley Unified
T6 Public Class 7 0.123221 0 Flat12Yr
APCD

234
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

EMFAC 2021 Vehicle Prediction

Air District Name Intercept Slope
Category Class
San Joaquin Valley Unified
T6 Utility Class 5 0.002252 0 Flat12Yr
APCD
San Joaquin Valley Unified
T6 Utility Class 6 0.002252 0 Flat12Yr
APCD
San Joaquin Valley Unified
T6 Utility Class 7 0.002252 0 Flat12Yr
APCD
San Joaquin Valley Unified
T6 Instate Tractor Class 7 0.02375 0 Flat12Yr
APCD
San Joaquin Valley Unified
T6 Instate Delivery Class 7 0.02375 0 Flat12Yr
APCD
San Joaquin Valley Unified
T6 Instate Other Class 7 0.02375 0 Flat12Yr
APCD
San Joaquin Valley Unified
T7 Public Class 8 0.115163 0 Flat12Yr
APCD
San Joaquin Valley Unified
T7 POLA Class 8 0.004902 0 Flat12Yr
APCD
T7 Single Concrete/Transit San Joaquin Valley Unified
0.035179 0 Flat12Yr
Mix Class 8 APCD
San Joaquin Valley Unified
T7 Single Dump Class 8 0.035179 0 Flat12Yr
APCD
San Joaquin Valley Unified
T7 Single Other Class 8 0.035179 0 Flat12Yr
APCD
San Joaquin Valley Unified
T7 Tractor Class 8 0.015355 0 Flat12Yr
APCD
San Joaquin Valley Unified
SBUS 0.229238 0 Flat12Yr
APCD
San Joaquin Valley Unified
All Other Buses 0.096803 0 Flat12Yr
APCD
T7 Single Concrete/Transit San Luis Obispo County
0.021498 0 Flat12Yr
Mix Class 8 APCD
San Luis Obispo County
T7 Single Dump Class 8 0.021498 0 Flat12Yr
APCD
San Luis Obispo County
T7 Single Other Class 8 0.021498 0 Flat12Yr
APCD
San Luis Obispo County
T7 Tractor Class 8 0.011624 0 Flat12Yr
APCD
San Luis Obispo County
All Other Buses 0.060606 0 Flat12Yr
APCD
Santa Barbara County
T6 Public Class 4 0.020833 0 Flat12Yr
APCD
Santa Barbara County
T6 Public Class 5 0.020833 0 Flat12Yr
APCD
Santa Barbara County
T6 Public Class 6 0.020833 0 Flat12Yr
APCD
Santa Barbara County
T6 Public Class 7 0.020833 0 Flat12Yr
APCD
Santa Barbara County
T6 Instate Tractor Class 6 0.001263 0 Flat12Yr
APCD

235
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

EMFAC 2021 Vehicle Prediction

Air District Name Intercept Slope
Category Class
Santa Barbara County
T6 Instate Delivery Class 4 0.001263 0 Flat12Yr
APCD
Santa Barbara County
T6 Instate Delivery Class 5 0.001263 0 Flat12Yr
APCD
Santa Barbara County
T6 Instate Delivery Class 6 0.001263 0 Flat12Yr
APCD
Santa Barbara County
T6 Instate Other Class 4 0.001263 0 Flat12Yr
APCD
Santa Barbara County
T6 Instate Other Class 5 0.001263 0 Flat12Yr
APCD
Santa Barbara County
T6 Instate Other Class 6 0.001263 0 Flat12Yr
APCD
Santa Barbara County
T6 Instate Tractor Class 7 0.005853 0 Flat12Yr
APCD
Santa Barbara County
T6 Instate Delivery Class 7 0.005853 0 Flat12Yr
APCD
Santa Barbara County
T6 Instate Other Class 7 0.005853 0 Flat12Yr
APCD
T7 Single Concrete/Transit Santa Barbara County
0.027401 0 Flat12Yr
Mix Class 8 APCD
Santa Barbara County
T7 Single Dump Class 8 0.027401 0 Flat12Yr
APCD
Santa Barbara County
T7 Single Other Class 8 0.027401 0 Flat12Yr
APCD
Santa Barbara County
T7 Tractor Class 8 0.004167 0 Flat12Yr
APCD
Santa Barbara County
SBUS 0.240605 0 Flat12Yr
APCD
T7 Single Concrete/Transit
Shasta County AQMD 0.024683 0 Flat12Yr
Mix Class 8
T7 Single Dump Class 8 Shasta County AQMD 0.024683 0 Flat12Yr
T7 Single Other Class 8 Shasta County AQMD 0.024683 0 Flat12Yr
T7 Tractor Class 8 Shasta County AQMD 0.007517 0 Flat12Yr
SBUS Shasta County AQMD 0.006944 0 Flat12Yr
T6 Public Class 4 Siskiyou County APCD 0.083333 0 Flat12Yr
T6 Public Class 5 Siskiyou County APCD 0.083333 0 Flat12Yr
T6 Public Class 6 Siskiyou County APCD 0.083333 0 Flat12Yr
T6 Public Class 7 Siskiyou County APCD 0.083333 0 Flat12Yr
T7 Tractor Class 8 Siskiyou County APCD 0.005952 0 Flat12Yr
T6 Public Class 4 South Coast AQMD 0.162558 0 Flat12Yr
T6 Public Class 5 South Coast AQMD 0.162558 0 Flat12Yr
T6 Public Class 6 South Coast AQMD 0.162558 0 Flat12Yr
T6 Public Class 7 South Coast AQMD 0.162558 0 Flat12Yr
T6 Utility Class 5 South Coast AQMD 0.005715 0 Flat12Yr
T6 Utility Class 6 South Coast AQMD 0.005715 0 Flat12Yr

236
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

EMFAC 2021 Vehicle Prediction

Air District Name Intercept Slope
Category Class
T6 Utility Class 7 South Coast AQMD 0.005715 0 Flat12Yr
T6 Instate Tractor Class 6 South Coast AQMD 0.007859 0 Flat12Yr
T6 Instate Delivery Class 4 South Coast AQMD 0.007859 0 Flat12Yr
T6 Instate Delivery Class 5 South Coast AQMD 0.007859 0 Flat12Yr
T6 Instate Delivery Class 6 South Coast AQMD 0.007859 0 Flat12Yr
T6 Instate Other Class 4 South Coast AQMD 0.007859 0 Flat12Yr
T6 Instate Other Class 5 South Coast AQMD 0.007859 0 Flat12Yr
T6 Instate Other Class 6 South Coast AQMD 0.007859 0 Flat12Yr
T6 Instate Tractor Class 7 South Coast AQMD 0.022221 0 Flat12Yr
T6 Instate Delivery Class 7 South Coast AQMD 0.022221 0 Flat12Yr
T6 Instate Other Class 7 South Coast AQMD 0.022221 0 Flat12Yr
T6 CAIRP Class 7 South Coast AQMD 0.000656 0 Flat12Yr
LinearGrow
T7 Public Class 8 South Coast AQMD -20.9769 0.010634
th
T7 CAIRP Class 8 South Coast AQMD 0.003645 0 Flat12Yr
T7 POAK Class 8 South Coast AQMD 0.002525 0 Flat12Yr
T7 POLA Class 8 South Coast AQMD 0.039228 0 Flat12Yr
T7 Single Concrete/Transit
South Coast AQMD 0.072393 0 Flat12Yr
Mix Class 8
T7 Single Dump Class 8 South Coast AQMD 0.072393 0 Flat12Yr
T7 Single Other Class 8 South Coast AQMD 0.072393 0 Flat12Yr
T7 Tractor Class 8 South Coast AQMD 0.026748 0 Flat12Yr
SBUS South Coast AQMD 0.724926 0 Flat12Yr
All Other Buses South Coast AQMD 0.249283 0 Flat12Yr
T7 Tractor Class 8 Tehama County APCD 0.005055 0 Flat12Yr
T7 Public Class 8 Ventura County APCD 0.041667 0 Flat12Yr
T7 CAIRP Class 8 Ventura County APCD 0.002506 0 Flat12Yr
T7 Single Concrete/Transit
Ventura County APCD 0.032515 0 Flat12Yr
Mix Class 8
T7 Single Dump Class 8 Ventura County APCD 0.032515 0 Flat12Yr
T7 Single Other Class 8 Ventura County APCD 0.032515 0 Flat12Yr
T7 Tractor Class 8 Ventura County APCD 0.002381 0 Flat12Yr
SBUS Ventura County APCD 0.037738 0 Flat12Yr
All Other Buses Ventura County APCD 0.083333 0 Flat12Yr
T6 Instate Tractor Class 6 Yolo/Solano AQMD 0.012401 0 Flat12Yr
T6 Instate Delivery Class 4 Yolo/Solano AQMD 0.012401 0 Flat12Yr
T6 Instate Delivery Class 5 Yolo/Solano AQMD 0.012401 0 Flat12Yr
T6 Instate Delivery Class 6 Yolo/Solano AQMD 0.012401 0 Flat12Yr
T6 Instate Other Class 4 Yolo/Solano AQMD 0.012401 0 Flat12Yr
T6 Instate Other Class 5 Yolo/Solano AQMD 0.012401 0 Flat12Yr
T6 Instate Other Class 6 Yolo/Solano AQMD 0.012401 0 Flat12Yr

237
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

EMFAC 2021 Vehicle Prediction

Air District Name Intercept Slope
Category Class
T7 Public Class 8 Yolo/Solano AQMD 0.004386 0 Flat12Yr
T7 Single Concrete/Transit
Yolo/Solano AQMD 0.045094 0 Flat12Yr
Mix Class 8
T7 Single Dump Class 8 Yolo/Solano AQMD 0.045094 0 Flat12Yr
T7 Single Other Class 8 Yolo/Solano AQMD 0.045094 0 Flat12Yr
T7 Tractor Class 8 Yolo/Solano AQMD 0.022099 0 Flat12Yr
All Other Buses Yolo/Solano AQMD 0.144792 0 Flat12Yr

238
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

6.5 Dynamometer Test Data of Light Heavy-Duty Trucks from

CARB Project 2R1702
Test Test Odo Test HC CO NOx PM CO2
Fuel
Veh ID Vehicle (mi) Cycle (g/mi) (g/mi) (g/mi) (mg/mi) (g/mi)
FTP-75 0.014 0.005 0.05 1.4 652
MAC1 0.022 0.025 0.63 7.3 1,351
2015 MAC3 0.002 0.007 0.10 3.1 518
Dodge MFC5 0.001 0.005 0.09 2.9 506
1 Diesel 51,380
Ram MFC6 0.002 0.016 0.08 3.6 529
2500 MFC7 0.001 0.008 0.17 2.6 570
HWFET 0.004 0.003 0.02 1.7 417
UC 0.004 0.007 0.17 2.3 663
FTP-75 0.007 0.015 0.03 2.1 604
MAC1 0.011 0.015 0.06 3.5 1,026
2017 MAC3 0.008 0.010 0.02 3.5 504
Daimler MFC5 0.003 0.013 0.06 1.8 547
2 Diesel 22,860
Sprinter MFC6 0.000 0.015 0.08 2.1 592
2500 MFC7 0.000 0.021 0.11 1.6 667
HWFET 0.005 0.007 0.06 2.2 482
UC 0.039 0.047 0.14 54.4 703
FTP-75 0.044 0.140 0.56 17.3 912
MAC1 0.149 1.606 0.57 6.5 1,532
MAC3 0.012 0.097 0.13 3.0 525
2015 GM MFC5 0.008 0.069 0.11 1.7 479
3 Silverado Diesel 64,600
2500 MFC6 0.004 0.016 0.09 3.2 502
MFC7 0.004 0.007 0.10 1.7 552
HWFET 0.005 0.004 0.27 1.3 411
UC 0.038 0.237 0.28 8.7 744
FTP-75 0.081 0.029 0.27 2.5 723
MAC1 0.044 0.131 0.77 1.8 1,176
MAC3 0.006 0.213 0.08 1.9 470
2015 MFC5 0.041 0.071 0.34 1.4 486
4 Ford Diesel 30,460
F250 MFC6 0.005 0.053 0.18 1.7 505
MFC7 0.004 0.026 0.18 1.5 553
HWFET 0.008 0.032 0.01 1.7 396
UC 0.031 0.051 0.28 1.7 612
FTP-75 0.019 0.016 0.16 3.6 661
2015 MAC1 0.029 0.387 2.34 2.6 1,289
5 Ford Diesel 70,630
F350 MAC3 0.006 0.020 0.24 1.9 528
MFC5 0.002 0.023 0.63 1.4 553

239
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Test Test Odo Test HC CO NOx PM CO2

Fuel
Veh ID Vehicle (mi) Cycle (g/mi) (g/mi) (g/mi) (mg/mi) (g/mi)
MFC6 0.001 0.018 0.81 1.3 598
MFC7 0.001 0.021 0.41 1.9 664
HWFET 0.080 0.052 0.71 8.6 528
UC 0.159 0.088 1.15 2.6 787
FTP-75 0.004 0.012 0.02 0.9 565
2017 MAC1 0.010 0.061 0.14 2.2 955
Daimler
7 Diesel 8,630 MAC3 0.002 0.017 0.08 1.3 480
Sprinter
3500 MFC7 0.000 0.023 0.32 1.1 645
UC 0.002 0.011 0.06 1.7 623
FTP-75 0.010 0.010 0.01 1.0 801
MAC1 0.028 0.025 0.36 13.0 1,508
2015 MAC3 0.005 0.188 0.02 2.0 606
Dodge MFC5 0.001 0.017 0.01 1.5 632
8 Diesel 139,340
Ram MFC6 0.084 0.024 0.73 227 776
3500 MFC7 0.000 0.023 0.01 2.5 729
HWFET 0.002 0.011 0.01 1.0 542
UC 0.028 0.072 0.08 2.1 832
FTP-75 0.007 0.014 0.01 0.3 766
MAC1 0.050 1.683 0.01 3.3 1,412
MAC3 0.006 0.159 0.01 0.4 535
2015 GM MFC5 0.006 0.179 0.01 0.7 526
9 Silverado Gasoline 42,400
2500 MFC6 0.009 0.910 0.03 0.7 544
MFC7 0.007 0.619 0.02 0.9 604
HWFET 0.003 0.271 0.01 0.4 456
UC 0.008 0.404 0.02 1.0 700
FTP-75 0.005 0.010 0.02 0.3 765
MAC1 0.027 0.327 0.02 2.4 1,431
MAC3 0.006 0.145 0.02 1.8 543
2015 GM MFC5 0.006 0.444 0.03 2.5 544
10 Sierra Gasoline 43,110
2500 MFC6 0.009 0.818 0.04 2.1 533
MFC7 0.006 0.425 0.02 2.5 580
HWFET 0.001 0.046 0.01 1.0 436
UC 0.006 0.334 0.02 2.2 714
FTP-75 0.04 0.24 0.01 3.5 862
MAC1 0.04 0.26 0.50 15.3 1,492
MAC3 0.00 0.15 0.09 0.4 551
2015 GM
11 Diesel 24,570 MFC5 0.01 0.08 0.06 1.4 556
Duramax
MFC6 0.00 0.02 0.33 2.1 587
MFC7 0.00 0.02 0.27 0.8 606
HWFET 0.00 0.04 0.02 0.8 458

240
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Test Test Odo Test HC CO NOx PM CO2

Fuel
Veh ID Vehicle (mi) Cycle (g/mi) (g/mi) (g/mi) (mg/mi) (g/mi)
UC 0.05 0.38 0.69 1.5 919
FTP-75 0.342 2.324 2.52 87.5 752
MAC1 0.553 4.621 5.59 241 1,496
MAC3 0.131 1.078 2.60 126 544
2006 MFC5 0.123 0.581 2.53 77.3 529
12 Ford Diesel 73,800
F250 MFC6 0.117 0.468 2.63 81.9 564
MFC7 0.119 0.461 3.07 91.1 616
HWFET 0.108 0.369 1.81 66.9 454
UC 0.190 0.954 3.22 144 731
FTP-75 0.353 1.813 4.73 139 671
MAC1 0.731 3.842 6.26 427 1,267
MAC3 0.215 1.159 3.13 87.0 478
2006 GM MFC5 0.184 0.723 3.29 53.6 455
13 Silverado Diesel 120,810
2500 MFC6 0.165 0.714 2.78 71.7 487
MFC7 0.171 0.809 2.29 114 560
HWFET 0.173 0.678 3.15 70.8 404
UC 0.308 1.282 3.80 151 642

241
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

6.6 Dynamometer Test Data of Heavy Heavy-Duty Diesel Trucks

from CARB Truck and Bus Surveillance Program
Test Engine Engine Odometer THC CO NOx PM CO2
Test Cycle
Vehicle Make MY (mi) (g/mi) (g/mi) (g/mi) (mg/mi) (g/mi)
UDDS 0.008 0.509 0.96 0.5 1,978
HHDT Creep 0.521 7.226 10.53 0.5 4,911
Drayage Near
0.078 1.244 4.12 2.0 2,127
K-1 Paccar 2013 144,683 Dock
Drayage Local 0.018 0.398 1.38 0.9 2,074
HHDT Cruise 0.005 0.025 0.17 1.3 1,261
ARB HS Cruise - - - - -
UDDS 0.008 0.082 2.65 8.6 2,005
HHDT Creep 0.627 6.808 21.91 8.5 4,378
Drayage Near
0.225 0.859 13.88 2.8 2,148
K-2 Paccar 2013 180,598 Dock
Drayage Local 0.102 0.320 7.87 2.0 2,028
HHDT Cruise 0.005 0.012 0.47 12.3 1,365
ARB HS Cruise - - - - -
UDDS 0.007 0.394 0.55 3.7 2,096
HHDT Creep 0.238 8.346 13.63 4.0 4,116
Drayage Near
0.050 1.036 2.87 2.4 2,379
K-3 Paccar 2013 248,095 Dock
Drayage Local 0.022 0.321 1.61 2.5 2,122
HHDT Cruise 0.004 0.037 0.22 6.7 1,367
ARB HS Cruise - - - - -
UDDS 0.017 0.003 4.92 1.7 2,048
HHDT Creep 0.270 0.518 11.33 0.0 7,908
Drayage Near
Cummin 0.079 0.013 4.11 2.1 2,949
L-1 2013 66,145 Dock
s
Drayage Local 0.059 0.003 5.22 3.1 2,303
HHDT Cruise 0.010 0.003 1.25 2.4 1,226
ARB HS Cruise 0.009 0.015 0.86 18.9 1,411
UDDS 0.022 0.014 9.65 4.9 2,098
HHDT Creep 0.309 0.060 22.32 12.4 5,762
Drayage Near
Cummin 0.137 0.033 10.70 3.5 2,511
L-2 2013 171,974 Dock
s
Drayage Local 0.082 0.077 9.42 4.0 2,383
HHDT Cruise 0.013 0.024 1.95 5.4 1,367
ARB HS Cruise 0.009 0.032 1.57 29.8 1,609
UDDS 0.021 0.043 6.01 7.1 2,217
HHDT Creep 0.355 0.119 16.62 11.1 6,790
Cummin
L-3 2013 336,120 Drayage Near
s 0.110 0.063 7.91 4.0 2,853
Dock
Drayage Local 0.038 0.043 5.70 3.2 2,461

242
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Test Engine Engine Odometer THC CO NOx PM CO2

Test Cycle
Vehicle Make MY (mi) (g/mi) (g/mi) (g/mi) (mg/mi) (g/mi)
HHDT Cruise 0.009 0.016 1.45 4.3 1,386
ARB HS Cruise 0.007 0.018 1.43 16.9 1,641
UDDS 0.009 0.439 4.02 4.1 2,120
HHDT Creep 0.193 8.693 38.65 2.8 4,319
Drayage Near
0.044 1.364 12.31 1.3 2,297
M-1 Volvo 2014 187,291 Dock
Drayage Local 0.029 0.520 10.60 1.5 2,173
HHDT Cruise 0.019 0.094 1.23 4.8 1,358
ARB HS Cruise 0.011 0.087 1.32 9.5 1,519
UDDS 0.046 0.494 9.63 2.7 2,025
HHDT Creep 0.332 11.541 43.82 2.7 3,792
Drayage Near
0.139 1.686 15.43 2.0 1,994
M-2 Volvo 2014 370,454 Dock
Drayage Local 0.127 1.032 15.59 3.1 2,092
HHDT Cruise - - - - -
ARB HS Cruise 0.020 0.099 6.04 - 1,443
UDDS 0.005 1.075 4.23 22.4 2,032
HHDT Creep 0.125 5.550 34.90 4.2 3,701
Drayage Near
0.029 1.640 9.56 24.8 2,061
M-3 Volvo 2014 72,055 Dock
Drayage Local 0.012 1.363 7.22 9.6 1,971
HHDT Cruise 0.005 0.108 1.01 27.1 1,325
ARB HS Cruise 0.003 0.105 0.84 16.5 1,470
UDDS 0.009 0.032 1.58 1.2 2,019
HHDT Creep 0.173 3.309 23.84 2.9 5,210
Drayage Near
0.042 0.296 6.51 1.8 2,029
N-1 DDC 2014 240,785 Dock
Drayage Local 0.023 0.053 3.42 0.7 1,928
HHDT Cruise 0.008 0.031 0.48 0.6 1,202
ARB HS Cruise 0.004 0.027 0.63 3.6 1,427
UDDS 0.006 0.004 1.02 2.3 2,072
HHDT Creep 0.124 0.451 17.77 0.0 4,550
Drayage Near
0.037 0.155 6.24 1.1 2,297
N-2 DD15 2014 177,394 Dock
Drayage Local 0.013 0.026 2.71 4.0 2,055
HHDT Cruise 0.005 0.009 0.29 1.1 1,300
ARB HS Cruise 0.002 0.004 0.16 4.9 1,482
UDDS 0.004 0.006 0.46 3.4 2,003
HHDT Creep 0.083 0.786 13.42 11.6 4,021
Drayage Near
N-3 DD15 2014 13,840 0.021 0.016 3.33 2.3 2,012
Dock
Drayage Local 0.010 0.009 1.75 1.6 1,874
HHDT Cruise 0.002 0.009 0.17 2.1 1,289

243
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Test Engine Engine Odometer THC CO NOx PM CO2

Test Cycle
Vehicle Make MY (mi) (g/mi) (g/mi) (g/mi) (mg/mi) (g/mi)
ARB HS Cruise 0.001 0.019 0.27 0.0 1,441
UDDS 0.020 0.028 3.16 12.9 2,170
HHDT Creep 0.413 0.488 8.19 4.5 9,864
Drayage Near
Cummin 0.055 0.084 1.61 8.0 3,622
O-1 2014 144,194 Dock
s
Drayage Local 0.043 0.156 1.60 5.4 3,059
HHDT Cruise 0.010 0.001 0.63 4.0 1,309
ARB HS Cruise 0.008 0.013 0.23 67.8 1,469
UDDS 0.019 0.011 3.96 6.6 2,114
HHDT Creep 0.303 0.000 14.17 4.9 8,969
Drayage Near
Cummin 0.055 0.001 3.11 3.8 3,347
O-2 2014 185,078 Dock
s
Drayage Local 0.094 0.165 6.50 2.7 2,577
HHDT Cruise 0.018 0.056 2.00 2.6 1,330
ARB HS Cruise 0.014 0.033 1.59 27.6 1,464
UDDS 0.020 0.014 4.73 10.6 2,317
HHDT Creep 0.263 0.001 11.48 6.6 8,796
Drayage Near
Cummin 0.056 0.001 3.65 9.9 3,411
O-3 2014 112,134 Dock
s
Drayage Local 0.038 0.002 2.68 6.3 2,825
HHDT Cruise 0.010 0.011 1.08 5.7 1,358
ARB HS Cruise 0.007 0.023 0.83 41.0 1,564
UDDS 0.008 0.023 0.36 9.0 2,121
HHDT Creep 0.387 5.343 9.32 7.9 5,094
Drayage Near
0.078 0.224 2.47 6.7 2,261
P-1 Navistar 2014 132,796 Dock
Drayage Local 0.038 0.482 0.81 4.1 2,178
HHDT Cruise 0.003 0.023 0.14 5.5 1,374
ARB HS Cruise - - - - -
UDDS 0.015 0.005 1.30 6.2 2,143
HHDT Creep 1.151 1.759 14.73 9.4 5,530
Drayage Near
0.175 0.282 4.39 3.0 2,342
P-2 Navistar 2014 179,350 Dock
Drayage Local 0.073 0.800 2.19 4.9 2,234
HHDT Cruise 0.008 0.066 0.50 5.1 1,371
ARB HS Cruise 0.007 0.011 0.28 38.0 1,648
UDDS 0.003 0.021 0.95 6.4 2,012
HHDT Creep 0.116 5.256 19.07 3.9 3,669
Drayage Near
0.064 0.137 4.33 4.8 1,924
Q-2 DDC 2015 219,059 Dock
Drayage Local 0.014 0.165 2.29 4.2 1,911
HHDT Cruise 0.006 0.027 0.32 4.3 1,322
ARB HS Cruise 0.001 0.011 0.26 14.7 1,623

244
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Test Engine Engine Odometer THC CO NOx PM CO2

Test Cycle
Vehicle Make MY (mi) (g/mi) (g/mi) (g/mi) (mg/mi) (g/mi)
UDDS 0.004 1.109 0.20 3.3 1,885
HHDT Creep 0.208 2.575 9.28 10.3 4,359
Drayage Near
0.065 1.086 3.52 6.9 1,892
R-1 Paccar 2014 290,981 Dock
Drayage Local 0.014 1.203 1.61 5.8 1,856
HHDT Cruise 0.001 0.137 0.05 2.5 1,429
ARB HS Cruise 0.001 0.017 0.01 2.6 1,691
UDDS 0.006 0.774 0.58 5.5 1,588
HHDT Creep 0.672 13.891 8.04 12.3 3,766
Drayage Near
0.088 0.993 3.24 8.1 1,953
R-2 Paccar 2014 275,565 Dock
Drayage Local 0.026 0.331 1.54 4.0 1,706
HHDT Cruise 0.003 0.120 0.07 1.4 1,012
ARB HS Cruise 0.003 0.000 0.15 7.3 1,154
UDDS 0.005 0.372 0.71 4.8 2,056
HHDT Creep 0.140 1.041 9.87 12.9 5,158
Drayage Near
0.030 0.376 4.54 7.7 2,087
R-3 Paccar 2014 234,326 Dock
Drayage Local 0.015 0.263 2.32 2.5 1,974
HHDT Cruise 0.004 0.028 0.11 1.7 1,430
ARB HS Cruise 0.002 0.001 0.06 1.3 1,578
UDDS 0.004 0.183 2.66 7.4 1,916
HHDT Creep 0.111 1.741 23.46 8.4 3,941
Drayage Near
0.017 0.640 6.58 5.6 2,024
V1-1 Volvo 2015 308,919 Dock
Drayage Local 0.014 0.359 6.13 1.0 1,968
HHDT Cruise 0.011 0.021 0.94 19.6 1,198
ARB HS Cruise 0.002 0.025 0.19 2.8 1,231
UDDS 0.007 0.893 6.33 2.9 2,013
HHDT Creep 0.098 3.439 25.32 2.6 3,655
Drayage Near
0.034 1.263 14.16 2.9 2,147
Dock
V1-2 Volvo 2015 511,406
Drayage Local 0.018 0.385 12.21 1.2 2,001
HHDT Cruise 0.012 0.082 1.53 1.8 1,315
Modified HS
0.007 0.263 1.75 n/a 1,484
Cruise
UDDS 0.004 0.338 3.76 1.5 1,647
HHDT Creep 0.072 0.854 52.59 7.9 6,418
Drayage Near
0.017 0.525 17.75 3.5 2,350
Dock
V1-3 Volvo 2015 128,370
Drayage Local 0.008 0.180 10.57 3.0 1,954
HHDT Cruise 0.002 0.069 0.74 0.7 927
Modified HS
0.003 0.194 0.90 2.0 986
Cruise

245
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Test Engine Engine Odometer THC CO NOx PM CO2

Test Cycle
Vehicle Make MY (mi) (g/mi) (g/mi) (g/mi) (mg/mi) (g/mi)
UDDS 0.005 0.012 3.06 13.4 2,247
HHDT Creep 0.240 0.041 14.04 14.6 6,849
Drayage Near
Cummin 0.094 0.007 6.07 9.6 2,719
V2-1 2015 241,304 Dock
s
Drayage Local 0.031 0.002 1.87 5.4 2,531
HHDT Cruise 0.006 0.013 0.36 18.2 1,603
ARB HS Cruise 0.004 0.016 0.33 26.7 1,959
UDDS 0.017 0.130 2.78 8.0 2,091
HHDT Creep 0.282 0.199 14.60 8.3 5,318
Drayage Near
Cummin 0.150 0.234 7.66 5.0 2,547
V2-2 2015 454,320 Dock
s
Drayage Local 0.076 0.120 4.96 3.4 2,253
HHDT Cruise 0.014 0.046 0.44 27.7 1,357
ARB HS Cruise 0.007 0.060 0.48 79.3 1,595
UDDS 0.004 0.024 0.43 3.3 2,154
HHDT Creep 0.060 0.393 10.47 3.9 3,935
Drayage Near
0.026 0.096 3.64 2.2 2,079
Dock
V3-1 DDC 2014 134,539
Drayage Local 0.020 0.038 2.16 2.3 2,044
HHDT Cruise 0.007 0.024 0.52 3.8 1,544
Modified HS
0.002 0.019 0.48 9.9 1,739
Cruise
UDDS 0.004 0.040 0.94 0.8 2,003
HHDT Creep 0.160 3.683 16.10 0.6 3,470
Drayage
0.035 0.282 5.43 n/a 2,068
NearDock
V3-2 DDC 2015 333,687
Drayage Local 0.025 0.076 2.76 0.4 1,799
HHDT Cruise 0.010 0.018 0.31 0.2 1,235
Modified HS
0.005 0.013 0.40 2.9 1,379
Cruise
UDDS 0.004 0.098 0.12 3.1 1,981
HHDT Creep 0.171 0.911 10.04 9.6 4,054
Drayage
0.051 0.351 3.54 2.1 2,084
NearDock
V4-1-2 Paccar 2015 194,575
Drayage Local 0.014 0.140 1.38 3.2 1,945
HHDT Cruise 0.001 0.008 0.01 1.6 1,307
Modified HS
0.001 0.004 0.05 1.9 1,480
Cruise
UDDS 0.004 0.052 0.31 2.0 1,840
HHDT Creep 0.680 3.949 17.32 71.8 4,321
Drayage
V4-2 Paccar 2015 128,288 0.084 0.405 3.91 2.5 1,872
NearDock
Drayage Local 0.018 0.190 1.94 0.7 1,786
HHDT Cruise 0.001 0.010 0.02 1.4 1,344

246
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Test Engine Engine Odometer THC CO NOx PM CO2

Test Cycle
Vehicle Make MY (mi) (g/mi) (g/mi) (g/mi) (mg/mi) (g/mi)
Modified HS
0.001 0.014 0.03 3.1 1,495
Cruise
UDDS 0.008 0.028 0.87 3.0 2,001
HHDT Creep 0.824 1.402 16.98 8.9 4,875
Drayage
0.121 0.198 4.63 33.4 2,484
NearDock
V6-1 Navistar 2016 13,769
Drayage Local 0.036 0.080 2.96 1.1 2,054
HHDT Cruise 0.002 0.003 0.34 0.9 1,314
Modified HS
0.003 0.002 0.25 1.7 1,464
Cruise
UDDS 0.004 0.291 0.36 2.0 1,871
HHDT Creep 0.189 2.137 15.13 0.0 4,442
Drayage
0.052 0.280 3.74 2.6 1,971
NearDock
V7-1 Paccar 2016 149,709
Drayage Local 0.016 0.296 2.00 1.4 1,899
HHDT Cruise 0.003 0.009 0.03 1.7 1,368
Modified HS
0.002 0.021 0.03 2.7 1,511
Cruise
UDDS 0.004 0.210 0.25 3.1 2,046
HHDT Creep 0.740 7.491 17.58 4.4 4,630
Drayage
0.085 0.793 4.67 5.8 2,036
NearDock
V7-3 Paccar 2016 212,460
Drayage Local 0.016 0.302 1.77 2.3 1,930
HHDT Cruise 0.001 0.035 0.04 1.5 1,577
Modified HS
0.001 0.002 0.06 3.9 1,765
Cruise
UDDS 0.018 0.043 2.55 4.2 1,719
HHDT Creep 0.247 0.010 6.60 24.8 9,171
Drayage
0.080 0.043 4.91 3.5 2,644
Cummin NearDock
V8-1 2016 98,594
s Drayage Local 0.029 0.087 3.14 1.9 2,294
HHDT Cruise 0.009 0.012 0.70 1.2 1,085
Modified HS
0.008 0.005 1.17 2.6 1,162
Cruise
UDDS 0.007 0.006 3.17 8.6 2,315
HHDT Creep 0.314 0.011 10.95 28.1 9,275
Drayage
0.075 0.018 4.11 5.0 3,345
Cummin NearDock
V8-2 2016 62,107
s Drayage Local 0.035 0.018 2.74 5.9 2,766
HHDT Cruise 0.003 0.010 0.36 6.9 1,624
Modified HS
0.003 0.009 0.37 34.6 1,823
Cruise
UDDS 0.014 0.138 4.05 2.3 2,053

V9-1 Volvo 2018 54,343 HHDT Creep 0.550 0.673 24.02 7.2 4,373
Drayage
0.140 0.223 9.67 2.7 2,230
NearDock

247
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Test Engine Engine Odometer THC CO NOx PM CO2

Test Cycle
Vehicle Make MY (mi) (g/mi) (g/mi) (g/mi) (mg/mi) (g/mi)
Drayage Local 0.044 0.058 6.16 1.5 2,153
HHDT Cruise 0.008 0.000 1.05 0.9 1,311
Modified HS
0.008 0.011 2.98 3.6 1,542
Cruise
UDDS 0.001 0.007 0.11 1.6 2,072
HHDT Creep 0.118 1.098 22.67 4.2 4,026
Drayage
0.068 0.393 4.87 8.2 2,076
NearDock
V10-1 DDC 2016 284,928
Drayage Local 0.023 0.338 1.83 12.0 2,020
HHDT Cruise 0.005 0.012 0.20 2.5 1,523
Modified HS
0.001 0.012 0.17 4.4 1,701
Cruise
UDDS 0.005 0.027 0.33 n/a 2,123
HHDT Creep 0.139 6.500 22.23 n/a 3,664
Drayage
0.049 0.354 5.14 n/a 2,027
NearDock
V10-2 DDC 2016 120,588
Drayage Local 0.017 0.279 1.55 n/a 1,945
HHDT Cruise 0.002 0.007 0.22 n/a 1,532
Modified HS
0.002 0.010 0.20 n/a 1,696
Cruise
UDDS 0.008 0.114 0.19 3.7 1,925
HHDT Creep 0.087 1.023 14.96 0.8 3,177
Drayage
0.067 0.205 5.18 1.6 1,870
NearDock
V11-1 DDC 2017 170,529
Drayage Local 0.032 0.102 2.33 1.2 1,814
HHDT Cruise 0.004 0.013 0.30 1.4 1,199
Modified HS
0.002 0.007 0.10 2.5 1,319
Cruise
UDDS 0.002 0.642 3.47 1.8 1,661
HHDT Creep 0.051 0.604 39.98 2.3 5,950
Drayage
0.007 0.417 16.24 6.1 2,437
NearDock
V12-1 Volvo 2016 101,767
Drayage Local 0.004 0.116 7.34 2.2 1,958
HHDT Cruise 0.001 0.026 0.70 1.7 959
Modified HS
0.001 0.340 0.28 3.2 1,053
Cruise
UDDS 0.005 0.000 1.05 1.8 1,100
HHDT Creep 0.064 0.614 9.43 5.5 3,029
Drayage
0.025 0.140 3.25 1.1 1,275
NearDock
V13-1 HINO 2013 151,150
Drayage Local 0.016 0.009 2.47 0.3 1,155
HHDT Cruise 0.006 0.000 0.34 0.1 780
Modified HS
0.005 0.000 0.24 1.3 855
Cruise
Cummin UDDS 0.024 0.021 1.92 3.5 1,999
V15-1 2019 24,228
s HHDT Creep 0.499 0.399 11.66 23.3 4,295

248
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Test Engine Engine Odometer THC CO NOx PM CO2

Test Cycle
Vehicle Make MY (mi) (g/mi) (g/mi) (g/mi) (mg/mi) (g/mi)
Drayage
0.359 0.102 5.71 5.7 2,143
NearDock
Drayage Local 0.170 0.054 3.58 3.4 1,969
HHDT Cruise 0.015 0.016 0.27 7.2 1,454
Modified HS
0.011 0.018 0.43 13.4 1,638
Cruise

249
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

6.7 Dynamometer Test Data of Medium Heavy-Duty Diesel Trucks

from CARB Surveillance Program for On-Road Class 4-6
Heavy-Duty Vehicles
Test Engine Engine Odometer THC CO NOx PM CO2
Test Cycle
Vehicle Make MY (mi) (g/mi) (g/mi) (g/mi) (mg/mi) (g/mi)
UDDS 0.005 0.115 0.46 0.7 1,153
HHDDT Creep 0.810 0.390 8.57 5.0 3,289
Parcel Delivery 0.003 0.130 0.89 2.4 1,538
Veh 1 Cummins 2014 143,200 HHDDT Transient 0.000 0.111 0.95 0.9 1,100
Local Cycle 0.000 0.055 0.56 2.5 989
HHDDT Cruise 0.000 0.052 0.31 0.7 821
Modified HS Cruise 0.001 0.043 0.52 2.1 948
UDDS 0.000 0.079 0.01 2.2 1,135
HHDDT Creep 0.285 0.487 6.18 0.3 3,719
Parcel Delivery 0.002 0.150 0.10 1.6 1,528
Veh 2 Cummins 2015 58,475 HHDDT Transient 0.000 0.091 0.03 1.2 1,081
Local Cycle 0.000 0.036 0.03 2.1 957
HHDDT Cruise 0.000 0.052 0.02 0.6 820
Modified HS Cruise 0.000 0.040 0.04 0.9 907
UDDS 0.000 0.084 0.05 1.1 1,142
HHDDT Creep 0.506 0.478 6.41 1.0 3,357
Parcel Delivery 0.000 0.160 0.11 1.1 1,517
Veh 3 Cummins 2015 92,914 HHDDT Transient 0.000 0.101 0.01 - 1,048
Local Cycle 0.001 0.054 0.05 1.5 978
HHDDT Cruise 0.000 0.061 0.01 0.3 826
Modified HS Cruise 0.000 0.046 0.03 0.8 954
UDDS 0.000 0.114 1.01 1.5 862
HHDDT Creep 0.000 1.420 9.36 5.1 2,499
Parcel Delivery 0.007 0.176 1.61 0.7 1,135
Veh 4 Ford 2014 69,309 HHDDT Transient 0.022 0.114 1.37 0.1 790
Local Cycle 0.002 0.062 0.24 5.3 774
HHDDT Cruise 0.002 0.066 0.05 1.5 695
Modified HS Cruise 0.000 0.047 0.18 9.1 778
UDDS 0.010 0.006 0.02 3.9 1,129
Creep 0.699 0.051 7.11 6.6 3,185
Drayage Near Dock 0.207 0.020 2.47 4.5 1,386
Veh 5 Cummins 2017 155,537
Drayage Local 0.069 0.011 1.09 2.7 1,337
Cruise 0.007 0.006 0.05 3.4 777
Modified HS Cruise 0.007 0.008 0.07 6.7 857
VJ-1 Isuzu 2013 96,562 UDDS 0.002 0.642 3.47 1.8 1,661

250
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Test Engine Engine Odometer THC CO NOx PM CO2

Test Cycle
Vehicle Make MY (mi) (g/mi) (g/mi) (g/mi) (mg/mi) (g/mi)
Creep 0.051 0.604 39.98 2.3 5,950
Drayage Near Dock 0.007 0.417 16.24 6.1 2,437
Drayage Local 0.004 0.116 7.34 2.2 1,958
Cruise 0.001 0.026 0.70 1.7 959
Modified HS Cruise 0.001 0.340 0.28 3.2 1,053
UDDS 0.005 0.000 1.05 1.8 1,100
Creep 0.064 0.614 9.43 5.5 3,029
Drayage Near Dock 0.025 0.140 3.25 1.1 1,275
V13-1 Hino 2013 151,150
Drayage Local 0.016 0.009 2.47 0.3 1,155
Cruise 0.006 0.000 0.34 0.1 780
Modified HS Cruise 0.005 0.000 0.24 1.3 855
UDDS 0.011 0.012 0.49 3.3 1,960
Creep 0.069 0.039 3.69 20.6 6,169
Drayage Near Dock 0.027 0.019 0.98 7.3 2,459
V16-1 Cummins 2017 29,204
Drayage Local 0.012 0.017 0.68 5.2 2,121
Cruise 0.003 0.008 0.37 2.1 1,585
Modified HS Cruise 0.004 0.009 0.39 4.3 1,764

251
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

6.8 Mapping of CEC Vehicle Classes to EMFAC Vehicle Classes

As mentioned in Section 0, CEC’s 18 vehicle classes (e.g., Car-Compact and Cross/Ut-
midsize) are mapped to EMFAC vehicle classes (i.e., LDA, LDT1, LDT2, and MDV),
utilizing 2019 DMV registration data as well as EMFAC and CEC’s classification guides.
Tables 1 shows the mapping for light-duty vehicles (sum of all fuel types) and table 2
shows the mapping of ZEVs in the base year 2019. The mapping for all fuel types and
for ZEVs are found to be very different for three CEC vehicle classes (i.e., cross/ut-
midsize, cross/ut-small-trk, and sport/ut-compact) in the base year. This is because
there are few ZEV models for these classes in the market and how they are classified
by EMFAC and CEC determines the result of the mapping. As the ZEV market is still
growing and introducing more models, ZEV’s vehicle class mapping is expected to be
closer to the mapping of conventional vehicles. Therefore, staff assumes that the ZEV’s
and LDV (of all fuel types)’s mapping for these 3 categories reach the same in 2030.
When mapping ZEV new sales from CEC vehicle classes to EMFAC vehicle classes,
their percentages were interpolated between 2019 (following the mapping for ZEV)
and 2030 (following the mapping for LDV, the sum of all fuel types).
Table 1. The percentage of CEC vehicle classes that are classified into different
EMFAC light-duty vehicle classes for the sum of all fuel types in the base year
2019.
Type LDA LDT1 LDT2 MDV
Car-Subcompact 100% 0% 0% 0%
Car-Compact 100% 0% 0% 0%
Car-Midsize 100% 0% 0% 0%
Car-Large 100% 0% 0% 0%
Car-Sport 100% 0% 0% 0%
Cross/Ut-Small-Car 100% 0% 0% 0%
Cross/Ut-Small-Trk 22% 1% 63% 14%
Cross/Ut-Midsize 36% 0% 42% 22%
Sport/Ut-Compact 13% 19% 59% 9%
Sport/Ut-Midsize 0% 0% 64% 36%
Sport/Ut-Large 19% 0% 0% 81%
Van-Compact 0% 14% 79% 7%
Van-Std 0% 0% 0% 100%
Van-Heavy 0% 0% 0% 17%
Pickup-Compact 0% 0% 100% 0%
Pickup-Std 0% 0% 46% 54%
Pickup-Heavy 0% 0% 0% 0%

252
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Table 2. The percentage of CEC vehicle classes that can be classified into EMFAC
light-duty vehicle classes for zero-emission vehicles in the base year 2019.
Type LDA LDT1 LDT2 MDV
Car-Subcompact 100% 0% 0% 0%
Car-Compact 100% 0% 0% 0%
Car-Midsize 100% 0% 0% 0%
Car-Large 100% 0% 0% 0%
Car-Sport 100% 0% 0% 0%
Cross/Ut-Small-Car 100% 0% 0% 0%
Cross/Ut-Small-Trk 36% 0% 61% 2%
Cross/Ut-Midsize 100% 0% 0% 0%
Sport/Ut-Compact 100% 0% 0% 0%
Van-Compact 0% 0% 0% 100%

253
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

6.9 Heavy Duty VMT Distribution by Hour Figures

Figure (a). VMT distribution by hour of Out-of-state HD Vehicles (including T6
CAIRP, T6 OOS, T7 CAIRP, T7 NOOS, T7 NNOOS, Motor Coach) in EMFAC2021
and EMFAC2017

Out-of-state HD Vehicles
10%
EMFAC2021
9%
EMFAC2017
Frequency of VMT

8%
7%
6%
5%
4%
3%
2%
1%
0%
0 1 2 3 4 5 6 7 8 9 101112131415161718192021222324
Hour of Day

Figure (b). VMT distribution by hour of SBUS in EMFAC2021 and EMFAC2017

SBUS
18%
EMFAC2021
16%
EMFAC2017
Frequency of VMT

14%
12%
10%
8%
6%
4%
2%
0%
0 1 2 3 4 5 6 7 8 9 101112131415161718192021222324
Hour of Day

254
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure (c). VMT distribution by hour of T6 Instate Delivery in EMFAC2021 and

EMFAC2017

T6 Instate Delivery
16%
EMFAC2021
14%
EMFAC2017
Frequency of VMT

12%
10%
8%
6%
4%
2%
0%
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Hour of Day

Figure (d). VMT distribution by hour of T6 Instate Tractor/Others in EMFAC2021

and EMFAC2017

T6 Instate Tractor/Others
16%
EMFAC2021
14%
EMFAC2017
Frequency of VMT

12%
10%
8%
6%
4%
2%
0%
0 1 2 3 4 5 6 7 8 9 101112131415161718192021222324
Hour of Day

255
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure (e). VMT distribution by hour of T7 POLA in EMFAC2021 and EMFAC2017

T7 POLA/POAK/Other Port
9%
EMFAC2021
8%
EMFAC2017
Frequency of VMT

7%
6%
5%
4%
3%
2%
1%
0%
0 1 2 3 4 5 6 7 8 9 101112131415161718192021222324
Hour of Day

Figure (f). VMT distribution by hour of T7 Single in EMFAC2021 and EMFAC2017

T7 Single
12%
EMFAC2021
10% EMFAC2017
Frequency of VMT

8%

6%

4%

2%

0%
0 1 2 3 4 5 6 7 8 9 101112131415161718192021222324
Hour of Day

256
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure (g). VMT distribution by hour of T7 SWCV in EMFAC2021 and EMFAC2017

T7 SWCV
20%
EMFAC2021
18%
EMFAC2017
Frequency of VMT

16%
14%
12%
10%
8%
6%
4%
2%
0%
0 1 2 3 4 5 6 7 8 9 101112131415161718192021222324
Hour of Day

Figure (h). VMT distribution by hour of T7 Tractor in EMFAC2021 and EMFAC2017

T7 Tractor
9%
EMFAC2021
8%
EMFAC2017
Frequency of VMT

7%
6%
5%
4%
3%
2%
1%
0%
0 1 2 3 4 5 6 7 8 9 101112131415161718192021222324
Hour of Day

257
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure (i). VMT distribution by hour of UBUS in EMFAC2021 and EMFAC2017

UBUS
9%
EMFAC2021
8%
EMFAC2017
Frequency of VMT

7%
6%
5%
4%
3%
2%
1%
0%
0 1 2 3 4 5 6 7 8 9 101112131415161718192021222324
Hour of Day

6.10 Heavy Duty VMT Distribution by Speed Figures

Figure (a). VMT distribution by speed of Out-of-state HD Vehicles (including T6
CAIRP, T6 OOS, T7 CAIRP, T7 NOOS, T7 NNOOS, Motor Coach) in EMFAC2021
and EMFAC2017

Out-of-state HD Vehicles
60%
EMFAC2021
50% EMFAC2017
Frequency of VMT

40%

30%

20%

10%

0%
5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90
Speed (mph)

258
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure (b). VMT distribution by speed of SBUS in EMFAC2021 and EMFAC2017

SBUS
20%
EMFAC2021
18%
EMFAC2017
Frequency of VMT

16%
14%
12%
10%
8%
6%
4%
2%
0%
5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90
Speed (mph)

Figure (c). VMT distribution by speed of T6 Instate Delivery in EMFAC2021 and

EMFAC2017

T6 Instate Delivery
25%
EMFAC2021
EMFAC2017
Frequency of VMT

20%

15%

10%

5%

0%
5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90
Speed (mph)

259
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure (d). VMT distribution by speed of T6 Instate Tractor/Others in EMFAC2021

and EMFAC2017

T6 Instate Tractor/Others
25%
EMFAC2021
EMFAC2017
Frequency of VMT

20%

15%

10%

5%

0%
5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90
Speed (mph)

Figure (e). VMT distribution by speed of T7 POLA in EMFAC2021 and EMFAC2017

T7 POLA/POAK/Other Port
20%
EMFAC2021
18%
EMFAC2017
Frequency of VMT

16%
14%
12%
10%
8%
6%
4%
2%
0%
5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90
Speed (mph)

260
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure (f). VMT distribution by speed of T7 Single in EMFAC2021 and EMFAC2017

T7 Single
35%
EMFAC2021
30% EMFAC2017
Frequency of VMT

25%
20%
15%
10%
5%
0%
5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90
Speed (mph)

Figure (g). VMT distribution by speed of T7 SWCV in EMFAC2021 and

EMFAC2017

T7 SWCV
12%
EMFAC2021
10% EMFAC2017
Frequency of VMT

8%

6%

4%

2%

0%
5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90
Speed (mph)

261
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure (h). VMT distribution by speed of T7 Tractor in EMFAC2021 and

EMFAC2017

T7 Tractor
40%
EMFAC2021
35%
EMFAC2017
Frequency of VMT

30%
25%
20%
15%
10%
5%
0%
5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90
Speed (mph)

Figure (i). VMT distribution by hour of UBUS in EMFAC2021 and EMFAC2017

UBUS
70%
EMFAC2021
60% EMFAC2017
Frequency of VMT

50%
40%
30%
20%
10%
0%
5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90
Speed (mph)

262
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

6.11 Engine Starts Distribution by Hour Figure

Figure (a). Engine starts distribution by hour of Out-of-states trucks in EMFAC2021

OOS
8%
7% EMFAC2021
frequency of starts

6%
5%
4%
3%
2%
1%
0%
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
hour of day

Figure (b). Engine starts distribution by hour of T6 Instate Delivery in EMFAC2021

T6 Instate Delivery
20%
EMFAC2021
frequency of vmt

15%

10%

5%

0%
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
hour of day

263
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure (c). Engine starts distribution by hour of T6 Instate Tractor/Other in

EMFAC2021

T6 Instate Tractor/Other
12%
EMFAC2021
frequency of starts

10%
8%
6%
4%
2%
0%
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
hour of day

Figure (d). Engine starts distribution by hour of T7 Port trucks in EMFAC2021

T7 POLA/POAK/Other Port
10%
EMFAC2021
frequency of starts

8%

6%

4%

2%

0%
1 2 3 4 5 6 7 8 9 101112131415161718192021222324
hour of day

264
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure (e). Engine starts distribution by hour of T7 Single trucks in EMFAC2021

T7 Single
EMFAC2021
6%
frequency of starts

5%
4%
3%
2%
1%
0%
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
hour of day

Figure (f). Engine starts distribution by hour of T7 SWCV in EMFAC2021

T7 SWCV
10%
EMFAC2021
frequency of starts

8%

6%

4%

2%

0%
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
hour of day

265
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure (g). Engine starts distribution by hour of T7 Tractor in EMFAC2021

T7 Tractor
14%
12% EMFAC2021
frequency of starts

10%
8%
6%
4%
2%
0%
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
hour of day

Figure (h). Engine starts distribution by hour of SBUS in EMFAC2021

SBUS
16%
EMFAC2021
14%
frequency of starts

12%
10%
8%
6%
4%
2%
0%
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
hour of day

266
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

6.12 Engine Starts Distribution by Soak Time Figure

Figure (a). Engine starts distribution by soak time of Out-of-states trucks in
EMFAC2021

60% OOS
50%
frequency of starts

EMFAC2021
40%

30%

20%

10%

0%
120
180
240
300
360
420
480
540
600
660
720
5

720+
10
20
30
40
50
60

soak time bin

Figure (b). Engine starts distribution by soak time of T6 Instate Delivery in

EMFAC2021

T6 Instate Delivery
20%
frequency of starts

EMFAC2021
15%

10%

5%

0%
120
180
240
300
360
420
480
540
600
660
720
5

>720
10
20
30
40
50
60

soak time bin

267
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure (c). Engine starts distribution by soak time of T6 Instate Tractor/Other in

EMFAC2021

T6 Instate Tractor/Other
18%
16% EMFAC2021
frequency of starts

14%
12%
10%
8%
6%
4%
2%
0%
120
180
240
300
360
420
480
540
600
660
720
5

>720
10
20
30
40
50
60

soak time bin

Figure (d). Engine starts distribution by soak time of T7 Port trucks in EMFAC2021

T7 POLA/POAK/Other Port
70%
EMFAC2021
frequency of starts

60%
50%
40%
30%
20%
10%
0%
120
180
240
300
360
420
480
540
600
660
720
5

>720
10
20
30
40
50
60

soak time bin

268
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure (e). Engine starts distribution by soak time of T7 Single trucks in

EMFAC2021

T7 Single
50%
EMFAC2021
frequency of starts

40%
30%
20%
10%
0%
120
180
240
300
360
420
480
540
600
660
720
5

>720
10
20
30
40
50
60

soak time bin

Figure (f). Engine starts distribution by soak time of T7 SWCV in EMFAC2021

T7 SWCV
40% EMFAC2021
frequency of starts

35%
30%
25%
20%
15%
10%
5%
0%
120
180
240
300
360
420
480
540
600
660
720
5

>720
10
20
30
40
50
60

soak time bin

269
EMFAC2021 Volume III Technical Document, Version 1.0.1
April, 2021

Figure (g). Engine starts distribution by soak time of T7 Tractor in EMFAC2021

T7 Tractor
45% EMFAC2021
40%
frequency of starts

35%
30%
25%
20%
15%
10%
5%
0%
120
180
240
300
360
420
480
540
600
660
720
5

>720
10
20
30
40
50
60

soak time bin

Figure (h). Engine starts distribution by soak time of SBUS in EMFAC2021

SBUS
60%
EMFAC2021
frequency of starts

50%
40%
30%
20%
10%
0%
120
180
240
300
360
420
480
540
600
660
720
5

>720
10
20
30
40
50
60

soak time bin

270