AVoided Emissions and geneRation Tool
(AVERT)
User Manual
Version 4.2
October 2023
U.S. Environmental Protection Agency
Office of Air and Radiation
Climate Protection Partnerships Division
v>EPA
-------
Acknowledgments
AVERT was developed by Synapse Energy Economics, Inc., under contract to EPA's
Climate Protection Partnerships Division, State and Local Climate and Energy Program and
under the direction of Robyn DeYoung and Colby Tucker of EPA.
EPA thanks the staff at Synapse who developed AVERT and the user manual, particularly
Jeremy Fisher, Ph.D., Patrick Knight, Ariel Horowitz, Ph.D., Caitlin Odom, Avi Allison,
Jamie Hall, Erin Camp, Jackie Litynski, Elizabeth A. Stanton, Ph.D., and Bruce Biewald.
Eastern Research Group, Inc. (ERG) provided production and logistical support and
developed the web-based version of AVERT, with thanks to Chris Lamie, Courtney Myers,
Bryan Neva, Emma Borjigin-Wang, Beverly Ge, Sydney Merrill, and several ERG
colleagues. ERG and Synapse provided these services under EPA contracts #EP-BPA-12-
H-0025, #EP-BPA-12-H0036, #68HERH19D0028, and #47QFHA20A0005 (GSA). EPA
also thanks Abt Associates staff David Cooley, Kait Siegel, Jonathan Dorn, Ph.D., and
Anna Belova, Ph.D., currently with Cognistx, who provided support for developing the PM2 5
emission rates under contracts #EP-W-11-003 and #EP-W-17-009.
EPA thanks the following individuals for participating in a formal peer review of the new
functionality and associated materials found in AVERT v4.0: T. Donna Chen, Ph.D.,
University of Virginia; Stephen Holland, University of North Carolina; Avi Mersky, Ph.D.,
American Council for an Energy-Efficient Economy; Jeremy Michalek, Ph.D., Carnegie
Mellon University; and Shawn Midlam-Mohler, Ph.D., The Ohio State University. The
information and views expressed in AVERT v4.2, user manual, and associated materials do
not necessarily represent those of the reviewers, who also bear no responsibility for any
errors or omissions. In addition, EPA thanks staff members at the United States
Department of Energy, National Renewable Energy Laboratory, and our colleagues in the
Office of Air Quality Planning and Standards and Office of Transportation and Air Quality
for their collaboration, assistance, and guidance.
Please contact avert@epa.gov with any inquires or requests for technical
assistance.
-------
AVERT User Manual Version 4.2
Contents
What's New in AVERT v4.1? 1
Key Abbreviations 2
1. Introduction 4
The Challenge of Estimating Changes in Emissions 5
Basic Method: eGRID Non-baseload Method. 6
Basic Method: Capacity Factor Approach 7
Intermediate Method: Historical Hourly Method 7
Sophisticated Method: Energy Modeling 7
Short-run and Long-run Power Sector Analysis 8
Using AVERT 8
Example Use A: Air Quality Planner Quantification of Changes in Emissions
Resulting from an Energy Efficiency Program for State Implementation Plan
Compliance 10
Example Use B: Stakeholder Review of Multiple Energy Efficiency Options for
Changing Emissions 10
Cautionary Note 10
Benefits of Using AVERT 11
2. The AVERT Analysis Structure 13
3. AVERT Main Module: An Overview 16
AVERT Regions 16
Energy Policy Characteristics 18
Scenario Results 20
4. AVERT Main Module: Step-by-Step Instructions 23
AVERT Welcome Page 23
Step 1: Load Regional Data File 25
Step 2: Set Energy Scenario 27
Manual User Input 29
Reduce Generation by a Percent in Some or All Hours 30
Reduce Generation by Annual GWh 32
Reduce Each Hour by Constant MW 32
Renewable Energy 32
Electric Vehicles 33
Step 3: Run Scenario 35
Step 4: Display Results 37
Summary Tables 38
Charts and Figures 42
Summary Tables, Charts, and Figures: Power Sector and Avoided Vehicle
Emissions Data 48
COBRA Text File 53
oEPA
State and Local Climate
and Energy Program
-------
AVERT User Manual Version 4.2
SMOKE Text File 54
Advanced Outputs 54
Appendix A: Installation Instructions 56
Main Module 56
System Requirements 56
Installation 56
Launching AVERTs Main Module 57
Technical Assistance 57
Statistical Module 57
System Requirements 57
Installation and Launching 57
Technical Assistance 57
Future Year Scenario Template 58
System Requirements 58
Installation 58
Launching and Working with the Future Year Scenario Template 58
Technical Assistance 58
Appendix B: Power Sector Data 59
Data from CAMD 59
Data from the National Emissions Inventory 63
Appendix C: Renewable Energy Hourly Profiles 65
Rooftop and Utility-Scale Photovoltaic 65
Onshore Wind 65
Offshore Wind 66
Appendix D: Overview of AVERT's Statistical Module 68
Parsing Generation Demand into Fossil-Fuel Load Bins 68
Collecting Statistical Information 71
Frequency of Operation by Fossil-Fuel Load Bin 71
Generation Level by Fossil-Fuel Load Bin 71
Heat Input and Emissions by Generation Level 74
Extrapolation to Higher and Lower Fossil-Fuel Loads 74
Extrapolating the Probability of Operation 75
Extrapolating the Generation Level 76
Statistical Analysis 78
Statistical Output 80
Appendix E: AVERT's Statistical Module: Step-by-Step Instructions 82
Step 1: Determine Your Windows Operating Environment 82
Step 2: Download the Statistical Module Executable 82
Step 3: Download the CAMD Database 83
Step 4: Install the MATLAB Compiler Runtime 83
Step 5: If Desired, Complete a Future Year Scenario Template 83
oEPA
State and Local Climate
and Energy Program
-------
AVERT User Manual Version 4.2
Step 6: Launch the AVERT Executable 83
Step 7: Choose a Base Year for Analysis 84
Step 8: Choose a Base- or Future Year Scenario 84
Step 9: Choose Region(s) of Interest 85
Appendix F: AVERT's Future Year Scenario Template 86
Retirement of Existing EGUs 86
Addition of Proxy EGUs 86
Pollution-Control Retrofits 88
Running Future Year Scenarios in AVERT 88
Appendix G: AVERT Regions and Instructions for States that Cross Regional
Boundaries 89
Appendix H: Frequently Asked Questions 92
AVERT Inputs 92
AVERT Results 97
AVERT Statistical Module 100
Future Year Scenarios 103
Electric Vehicles 104
Appendix I: Web-Based AVERT 106
Differences Between the Web Edition and the Excel Main Module 106
Web AVERT State Analysis 107
Input Validation 107
Appendix J: Electric Vehicles in AVERT 109
Inputs and Assumptions 109
EV Detailed Inputs - Excel Main Module Only 109
Background Assumptions 112
Calculations 114
Power Sector 115
Vehicles 116
How to: Analyzing Emissions Impacts of Electric Vehicles 117
Example 1: What is the impact of deploying 39,000 new battery-powered
electric vehicles in 2022 in North Carolina? 118
Example 2: What if Florida were evaluating the impacts of a proposed policy
that increased the sales of electric vehicles by 5 percent each year from 2022
to 2024? 120
Example 3: How might different charging profiles for transit buses affect
emissions from additional electric vehicles in New York? 122
Appendix K: Caveats and Limitations 125
Caveats and Limitations: Power Sector 125
Caveats and Limitations: Modeling Electric Vehicles 128
Appendix L: Version History 130
oEPA
State and Local Climate
and Energy Program
-------
JtVert
AVoid«d Emissions and q»rwRatlon Tool^^
wwwepagovTavSt'^W
AVERT User Manual Version 4.2
What's New in AVERT v4.2?
AVERT v4.2 is the latest version of AVERT. Key updates in AVERT v4.2 include:
• MOVES4 data: Version 4.2 includes updated vehicle emissions rates from MOVES4. As
part of this update, users can now analyze vehicle model years 2023-2028 (updated from
2020-2025). Additionally, the source for data on vehicle miles traveled (VMT) by county
was updated from the National Emissions Inventory (NEI) to MOVES4.
Users should note that parameters used to define electric vehicles in AVERT v4.2 are current as of
January 2023 when AVERT v4.0 was first published. There are several updates to internal
combustion engine vehicle representation in AVERT v4.2, particularly with respect to emission
rates and county-level VMT. The transportation sector is rapidly evolving, as are the tools available
to help analyze their impact on energy and emissions. EPA strives to keep AVERT as current as
possible and will continue to make periodic updates to improve model assumptions.
To see changes and updates from previous versions of AVERT, including recent changes released
as part of v4.0 and v4.1, refer to the version history in Appendix L.
SEPA
State and Local Climate
and Energy Program
-------
AVERT User Manual Version 4.2
Key Abbreviations
AVERT AVoided Emissions and geneRation Tool
BEV battery-powered electric vehicle
BOEM Bureau of Ocean Energy Management
CAMD EPA Clean Air Markets Division
CHP combined heat and power
CO2 carbon dioxide
COBRA CO-Benefits Risk Assessment Health Impacts Screening and Mapping Tool
DOE U.S. Department of Energy
EE energy efficiency
EGU electric generating unit
EIA Energy Information Administration
EPA U.S. Environmental Protection Agency
EV electric vehicle
GWh gigawatt-hour
ICE internal combustion engine
ISO Independent System Operator
lb pound
LDV light-duty vehicle
LRMER long-run marginal emission rate
MMBtu million British thermal units
MOVES MOtor Vehicle Emission Simulator
MW megawatt
MWh megawatt-hour
NEI National Emissions Inventory
NAAQS National Ambient Air Quality Standards
NH3 ammonia
NOx nitrogen oxides
PHEV plug-in hybrid vehicle
PM2.5 particulate matter with a diameter of 2.5 microns or less
PV photovoltaic
RDF Regional Data File
State and Local Climate
and Energy Program
-------
AVERT User Manual Version 4.2
RE
renewable energy
RTO
Regional Transmission Organization
SIP
State Implementation Plan
SRMER
short-run marginal emission rate
SMOKE
Sparse Matrix Operator Kernel Emissions Model
S02
sulfur dioxide
TWh
terawatt-hour
VOCs
volatile organic compounds
State and Local Climate
and Energy Program
-------
AVERT User Manual Version 4.2
1. Introduction
The U.S. Environmental Protection Agency (EPA) recognizes that many state and local
governments are adopting, implementing, and expanding cost-effective energy efficiency (EE),
renewable energy (RE), electric vehicle (EV), and energy storage policies and programs. States
are investing in policies and programs to achieve benefits including lowered customer costs,
improved electric supply reliability, and diversified energy supply portfolios.1 Certain energy policies
can also reduce pollution of criteria air pollutants and greenhouse gases, especially on high-
electricity-demand days that typically coincide with poor air quality.
EPA's Roadmap for Incorporating Energy Efficiency/Renewable Energy Policies and Programs in
State and Tribal Implementation Plans describes basic, intermediate, and sophisticated methods
for quantifying the emissions changes resulting from energy policies.2 Basic methods entail a
simple calculation: multiplying the amount of generation or electricity consumption changed by the
policy or program by the "non-baseload" emission rate indicated for a specific pollutant in a region
(e.g., eGRID subregion or electricity market area). Intermediate methods, like the AVERT tool
described in this manual, offer more temporal resolution and greater functionality than basic
methods, while being transparent, credible, free, and accessible. Sophisticated methods offer the
highest level of detail, but cannot be implemented without a detailed understanding of the electricity
grid and electric generator dispatch dynamics, and/or energy modeling expertise. EPA is
committed to helping state and local air quality planners and other agencies calculate the
emissions impacts of energy policies and programs so that these emissions reductions can be
incorporated in Clean Air Act plans to meet National Ambient Air Quality Standards (NAAQS) and
other clean air goals.
AVERT estimates the change in generation from one or more energy policy scenarios. These
scenarios could be EE savings or RE deployments that reduce the amount of generation needed,
or policies and programs that increase the amount of generation needed (e.g., EVs). AVERT
applies this change in generation and predicts changes in hour-by-hour generation and emissions
for individual power plants, called electric generating units (EGUs). AVERT is therefore indirectly
estimating the change in emissions from these interventions, which is in contrast to direct
measurements, like emissions reductions resulting from stack controls in an EGU's smokestack.
Energy policies may be implemented though specific programs and technologies that have hourly
load3 profiles, which are hour-by-hour schedules of expected reductions or increases in electricity
demand or electricity production for a year. Understanding the hour-by-hour relationship between
specific energy programs and the dispatch of fossil fuel EGUs is essential to the estimation of the
magnitude and location of changes in emissions resulting from energy policies.
EPA has developed a credible, free, user-friendly, and accessible tool to estimate emissions
changes resulting from energy policies and programs so that air quality planners can incorporate
1 A variety of technologies, policies, programs, and specific projects can be modeled in AVERT. These activities
increase or decrease electricity generation, electricity demand, and/or electric sector emissions in at least one
hour of the year. For simplicity, the term "energy policies" is used in this document to encompass all these types
of activities.
2 See Appendix I at https://www.epa.aov/enerav-efficiencv-and-renewable-enerav-sips-and-tips/enerav-
efficiencvrenewable-enerav-roadmap.
3 "Load" is the term used throughout this manual to describe regional demand for electricity.
State and Local Climate
and Energy Program
-------
AVERT User Manual Version 4.2
those impacts into their NAAQS State Implementation Plans (SIPs).4 The AVoided Emissions and
geneRation Tool (AVERT) quantifies the changes in emissions of sulfur dioxide (SO2), nitrogen
oxides (NOx), carbon dioxide (CO2), particulate matter with diameter of 2.5 microns or less (PM25),
volatile organic compounds (VOCs), and ammonia (NH3) associated with energy policies and
programs within the contiguous United States (Alaska, Hawaii, and U.S. territories are not
modeled). AVERT captures the actual historical behavior of EGUs' operation on an hourly basis to
predict how EGUs will operate with these energy changes in place.
AVERT users can analyze how different types of EE, RE, and EV policies and programs affect the
magnitude and location—at the county, state, and regional level—of emissions. AVERT is a flexible
modeling framework with a simple user interface designed specifically to meet the needs of state
air quality planners and other interested stakeholders.
The Challenge of Estimating Changes in Emissions
Estimating the location of changes in generation and associated changes in emissions presents
several challenges:
• The balance of electricity supply and demand varies by hour and by season.
• Multiple EGUs are dispatched to supply demand for electricity over a broad region.
• Different programs and technologies save or generate energy at varying times throughout
the day and seasonally.
Within each region across the country, system operators decide when, how, and in what order to
dispatch generation from each power plant in response to customer demand for electricity in each
moment and the variable cost of production at each plant.5 Electricity from the power plants that
are least expensive to operate is dispatched first, and the most expensive plants are dispatched
last. That is, given a cohort of EGUs, the lowest-variable-cost units are brought online first; as the
load increases through peak (high-demand) hours, increasingly expensive units are brought online.
(Ideally, given no other constraints—e.g., transmission, voltage support, ramp rates, maintenance
outages—EGUs will dispatch into an electric system in a regular economic order based on the cost
of fuel, the units' heat rates, and other variable costs of production.) In this "economic dispatch"
decision-making process, EE and RE resources generally have low variable costs or are
considered "must-take" resources, the operation of which is determined by sun, wind, the flow of a
river, or efficiency program designs.6 EE and RE resources typically displace higher-variable-cost,
higher-emission-producing fossil-fuel generation. While electricity planners typically think about a
4 See Appendix I of the Roadmap for Incorporating Energy Efficiency/Renewable Energy Policies and Programs
(https://www.epa. aov/enerav-efficiencv-and-renewable-enerav-sips-and-tips/enerav-efficiencvrenewable-enerav-
roadmap) for details on how this approach can be used in the different NAAQS SIP pathways.
5 "Variable costs" are the costs realized in the hour-to-hour operation of a generator that vary with the amount of
energy that the unit produces. They typically include the cost of fuel, maintenance costs that scale with output,
and the cost of emissions. Power plants also have "fixed costs"—such as staffing and regular maintenance—that
must be met regardless of whether or not their units are generating power.
6 "Must-take" resources are so named because they cannot generally be centrally dispatched (i.e., a controller
cannot determine when they provide power), and as such they must be taken when they provide power. These
resources can be curtailed under unusual circumstances, such as during periods of excess supply. These
periods are not the norm; for the most part, controllers operate dispatchable resources around the must-take
resources.
State and Local Climate
and Energy Program
-------
JtVert
AVoid«d Emissions and q»rwRatlon Tool^^
wwwepagovTavSt'^W
AVERT User Manual Version 4.2
single marginal resource,7 there are often several EGUs with similar variable costs that are
simultaneously on or near the margin. EE load reductions and low-variable-cost RE generation can
change the level of generation dispatched at multiple marginal EGUs at the same time across state
boundaries.
Different resources impact different generation in different hours or seasons. Hourly energy profiles
describe the hourly changes in customer demand resulting from a program or measure, or the
combined impact of a set of programs or measures. For example, hourly energy profiles can
represent a portfolio of programs used to meet a policy target, such as the Energy Efficiency
Resource Standards adopted by 27 states.8 Generation profiles for RE technologies, such as wind
or solar photovoltaic (PV), also vary by hour and season.
Determining which cohorts of EGUs are most likely to be impacted during particular hours or under
certain conditions is a complex endeavor. It is not possible to definitively predict how resources
from an energy policy will affect any given power plant. There are, however, several ways to
estimate which EGUs would be impacted when and by how much based on new resources' hourly
energy profiles, EGUs' historical operational behavior, projected information on cost, and other
factors affecting dispatch in each regional electricity market.
As noted above, methods for
estimating projected changes
to emissions range from
basic to sophisticated (see
Figure 1). The first three
methods (two basic methods
and the AVERT intermediate-
complexity method described
in this manual) use historical
operations and profiles to
estimate likely changes to
emissions. The more
sophisticated fourth approach
predicts future electricity market conditions and emissions changes. Each method attempts to
identify the group of EGUs whose generation activity would change as a result of new programs or
measures that vary in terms of their sizes, geographic areas, and timing during the day and the
year.
Basic Method: eGRID Non-baseload Method
This basic method calculates the average non-baseload changes in emissions resulting from
energy policies and programs in one of EPA's "eGRID" subregions.9 Annual electricity generation
or sales increased or displaced by a program or measure are multiplied by the "non-baseload"
Figure 1. Emissions quantification methods.
4
Assumptions are Simpler
Capacity
Factor
Method
Hourly
Historical
Method
Energy
Modeling
Method
Methods are More Sophisticated
7 The highest-cost unit that is required to meet customer demand at any particular time.
8 U.S. EPA. 2015. Energy and Environment Guide to Action. Chapter 4.1.
https://www.epa.aov/statelocalenerav/enerav-and-environment-auide-action.
9 eGRID data can be found at https://www.epa.aov/enerav/egrid.
State and Local Climate
and Energy Program
-------
AVERT User Manual Version 4.2
emission rate for each pollutant in each eGRID subregion.10 The non-baseload emission rates for
an eGRID subregion are appropriate to represent the average emission rate for the EGUs most
likely to be impacted by EE or RE.
Basic Method: Capacity Factor Approach
The capacity factor approach estimates emissions impacts for an EGU based on its current
capacity factor (i.e., its production of electricity in the most recent year as a percentage of the
maximum energy that it can produce).11 The capacity factor is used as a proxy for the likelihood
that any given EGU will be impacted by new resources resulting from energy policies. Infrequently
dispatched EGUs with low capacity factors are more likely to be impacted than EGUs with higher
capacity factors.
Intermediate Method: Historical Hourly Method
The AVERT method described here uses historical hourly emission rates based on recent EPA
data on EGUs' hourly generation and emissions reported through EPA's Acid Rain Program.12 This
method couples historical hourly generation and emissions with the hourly load profiles of energy
resources to determine hourly marginal emission rates and hourly changes in emissions. AVERT
can be used to predict emissions changes in a current or near-future year—though it is based on
historical behavior rather than predicted economic behavior and, therefore, does not use
projections of future fuel or electricity market prices.
Sophisticated Method: Energy Modeling
The most sophisticated method, energy modeling, is the use of highly complex simulation models
that predict individual EGU dispatch, commitment, and emissions based on economic dispatch.13
Energy models that simulate unit-by-unit dispatch and attempt to replicate decisions made by
controllers and operators are called "production-cost" models, and will often include operational
and transmission constraints. Operating economic dispatch models require the modification and
validation of extensive input datasets, significant expertise to operate proprietary models, and
ultimately a fairly high cost to evaluate individual scenarios. Other models called "capacity
expansion" models are designed to optimize resource build-out (i.e., new capacity additions), and
may be appropriate for examining long-term impacts of specific new resources.14
10 Grid loss factors approximate the line losses that occur between the electric generating facilities and the buildings
that purchase the electricity. They should be included in this calculation.
11 See, for example, the eCALC model, documented in Texas A&M Energy Systems Laboratory. 2004. Texas
Emissions and Energy Calculator (eCALC). http://oaktrust.librarv.tamu.edU/handle/1969.1/2079.
12 See EPA's Power Sector Emissions Data at https://ampd.epa.gov/ampd/.
13 Fisher, J., C. James, N. Hughes, D. White, R. Wilson, and B. Biewald. 2011. Emissions Reductions from
Renewable Energy and Energy Efficiency in California Air Quality Management Districts. https://www.svnapse-
enerqv.com/sites/default/files/Emissions%20Reductions%20from%20Renewable%20Enerqv%20and%20Enerqy
%20Efficiencv%20in%20California%20Air%20Qualitv%20Manaqement%20Districts%2008-016.pdf.
14 A variety of utility-standard models are available to estimate the impact of new energy changes on existing plant
dispatch. Generally, the models best suited for this purpose in near-term years are the production-cost models,
including such systems as Market Analytics—Zonal Analysis, PROMOD IV, and PLEXOS
fhttp://www.enerqvexemplar.com). Some examples of capacity expansion models include such platforms as
EGEAS fhttps://www.epri.comhttps://www.epri.com) and System Optimizer and Strategist. Other models, like
State and Local Climate
and Energy Program
-------
JtVert
AVoid«d Emissions and q»rwRatlon Tool^^
wwwepagovTavSt'^W
AVERT User Manual Version 4.2
Short-run and Long-run Power Sector Analysis
Electric sector interventions can have both operational as well as structural impacts on the electric
grid. For example, the charging of new EVs might initially come from existing generators (i.e.,
operational impacts only), but after time, the existence of the new load may influence the
deployment or retirement of generators (a structural change). Methodologies for estimating
emissions changes can be classified as short-run (operational only, holding the structure of the grid
fixed) or long-run (incorporating both operational as well as structural responses to an intervention).
Each approach is appropriate for distinct purposes. As structural impacts often take time to
materialize, short-run approaches are appropriate for characterizing the near-term impacts of an
intervention prior to the point where structural impacts are expected to occur. As a guiding
principle, users can consider that it takes five years for an intervention to create structural impacts
(although it can be shorter if the intervention was anticipated, such as a large EE campaign that the
local utility was involved in and incorporated into their resource plans).
Depending on the interests of the analyst, short-run approaches may be sufficient. However, they
are generally inappropriate for characterizing the long-term impacts of interventions due to the fact
that they omit the intervention's impact on the structure of the grid.15 AVERT is considered a short-
run model, and it produces estimates of emission changes of scenarios representing the near-
future and estimates of short-run marginal emission rates (SRMER).
Using AVERT
AVERT is a free tool that allows users with minimal electricity-system expertise to easily evaluate
county-level changes in emissions resulting from energy policies. AVERT is primarily designed to
estimate the impact of new energy policies and programs on emissions from large (greater than 25-
megawatt [MW]), stationary fossil-fired EGUs. It uses public data, is accessible and auditable, and
can be used for quantifying emissions impacts in NAAQS SIPs.16 Of the six criteria pollutants
governed by NAAQS, AVERT provides analytical capability for three pollutants from direct
emissions (NOx, SO2, and PM2 5) and two from a precursor basis (ozone and PM2.5).17,18
Anchor Power's EnCompass model, are increasingly blurring the difference between capacity expansion and
production-cost models, allowing analysts to evaluate both simultaneously. These production-cost and capacity
expansion models are generally proprietary and usually require either licensure or specific project contracts to
operate for most users. Large-scale, integrated assessment models such as ICF's Integrated Planning Model, or
IPM (https://www.epa.oov/airmarkets/clean-air-markets-power-sector-modelino): the National Renewable Energy
Laboratory's Regional Energy Deployment System, or ReEDS (https://www.nrel.gov/analvsis/reeds/): and DOE's
National Energy Modeling System, or NEMS (https://www.eia.gov/forecasts/aeo/) are appropriate for testing the
implications of large-scale policies and initiatives over longer periods, but use simplified representations of
electricity dispatch and generally aggregate units for computational efficiency. These models include such
platforms as EGEAS (https://www.epri.comhttps://www.epri.com) and System Optimizer and Strategist.
15 For more information about short-run and long-run power sector analysis, see: Gagnon, P., and W. Cole. 2022.
Planning for the evolution of the electric grid with a long-run marginal emission rate, iScience 25(3).
https://www.sciencedirect.com/science/article/pii/S25890042220Q1857.
16 See EPA's Roadmap for Incorporating Energy Efficiency/Renewable Energy Policies and Programs in SIPs
(https://www.epa.gov/energv-efficiencv-and-renewable-energv-sips-and-tips/energv-efficiencvrenewable-energy-
roadmap) for details on other regulatory requirements.
17 For more information about SIPs, see https://www.epa.gov/sips/basic-information-air-gualitv-sips.
18 The ozone precursors that AVERT models are VOCs and NOx. The PM2.5 precursors that AVERT models are
NOx, S02, and NH3.
State and Local Climate
and Energy Program
-------
JtVert
AVoid«d Emissions and q»rwRatlon Tool^^
wwwepagovTavSt'^W
AVERT User Manual Version 4.2
To estimate changes in emissions using AVERT, users will need to know the type of program or
measure to be analyzed or the program's energy profile. An annual energy profile can be
presented in 8,760 hourly intervals, or more coarsely in a few intervals (for example, peak, off-
peak, and shoulder periods). For EE policies and programs, users will need the expected annual
load reduction and an understanding of the temporal profile (e.g., would the EE program save
energy throughout the year or primarily during peak periods). For RE programs, users will need to
know the capacity of the solar or wind resource they are analyzing. These annual profiles are used
to identify more precisely what specific generation resources will experience a change in output as
a result of specific programs or measures.19 In the absence of specific data on energy changes,
planners need to use their judgement to approximate the timing of these changes.
Using these inputs, AVERT automatically estimates emission changes in a region. The user then
can view various outputs, maps, charts, and tables useful for many different types of analysis.
Users can choose outputs that show regional-, state-, and county-level changes in emissions, with
the option of highlighting high-electric-demand days. Expert air quality modelers assessing
changes in PM2 5, NOx, and SO2 emissions can use the SMOKE (Sparse Matrix Operator Kernel
Emissions) output function to produce hourly, EGU-specific air-dispersion-model-ready data.
AVERT users assessing the public health impacts of the criteria pollutant reductions can use the
COBRA (CO-Benefits Risk Assessment Health Impacts Screening and Mapping Tool) output
function to produce model-ready county-level emissions impact data.20
AVERT is best suited to analyze the emissions changes resulting from state-wide or multi-state
energy policies and programs. Since AVERT modeling is conducted in one of 14 large regions that
represent electricity markets and does not account for transmission constraints within each region,
this tool is not recommended for estimating the change in emissions under small local programs.
Smaller programs can use AVERT-generated emission rates to estimate emission changes within
an AVERT region; these rates are available at www.epa.qov/avert. (See Appendix H for more
details on determining the upper and lower bounds for load changes to be modeled in AVERT.) In
addition, the tool is equipped to predict changes in near-term future years by estimating each unit's
generation, heat input, and emissions in the event that other EGUs are retired, newly brought
online, or retrofitted with pollution controls.
AVERT provides useful information to expert energy or air quality planners, and also to interested
stakeholders. In 2018, EPA released a web-based version of the AVERT Main Module on EPA's
website (www.epa.gov/avert). It provides a streamlined interface with much of the same
functionality as the downloadable Excel-based Main Module, without the need to use Excel
software or upload separate Regional Data Files (RDFs). The Web Edition relies on the most
recent year of input data. Refer to Appendix I for a comparison between the Excel- and web-based
Main Modules' functionality and available display outputs.
19 U.S. EPA. 2010. Assessing the Multiple Benefits of Clean Energy: A Resource for States. Chapter 3, page 64.
20 At this time, the online version of COBRA is able to read in changes to NH3 and VOC emissions from AVERT, but
only through a direct connection from the AVERT web edition. The downloadable version of COBRA is able to
read the output file generated by AVERT, but only for NOx, SO2, and PM2.5. See page 53 for more information on
how to account for NH3 and VOC emission changes in the downloadable version of COBRA. At this time, AVERT
is not capable of producing SMOKE-readable outputs for VOCs or NH3.
SERA
State and Local Climate
and Energy Program
-------
JtVert
AVoid«d Emissions and q»rwRatlon Tool^^
wwwepagovTavSt'^W
AVERT User Manual Version 4.2
Example Use A: Air Quality Planner Quantification of Changes in Emissions
Resulting from an Energy Efficiency Program for State Implementation Plan
Compliance
Air quality planners can use AVERT to quantify the expected emissions of a new wind farm, solar
initiative, or EE program for the purposes of Regional Haze Rule or NAAQS SIP/Tribal
Implementation Plan compliance. For example, planners can evaluate a program that could
displace NOx during the summer ozone season, efforts to bolster a state wind energy program, or
proposed additional incentives for an EE program that targets peak energy usage (e.g., high-
efficiency air conditioner replacement).
Using AVERT, planners can input estimates for the amount of energy a wind farm of a particular
size and output typical of the region can produce or the amount of energy that could be avoided
from an EE program. Among other outputs, AVERT can estimate annual SO2, PM2 5, and VOCs
emissions as well as ozone-season NOx emissions reductions or a pounds (Ib)-per-day 10-day
average of NOx emissions in counties selected, allowing a comparison of the effectiveness of these
programs against other SIP measures. Advanced AVERT users can incorporate expected
retirements and changes in emission rates expected in future years, and establish new baseline
conditions. Because AVERT can output SMOKE-formatted emissions estimates for each EGU in
the region in each hour of the year, planners can also assess the air quality improvements using an
air dispersion model. Following EPA guidance, this information could be incorporated into a SIP.21
Example Use B: Stakeholder Review of Multiple Energy Efficiency Options for
Changing Emissions
Stakeholders can use AVERT to develop and test multiple types of EE programs in a state or group
of states within an AVERT region to compare potential reductions in PM2 5, NOx, SO2, VOC, NH3, or
CO2 emissions. Using AVERT, they can quickly test different types of EE load profiles and estimate
the resulting displaced emissions. Users would modify input parameters to simulate baseload, peak
load, or total EE portfolio reductions, or hour-by-hour load reduction profiles. This type of analysis
allows stakeholders to review estimated emissions benefits from a wide variety of programs, which
can help them consider adopting and/or implementing programs with the greatest improvements to
air quality.
Cautionary Note
AVERT should only be used to assess changes to emissions resulting from energy programs—not
to assess changes to an EGU fleet. For example, AVERT is not equipped to examine the changes
in emissions that result from retirements, changes to heat rates, or specific fuel changes. AVERT
uses data based on historical dispatch patterns and cannot credibly estimate changes in emissions
resulting from changes to the overall pattern of dispatch.
21 For more information on EPA's guidance, refer to EPA's Roadmap for Incorporating Energy
Efficiency/Renewable Energy Policies and Programs in SIPs/TIPs at https://www.epa.qov/enerqy-efficiencv-and-
renewable-enerqv-sips-and-tips/enerqv-efficiencvrenewable-enerqy-roadmap.
State and Local Climate
and Energy Program
-------
AVERT User Manual Version 4.2
Benefits of Using AVERT
AVERT combines historical hourly generation data with energy profiles, making it possible for users
to:
• Compare the emissions impacts of different types of energy polices, programs, or
technologies.
• Incorporate energy policies and programs into air quality models and public health impact
tools, such as EPA's COBRA Screening Model; and identify opportunities within the power
sector for SIPs to demonstrate Clean Air Act compliance.22
• Estimate emission changes during peak energy demand periods in a historical year or
near-term future year.
For example, wind and solar power have
different hourly and seasonal operational
profiles. AVERT can compare the change in
emissions between these two RE technologies
at different times of the year. Similarly, various
EE programs have different hourly load
profiles. AVERT can also help users analyze
different EE programs or portfolios of
programs that offer different energy savings
and emissions changes throughout the year.
Using this information, air quality planners
could, for example, assess which EE programs
provide the greatest air quality improvement
on high ozone days. For smaller programs,
users can use AVERT-generated emission
rates to get a general estimate within an
AVERT region. (See Appendix H for more
details on determining the upper and lower
bounds for load changes to be modeled in
AVERT.)
AVERT is driven entirely by historical, publicly available data reported to EPA and the U.S.
Department of Energy's (DOE's) Energy Information Administration (EIA). It uses statistically driven
"behavior simulation" to estimate near-term future emissions changes based on the recorded
historical behavior of existing EGUs in the recent past. Using this dataset alone, the model derives
unit generation behaviors (i.e., how these EGUs respond to load requirements), EGUs that have a
must-run designation,23 and forced and maintenance outages. In addition, AVERT accurately
represents the recent historical relationship between unit generation and emissions, with
characteristics such as a decreasing heat rate (i.e., increasing efficiency) at higher levels of output,
higher emissions from EGUs that are just warming up, and seasonally changing emissions for
22 AVERT may not be used for mobile source regulatory analyses, including SIP and transportation conformity
analyses. Consult the most recent EPA guidance document for applying EPA's MOVES model
at: https://www.epa.qov/moves/latest-version-motor-vehicle-emission-simulator-moves.
23 A must-run designation indicates that a unit is required to operate for reliability reasons; such units often operate
at minimum levels to maintain the ability to meet higher load in subsequent hours.
State and Local Climate
and Energy Program
Emission Rates from AVERT
EPA has used AVERT to produce
approximations of marginal emission rates for
each AVERT region and for a national
weighted average. Current and historical
emission rates are available at
https://www.epa.gov/avert/avoided-emission-
rates-qenerated-avert. These emission rates
were calculated by assuming a 0.5%
reduction in the regional fossil generation and
are divided into six categories: onshore wind,
offshore wind, utility PV, distributed PV,
portfolio EE, and uniform EE. These emission
rates can be used for quick estimates of
avoided emissions under specific scenarios,
especially for very small energy policies.
-------
JtVert
AVoid«d Emissions and q»rwRatlon Tool^^
wwwepagovTavSt'^W
AVERT User Manual Version 4.2
EGUs with seasonal environmental controls. The derivation of unit behavior and its application to
AVERT is described in detail in Appendix D of this user manual.
AVERT has many advantages but requires several simplifying assumptions. Unlike traditional
electricity system simulation dispatch or production cost models, AVERT does not use operating
costs to estimate how and when an EGU dispatches to meet load requirements. As a result, there
are important electric system dynamics that AVERT cannot capture: temporal characteristics (i.e.,
EGU minimum maintenance downtime and ramp-rates), changing economic conditions (i.e., rising
or falling fuel or emissions prices), and explicit relationships between EGUs (i.e., units that
substitute for one another). AVERT should not be used to assess these types of changes in the
electric system or overall dispatch. These limitations are discussed in more depth in Appendix K.
AVERT operates in a basic computer environment and leads users through the process step-by-
step. Detailed instructions for the Excel version of the Main Module can be found in Section 4 of
this manual or EPA's AVERT online tutorial.24 In 2018, EPA launched a simplified web-based
version of the Main Module. Refer to Appendix I for a comparison between the web and Excel
versions.
24 AVERT's online tutorial provides video demonstrations and information about how to run AVERT:
https://www.epa.gov/avert.
State and Local Climate
and Energy Program
-------
AVERT User Manual Version 4.2
2. The AVERT Analysis Structure
AVERT has three components:
• An Excel- and web-based Main Module allows users to estimate the changes in emissions
likely to result from new energy programs, policies, or projects. The Excel-based version
requires users to select RDFs generated by the Statistical Module to analyze scenarios in
reference to either a historical base year or a future year. (See Sections 3 and 4 for a
detailed description of the Main Module.) The web-based version of the Main Module
provides a streamlined interface with much of the same functionality as the downloadable
Excel tool, without the need for Excel software or separate RDFs. The Web Edition relies
on a single year of data and generates a subset of display outputs of state and county level
emission changes. Refer to Appendix I for a full comparison of the Excel- and web-based
Main Modules. Except for this appendix and where noted otherwise, this user manual
describes features available in the Excel version.
• The MATLAB®-based Statistical Module performs statistical analysis on historical
generation, heat input, and emissions data collected by the EPA Clean Air Markets
Division (CAMD)25 to produce the statistical data files used by AVERT's Main Module. The
Statistical Module is available to users as a stand-alone executable. (See Appendices D
and E for a detailed description of the Statistical Module.)
• The Excel-based Future Year Scenario Template allows users to modify base-year
CAMD data with specified retirements and additions of power plants, as well as changes in
emission rates due to pollution controls. This modified data can be input into AVERT's
Statistical Module to produce scenario-specific statistical data files, which are then fed into
the Main Module. (See Appendix F for a detailed description of the Future Year Scenario
Template.)
AVERT analyzes how hourly changes in demand in a user-selected historical base year change
the output of fossil EGU.26 Using detailed hourly data from CAMD, AVERT probabilistically
estimates the operation and output of each EGU in a region based on a region's hourly demand for
fossil-fired generation. This statistical information is used to predict EGUs' likely operation in
response to energy changes from modeled resources. Figure 2, below, shows the flow of data from
its source, through various processing tools, to its end-use in the Main Module.
In general, hourly "prepackaged" CAMD data are input into AVERT's Statistical Module. Hourly
generation, heat input, and emissions of SO2, NOx, and CO2 from each EGU reporting to CAMD (a
requirement for fossil-fuel EGUs 25 MW and greater) are read from monthly or quarterly files.
25 Power Sector Emissions Data are available from EPA's Air Markets Programs Data website:
https://ampd.epa.gov. For more information, see "CAMD's Power Sector Emissions Data Guide":
https://www.epa.aov/sites/production/files/2020-
02/documents/camds power sector emissions data user quide.pdf.
26 AVERT's "base year" is modeled from recent, detailed hourly generation and emissions data collected by EPA for
each U.S. fossil-fired generator with capacity greater than 25 MW. Base-year data are usually made available in
the second quarter of the following year.
State and Local Climate
and Energy Program
-------
cndVEi
AVERT User Manual Version 4.2
Figure 2, Schematic of AVERT.
Raw Hourly
Generation and
Emissions Data
from Power Sector
Emissions Data
from CAMD
Text files
5
Future Year
Scenario Template
User interface for
retirements,
additions, and
retrofits
0
AVERT Main
Module
AVERT Statistical
Regional Data
Module
Files
User interface for
creating energy
Inputs CAMD
Contain annual
policy load
data, performs
r-\
hourly load data
curves.
statistical
h/
and unit-level
performing
analysis, outputs
statistics on
analyses of
new Regional
generation and
emissions
Data Files
emissions data
changes, and
creating output
charts and tables
Excel workbook
MATLAB code
Text files
Excel workbook or
web-based version
AVERT's Statistical Module can analyze either raw data for a base year or modified data created in
the AVERT's Future Year Scenario Template.27 AVERT's Future Year Scenario Template allows
users to create a scenario for a year in the near future28 by modifying data representing a historical
year. Users designate existing fossil-fuel EGUs that will no longer be in operation or will have
different emission rates as a result of pollution-control retrofits, and add new fossil-fuel EGUs
based on the characteristics of proxy existing EGUs. (See Appendix F for a more detailed
description of AVERT's Future Year Scenario Template.)
After receiving either base- or future year scenario data, AVERT's Statistical Module performs a
statistical analysis of how each EGU responds to variations in regional fossil load, and simulates
the average generation, heat input, and emissions of each EGU across a range of possible load
levels, from zero to the maximum coincident fossil capacity.29 These EGU and load-level-specific
averages are stored in the AVERT RDFs, which are the input files used into AVERT's Main
Module. (See Appendices D and E for a more detailed description of AVERT's Statistical Module.)
AVERT's Main Module is accessible as an online tool or a downloadable Excel workbook. It allows
users, regardless of their level of electricity modeling expertise, to quickly estimate the changes in
emissions likely to result from energy policies in a chosen year.30 AVERT's Main Module provides a
simple interface that guides users through inputting an hourly energy profile depicting electricity
demand in every hour of a year. The user is then prompted to launch automatic calculations that
27
28
29
30
Versions of the Future Year Scenario Template are available for 2017 through the present It is expected that
data for future years will continue to be provided as additional data from CAMD are released.
it is recommended that future year scenarios be designed for no more than five years forward from the base data
year to account for changing emissions control technologies, changing fuel prices, and retirements and additions
into the system. Caution should be exercised in reviewing future year scenarios to ensure reasonable results in
light of known or expected system changes.
Maximum coincident fossil capacity is equivalent to the sum of each and every fossil generator producing its
maximum output in a single hour.
The web-based version of AVERT's Main Module only has a single data year available and limited result formats.
Refer to Appendix I for a complete comparison.
14
SEPA
State and Local Climate
and Energy Program
-------
AVERT User Manual Version 4.2
AVoid«d Emissions and q»rwRatlon Tool^^w
www!epagov7S5<5t'^»
result in final results tables and charts for one of the 14 AVERT regions and, if desired, smaller
areas within the region. In addition, AVERT's Excel-based Main Module also outputs SMOKE-
formatted data for advanced air modeling applications. Both the Excel and web versions output
COBRA-formatted data for public health modeling applications. (See Sections 3 and 4 for a more
detailed description of AVERT's Main Module and instructions for its use.)
15
State and Local Climate
and Energy Program
-------
AVERT User Manual Version 4.2
3. AVERT Main Module: An Overview
This section provides a simplified overview of user inputs and model results. See Section 4 for
detailed, step-by-step instructions and Appendix A for detailed installation instructions. Appendix D
describes, in detail, how AVERT performs its calculations. Appendix I provides a comparison
between the web- and Excel-based versions of AVERT's Main Module.
AVERT's Excel-based Main Module estimates the emission changes resulting from user-entered
energy policy scenarios. AVERT predicts emissions changes for every individual fossil-fuel EGU in
a region and aggregates these changes to the county level.31 The Main Module uses two key
pieces of information:
• Expected emissions at every load level—the likely generation level and emissions of all
but the smallest fossil-fuel EGUs in a region in a base- or future year scenario (as modeled
in AVERT's Statistical Module and input automatically into the Main Module).
• A change in load level for every hour of the year—a user-created energy profile
depicting user-created changes in the regional fossil-fuel load for every hour of the year.
The Main Module estimates how much each fossil-fuel EGU changes its generation output and
emissions in response to an energy policy (e.g. a scenario) as compared to the base- or future year
without the program (e.g. the baseline). The Main Module presents emissions changes between
the scenario and the baseline in summary tables for quick comparison, and in graphs and maps for
rapid visual assessment.
Section 4 presents detailed, step-by-step instructions on the process of identifying a region for
analysis; obtaining and importing data; designing an energy profile; calculating the changes
resulting from this profile; and accessing tables, graphs, and maps summarizing the results. Once
an energy profile has been designed, typical processing takes 1 to 10 minutes depending on the
size of the region of interest and the processing speed of the computer.
AVERT Regions
Because customers' electricity demand is met jointly by generation resources throughout a region,
emissions changes from energy policies occur region-wide. All AVERT analysis, therefore, is
conducted at a regional level. For users, designating one of the 14 AVERT regions for analysis is
"Step 1" in using AVERT's Main Module. A map of these regions is shown in Figure 3.
Twenty-four of the contiguous U.S. states are split between two or more regions each; the other 24,
and the District of Columbia, are not split. Table 1 describes each region in detail.
Generally, air quality managers for states that are split between more than one AVERT region
should evaluate the emission changes for all regions that state is a part of. Appendix G includes
further discussion of the regions and more complete instructions for users analyzing expected
changes from energy policies in split states.
31 Excludes EGUs smaller than 25 MW that do not report to CAMD.
State and Local Climate
and Energy Program
-------
J/VeW
AVoidod Einissiortt and qenaRailori Tool^^
www.epa.gov/avarl
AVERT User Manual Version 4.2
Figure 3. Map of AVERT's regions.
Table 1. AVERT regions, abbreviations, and states.
AVERT region
Full states
Partial states
California
California
—
Carolinas
South Carolina
North Carolina
Central
Kansas
Arkansas, Iowa, Louisiana, Minnesota, Missouri,
Montana, North Dakota, Nebraska, New Mexico,
Oklahoma, South Dakota, Texas
Florida
—
Florida
Mid-Atlantic
District of Columbia, Delaware,
Maryland, New Jersey, Ohio,
Pennsylvania, Virginia, West Virginia
Illinois, Indiana, Kentucky, Michigan, North
Carolina, Tennessee
Midwest
Wisconsin
Arkansas, Iowa, Illinois, Indiana, Kentucky,
Louisiana, Michigan, Minnesota, Missouri,
Mississippi, North Dakota, Oklahoma, South
Dakota, Texas
New England
Connecticut, Massachusetts, Maine,
New Hampshire, Rhode Island,
Vermont
—
New York
New York
—
Northwest
Idaho, Nevada, Oregon, Washington
Montana, Utah, Wyoming
Rocky
Mountains
Colorado
Montana, Nebraska, New Mexico, South Dakota,
Utah, Wyoming
Southeast
—
Alabama, Florida, Georgia, Mississippi
Southwest
Arizona
New Mexico, Texas
Tennessee
—
Alabama, Georgia, Kentucky, Mississippi,
Tennessee
Texas
—
Texas
^ ks'EFW
State and Local Climate
and Energy Program
-------
AVERT User Manual Version 4.2
Energy Policy Characteristics
AVERT users should understand the characteristics of the energy scenario that they want to
analyze in terms that can be input into the model; this is Step 2 in the AVERT Main Module.
AVERT analyzes the difference in hourly load (regional electricity demand) between a baseline
scenario and some intervention scenario created by the energy policy, resulting in an energy profile
consisting of hourly changes to fossil-fuel EGU generation. This in turn leads to a change in
emissions. Starting with a baseline schedule of load for every hour of a reference year (the "load
profile"),32 the Main Module guides the user to create or input an energy profile to represent their
energy policy—i.e., the amount that load that will be reduced or increased by the energy policy on
an hourly basis. For details and instructions, refer to Step 2 in Section 4.
Users are encouraged to create and adopt their own energy profiles, representing energy projects,
programs, or policies specific to their interest or area of concern. The following energy profiles are
built into the Main Module:
• Reduction of fossil-fuel generation by a chosen percent in some or all hours. This
option is recommended to represent a mix of EE programs that target some or all hours of
the year, but preferentially target higher hours with greater demand. Users can also use
this option to increase fossil-fuel generation by entering a negative number for the percent
generation reduction.
• Reduction of annual fossil-fuel generation by total gigawatt-hours (GWh) or by a
constant MW each hour. This option is recommended to represent a rough approximation
of baseload-only reductions where the total number of GWh reduced over the course of a
year is known and is expected to be equally distributed over all hours of the year. Users
can also use this option to increase fossil-fuel generation by entering a negative number
for the amount of generation reduction.
• Renewable energy. With this option, users can model onshore wind, offshore wind, utility
solar, and rooftop solar resources that are broadly representative of the selected region.
• Electric vehicles. Users can model the impact of EVs on power sector generation
changes and associated changes in emissions from the power sector. EVs are motor
vehicles that obtain some or all of their power supply from batteries, which are charged by
power plants on the electric grid. Users can model light-duty vehicles (LDVs), school
buses, and transit buses using a default charging profile for LDVs or buses. Users can also
define their own 8,760 hour charging profile. AVERT also estimates the avoided internal
combustion engine (ICE) vehicle emissions.
• Combination of energy policies. Users can also layer the above options together, as well
as combine the pre-set options with manually entered energy changes.
AVERT combines all of the user's inputs to generate a single energy profile with 8,760 hourly
values.33 This profile feeds into the calculations in the next step. Since the release of version 1.5 of
AVERT's Main Module, AVERT adjusts the energy profile to account for avoided transmission and
distribution line losses associated with certain resources that avoid the need for long-distance
32 Technically, within AVERT, the load profile represents aggregate fossil generation for a region, and not end use
consumption.
33 Or 8,784 in the case of leap years.
State and Local Climate
and Energy Program
-------
AVERT User Manual Version 4.2
transmission: specifically, EE, distributed PV systems, and EVs.34 The magnitudes of hourly load
changes associated with each of these resources are adjusted upward by the following formula:
adjusted load change = user's input / (1 - x),
where x is the regional average line loss percentage. Starting with Main Module version 2.3
(released in spring 2019), AVERT uses line loss factors from the Annual Energy Outlook, published
by EIA.35 AVERT uses the historical line loss factors that correspond to the year and region of the
user's analysis, as shown in Table 2.
Table 2. Transmission and distribution line loss factors used in AVERT.
Data year
Texas
Eastern
Interconnect
Western
Interconnect
2017
5.60%
7.00%
8.13%
2018
4.83%
6.74%
8.54%
2019
5.38%
7.20%
8.60%
2020
5.17%
7.58%
8.28%
2021
4.95%
7.50%
8.39%
2022
4.58%
7.23%
8.67%
The Eastern Interconnect corresponds to the following AVERT regions: Carolinas, Central, Florida,
Mid-Atlantic, Midwest, New England, New York, Southeast, and Tennessee. The Western
Interconnect corresponds to the following AVERT regions: California, Northwest, Rocky Mountains,
and Southwest. The Texas region in Table 2 corresponds to the AVERT Texas region.
This adjustment has the effect of increasing the magnitude of fossil load change and emissions
change associated with distributed RE generation and energy policies and programs that change
electricity consumer demand (e.g., EE and EVs). The adjustment provides more accurate results.
Without it, AVERT would underestimate changes to emissions. For example, AVERT without
adjustments would assume that 100 megawatt-hours (MWh) of EE or 100 MWh of onsite
(distributed) PV generation in the Eastern Interconnect in 2018 avoids 100 MWh of fossil
generation, whereas it actually avoids approximately 107 MWh of fossil generation after accounting
for the additional power that would have been generated and lost during transmission in order to
deliver 100 MWh to the end-user. Similarly, for this region and year, a fleet of EVs that requires 100
MWh to charge would be modeled at approximately 107 MWh of additional load due to
transmission losses.
Once an energy profile has been designed and the appropriate transportation characteristics
selected, if necessary, the user is prompted to begin the model run in Step 3 of the AVERT Main
Module.
34 AVERT treats wind and utility-scale PV as centralized resources that still require transmission and distribution to
end-users; thus, while they displace fossil generation, a simple assumption is that they do not avoid any line
losses.
35 Annual Energy Outlook data can be downloaded from https://www.eia.qov/outlooks/aeo/. Each Outlook provides
historical data for the preceding year. The transmission and distribution line loss factors are calculated as ((Net
Generation to the Grid + Net Imports - Total Electricity Sales)/Total Electricity Sales).
SERA
State and Local Climate
and Energy Program
-------
AVERT User Manual Version 4.2
Scenario Results
Step 4 of AVERT's Main Module reports the difference between the baseline and modeled energy
policy scenario through the following outputs:
• Summary tables - Power sector only:
o Annual changes to generation and emissions
o Emissions changes for top 10 fossil-fuel generation days
o Annual emissions changes by county
o Monthly emissions changes by county
o Daily NOx emissions changes by county
• Charts and figures - Power sector only:
o Map of generation and emissions changes
o Hourly emissions changes by week
o Monthly emissions changes by selected geography (region, state, or county)
o Signal-to-noise diagnostic
• Summary tables, charts, and figures - Power sector and avoided vehicle emissions data:
o Annual changes to generation and emissions - with vehicles
o Annual emissions changes by county - with vehicles
o Monthly emissions changes by county
o Annual emissions results by selected geography
o Projected CO2 emission rates over time
• COBRA text file generation (for public health impact modeling)
• SMOKE text file generation (for air quality modeling)
For assessing the air quality implications of energy policies, the location of air emissions changes
resulting from either load increases or decreases can be critical. The example shown in Figure 4
represents a 2,000 MW onshore wind program in the Midwest AVERT region compared with 2019
base-year data. The map displays annual changes in SO2 emissions from specific EGUs as blue
circles; larger circles indicate greater changes to emissions. Where multiple EGUs overlap on the
map (i.e., multiple units at one plant or several plants close together), the circles appear darker.
State and Local Climate
and Energy Program
-------
JtVert
AVoid«d Emissions and q»rwRatlon Tool^^
wvvw.epa.gov/avert
AVERT User Manual Version 4.2
Figure 4. Map of annual SO2 emissions reductions from an example wind program in the Midwest
region.
Annual Change in S02 (lbs)
Midwest
The diameter of each circle indicates the magnitude of an EGU's change in emissions. Circles are semi-
transparent; darker areas occur in regions with overlapping EGUs. Emissions reductions are indicated
with blue circles; increases in emissions are indicated with black-bordered white circles.
Increases in emissions are shown as black-bordered white circles. Increasing emissions may occur
because higher load is programmed into AVERT (e.g., for testing a higher baseline if reviewing the
impact of existing renewable portfolio standards, if shifting load from peak to trough hours, or
increases due to EVs), or due to aberrations in the statistical dataset.36
Many users will be interested in changes to emissions for smaller areas within a region. Monthly
results can be displayed by region, state, or county. Figure 5 displays monthly SO2 emissions
changes within a single county in Illinois, continuing the same example shown in Figure 4.
Figure 5. Monthly SO2 emissions reductions for Madison County, Illinois, from an example wind
program in the Midwest region.
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
s 11111 111 I I 1
-80,000 ^ ¦ I
-100,000
Some users may wish to instead view hourly changes to generation or emissions to understand the
behavior of the system at a finer scale or during specific hours of the year. Using the same 2,000
MW onshore wind program, Figure 6 displays the hour-by-hour fossil generation displaced in the
week of August 1-7 in the Midwest AVERT region from the wind project. Individual EGUs are color-
coded along a gradient from dark blue (baseload EGUs) to light blue (peaking EGUs).37 The EGUs'
36 Some units show increasing generation even as regional load decreases. This usually occurs with baseload units
during trough hours. It is primarily explained by maintenance outages that occur during periods of low demand,
but not necessarily at the lowest demand hour of the year. Statistically, these units appear likely to generate
slightly more at the lowest-load hours than at medium-low-load hours. Therefore, reducing demand from a
medium-low load to a very low load could appear to increase the output of these units.
37 The color coding is illustrative only. It is not portrayed on an absolute scale, and is relative to all units in the
region (i.e., if a region comprises 200 units, they will all be parsed along the gradient from light blue to dark blue;
if a region comprises 1,000 units, they will similarly be parsed along the same gradient). Units are sorted based
on annual capacity factor. Units with uniform generation across high and low load levels are closer to "high-
capacity-factor" behavior, while units with high output at high load levels and none at low load levels are closer to
"low-capacity-factor" behavior.
State and Local Climate
and Energy Program
-------
JtVert
AVoidod Einissiortt and qenaRailori Tool^^
www.epa.gov/avarl
AVERT User Manual Version 4.2
individual generation reductions are shown in stacked bar plots and the net total contribution is
shown with a yellow line. The yellow line represents the hourly energy displaced by the wind
project. Note that peaking (gas) EGLIs are primarily displaced during daytime hours, while
baseload (usually coal) EGUs are displaced during off-peak hours.
Figure 6. Hourly generation reductions in the week of August 1 from an example wind program in the
Midwest region.
1,000
500
-1,000
Change in Generation (MW) in Week of 8/1
Midwest
JluUL- kniudL* *jAflUkjJL JLkJMaM nihi il kMjLui a.
~ -1,500
~i
| -2,000
u
-2,500
-3,000
Negative numbers indicate
displaced generation and
emissions.
Total Change in
Generation (MW)
_Total fossil-fuel load,
pre impact
~
High capacity factor units
Low capacity factor units
Figure 7 displays SO2 emissions displaced in the same week from the same wind project. Note that
almost all of the EGUs portrayed in this figure are dark blue in color; SO2 reductions are primarily
captured from reductions at high-capacity-factor coal EGUs, which are displaced during off-peak
hours.38
Figure 7. Hourly emissions reductions in the week of August 1 from an example wind program in the
Midwest region.
Negative numbers indicate
displaced generation and
emissions.
Total Change in S02
(lbs)
Total fossil-fuel load,
pre impact
High capacity factor units
Low capacity factor units
3,000
2,000
1,000
I 0
rv
£ -1,000
c
Z -2,000
=
5 -3,000
o
-4,000
-5,000
-6,000
Change in S02 (lbs) in Week of 8/1
Midwest
38 High-capacity-factor baseload units operate during most hours of the year, and are (by definition) the bulk of the
units (or only units) online during off-peak hours, and therefore are the units that will be displaced in off-peak
hours when modeling a decrease to fossil generation. This analysis indicates that baseload units are displaced
during off-peak hours, but not during the daytime. During the day, gas plants (with no, or negligible, SO2
emissions) are displaced, and therefore do not appear in this graphic.
&EPA
State and Local Climate
and Energy Program
-------
cndVEi
AVERT User Manual Version 4.2
4. AVERT Main Module: Step-by-Step Instructions
This section provides step-by-step instructions for using AVERT's Excel-based Main Module to
estimate the change in emissions resulting from energy policies.
To begin, download two files and save them to your computer:
• The Main Module workbook: "AVERT Main Module.xlsb." Download the workbook at
https://www.epa.gov/avert.
• The RDF for the region under analysis.
o Default RDFs developed for use by EPA are labeled
"AVERT RDF [DataYear] EPA_NetGen25 ([Region]).xlsx"; they can be
obtained at https://www.epa.qov/avert.
o Regional analyses developed by advanced users using AVERT's Statistical
Module will be saved, by default, in a folder of the Statistical Module titled
"AVERT Output." These files use the following naming convention:
"AVERT RDF [DataYear] [RunName] ([Region]) [RunDateTime].xlsx."
For more detailed installation instructions and model specifications, see Appendix A of this manual.
AVERT Welcome Page
When launched in Excel, the Main Module opens to its "Welcome" page as shown in Figure 8.
Figure 8. AVERT Main Module "Welcome" page.
Welcome to AVERT's Main Module
AVERT is an EPA tool that quantifies the generation and emission changes of energy
policies and programs in the contiguous United States Please refer to the AVERT user
manual for details on step-by-step instructions, appropriate uses, and assumptions built
into the tool
AVERT
v>EPA
AVERT
NOTE
Please ensure macros are enabled on your computer
AVERT requires Excel 2007 of higher in Windows and Excel 2011 or higher on Mac
Synapse
Cccnomci, mc
AVERT V4.1
Developed by Synapse Energy Economics, Inc., March 2023
Use the blue entry to describe each scenario and keep track of multiple versions of AVERT
Editor:
Date edited:
Edition name:
Edition description:
Click here
to begin
Click here to hide
default Excel
functionality
23
SEPA
State and Local Climate
and Energy Program
-------
JtVert
AVoid«d Emissions and q»rwRatlon Tool^^
wwwepagovTavSt'^W
AVERT User Manual Version 4.2
The Main Module is primarily driven by macros to conserve memory and processing time.39 Before
you make any selections or begin calculations, macros must be enabled on your computer. This
must be done first, before any other steps; if macros are not enabled before you load an RDF (Step
1), you will need to close the workbook, re-open it, and then enable macros to continue.
To enable macros in Excel 2007 for Windows:
1. Click the Microsoft Office Button - , then click "Excel Options."
2. Click "Trust Center," click "Trust Center Settings," then click "Macro Settings."
3. Click on "Enable all macros."
To enable macros in Excel 2010 or later for Windows:
1. Click the "File" menu (Office Backstage), then click "Options" in the left sidebar.
2. Click "Trust Center" in the left sidebar, then click the "Trust Center Settings" button in the
main window.
3. Click "Macro Settings" in the left sidebar.
4. Choose the "Enable all macros" option and hit "OK."
To enable macros in Excel 2011 or later for Mac, select "Enable macros" in the dialog box that
appears when opening the file.
Next, we recommend that you personalize the Welcome page with details about the user, the date
of use, and the energy policy for which emissions changes are to be estimated. This version
specification is very useful for keeping track of multiple versions of AVERT saved to the same
drive. Please note that the version of AVERT available from EPA does not contain an RDF, and is
thus fairly small (<12 MB). When an RDF is loaded into AVERT and changes are calculated, the
file grows substantially to 50 to 170 MB.
A final note on the Welcome page: The Main Module has been designed to provide a clear user
interface. While it is an Excel file, the tabs that drive the calculations for AVERT are hidden to
enhance the usability and appearance of the tool, making it more similar to a web-based or
executable program. For users who prefer full Excel functionality while using AVERT, there is a
button at the lower right-hand side of the Welcome page that reads "Click here to restore default
functionality." You can complete the steps required to estimate changes to emissions regardless of
whether or not full Excel functionality is visible.
Click on the button labeled "Click here to begin" to move on to AVERT's first step.
39 Due to the large number of calculations—typically several hundred EGUs times 8,760 hours times five output
variables, anywhere from 4 to 40 million complex calculations—storage in a dynamic Excel environment would be
analytically burdensome and space-intensive. Therefore, most calculations are performed once in a Visual Basic
environment, and are not stored in memory.
State and Local Climate
and Energy Program
-------
cndVEi
AVERT User Manual Version 4.2
Step 1: Load Regional Data File
Next, choose the region of analysis in Step 1 (as shown in Figure 9) and load the corresponding
RDF. There are 14 AVERT regions to choose from, described in detail in Section 3. (For details on
how the regions were developed, refer to Appendix B.)
Figure 9. Image of the AVERT Main Module's "Step 1: Import Regional Data File" page.
Step 1: Import Regional Data File
Select region
Select a region for analysis by using the
dropdown or by clicking the map.
If vou haven't vet downloaded a Regional Data
File, click here
Enter filepath
Double-click below to enter the location
of the Regional Data File.
Load data
Click here to load the
Regional Data File
Welcome
1. Regional Data
File
2. Set Energy
Impact Profile
3. Run Impacts
4. Display Outputs
The choice of a region determines both which EGUs are included in the analysis and which default
renewable resources are available for modeling the effects of energy policies.
The 14 AVERT regions are:
California
Caroiinas
Central
Florida
Mid-Atlantic
Midwest
New England
New York
Northwest
Rocky Mountains
Southeast
Southwest
Tennessee
Texas
If you have not yet downloaded an RDF, you can either select a region of interest from the
dropdown menu or by clicking on the region in the map, then click the hyperlink in the bottom left of
the page.
Then double-click on the filepath box and navigate to the folder where the RDF has been saved.
Once these steps have been completed, click on the green button labeled "Click here to load the
Regional Data File." This step may take several minutes to complete, depending on the size of the
region and the speed of the computer used. When the RDF has been loaded, a pop-up box
appears containing information about the loaded AVERT region and indicating "Import complete."
25
SEPA
State and Local Climate
and Energy Program
-------
JtVert
AVoid«d Emissions and q»rwRatlon Tool^^
wwwepagovTavSt'^W
AVERT User Manual Version 4.2
The pop-up box confirms the region and data year loaded by the user and indicates the number of
reporting units in the analysis. The pop-up box also specifies the states that are fully represented
and partially represented in a loaded AVERT region, as shown in Figure 10.
Figure 10. RDF import pop-up box example for the 2019 Midwest AVERT region.
AVERT X
Import complete.
You have loaded the 2019 Midwest (MIDW) Regional Data File, This region
contains 643 fossil units,
Load from the following states is fully represented in this AVERT region: Wl
Load from the following states is only partially represented in this AVERT region:
AR, IA IL IN, KY, LA Ml, MN, MO, MS, ND, OK, SD, TX
Appendix G of the User Manual describes the methodology for assessing the
changes resulting from of energy policies and programs that are represented in
multiple AVERT regions.
Click the red "Next" button to continue.
OK
Some states are divided by AVERT regions. For example, parts of Illinois are in the Midwest
Region, while other parts are in the Mid-Atlantic Region.40 Appendix G of this document describes
the process that one should use to determine the emission changes resulting from energy policies
in states that are partially represented in any one AVERT region. Users analyzing these states
should assign the energy policies proportionally based on the fraction of the state's electricity sales
within each relevant AVERT region. These fractions are shown in Table 6 in Appendix G and, as a
reminder and for clarity, in the pop-up box.
After you load an RDF, the blue AVERT header bar will indicate the region and data year in the top
left corner (e.g., "Midwest, 2019"). The blue footer bar will also indicate the name of the AVERT run
that has been loaded in the RDF (e.g., "EPA_NetGen_PMVOCNH3"). When you run the Statistical
Module, you will be able to give runs alternative names.
All four of AVERT's steps have a navigation panel on the right-hand side. You can click on any step
or click "Next" or "Back" to move through the steps of the model. To move on to the next step, click
"Next."
40 In this case, the division in Illinois is primarily a function of the eastern part of the state falling under the PJM
Regional Transmission Organization, while the rest of the state falls into MISO. Other states may have similar
situations, where one part of the state is associated with one set of balancing authorities (and therefore one
AVERT region) while another state is associated with a different set of balancing authorities and is thus placed in
a different AVERT region.
-------
AVERT User Manual Version 4.2
Step 2: Set Energy Scenario
In this step, you will create an energy profile (schedule of changes to electricity demand for every
hour of a year) depicting the change in load expected from an energy policy as shown in Figure 11.
In this example, the user has input a 5,000 MW onshore wind program with 50,000 battery EVs, the
energy profile of which is being displayed in the chart on the right side of the screen.
Figure 11. Image of the AVERT Main Module's "Step 2: Set Energy Impacts" page.
Step 2: Set Energy Scenario
r
DIRECTIONS: Enter the energy efficiency and/or renewable energy and/or electric vehicle changes for
one or more policies, programs, and/or projects
Each entry is additive creating a single energy change profile
To modify each hour manually click the button on the nght
For further instructions, consult Section 4 of the AVERT User Manual.
Enter EE based on the % reduction of regional fossil generation
Enter detailed load
^change^
data by hour
Reduce generation by a percent in some or all hours
Apply reduction to top X% hours:
Reduction % in top X% of hours:
% ol top hours
% reduction
Changes m Hourly Energy
i i J ! ! f r ? I s s ? f
And/or enter EE distributed evenly throughout the year
Reduce generation by annual GWh
o
GWh
OR
Reduce each hour by constant MW
0.0
MW
And/or enter annual capacity of RE resources
Onshore wind capacity
5000
MW
MW
MW
MW
Kdit capacity
factors
Offshore wind not available
Utility solar PV capacity
0
Rooftop solar PV capacity
0
The current!, entered reduction profile equals 13,816 GWTi. or 2.7% ol regional fossil generation
Table 1. Sales and
Holann
al vehicle sales
% of registered vehicles
S of EVs sold in 2021 in
Battery EVs
50,000
•lock comparison
rtire region
in entire region
Midwest
Enter number Plua-in hvbnd EVs
0
22*
01%
100 0%
of vehicles Electric transit buses
0
TrartM tu«5
0.0%
0.0%
Electric school buses
0
data
School tuKi
00%
00%
Select location of EV deployment
Entire Region
¦
table 3. EERE IV
Historical additions
CERE required to offset
EERE required divided by
comparison for
•vg »19 J0?1|
EV demand
historical additions
UW Grtft
MW GVVft
EE i retail
606
5J12
8 67
1% 1%
Onsnore Wr>s
3,457
9 586
40 113
1% 1%
Utnty Soiar
657
US9
8 16
1% 1%
Total
4.768
16.770
56 195
P.Vld;M
3. Run Scenario
4. Display Results
Next ->
-------
cndVEi
AVERT User Manual Version 4.2
Figure 12. Image of the AVERT Main Module's "Step 2: Set Energy Impacts" page with program size
resulting in a more than 15 percent reduction in fossil load in some hours.
Step 2: Set Energy Scenario
DIRECTIONS: Enter the energy efficiency and/or renewable energy and/or electric vehicle
changes for one or more policies, programs, and/or projects.
Each entry is additive, creating a single energy change profile
To modify each hour manually, click the button on the right.
For further instructions consult Section 4 of the AVERT User Manus
_EnterdetailecHoa<^hange_dat^>yJ>ou^
Enter EE based on the % reduction of regional fossil generation
Reduce generation by a percent in some or all hours
Apply reduction to top X% hours:
Reduction % in top X% of hours:
% of top hours
% reduction
n! Energy change piofile exceeds 15X of fossil generation in or
Changes in Hourly Energy:
l,S.£££<£2i.l.:
or more hours (see below).
i & a
And/or enter EE distributed evenly throughout the year
Reduce generation by annual GWh[
Reduce each hour by constant MW] 0 o
And/or enter annual capacity of RE resources
Onshore wind capacity:
12000
MW
Offshore wind not available
MW
Edit capacity
Utility solar PV capacity.
0
MW
factors
Rooftop solar PV capacity
0
MW
And/or enter electric vehicle (EV) date
The currently entered reduction profile equals 33,412 GWh, or 6.5% of regional fossil
generation.
salelfnenMieTeglin
V. of leglsteied
Enter Battery EVs:
. r . Plug-in hybrid EVs
vehicles Electric transit buses.
Electric school buses
50.000
View detailed EV
data
0
Light-duty vehicles
Transit buses
School buses
2. OX
ooy.
0.0 X.
0.1X
OOY.
0.0V.
0
0
Select location of EV deployment
Entire Region
¦
Table 2. EERE EV
(Annual Aug. 2018-
?n?m
EERE lequiied to EERE lequiied divided
offset EV demand by historical additions
Entile Region
/W GUft
!W &-W GWh
EE (retail)
626 5,481
10 83 2V. 2v.
Onshore Wind
3,074 8,615
43 121 iy. r/.
Utility Solar
393 813
6 11 ti 1X
Total
4,142 15,334
58 216
EPA NetGen PMVOCNH3
1. Regional Data
File
2. Set Energy
Scenario
3. Run Scenario
KJ
4. Display
Results
I Next -> I
I Back I
The Step 2 page allows you to estimate a change in load from basic characteristics:
Enter hourly data manually (see green button in right-hand corner)
Reduce fossil-fuel generation by a percent in some or all hours
Reduce fossil-fuel generation by total GWh (flat)
Reduce fossil-fuel generation by a constant MWin each hour
Add RE capacity (utility solar, distributed solar, onshore wind, and offshore wind)
Add EV charging load by specifying the number of EVs (light-duty battery-powered EVs
[BEVs], light-duty plug-in hybrid vehicles [PHEVs], electric transit buses, and electric
school buses), and location of EV deployment
• Combination of energy policies including combining pre-set options with manual entry
Choose the option(s) that works best for your modeled energy policy and fill in the necessary data.
Each option is described in more detail below. If you choose more than one option (including
manual entry), the selected options will be combined into a portfolio of programs. For any program,
or combination or programs, the Step 2 page returns a total annual energy change (in GWh)
achieved by the modeled energy policy inclusive of appropriate T&D losses. Total hourly changes
can be found in the manual input page.
Note that it is not recommended for energy profiles to exceed 15 percent of fossil load in any given
hour. If a user-entered energy profile exceeds these recommended limits, a caution message will
appear. The graph titled "Changes in Hourly Energy" will also indicate the affected hours.
Exceeding the 15 percent threshold in one or more hours does not prohibit the user from
proceeding with the calculations.
28
SEPA
State and Local Climate
and Energy Program
-------
JtVert
AVoid«d Emissions and q»rwRatlon Tool^^
wwwepagovTavSt'^W
AVERT User Manual Version 4.2
Manual User Input
If the hourly energy changes expected from a particular energy policy, program, or measure are
known, a customized profile can be created (consisting of 8,760 hourly values for a non-leap year
or 8,784 hourly values for a leap year). For example, you might use this approach to test the
measured or modeled emission changes from a particular known wind farm or EE program. To
enter data manually or cut and paste data from another source, click on the green button that reads
"Enter hourly data manually" to get to a manual energy impact data entry screen (Figure 13). Data
are entered as a single column of values from midnight on January 1 through 11 p.m. on December
31. On this page, as with other AVERT inputs, positive values represent displacements. Users who
wish to model scenarios with increases in load (for example, a retroactive "what-if scenario to see
what would have happened if a particular energy policy or program had not been implemented) can
enter negative displacement values to represent increased loads.
This page also includes two columns that indicate whether a user has exceeded the recommended
and/or calculable ranges of hourly load changes. Alerts will appear in these two columns if the data
entered by the user a) produces a cumulative load change that exceeds 15 percent beyond the
upper and lower limits of each hour's original fossil load or b) produces a new load that is outside
the range of AVERT's ability to calculate generation and emissions changes.41
The "Total Change" column on the manual input page shows the total aggregate hourly energy
change from the programs input or selected by the user.
41 AVERT can estimate generation and emissions changes within the range of actual observed load for a certain
year. In addition, it uses extrapolated data to estimate the changes that could occur down to a load level of 0 MW
and up to a maximum load level associated with all plants within a single region running at full capacity. It is
unable to estimate changes outside of this maximum and minimum range.
SERA
State and Local Climate
and Energy Program
-------
Bsd^SRT
missions and qenaRaiion Tool^^
www.epa.gov/avan
AVERT User Manual Version 4.2
AVoidad Emissions.
Figure 13. Image of the AVERT Main Module's "Manual Energy Impact Data Entry" screen.
Manual Energy Impact Data Entry
When complete, click here to return to
^^_^^^tep^EnteMmpacts^^_^_
Delete all manual data
Date »
Hour *
Day of Wee '
Regional Fossil Load (MW) -
Manual Profile (MW)
Total Change (MW)
Larger than 15%? »
Outside of Range? »
1/1/2019
1
Tuesday
38,709
0
1/1/2019
2
Tuesday
37,264
0
1/1/2019
3
Tuesday
37,166
0
1/1/2019
4
Tuesday
37,596
0
1/1/2019
5
Tuesday
38,897
0
1/1/2019
6
Tuesday
40,849
0
1/1/2019
7
Tuesday
42,614
0
1/1/2019
8
Tuesday
44,490
0
1/1/2019
9
Tuesday
46,857
0
1/1/2019
10
Tuesday
50,031
0
1/1/2019
11
Tuesday
52,298
0
1/1/2019
12
Tuesday
53,460
0
1/1/2019
13
Tuesday
54,975
0
1/1/2019
14
Tuesday
55,452
0
1/1/2019
15
Tuesday
55,933
0
1/1/2019
16
Tuesday
56,603
0
1/1/2019
17
Tuesday
57.998
0
1/1/2019
18
Tuesday
59,138
0
1/1/2019
19
Tuesday
60,148
0
1/1/2019
20
Tuesday
60,750
0
1/1/2019
21
Tuesday
58,115
0
1/1/2019
22
Tuesday
55,648
0
1/1/2019
23
Tuesday
54,864
0
1/1/2019
24
Tuesday
51,120
0
1/2/2019
1
Wednesday
47,477
0
1/2/2019
2
Wednesday
45,974
0
1/2/2019
3
Wednesday
45,030
0
1/2/2019
4
Wednesday
45,169
0
1/2/2019
5
Wednesday
46,412
0
1/2/2019
6
Wednesday
49,689
0
1/2/2019
7
Wednesday
54,014
0
1/2/2019
8
Wednesday
56,851
0
1/2/2019
9
Wednesday
57,885
0
Reduce Generation by a Percent in Some or All Hours42
To estimate the emission changes expected from a broad-based EE or demand response program
targeting high-cost peak fossil-consumption hours, enter two values: a) the fraction of hours to
which load reduction is applied (with reductions applied to the highest fossil-fuel generation hours
first) and b) the percent by which those hours should be reduced. For a broad-based efficiency
program, the fraction of hours that experience reductions is very high; for a demand response
program, it is very low.
Note that, when the percentage options are used, reductions are a share of fossil-fuel generation
and not total system load (i.e., consumption). Therefore, using a 2 percent reduction per hour
effectively reduces load by 2 percent of fossil-fired generation.
To simulate a broad-based efficiency program, enter 100 percent as the "% of top hours" and an
estimated load reduction fraction in the cell labeled "% reduction." A graph of the selected energy
profile—which combines both manual entries and user selections made on the Step 2 page—is
shown on this page.
To simulate a peak-reduction targeting program such as demand response, enter a fraction of high-
demand hours that the program is expected to affect, and the load reduction (as a fraction of
peaking load) that would be targeted in those hours. This type of scenario is recommended for
programs that emphasize reductions in peak hours. Generally, reductions exceeding 15 percent of
fossil-fired generation are not recommended.43
42 Users wishing to model load increases (e.g., as a result of hourly load shifting) can enter negative numbers. This
applies to all input fields.
43 AVERT is designed to review marginal changes in a system, not large-scale changes. At some point, significant
changes in load will result in non-marginal changes, such as the decommitment of units.
SERA
State and Local Climate
and Energy Program
-------
JtVert
AVoid«d Emissions and q»rwRatlon Tool^^
wwwepagovTavSt'^W
AVERT User Manual Version 4.2
Figure 14 shows a baseline energy profile (in black) and two example load reductions for three
days in August. The blue, dotted line represents a 3 percent reduction in fossil-fired generation
requirements in every hour of the year. The red, dashed line represents an 8 percent reduction in
the 5 percent of hours with the highest demand for fossil-fired generation.
Figure 14. Examples of two load reduction programs as a percentage of some or all hours.
70,000
60,000
TJ
c
rt
E
cu
Q
~rt
c
O
"w>
cu
OL
30,000
20,000
10,000
¦8% reduction in top 5th percentile hours
3% reduction in all hours
¦ Original fossil-fuel load
Aug I,
12 AM
Aug I,
12 PM
Aug 2,
12 AM
Aug 2,
12 PM
Aug 3,
12 AM
Aug 3,
12 PM
It is important to note that reducing load by a percent in some or all hours measures a reduction
relative to hourly fossil-fuel load, and not system demand (i.e., consumption). It is important to
ensure that the total reduction (in annual GWh) or peak reduction (in MW) comports with
expectations of a reasonable fractional reduction. For example, if fossil-fuel load only accounts for
50 percent of regional generation, a user attempting to find the emissions changes of a regional
load reduction of 3 percent should double the size of the reduction specified in AVERT's Main
Module (i.e., to 6 percent) to scale from total fossil-fuel load to regional load. Similarly, to estimate
the effect of a portfolio reduction (i.e., a percentage change) where you have an expectation of the
total annual reduction (in MWh or GWh), find the correct percentage reduction such that the total
amount of energy reduced is equal to the expected quantity.44 In all situations where EE is being
modeled, AVERT increases the value entered to reflect a transmission and distribution loss factor,
under the assumption that the user is intending to model reductions in sales, even though the
impact modeled is calculated based on the quantity of fossil generation in that region.
44 For a program expected to accomplish an annual target reduction through a portfolio efficiency program, the user
should find the percent reduction that will accomplish this target. For example, using Texas data from 2012, a
portfolio EE program expected to reach 10,000 GWh of annual reduction (Ereq) would require a fractional
reduction of 3.79% in all hours of the year (Freq). To find this percentage value (Freq), the user can choose an
arbitrary estimated fraction (e.g., 2%, Fest), and review the text below the graphic on the Step 2 page indicating
the GWh reduction (Eest)- Divide the required annual reduction (Ereq) by the achieved estimated reduction (Eest),
then multiply this value by the estimated fraction (Fest)- The resulting value is the percentage that should be
entered into the "% reduction" box to achieve the desired energy reduction. Check to ensure that the new total
energy reduction value (in the text below the graphic) is consistent with the desired results.
State and Local Climate
and Energy Program
-------
JtVert
AVoid«d Emissions and q»rwRatlon Tool^^
wwwepagovTavSt'^W
AVERT User Manual Version 4.2
Reduce Generation by Annual GWh
You may have an estimate of the total amount of energy that is targeted or required to be reduced
by a program in a given year, but lack information about the distribution of those reductions over
the course of the year. "Reduce generation by annual GWh" simply distributes those savings
evenly over all hours of the year. The user inputs a total number of GWh expected to be saved in a
single year. This may be a highly erroneous assumption if savings are targeted from residential or
commercial customers, for whom EE measures tend to target peak use reductions. However, an
industrial or refrigeration efficiency program may be well represented by a constant reduction
across most hours of the year. Use this option with close attention to the types of programs
assumed in your analysis.
Reduce Each Hour by Constant MW
This option is identical in effect to the annual load reduction by total GWh above. The user selects
a constant reduction for every hour of the year in MW.
Renewable Energy
This option allows you to model onshore wind, offshore wind, utility-scale solar PV, and rooftop-
scale solar PV using regionally specific energy profiles. Select the annual capacity (maximum
potential electricity generation) for each type of resource, measured in MW. The model applies
these values to "hourly capacity factors" that vary by resource type and region. Hourly capacity
factors are the likelihood that a resource is generating at its full capacity in a given hour of the year.
For example, solar panels might have a 90 percent or higher capacity factor on an August
afternoon, but a 0 percent capacity factor at midnight any day of the year. The data and
methodology used to develop these capacity factors are described in Appendix C.
You can scale the capacity factors used for each RE technology by clicking on the "Edit capacity
factors" button to get to a manual capacity factor entry screen (Figure 15). For each resource, the
default average annual capacity factor on this page reflects the average capacity factor across all
hours of the year. The user can input their desired average annual capacity factor and the hourly
values used in the model will scale accordingly. For example, if the default capacity factor for
onshore wind in a region is 25 percent and a user enters a value of 30 percent, all hourly capacity
factors for that resource in that region are then scaled up by 20 percent.45 Capacity factors can be
revised upwards or downwards.
45 In some situations, this change may result in capacity factors that are higher than 100% in some hours. In these
cases, capacity factors are bounded at 0 and 100%. If this occurs for many hours, the "modeled" capacity factor
may not equal the "input" capacity factor. In these situations, the model will display a warning that the desired
annual capacity factor is unable to be modeled.
State and Local Climate
and Energy Program
-------
JtVert
AVoid«d Emissions and q»rwRatlon Tool^^
wwwepagovTavSt'^W
AVERT User Manual Version 4.2
Figure 15. Image of AVERT's "Manual Renewable Energy Capacity Factor (CF) Entry" screen.
Midwest, 2019 AVERT
Manual Renewable Energy Capacity Factor (CF) Entry
DIRECTIONS: Enter the specific CF for renewable energy project
Enter annual average capacity factor (%}: | Default
Input
Modeled
Onshore Wind: 41.22%
Offshore Wind: 0.00%
Utility Solar PV: 23.60%
Rooftop Solar PV: 18.67%
When complete, click here to return to
^^teg^_Enter_Enej2^_[m£acts__
Electric Vehicles
This option allows users to set the number of EVs they would like to model, as well as where the
vehicles will displace emissions from ICE vehicles. More detail on the calculations related to EV
impacts can be found in Appendix J. For information on limitations related to modeling the emission
impacts of EVs in AVERT, see Appendix K.
Number of Electric Vehicles
In Step 2, users can specify a number of EVs to model, which is the number of EVs they expect to
be added to the road in their scenario. Users have the option of entering inputs for four different
vehicle types:
• Light-duty BEVs46
• Light-duty PHEVs
• Electric transit buses
• Electric school buses
More detailed inputs related to the type of BEVs being modeled (e.g., cars versus trucks) and the
fuel type of conventional vehicles being displaced (e.g., gasoline ordiesel) are described in
Appendix J.
Next to the number of vehicle input cells in Step 2 are two tables that serve as guides for the user.
Table 1 translates the inputted number of EVs into shares of vehicle sales and shares of vehicles
on the road (i.e., vehicle stock). These shares are based on recent historical data aggregated for
the location of EV deployment selected by the user.47 For example, if a user models 10,000 BEVs
46 A "light-duty vehicle" or "LDV" is any vehicle weighing less than 8,500 lb. It includes smaller vehicles that may
colloquially be called cars, such as sedans, compacts, hatchbacks, and small or compact SUVs. It also includes
larger vehicles that may colloquially be called trucks, such as pickup trucks, minivans, or medium or large SUVs.
Broadly speaking, an LDV can be thought of as any type of vehicle that a person or household might purchase
for personal use (although it could of course be used for business or governmental use).
47 2021 sales: Alliance for Automotive Innovation. 2022. Advanced Technology Vehicle Sales Dashboard. Data
compiled by the Alliance for Automotive Innovation using information provided by IHS Markit (2011-2018, Nov
2019-2022) and Hedges & Co. (January 2019-October 2019). Data last updated December 21, 2022. Retrieved
March 23, 2023. Available at: https://www.autosinnovate.ora/resources/electric-vehicle-sales-dashboard. 2021
SERA
State and Local Climate
and Energy Program
-------
JtVert
AVoid«d Emissions and q»rwRatlon Tool^^
wwwepagovTavSt'^W
AVERT User Manual Version 4.2
in the New York region, which has about 100,000 LDV sales per year and about 10 million LDVs on
the road, the shares reported will read 10 percent and 0.1 percent, respectively.
Table 2 provides a comparison between the total annual energy impact of the EVs entered and
recent trends in RE capacity installation and EE programs. Using the calculations described in the
Calculations section of Appendix J. AVERT converts the number of inputted vehicles into an annual
GWh quantity. AVERT then compares this GWh quantity against the average first-year GWh
generation from wind, solar, and EE resources deployed in the selected state or region between
2019 and 2021. For example, users modeling an addition of 10,000 BEVs in New York will find that
the load impact of these vehicles charging is about 50 GWh. Meanwhile, the average first-year
GWh generation of wind and solar projects deployed in 2019-2021 is about 400 GWh. In this
example, Table 2 helps a user note that the generation required to power 10,000 BEVs is about
one-eighth of the annual energy generated from the recent additions from wind, solar, and EE.
Table 2 helps users build more likely scenarios combining EVs, EE, and RE. See the Calculations
section in Appendix J for more information about how impacts from different resources are
combined in AVERT.
Location of EV Deployment
The second primary input is the location of EV deployment. EVs may be deployed in the "Entire
Region" or in one of the states that is a component of the AVERT region (see Table 1 for
information on which states are in each AVERT region). While AVERT's power sector modeling
algorithm is agnostic to where electricity load changes occur within an AVERT region, this input
parameter determines where emission decreases from ICE vehicles occur. The default selection is
Entire Region. If the default is selected, vehicle emission reductions are assumed occur throughout
the region. Emission reductions are allocated to each county based on the vehicle miles traveled
(VMT) in each county relative to the total VMT in the AVERT region. If instead a state is selected,
vehicle emission reductions will occur only in the specified state (or portion thereof for states that
exist in more than one AVERT region). Emission reductions within the state are allocated to each
county based on the VMT in each county, relative to the total VMT in that state (or the portion of
that state that lies within the selected AVERT region).48
This parameter also informs Table 1 on Step 2 of the Main Module, as described above in the
Number of Electric Vehicles section.
After the energy profile has been designed, click "Next" to move to the next step.
stock (cars): Federal Highway Administration. 2023. State Motor-Vehicle Registrations - 2021. MV-1. Available at:
https://www.fhwa.dot.gov/policvinformation/statistics/2021/mv1.cfm. 2021 stock (LDV trucks): Federal Highway
Administration. 2023. Truck and Truck-Tractor Registrations - 2021. MV-9. Available at:
https://www.fhwa.dot.gov/policvinformation/statistics/2021/mv9.cfm. Total LDV sales are calculated based on EV
sales and the EV market share. LDV car stock data include private, commercial, and publicly owned automobiles.
LDV truck stock data include private and commercial pickups, vans, sport utilities, and other light vehicles.
Federal Transit Administration. National Transit Database. "2021 Vehicles" available at
https://www.transit.dot.qov/ntd/data-product/2021-vehicles. School bus fleet. 2023. "Fact Book 2023." Available
at: https://stavwell.mvdiqitalpublication.com/publication/?m=65919&i=771183&p=16&ver=html5.
48 For example, if an analyst working in the Mid-Atlantic RDF is interested in modeling EV deployment in only the
state of Illinois, the analyst will find that emission reductions associated with EVs will be allocated only to the 27
Illinois counties that are also assigned to the Mid-Atlantic RDF (there are 102 counties in Illinois). The total of
these vehicle emission reductions for this scenario is apportioned to each county relative to the 27 counties in the
state and AVERT region.
State and Local Climate
and Energy Program
-------
cndVEi
AVERT User Manual Version 4.2
Prior to advancing to the next step, an alert box may pop up (see Figure 16). This alert is triggered
when users' scenarios include EVs but do not include any amount of energy efficiency or
renewable energy (EERE). The alert, as seen in Figure 16, recommends that users include EERE
when modeling EVs with AVERT and points to Table 2 (as seen in Figure 12) as a reference.
When users see this message, they may click "Ok" to continue to Step 3, or "Cancel" to return to
Step 2 to add EE and RE estimates.
Figure 16. Pop-up message regarding modeling EERE alongside EVs.
AVERT
Alert! You have entered a quantity of EVs, but have not entered any
energy efficiency or renewable energy.
Recent trends show significant amounts of energy efficiency and
renewables coming online. Consider adding these resources alongside
EVs in orderto examine the portfolio effects of adding multiple
resources at the same time.
Table 2 summarizes average annual additions of EE and RE in recent
years and compares these with the EERE required to offset your entered
EV demand,
For more ideas on how to model EVs in AVERT, see Appendix J in the
AVERT user manual.
Click OK to proceed or click Cancel to return to Step 2,
X
OK
Cancel
Step 3: Run Scenario
Step 3 launches the automatic calculation of hourly changes in generation, heat input, and
emissions for each EGU in the region as shown in Figure 17.
35
SEPA
State and Local Climate
and Energy Program
-------
Bsd^SRT
missions and qenaRaiion Tool^^
www.epa.gov/avan
AVERT User Manual Version 4.2
AVoidad Emissions.
Figure 17. Image of the AVERT Main Module's "Step 3: Run Impacts" page.
Midwest, 2019
Step 3: Run Impacts
Click below to calculate changes to generation and emissions.
NOTE
Please be patient.
This calculation may take up to ten minutes to run on older machines.
During this time your screen may go blank or a "not responding" error
may occur - please disregard and allow the calculation to continue.
Click here to calculate changes to generation and
emissions
AVERT
Welcome
1. Reqional Data
File
2. Set Energy
Impact Profile
3. Run Impacts
I J
4. Display Outputs
Next -»
<- Back
EPA NetGen PM25
Click on the box labeled "Click here to calculate changes to generation and emissions." Because of
the large amount of data being processed, this calculation may take several minutes depending on
the size of the region and the processing speed of the computer. A status bar in the lower left
corner indicates the share of processing completed.
Note that two separate alert boxes may pop up when this box is clicked. The first alert informs the
user that for at least one of the hours under consideration, the load decrease or increase is more
than 15 percent of the original hourly load (see Figure 18). Users can click "OK" and modify their
load changes to be within the recommended range, or click "Cancel" to proceed with the
calculation.
Figure 18. Pop-up alert for an energy profile that exceeds a 15 percent change in load.
You have entered an energy impact profile that exceeds the
recommended range. Press OKto adjust the magnitude of the energy
impact, or Cancel to continue with the calculation.
OK | Cancel
The second alert informs the user that in at least one of the hours where load has been adjusted,
the energy profile will be outside the calculable range for AVERT (see Figure 19). In these
6 SrEPA
State and Local Climate
and Energy Program
-------
JtVert
AVoid«d Emissions and q»rwRatlon Tool^^
wwwepagovTavSt'^W
AVERT User Manual Version 4.2
situations, the user must return to Step 2 and reduce his or her load adjustments to avoid
producing this alert.49
Figure 19. Pop-up alert for an energy profile that exceeds AVERT's calculable range.
AVERT
X
You have entered an energy impact profile that exceeds the calculable
range for hourly load impacts. Change your hourly load impact to
eliminate any errors in the Outside of Range column.
OK
Cancel
A pop-up box that reads "Calculation complete" will appear once the calculations are complete.
Click "Next" to move on to the final step.
Step 4: Display Results
Step 3 of AVERT's Main Module generates raw data in the form of hourly estimated changes in
generation, heat input, and emissions of PM2 5, SO2, NOx, VOCs, NH3, and CO2 for each EGU in
the region. These data are aggregated in charts and tables, which are then accessed through the
"Step 4: Display Results" page, as shown in Figure 20.
49 To remedy this error, users may find it useful to review the rightmost column on the "Manual User Input" page,
which produces an alert for specific hours where the load change has exceeded AVERT's calculable range.
State and Local Climate
and Energy Program
-------
osd^ERT
AVoidod Emission* and qenaRailon Tool^^
www.epagov/avert
AVERT User Manual Version 4.2
Figure 20. Image of the AVERT Main Module's "Step 4: Display Results" page.
Midwest, 2019
Step 4: Display Results
Summary tables - Power sector only
Annual regional results
Results for top ten
peak days
Annual results by
Monthly results by
county
county
Daily NOx results by county
Charts and figures - Power sector only
Map of generation and
emissions changes
Hourly results by week
Monthly results by
selected geography
Signal-to-noise
diagnostic
COBRA text file generation
Enter a filepath, then click
the button to save a
COBRA text file.
Please be patient.
This calculation may take
up to twenty minutes to
run on older machines.
Generate COBRA
text files
SMOKE text file generation
Enter a filepath, then click
the button to save
SMOKE text files.
Please be patient.
This calculation may take
up to twenty minutes to
run on older machines.
Generate SMOKE
text files
Welcome
1. Regional Data
File
2. Set Energy
Scenario
3. Run Scenario
4. Display
Results
Back
4- Start new
scenario
Summary tables, charts, and figures -
Power sector and avoided vehicle emissions data
Annual regional results -
with vehicle
Annual county results -
with vehicle
Results by selected
geography
Results by month
Emission rates over time
EPA_NetGen_PMVOCNH3
To display outputs, click on each of the green boxes under the headings: summary tables, charts
and figures, COBRA text file generation, and SMOKE text file generation. The sample tables and
figures shown below use the same example of a 2,000 MW onshore wind installation in the
Midwest AVERT region.
Summary Tables
Annual regional results
Figure 21 shows a high-level summary of the results of the analysis. This table displays the total
annual generation, heat input, and emissions from the region's fossil generation fleet as reported in
the base year ("Original") and as calculated by AVERTS Main Module after the modeled energy
change ("Post Change"). The last data column, labeled "Change," is the delta of "Original" and
"Post Change" and is the total annual expected change from the user-defined scenario. The chart
also shows total annual fossil-fuel emissions and emission rates for PM: s, SO2, NO*, VOCs, NH3,
and CO2. Emissions and emission rate data is shown separately for total annual NOx and ozone
season NOx, a subset of total annual NOx emissions occurring May 1 to September 30. Emission
rate data in the first column (under the "Original" heading and labeled as "Average Fossil")
describes the average emission rate associated with fossil-fired plants in the selected AVERT
SERA
State and Local Climate
and Energy Program
-------
bJ^ERT
AVoidod Einissiortt and qenaRailori Tool^^
www.epa.gov/avarl
AVERT User Manual Version 4.2
region in the original baseline of the selected year's data.50 Fossil-fuel emission rates presented in
the third column (under the "Change" heading and labeled as "Marginal Fossil") are the change in
emissions divided by the change in generation, resulting from the user-specified scenario.
Emission rates are not shown for the second column (under the "Post Change" heading) because
these are frequently very similar to the emission rates calculated in the first column. Users may
select from among several different units (lb, short tons, kg, metric tons) to display emissions.
All numerical results are shown rounded to the nearest 10 unit.51 Dashes ("—") indicate that
AVERT reported a value greater than zero, but lower than the level of reportable significance. True
zeros are reported as zero (0) values, increasing the size of the energy policy will increase the
amount of output above the reportable significance level (i.e., reduce the number of dashes in the
results datasets).
Figure 21. Image of annual regional results table for an example wind program in the Midwest region.
Midwest, 2019 AVERT
Output: Annual Regional Results
I Click here to return to Step 4: Display Outputs'
Original
Post Change
Change
Generation (MWh)
510.511.950
496,744.320
-13.767.630
Heat Input (MMBtu)
4,917,277,270
4,784,950.300
-132,326,970
Total Emissions from Fossil Generation Fleet (lb)
S02
710.791,670
689,174,570
-21,617.100
NOx
528.845,720
512,829,990
-16,015,740
Ozone season NOx
224.707,220
219.441.050
-5,266,180
C02
881,078,630,540
857,291,333,570
-23,787,296,370
PM;.s
47.484,740
46,219,000
-1,265,740
VOCs
15.329,350
14.892,980
436,370
NHj
10,163,730
9,902,340
-261,390
AVERT-derived Emission Rates (Ib/MWh)
Average Fossil
Marginal Fossil
S02
1.392
0
NOx
1.036
0
Ozone season NOx
0.976
0
C02
1,725.873
0
PM2.5
0.093
0
VOCs
0.030
0
NHj
0.020
0
Select unit for emissions lb
Ozone season is defined as May 1 - September 30. Ozone season emissions are a subset of annual emissions.
Negative numbers indicate displaced generation and emissions.
All results are rounded to the nearest ten. A dash indicates non-zero results,, but within +/-10 units.
When users evaluate a portfolio scenario including EVs and EE or RE, marginal fossil values are not reported and a null sign ("0") is shown.
Data on this page does not include changes to ICE vehicle emissions (e.g.,. emissions from tailpipes).
50 This value should not be confused for an approximation of the marginal emission rate. It is also not a power-
sector-wide average emission rate, as that emission rate would also incorporate generation from other resources
(e.g., nuclear, wind, solar, hydro) not modeled in AVERT.
51 The Power Sector Emissions Data are reported in integer units of MWh (generation), lb (PM2.5, NOx SO2, VOCs,
and NH3), tons (CO2), and MMBtu (heat input). Results in AVERT are rounded to the closest 10 MWh, lb PM2.5,
NOx, SO2, VOCs, and NH3, tons CO2, and MMBtu fuel input.
SEFA
State and Local Climate
and Energy Program
-------
JtVert
AVoid«d Emissions and q»rwRatlon Tool^^
wwwepagovTavSt'^W
AVERT User Manual Version 4.2
Annual results by county
Figure 22 shows a summary of the changes in generation and emissions for each of the counties
from each of the states in the region. The Midwest region, for example, contains EGUs in
Arkansas, Iowa, Illinois, Indiana, Kentucky, Louisiana, Michigan, Minnesota, Missouri, Mississippi,
North Dakota, Oklahoma, South Dakota, Texas, and Wisconsin. A line for each county containing
an EGU is displayed.
For each county, the following annual output statistics are given:
• Peak Generation Post Energy Change (MW): The peak (maximum) hourly generation
produced by an EGU in the base- or future year scenario after energy changes have been
applied.52
• Annual Generation Post Energy Change (MWh): The total annual generation of an EGU
in the base- or future year scenario after energy changes have been applied.
• Annual Change in Generation (MWh): The EGUs' estimated change in generation from
baseline conditions to post-energy changes over a full year (i.e., the annual increased or
decreased generation of this unit).
• Annual Change in Heat lnput/PM2.5/S02/N0x/C02/V0Cs/NH3 (MMBtu, lb, or tons): The
EGUs' estimated change in heat input or emissions from baseline conditions to post-
energy changes conditions over a full year (i.e., the annual increased or decreased heat
input or emissions of this unit).53
• Ozone Season Change in SO2/NOX/PM2.5 (lb): The EGUs' estimated change in emissions
from baseline conditions to post-energy changes during the ozone season (May to
September, inclusive).
• Ozone Season, 10 Peak Days Change in SO2/NOX/PM2.5 (lb): The EGUs' estimated
change in emissions from baseline conditions to post energy changes during the 10
highest fossil generation days in the ozone season (May to September, inclusive).
All results (except for peak generation) are shown rounded to the nearest 10. Dashes indicate
results greater than zero, but lower than the level of reportable significance. Negative numbers
indicate displaced generation and emissions.
52 Note that generation is counted on a "net" basis. Generation at the level of the boiler, prior to parasitic use by the
plant or generator, is corrected to "net" generation exported to the grid using technology-specific loss factors.
Parasitic use may include use for fans, pumps, and heating and cooling, and emissions control equipment.
AVERT uses different parasitic loss factors for natural-gas-fired combined cycle units, combustion turbines, and
coal-fired steam units with and without controls for sulfur emissions.
53 Note that summary data for VOC and NH3 emissions is not shown on any other output page.
State and Local Climate
and Energy Program
-------
bJ^ERT
AVoidod Einissiortt and qenaRailori Tool^^
www.epa.gov/avarl
AVERT User Manual Version 4.2
Figure 22. Annual results by county for an example wind program in the Midwest region.
Midwest. 2019
Output: Annual Results by County
^^£ncMTerM^eturr^^Ste^^is£|a^Dut£ut^^_^J
State
County F]
Peak Generation, Annual Generation. Annual Change in
Post-Change (MV>» Post-Change (MWpf Generation (MWh »
Annual Change in
S02 (lbs) 0
Annual Change in
NOx (lbs) Fl
Annual Changi
C02 (tons)
AR
Craighead County
74
25,260
-860
-10
-1,370
-540
AR
Hot Spring County
1,249
4.770,860
-36,910
-140
-12,640
-16,560
AR
Independence County
1,402
5,555,760
-55,770
-302,340
-74,740
-64,210
AR
Jefferson County
1.719
8,527,760
-68,760
-337,680
-90,340
-73,790
AR
Mississippi County
1,143
7,155,230
-31,150
-32,890
-18,860
-27,930
AR
Pulaski County
404
349,930
-9,380
-10
-12,750
-4,430
AR
Union County
1,787
11,077,780
-45,790
-150
-2,990
-19,210
IA
Allamakee County
213
449,080
-7,280
-3,730
-2,890
-8,000
IA
Audubon County
84
109,290
-2,260
-150
-2,240
-1,380
IA
Black Hawk County
10
4,260
-110
-
-530
-110
IA
Cerro Gordo County
461
2,525,480
-21,530
-80
-810
-9,490
IA
Des Moines County
193
1,142,450
-4,280
-24,920
-7,370
-5,280
IA
Louisa County
678
3,265,020
-43,410
-137,970
-73,630
-46,360
IA
Marshall County
710
3,742,250
-25,490
-80
-3,320
-11,210
IA
Muscatine County
150
943,160
-4,610
-6,240
-13,000
-5,490
IA
Polk County
313
290,620
-6,310
-40
-830
-4,190
IA
Pottawattamie County
1,351
7,858,440
-59,010
-116,470
-91,640
-62,640
IA
Scott County
39
5,760
-310
-
-590
-210
IA
Story County
27
135,290
-860
-90
-1,210
-580
IA
Union County
35
6,000
-320
-
-3,240
-290
IA
Wapello County
685
3,732,100
-29,180
-22,980
-32,400
-34,580
IA
Woodbury County
1,087
3,140,390
-53,680
-187,060
-107,380
-57,000
IL
Ford County
158
72,750
-2,420
-20
-1,890
-1,550
IL
Fulton County
386
2,253,230
-11,540
-920
-16,720
-13,140
IL
Jackson County
347
989,820
-9,970
-40
-5,580
-4,770
IL
Jasper County
557
3,216,150
-24,740
-82,670
-24,770
-24,300
IL
Madison County
150
25,100
-1,400
-10
-560
-780
IL
Marion County
31
1,840
-170
-
-110
-110
IL
Mason County
402
1,522,570
-16,470
-13,160
-14,720
-18,710
IL
Massac County
900
4,469,360
-38,030
-167,850
-39,640
-38,830
IL
Montgomery County
651
2,728,810
-23,690
-260
-23,310
-24,060
IL
Peoria County
500
3,181,280
-16,200
-71,090
-16,900
-17,150
IL
Perry County
174
79,730
-2,760
-
-2,390
-1,520
Results for top 10 peak days
Figure 23 shows a summary of the 10 days in the region featuring the highest level of fossil fuel
load. Separate columns show the total fossil generation in each day, the expected changes in
generation, and the simulated changes in generation and emissions. All results are shown rounded
to the nearest 10. Dashes indicate results greater than zero, but lower than the level of reportable
significance.
Figure 23. Results for the top 10 load days for an example wind program in the Midwest.
| Midwest. 2019 AVERT |
Output: Results for Top Ten Peak Days
^^licMier^^^tur^^steg^^Disgla^^jutgiit^J
Total Fossil Generation
Expected Change in
Change in Generation
Change in NO,
Change in S02
Change in C02
Change in PM25
Day Rank
Date
(MWh)
Generation (MWh)
(MWh)
(lbs)
(lbs)
(tons)
(lbs)
1
Jul 19
1,902.830
-17.190
-16.930
-21,110
-19.060
-12.930
-2.490
2
Jan 30
1,868,580
-23.980
-24,250
-21.260
-28,070
-17,680
-3,410
3
Jul 02
1.835,170
-9,510
-9.580
-15.200
-12,200
-7.880
-1,640
4
Jan 31
1,833,870
-29,380
-28.820
-19,490
-32.280
-19,380
-5,010
5
Jul 17
1.833,710
-15.420
-15,580
-24.180
-18,590
-13,690
-2,570
6
Jul 18
1,825.060
-13,490
-13,360
-15,380
-15.450
-11.080
-2,170
7
Aug 06
1,818,660
-10.260
-10,270
-13,940
-12,580
-8,280
-1,620
8
Aug 12
1.808,150
-7,820
-7,800
-10,240
-9,010
-6,570
-1.300
9
Aug 07
1.781,340
-8,520
-8,340
-11,860
-9,500
-7,310
-1,570
10
Auq 19
1,780,230
-9.510
-9,310
-13,820
-11,520
-8.080
-1.510
Negative numbers indicate displaced generation and emissions.
All results are rounded to the nearest ten. A dash indicates a result greater than zero, but lower than the level of reportable significance
Monthly results by county
Figure 24 shows a summary of the change in generation and emissions for each of the counties
from each of the states contained within the region, broken out by month and with an annual total.
&EPA
State and Local Climate
and Energy Program
-------
Jmert
AVoidod Einissiortt and qenaRailori Tool^^
www.epa.gov/avarl
AVERT User Manual Version 4.2
All results (except for change in generation) are shown rounded to the nearest 10. Dashes indicate
results greater than zero, but lower than the level of reportable significance.
Figure 24. Monthly results by county for an example wind program in the Midwest region.
Output: Monthly Results by County
|^_^_CHckjTeretOBreturnJoiSte|j4^Djs£la^HOut£uts^_^J
Negative numbers indicate displaced generation and emissions. All results are rounded
to the nearest ten. A dash ("—') indicates a result greater than zero, but lower than the
level of reportable significance. Counties are displayed on// if they contain power plants.
State
0. County 0
Month
Change in
Generation
(MWh)
H
Change in S02
(lbs) 0
Change in NOx
(lbs) Fll
Change in C02
(tons) F
Change in I
(lbs)
AR
Craighead County
1
-120
-130
-80
-10
AR
Craighead County
2
-150
-
-150
-90
-10
AR
Craighead County
3
-90
-
-100
-60
-10
AR
Craighead County
4
-60
-
-60
-40
-
AR
Craighead County
5
-50
-
-60
-30
-
AR
Craighead County
6
-160
-
-270
-100
-10
AR
Craighead County
7
-490
-10
-990
-310
-30
AR
Craighead County
8
-280
-
-500
-180
-20
AR
Craighead County
9
-110
-
-190
-70
-10
AR
Craighead County
10
-80
-
-90
-50
-10
AR
Craighead County
11
-80
-
-80
-50
-10
AR
Craighead County
12
-80
-
-90
-60
-10
AR
Craighead County
Annual
-1,750
-20
-2,720
-1,120
-120
AR
Hot Spring County
1
-11,430
-40
-4,260
-4,940
-920
AR
Hot Spring County
2
-8,670
-30
-2,740
-3,780
-680
AR
Hot Spring County
3
-5,800
-20
-2,010
-2,510
-440
AR
Hot Spring County
4
-7,680
-20
-1,520
-3,350
-550
AR
Hot Spring County
5
-8,100
-30
-2,330
-3,620
-650
AR
Hot Spring County
6
-7,810
-30
-2,980
-3,640
-680
AR
Hot Spring County
7
-5,800
-30
-3,110
-2,750
-500
Daily NOx results by county
Figure 25 shows a summary of changes in NO. emissions in the counties of each state contained
within the region, broken out by day and with a daily average. Users enter up to 10 days for the
analysis year in MM/DD format. All results are shown rounded to the nearest 10. Dashes indicate
results greater than zero, but lower than the level of reportable significance.
Figure 25. Daily NOx results by county in the Midwest region.
Output: Daily NOx Results (lbs)
| | We
| Click here to return to Step 4: Display Outputs |
displaced generation and <
re rounded to the.
Enter up to ten dates in the header column This page will calculate the NOx changes associated with each day in each county, as well as the average for each county Use the filters to select individual states or counties
State
Enter dates of int
erest (MM/DD)
County i * j
1-Jan
2-Mar
3-Apr |
4-Jul
5-Aug |
6-Sep
7-Oct
8-Nov i
9-Dec
| 10-Jan
Average
AR
Craighead County
-13
0
-7
-13
-38
-5
-16
-3
-4
-8
-11
AR
Hot Spring County
-167
34
-70
-86
-57
-189
-244
-74
-30
-113
-100
AR
Independence County
-1,217
-172
-499
-671
-94
-684
-698
-1,001
-531
-589
-616
AR
Jefferson County
-1,118
-199
-823
-973
-100
-623
-853
-884
-816
-1,084
-747
AR
Mississippi County
-445
-89
-254
-60
-16
-94
-230
-262
-118
-128
-170
AR
Pulaski County
-76
-48
-138
-165
-94
-109
-35
-70
-27
-37
-80
AR
Union County
-225
21
-686
-10
-30
-364
-32
-369
-38
-174
IA
Allamakee County
-70
3
-22
-82
-15
-15
-14
-45
-45
-37
-34
IA
Audubon County
-29
-7
-21
-26
-10
-17
-28
-2
-13
-18
-17
IA
Black Hawk County
-1
-2
-3
-18
-2
-4
7
-3
-4
-2
-3
IA
Cerro Gordo County
-21
4
-16
9
-2
-1
-31
-8
-20
4
-8
IA
Des Moines County
-84
-26
-158
-10
4
-64
-93
11
18
16
-39
IA
Louisa County
-1,665
-858
-816
-470
-159
-82
-883
-1,039
-605
-476
-705
IA
Marshall County
-22
-22
-27
-55
-58
-4
-27
-13
-27
-46
-30
IA
Muscatine County
-113
-167
-104
18
-89
-16
-179
-195
-214
-186
-125
IA
Polk County
-9
0
-11
-20
-11
-7
-5
-12
-4
-6
-8
IA
Pottawattamie County
-1,907
-736
-1,298
-675
-31
-154
-768
-1,208
-699
-662
-814
IA
Scott County
0
1
0
-12
-18
-1
0
0
0
-1
-3
IA
Story County
-39
-16
-10
-8
-12
-11
-2
-6
-27
-17
-15
IA
Union County
1
-7
-1
-55
-72
-3
-5
-4
1
-60
-21
IA
Wapello County
-499
-309
-258
-355
-116
-125
9
-477
-199
-218
-255
IA
Woodbury County
-1.589
-942
-1,428
-992
-298
-285
-992
-1,347
-1,027
-1,126
-1,003
IL
Ford County
-13
-1
-16
-58
-17
-5
-28
-9
-9
-16
-17
IL
Fulton County
-114
-167
-150
-39
-46
-60
-220
-172
-214
3
-118
Charts and Figures
Map of generation and emissions changes
This dynamic map (shown in Figure 26) shows where emissions change within the selected region.
You can choose from the following options in a dropdown menu:
• Annual Change in Generation (MWh)
State and Local Climate
and Energy Program
-------
JtVert
AVoid«d Emissions and q»rwRatlon Tool^^
wwwepagovTavSt'^W
AVERT User Manual Version 4.2
• Annual Change in Heat Input (MMBtu)
• Annual Change in SO2 (lb)
• Annual Change in NOx (lb)
• Annual Change in PM2 5 (lb)
• Annual Change in CO2 (tons)
• Ozone Season Change in SO2 (lb)
• Ozone Season Change in NOx (lb)
• Ozone Season Change in PM2 5 (lb)
• Ozone Season, 10 Peak Days Change in SO2 (lb)
• Ozone Season, 10 Peak Days Change in NOx (lb)
• Ozone Season, 10 Peak Days Change in PM25 (lb)
Click on "Refresh map with selected variable" after making a selection. The map displays the
annual, seasonal, or peak change in emissions at specific EGUs in the region. The size of the
circles indicates the relative change of each EGU. Circles are semi-transparent. If multiple sources
are near the same location, the circle is darker. Emissions increases are shown with black outlines
and white interiors; these can occur when the user is modeling an increase in load or can be the
result of the timing of maintenance outages in the base-year data (see Appendix H for details).
43
State and Local Climate
and Energy Program
-------
cndVEi
AVERT User Manual Version 4.2
Figure 26. Map of generation and emissions changes for an example wind program in the Midwest
region.
Midwest, 2019
Output: Map of Generation and Emissions Changes
I Click here to return to Step 4: Display Outputs I
Select variable to display: |Annual Change in CQ2 (tons) B1
Note: The diameter of each circle indicates the
magnitude of a unit's change in generation /
emissions. Circles are semi-transparent; darker
areas occur in regions with overlapping units.
Reductions in emissions or generation are
indicated with blue circles; increases are
indicated with black-bordered white circles.
Annual Change in C02 (tons)
Midwest
503
130,000 tons
The diameter of each circle indicates the magnitude of an EGU's change in emissions. Circles are semi-
transparent; darker areas occur in regions with overlapping EGUs. Emissions reductions are indicated with
blue circles; increases in emissions are indicated with black-bordered white circles.
Monthly results by selected geography
Monthly results can be viewed over the entire region or a specific state or county within the region
(see examples in Figure 27 and Figure 28). First select "Region," "State," or "County" in the top
dropdown menu. If selecting a state, choose the state in the next dropdown menu. If selecting a
county, choose both the state and the county in the next two dropdown menus.
44
SEPA
State and Local Climate
and Energy Program
-------
Bsd^SRT
missions and qenaRaiion Tool^^
www.epa.gov/avan
AVERT User Manual Version 4.2
AVoidad Emissions.
Figure 27. Monthly results (chart) for an example wind program in the Midwest region.
Midwest, 2019
Output: Monthly Results by Selected Geography
|^_^_Click_hereitoreturnJo^te£4^iDisgla^OiJtgiits^_^_J
Counties are displayed only if they contain
power plants
Select level of aggregation
Select state:
State
O
(f)
O
o
J an
0
-50,000
-100,000
-150,000
-200,000
0
-50,000
-100,000
-150,000
0
-20:000
-40,000
-60,000
-80.000
-5,000
-10,000
Monthly Emission Changes, Midwest (IA)
Feb Mar Apr May Jun Jul Aug Sep
Oct Nov Dec
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
LI LJ u u
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
—
Jan
Feb
Mar
Apr
May
Jun
Jul Aug
Sep
Oct
Nov
Dec
'U U
—
—
—
Figure 28. Monthly results (table) for an example wind program in the Midwest region.
Monthly Generation and Emission Changes, Midwest (IA)
Gen (MWh)
S02 (lbs)
NOx (lbs)
C02 (tons)
PM2 5 (lbs)
Jan
-64.190
-124.040
-83,740
-60,870
-6,970
Feb
-55,270
-112,620
-77.100
-53,680
-6,240
Mar
-68,330
-137,220
-90,380
-65.380
-7,630
Apr
-78 080
-155.820
-98.360
-73.760
-8,550
May
-74.980
-139,810
-98.070
-71.870
-8.270
Jun
-53.580
-98,130
-70.630
-50,970
-6.120
Jul
-32,210
-54.000
-42,890
-30,080
4,070
Aug
¦34.750
-59.790
44.850
-32,670
4,130
Sep
48.000
-89.250
-63,130
45.560
-5,390
Oct
-69.880
-135,670
-87,600
-65,770
-7,680
Nov
-70.960
-145.820
-94.300
-68.380
-8.060
Dec
-61,980
-121.650
-80.040
-59.030
-6.850
Total
-712,000
-1,374,000
-931,000
-678,000
-80,000
Negative numbers indicate displaced
generation and emissions. All results are
rounded to the nearest ten. A dash
indicates a result greater than zero, but lower
than the level of reportable significance.
State and Local Climate
and Energy Program
45
SEPA
-------
AVERT User Manual Version 4.2
Hourly results by week
Figure 29 is a dynamic representation of hourly changes in EGU operation. Individual plants are
stacked as gradated bar plots, from high-capacity-factor "baseload" EGUs in dark blue to low-
capacity-factor peaking EGUs in light blue.54 The total contribution of all EGUs sums to the yellow
iine. As noted above, some EGUs can show a net increase in emissions as regional load is
reduced, often due to the timing of maintenance outages in the base-year data. The second chart
in Figure 29 shows the same week-long energy profile as above, but presents the change in
generation in reference to the total fossil-fuel generation. This chart illustrates the degree of change
represented by the energy policy relative to the baseline. The solid line represents the total fossil-
fuel load by hour in the baseline; the dashed line represents the fossil-fuel load after the user-
specified energy change has been modeled.
Figure 29. Hourly results for an example wind program in the Midwest region.
Output: Hourly Results by Week
j Hick here to return to »tep 4: U.splay Outputs"
Select variable to display:
First day to display (MM-DD): | August 1
I Change in Generation (MW)
n
| August 1
r
Change in Generation (MW) in Week of 8/1
Midwest
iivjasf
A'egsf/l'e numbers indicate
displaced generation and
Total Change in
Generation (MW)
_ Total fossil-fuel load,
pre change
High capacity factor ui
Low capacity factor ur
Fossil-fuel Load, Pre- and Post-Change, in Week of 8/1
Midwest
100.000
1 80,000
g 70,000
i; 60,OX)
/\ /\ /\ /A
f\ \
/ \
/\ /\ A f\
/ \ /
/ \
/ \ / \ / \ /
w \
I / ^
\ \J w
\ V
V
S. 40,000
1" 30,000
1 20,000
10,000
1 1 1 1
s 3 3 s
1 S 5 1
? 1
8 8
s t
5 1-
1
1
Total fossil-fuel load,
pre change
Total fossil-fuel load,
post change
^ Change in load
Signal-to-noise diagnostic
The signal-to-noise diagnostic shown in Figure 30 has a different structure from the time-series
images shown previously. This chart is a scatterplot of every hour of the year (8,760 or, in a leap
54 Gradations are relative. Within a region, the unit with the highest capacity factor sets the darkest gradation end
(baseload), while the unit with the lowest capacity factor sets the lightest gradation end (peaking). Units are
sequentially partitioned into color blocks. Most regions include several hundred units, so this gradation will likely
be similar in most regions, with true baseload units at the darkest end and true peaking units at the lightest end.
46
SEPA
State and Local Climate
and Energy Program
-------
bJ^ERT
AVoidod Einissiortt and qenaRailori Tool^^
www.epa.gov/avarl
AVERT User Manual Version 4.2
year, 8,784 points), showing calculated total change in generation in each hour (y-axis) against
user-input, expected energy changes in each hour (x-axis). Ideally, AVERT perfectly matches
modeled change in unit generation to the amount of energy changes requested by the user. This
graphic shows where that assumption holds and where it does not hold, and to what extent. If the
generation change is well-matched to the user-input energy change, the graphic will show a
straight line with little scatter. If the changes are not well matched, the line will have significant
scatter. Overall, the quality of fit (i.e., how well the model captures the requested energy change)
can be judged from the R2 metric shown in the chart title.55 Highly scattered data points should be
viewed with less weight than well-constrained data points.
Figure 30. Signal-to-noise diagnostic for an example wind program in the Midwest region.
Output: Signal-to-noise diagnostic
Click here to return to Step 4: Display Outputs
Change in Total Unit Generation Relative to Expected Energy
Change (MW) for All Hours, R2=0.98
Expected Change in Generation (MW)
-2,000 -1,800 -1,600 -1,400 -1,200 -1,000 -800 -600 -400 -200 0
i i i i i i i i i i t. n
-200
-400
-600 §
e
-800 %
a
®
e
-1,000 5
-1,200 |
-C
O
-1,400 .f
-1,600 &
-1,800
-2,000
The above chart is a scatterplot of every hour of the year it contains either 8,760 or 8,784 data points. one for each hour. Charted points show the resulting total change
in generation (y-axis) versus the user-input (expected) energy change (x-axis) Ideally, A VERT perfectly matches unit generation changes to the amount of energy
change requested by the user If the generation changes are well-matched to the energy change, the graphic will show a straight line with little scatter. If the changes are
not well-matched, the line will have significant scatter. Overall, the quality of fit (i.e., how well the modeled change in generation captures the requested change) can be
judged from the R2 metric shown in the chart's title. Consult the user manual for further details.
Note that flat load changes (i.e., the same MW increase or decrease in every hour of the year) will
result in a very different pattern than shown here. In this unique circumstance, the expected energy
changes (the x-axis in this graphic) will be a single value, while there will be variance along the
"resulting change in generation" (the y-axis in this graphic). In this case, the R2 value will have
55 R2 values indicate the quality of fit of a line, or how well dependent variables describe independent variables (i.e.,
how well the y-axis value describes the x-axis). Random points have an R2 value of zero (0), while perfectly
matched data have an R2 value of 1. The R2 value of 0.99 shown in this figure indicates that AVERT captures
99% of the energy change expected by the user (i.e., noise accounts for 1 % of the observed variability).
&EPA
State and Local Climate
and Energy Program
-------
JtVert
AVoid«d Emissions and q»rwRatlon Tool^^
wwwepagovTavSt'^W
AVERT User Manual Version 4.2
limited value, but you can visually review the scatter in the plot to determine the reasonableness of
the results.
Summary Tables, Charts, and Figures: Power Sector and Avoided Vehicle
Emissions Data
Outputs related to vehicle impacts are confined to four output pages accessed through Step 4:
Display Results and discussed below.56 All other output pages in AVERT do not describe any
vehicle impacts.
Annual Regional Results, Including Vehicles
This page describes the total regionwide emissions impacts from the electric power sector ("From
Fossil Generation") and from vehicles ("From Vehicles") for each of the six pollutants modeled in
AVERT (see Figure 31). This page also includes a column for "Net Changes," which combine the
emissions impacts from the power sector and vehicles. Users may select from among several
different units (lb, short tons, kg, metric tons) to display emissions. A separate table at the bottom
of the page provides additional context on how many emissions are produced from the fossil
generation sector and from vehicles in the selected region. Throughout this page, negative
numbers indicate reductions in emissions.
56 For the purposes of AVERT, "vehicle emission impacts" refer to the impacts associated with emissions from
vehicle tailpipes and other emissions closely related to the driving and fueling of vehicles. Specifically, these
include emissions of NOx, SO2, VOCs, and NH3 from vehicle exhaust, emissions of VOCs from vehicle exhaust,
evaporation, and refueling, and emissions of PM2.5 from vehicle exhaust (but not PM2.5 emissions related to brake
wear or tire wear).
State and Local Climate
and Energy Program
-------
bJ^ERT
AVoidod Einissiortt and qenaRailori Tool^^
www.epa.gov/avarl
AVERT User Manual Version 4.2
Figure 31. Annua! regional results table, including vehicles, for an example EV and renewable program
in the Midwest region.
Midwest, 2019
AVERT
Output: Annual Regional Results, Including
Vehicles
[
i^lickjTereJo_retum^^iSteg(4|iiDis£la^>Outguts>i
From Fossil Generation
From Vehicles
Net Changes
Total Emission Changes (lb)
S02
-21,613,990
-2,810
-21,616,800
NO*
-16.018.110
-43,260
-16.061,370
o
o
-23.786.769.130
-423,058.990
-24,209,828,120
PM2.5
-1,265,830
-3,160
-1,268,990
VOCs
-436,430
-91.590
-528,020
NHj
-261,540
-24,410
-285,950
Select unit for emissions:
lb
Negative numbers indicate displaced emissions.
All results are rounded to the nearest 10. A dash ("¦—") indicates non-zero results, but within +/- 10 units.
Fossil results include combined changes from all modeled resources (including EVsj.
Background: Total emissions for the Midwest region
Fossil Generation
Vehicles
Total Emissions (lb)
S02
710,791,670
4,987,170
NO*
528.845.720
563,103,360
O
O
w
881,078,630,540
754.890,452,890
PM2.5
47,484,740
10,478.450
VOCs
15,329,350
556,445,350
NHs
10,163,730
53,456,840
Total emissions as depicted above are calculated only for the resources and vehicles covered in A VERT.
For fossil generation, this includes units that report to EPA's AMP dataset and are larger than 25 MW.
For vehicles, this is based on modeled MOVES data for light-duty cars and trucks, transit buses, and school buses, for
the year 2021.
Annual Results by County, Including Vehicles
This page describes the emissions impacts from the electric power sector ("From Fossil
Generation") and from vehicles ("From Vehicles") for each of the six pollutants modeled in AVERT
(see Figure 32). On this page, emissions are summarized by state and county. This page also
includes a column for "Net Changes," which combine the emissions impacts from the power sector
and vehicles. Throughout this page, negative numbers indicate reductions in emissions. Finally,
this page includes a column identifying the FIRS code for each county appearing in the selected
analysis to facilitate mapping and visualizing geospatial results.
&EPA
State and Local Climate
and Energy Program
-------
bJ^ERT
AVoidod Einissiortt and qenaRailori Tool^^
www.epa.gov/avarl
AVERT User Manual Version 4.2
Figure 32. Annual results by county table, including vehicles, for an example EV and renewable
program in the Midwest region.
/>. . xa ¦ rt « ¦ II- i i ¦ - ¦ All results are rounded to the nearest ten. A dash
Output: Annual Results by County, Including Vehicles ^Mmm*«
numbers indicate displaced emissions. Fossil results include
1 combined changes from all modeled resources (including
EVs).
State
County
FIPS Code
Pollutant
From Fossil Generation
From Vehicles
Net Changes
AR
Arkansas County
05001
S02 (lb)
0
0
0
AR
Ashley County
05003
S02 (lb)
0
0
0
AR
Baxter County
05005
S02 (lb)
0
0
0
AR
Boone County
05009
S02 (lb)
0
0
0
AR
Bradley County
05011
S02 (lb)
0
0
0
AR
Calhoun County
05013
S02 (lb)
0
0
0
AR
Carroll County
05015
S02 (lb)
0
0
0
AR
Chicot County
05017
S02 (lb)
0
0
0
AR
Clark County
05019
S02 (lb)
0
0
0
AR
Clay County
05021
S02 (lb)
0
0
0
AR
Cleburne County
05023
S02 (lb)
0
0
0
AR
Cleveland County
05025
S02 (lb)
0
0
0
AR
Columbia County
05027
S02 (lb)
0
0
0
AR
Conway County
05029
S02 (lb)
0
0
0
AR
Craighead County
05031
S02 (lb)
-30
-10
-40
AR
Crittenden County
05035
S02 (lb)
0
0
0
AR
Cross County
05037
S02 (lb)
0
0
0
AR
Dallas County
05039
S02 (lb)
0
0
0
AR
Desha County
05041
S02 (lb)
0
0
0
AR
Drew Onuntv
05043
SO? nh"i
0
0
0
County-level values on this page are used to generate the CSV that is used for COBRA analysis.
See the COBRA Text File section, below, for more information about performing analyses in
COBRA using AVERT data.
Emission Results by Selected Geography, Including Vehicles
This page features a bar chart that compares emission impacts from power generation and
vehicles (see Figure 33). A "Net" bar is also shown, which describes the aggregate emission
impacts from both the power sector and vehicles. Results are also shown in tabular form.
Throughout this page, negative numbers indicate reductions in emissions.
On this page, users may select one of the six pollutants modeled in AVERT. Users may also select
different geographic levels of aggregation: They may choose to aggregate all emission impacts
together under the "Regional" selection or view emission impacts for just one state or county. Note
that some states and counties may have a zero value for emission impacts from either the power
sector or vehicles. Emission impacts in the power sector will only be present if that state or county
features power plants that are affected by the user-inputted load change. Emission impacts from
vehicles will only be present if the user has selected the area in question as a location where EVs
are being deployed. Note that changes in emissions within counties reflect point source changes
(as in power plants) or estimated locations where vehicle emissions have changed. AVERT results
should be exported to an air quality model to estimate changes in ambient concentrations of
pollutants within counties or regions due to regional air pollution transport.
SERA
State and Local Climate
and Energy Program
-------
cndVEi
AVERT User Manual Version 4.2
Figure 33. Emission results by selected geography table and chart, including vehicles, for an example
EV and renewable program in the Midwest region.
Midwest, 2019
AVERT
Output: Emission Results by Selected
Geography, Including Vehicles
Annual Emission Changes, Midwest
o
o
-2,000 000
-4,000,000
-6,000,000
-8,000,000
-10,000,000
-12,000,000
-14,000,000
Power Generation
Vehicles
Net
C02 (tons)
Power Generation
-11.893,700
Vehicles
-211,270
Net
-12,104,970
Negative numbers indicate displaced emissions; positive numbers indicate increased generation and emissions. Fossil
results include combined changes from all modeled resources (including EVs). Values may not match those on "Output:
Annual Regional Results, Including Vehicle" due to rounding.
Results by Month, Including Vehicles
This page features a bar chart that compares monthly regionwide emission impacts from power
generation and from vehicles (see Figure 34). Results are also shown in tabular form. Throughout
this page, negative numbers indicate reductions in emissions.
51
SEPA
State and Local Climate
and Energy Program
-------
Bsd^SRT
missions and qenaRaiion Tool^^
www.epa.gov/avan
AVERT User Manual Version 4.2
AVoidad Emissions.
Figure 34. Emission results by month table, including vehicles, for an example EV and renewable
program in the Midwest region.
Midwest. 2019
Output: Results by Month,
Including Vehicles
Click here to return to Step 4: Display Outputs
Select Pollutant:
C02
Monthly Emission Changes, C02
3 4 5 6 7
¦ Power Generation Vehicles
10 11 12
C02 (tons)
Power Generation
Vehicles
1
-1.150.920
-15.707
2
-972 390
-14.787
3
-1.079.230
-17.087
4
-1.099.990
-17.057
5
¦1.003.890
-18.211
6
-779,040
-18.875
7
-726.060
-20.065
8
-748.780
-20.207
9
-925,750
-17,769
10
-1.099,270
-17,913
11
-1,122,790
-16,741
12
-1,185,490
-16.982
Total
-11.893.600
-211.401
Negative numbers indicate displaced emissions; positive numbers indicate increased
generation and emissions Fossil results include combined changes from all modeled
resources (including EVs).
Reference: Modeled Marginal Emission Rates Over Time
This page, shown in Figure 35, features a graph of marginal emission rate projections overtime.
This graph includes AVERT's modeled marginal emission rates from 2018 through 2021. Emission
rates shown for future years includes projected SRMER and long-run marginal emission rates
(LRMER) from the National Renewable Energy Laboratory's (NREL's) 2021 Standard Scenarios,
SEFA
State and Local Climate
and Energy Program
-------
cndVEi
AVERT User Manual Version 4.2
as shown in NREL's Cambium data set.57 AVERT's calculations assume no additions or
retirements of power plants in response to the modeled change in load. This is analogous to the
SRMER trajectory and is useful for understanding emission impacts as they occur over a relatively
short time horizon (e.g., 5 years), before a structural response can occur. Users who wish to
understand how load changes affect emissions over a longer time period would be better served by
the LRMER trajectory, which takes into account the fact that power plants may be added or retired
in response to the modeled load change. SRMER from the Cambium data set shown for future
years do not correspond perfectly to historical AVERT emission rates due to differences in
modeling methodology and topology. For more information on SRMER and LRMER, and other
approaches to calculating marginal emissions, see "Short-run and Long-run Power Sector
Analysis" on page 8.
Users are able to select which of these three emission rates (AVERT's SRMER, Cambium
SRMER, and Cambium LRMER) they would like to display on the graph.
Figure 35. Reference page for modeled emission rates over time.
I Midwest. 2019
AVERT I
Reference: Modeled Marginal Emission Rates Over Time
Click here to return to Step 4: Display
Marginal emission rates will likely decrease in Mure jears; by bow much depends on die time horizon and
framing of the analysis.
2.000
1.800
» 1.000
ft
c 1.400
li 1200
1.000
O® 800
000
1 400
200
dbbhd
E AVERT
D
¦ Cambium Stan-
ly run Marginal
£me»a»onRate
i
1 1
i .
¦
1
1
(SftUER)
¦
¦ Camttum u>ng-
HI
i 1
k run Marginal
Emruion Rate
¦
3 1
< 1
1
(LRMER)
2015 2020 2025 2030 2035 2040 2045
Stars shown in the chart represent a range of vetoes avaHaNefor each source.
For AVERT rates the central vatue fa circle I represents the marginal emission rate for that region,
as cafauiatedbji the average of the emission rates reportedin the "AvoidedEmission Rates Generated
from A VERT"resource. The upper end/over ranges fa hat) represent the highest end fairest rates for that
region in that dataset.
For Camhknn rates, the centra/ value (a circle) and upper and fover ranges fehar) represent
marginal emission rates from three of AtREL's scenarios-A'tid-Case, L ovRenevabie Energy Cost, andAUd-
Case SfijfJVSXt HepefxAng on thereupon, scenario andemissionrere. the reiative order of these
scenarios mag vary
Show series on chart?
The AVERT rates shown represent the range
of resources found in the AVERT avoided
emission rates spreadsheet (available at
https:Hwww.epa.govf avertfavotded-errossiori-
rates-generated-avert).
Cambium SRNtffR
Cambium LRMER
SRMER rates describe the marginal emission
raie estimated in a single year, assuming no
changes in additions or retirements of power
plants in response to the modeled load
change.
LRMER rates describe the marginal emission
rate for a single year, assuming that power
plants are allowed to be added or retired in
response to changes in load.
Recent historical marginal emission rates for the Midwest region range from:
IS27 to 18021> C02fMWh f of central values and
1.551 to I860 lb C02
-------
JtVert
AVoid«d Emissions and q»rwRatlon Tool^^
wwwepagovTavSt'^W
AVERT User Manual Version 4.2
from a scenario analyzed in AVERT. Go to "Step 4: Display Results" sheet and double-click on the
blue box to enter a file path under "COBRA text file generation." Then click the "Generate COBRA
text files" green button. One CSV file will be saved in the selected file folder. It will include county-
level emission impacts for NOx, SO2, PM25, VOC, and NH3 for both the power sector and vehicles.
This file can be easily uploaded into COBRA. For COBRA input instructions, refer to COBRA'S
main webpage at https://www.epa.gov/cobra and EPA's COBRA user manual.58 Note that the web
version of COBRA does not presently accept file uploads, but it can read in all pollutants through
an automated connection from the output screen of the AVERT Web Edition.
SMOKE Text File
AVERT allows you to create files for use in SMOKE: enter a filepath in the bottom center of the
"Step 4: Display Results" sheet, then press the "Generate SMOKE text files" button. Twenty-four
text files are then generated in this filepath, two for each month. One of these files is for the
selected region and year, pre-energy change, while the second set of files details information post-
energy change.59 For more detailed instructions on how to interpret and use SMOKE outputs from
AVERT, see the tutorial and step-by-step instruction slides available on EPA's website at
https://www.epa.gov/avert.
Advanced Outputs
Though AVERT does track the estimated output and emissions from specific EGUs, we
recommend only using these outputs for SMOKE processing and/or quantitative validation
purposes.
To access annual results on a unit-by-unit basis, restore default Excel functionality via the button
on the Welcome page. Data are available for each EGU in the region of interest in the "Summary"
worksheet. EGUs are identified by their ORISPL number,60 unit number, and unit name, fuel type,
state, county, and geographic location (latitude/longitude). Summary data are provided for each
EGU in a similar format to the county data, as described previously.
In addition, detailed data are available in the worksheets labeled "Gen" (generation), "Heatlnput,"
"SO2," "NOx," "CO2," "PM2 5", "VOCs", and "NH3." These worksheets record emissions and
generation changes for each EGU in the region for each hour of the modeled year. Hours are
arrayed vertically; EGUs are arrayed horizontally.
58 Users may also be interested in using AVERT outputs in a variety of other models for the purposes of analyzing
health impacts resulting from changes to emissions. Such models include APEEP (available at
https://public.tepper.cmu.edu/nmuller/APModel.aspx). EASIUR (available at
https://barnev.ce.cmu.edu/~iinhvok/easiur/). and INMAP (available at http://spatialmodel.com/inmap/). More
information on the strengths and advantages of these and other models can be found at the Center for Air,
Climate, & Energy Solutions, a multi-university research center created through a partnership with U.S. EPA (see
https://www.caces.us/).
59 At this time, AVERT is not capable of producing SMOKE-readable outputs for VOC or NH3 emissions from the
power sector, or any emissions from the transportation sector.
60 The ORIS or ORISPL number is a code used by DOE and EPA to identify specific generating plants, where a
"plant" is a site that may include multiple EGUs. Each ORIS number is unique and (usually) persistent. "Unit
numbers" are assigned to generators and boilers by DOE and EPA, respectively, and are subject to change or
modification by accounting agency or reporting entity.
State and Local Climate
and Energy Program
-------
AVERT User Manual Version 4.2
In contrast to the results shown in the summary tables, charts, and figures, which have been
rounded to the nearest 10, results shown in the advanced outputs have not been rounded.
Unrounded results should only be used after due consideration of their significance.
55
SEPA
State and Local Climate
and Energy Program
-------
AVERT User Manual Version 4.2
Appendix A: Installation Instructions
AVERT is divided into three components: an Excel-based platform for user-specified analysis of
generation and emission changes (called the Main Module), a MATLAB®-based statistical
analysis program (called the Statistical Module), and a second Excel-based spreadsheet for
creating user-specified future year scenarios (called the Future Year Scenario Template). This
section provides installation instructions for each AVERT module. More detailed information on
AVERT components is provided in Section 2 of this manual, "The AVERT Analysis Structure."
Main Module
AVERT's Main Module estimates the change in emissions likely to result from energy policies in
reference to a base-year or future year scenario.
System Requirements
The Main Module requires Excel 2007 or newer to run in Windows. The Main Module can also be
used in Excel 2011 or newer for Mac; it has been verified to work up to Excel for Mac v16.49.
Macros must be enabled. You do not need to install the Statistical Module and the Future Year
Scenario Template to use the Main Module to estimate change in emissions for energy policies
modeled in reference to a historical base year; however, you will need all three AVERT modules to
model change in emissions with reference to user-created future years.
The Main Module has no special requirements for hard drive space or RAM on the computer
running it, but it will run faster on computers with more RAM and higher-speed processors. Excel
files generated in the Main Module can exceed 100 MB in size, depending on the number of EGUs
in the region of analysis. Analyzing data for large regions may take over 10 minutes on some
computers. EPA recommends that users not use any other computer functions (e.g., copy-paste)
during this time in order to speed the calculation and avoid errors.
Installation
To use the Main Module, download two files and save both to the same folder on a local computer
or drive:
• The Main Module workbook: "AVERT Main Module.xlsm." Download the workbook at
https://www.epa.gov/avert.
• The RDF for the region under analysis.
o Default RDFs developed for use by EPA are labeled
"AVERT RDF [DataYear] EPA_ EPA_NetGen_PMVOCNH3 ([Region]).xlsx";
they can be obtained at https://www.epa.gov/avert.
o Regional analyses developed by advanced users using AVERT's Statistical
Module will be saved, by default, in a folder of the Statistical Module titled
"AVERT Output." These files use the following naming convention:
"AVERT RDF [DataYear] [RunName] ([Region]) [RunDateTime].xlsx."
In the RDF:
• "Region" refers to one of 14 regions defined for the purposes of this tool. AVERT's regions
are described in Section 3 of this manual, under "AVERT Regions" (page 16). The
dS ®EPA
State and Local Climate
and Energy Program
-------
AVERT User Manual Version 4.2
"BaselineYear" tag indicates the base year (the year upon which the analysis is based).
Generally, for contemporary or forward-looking analyses, this should be the most recent
full year of data available from CAMD's Air Markets Program (2022), although older data
years 2017 through 2021 are also currently available for input.
• "RunDateTime" indicates when the data file was generated by the Statistical Module.
Launching A VERT's Main Module
To launch the model, open the Main Module workbook in Excel and follow the step-by-step
instructions in Section 4 of this manual.
Technical Assistance
For more information, please contact EPA's State and Local Climate and Energy Program at
avert@epa.qov.
Statistical Module
AVERT's MATLAB®-based Statistical Module performs statistical analysis on Power Sector
Emissions Data to generate output files used to model emissions changes in the Main Module.
Running the Statistical Module is not required to operate the Main Module; it is anticipated that
most AVERT users will not run it. Users creating specific future year scenarios, however, will need
to run the Statistical Module.
For more information on AVERT's Statistical Module, see Appendix D.
System Requirements
The Statistical Module requires a machine capable of running Windows XP or higher.
It is recommended that computers operating the Statistical Module have at least 2 GB of memory
available. Processing time for individual regions depends on the number of EGUs in the analysis
and the number of processors available for use by the MATLAB® platform. In development of
AVERT, it was found that for full-scale runs, larger regions could take over two hours to analyze
with four processers dedicated to the operation.
The Statistical Module can perform analysis either with a base-year dataset from 2017 through
2022, or with a revised electricity generation fleet created in the Future Year Scenario Template,
used in conjunction with a pre-loaded base-year dataset.
Installation and Launching
To use the Statistical Module, follow the instructions in Appendix E.
This output file can be used directly in the Main Module to analyze change in emissions from
energy policies.
Technical Assistance
For more information, please contact EPA's State and Local Climate and Energy Program at
avert@epa.qov.
State and Local Climate
and Energy Program
-------
AVERT User Manual Version 4.2
Future Year Scenario Template
AVERT's Future Year Scenario Template allows the user to modify the list of EGUs analyzed by
the Statistical Module. EGUs can be added or retired, or have their emission rates modified. Newly
added EGUs are copied from existing EGUs (proxy units), but can be scaled to a desired capacity
and given a location (county or latitude/longitude) in a different location.
System Requirements
The Future Year Scenario Template requires Excel 2007 or newer to run. It has been designed for
Windows and has not been tested on a Mac, as the companion Statistical Module requires
Windows. You do not need to install the Statistical Module to design scenarios within the Future
Year Scenario Template, but you will need it to analyze those scenarios and estimate their future
emissions in the Main Module.
The Future Year Scenario Template has no special requirements for hard drive space or RAM, but
it will run faster on computers with more RAM and higher-speed processors. Scenarios saved by
the Future Year Scenario Template are likely to be between 14 and 25 MB in size, depending on
the number of EGUs being added in a new scenario.
Installation
The Future Year Scenario Template is packaged with the Statistical Module executable package.
Instructions on obtaining this package can be found in Appendix E.
On downloading and unpacking the package, you will be presented with a folder entitled "AVERT
Future Year Scenarios." This folder contains a number of example templates illustrating
retirements, additions, and retrofits.
Launching and Working with the Future Year Scenario Template
To launch the Future Year Scenario Template, open the workbook in Excel and follow the step-by-
step instructions in Appendix F.
Technical Assistance
For more information, please contact EPA's State and Local Climate and Energy Program at
avert@epa.qov.
State and Local Climate
and Energy Program
-------
AVERT User Manual Version 4.2
Appendix B: Power Sector Data
AVERT primarily relies on Power Sector Emissions Data from CAMD, with supplemental data
obtained from the National Emissions Inventory (NEI).61 Table 3 identifies the data source for each
pollutant in AVERT and the related data year of each source in AVERT v4.2's 2022 dataset.
Table 3. Data sources used for each pollutant modeled in the power sector.
Pollutant
Source
Data year
CO2
CAMD
2022
NOx
CAMD
2022
SO2
CAMD
2022
PM2.5
NEI
2020
VOCs
NEI
2020
NHs
NEI
2020
Data from CAMD
For the purposes of tracking and verifying emissions, and monitoring emissions trading programs,
CAMD collects extensive operational data from nearly all operating fossil-fuel EGUs with
generating capacities greater than 25 MWin the lower 48 states (i.e., excluding Alaska and
Hawaii).62 CAMD data include reported gross generation (in MWh per hour),63 steam output (in
tons, from combined heat and power [CHP] facilities), heat input (in million British thermal units, or
MMBtu), and emissions of SO2, NOx, and CO2. (Note that emission rates for PM2 5, VOCs, and NH3
come from a different source, the NEI, as described later in this section.) Each quarter, CAMD
consolidates information from the previous quarter (i.e., there is typically a three-month delay in
releasing data) and produces text-based datasets for each of these factors for each fossil-fuel EGU
in each state.64
Each power plant reports a "method of determination code" (MODC) for each pollutant for each
hour. These MODCs reflect how emissions data were determined, which can vary based on power
plant operation and emissions monitoring circumstances. Generally, the data are either classified
as "measured" data or "substitute" data. While the vast majority of these data are classified as
measured, occasionally power plants report substitute data when measured data are not available.
A subset of these substitute data is the "maximum potential concentration" (MPC), the most
conservative value power plants are required to report in certain circumstances (e.g., when the
61 https://ampd.epa.gov/ampd/ and https://www.epa.qov/air-emissions-inventories/national-emissions-inventorv-nei.
62 For the purposes of CAMD data collections, "units" are typically individual boilers, but sometimes represent either
a single emissions source (i.e., smokestack) from several attached boilers or the consolidated output of a single
generator with multiple boilers. CAMD unit designations are often, but not always, the same as DOE unit
designations. CAMD's data guide with more information can be accessed at
https://www.epa.qov/sites/production/files/2020-
02/documents/camds power sector emissions data user quide.pdf.
63 Gross generation is measured at the level of the boiler, prior to parasitic use by the plant or generator. Parasitic
use may include use for fans, pumps, heating and cooling, and emissions control equipment. Therefore,
generation seen by the grid may differ from the values in this database by 0 to 10%, depending on the unit.
Emissions, however, are "at stack" and represent total emissions released to the atmosphere.
64 CAMD collects data from most fossil-fired electrical generating stations over 25 MW in the lower 48 contiguous
states (i.e., excludes Alaska, Hawaii, and territories). This dataset generally does not include data from biomass
generation or most small diesel backup generators.
SERA
State and Local Climate
and Energy Program
-------
JtVert
AVoid«d Emissions and q»rwRatlon Tool^^
wwwepagovTavSt'^W
AVERT User Manual Version 4.2
emissions monitor is bypassed). The MODCs that correspond to MPC are MODC 12, 18, and 23.
To improve reliability of AVERT's results, data in hours and for pollutants coded with MODC 12, 18,
and 23 have been removed from the text-based dataset.65
A MATLAB®-based preprocessing engine converts these hourly text files into compact data arrays
and a reference EGU records file.66 The preprocessing engine calls an Excel-based spreadsheet
populated with ancillary information about each EGU, with most information gathered from the
CAMD "facility information" records. The spreadsheet is populated with ancillary lookup information
about each EGU that has reported to CAMD.
Gross generation is converted to net generation within the preprocessing engine using unit-specific
parasitic loss factors. These factors were calculated based on a comparison of by-plant gross
generation67 and by-plant net generation68 using 2015 data.69 Different loss factors are used for
coal-fired steam units with and without sulfur controls (8.3 percent and 6.9 percent, respectively),
natural-gas-fired combined cycle units (3.3 percent) and combustion turbines (2.2 percent), and
natural-gas- or oil-fired steam units (7.7 percent). For example, a sulfur-controlled coal steam unit
with an annual gross generation of 100 GWh is assumed to export a total of 91.7 GWh to the grid,
while a natural-gas-fired combined cycle unit with the same gross generation is assumed to export
96.7 GWh.
The six data arrays store two-dimensional matrices of net generation, steam output,70 heat input,
SO2, NOx, and CO2 organized by EGU and by hour of the year. Figure 36 shows an example two-
dimensional data array for base-year hourly generation (8,760 or 8,784 hours across the horizontal
axis) for each of the 4,734 fossil-fuel EGUs (down the vertical axis) for which CAMD collected
emissions data in 2011. Black areas represent hours during which particular EGUs are not in
operation (or are operating at very low levels, i.e., less than 10 MW). Figure 36 also includes detail
from the data array that focuses in on 10 EGUs and hours 3,000 through 4,000 in the base year.
65 To learn more about Method of Determination Codes, see 40 CFR Part 75.57, Table 4a.
66 https://www.mathworks.com/products/matlab.html.
67 As reported to CAMD.
68 As reported to EIA on Form 923 fhttps://www.eia.gov/electricitv/data/eia923/).
69 Empirical parasitic loss factors were found to be comparable to those published in the literature.
70 AVERT does not use steam output.
SEPA
State and Local Climate
and Energy Program
-------
Bsd^SRT
missions and qenaRaiion Tool^^
www.epa.gov/avan
AVERT User Manual Version 4.2
AVoidad Emissions.
Figure 36. 2011 gross generation output (in MW) for each EGU in each hour of the year.
4,000 5,000
Hour of Year
6,000
7,000
400 500 600
EGU Level of Output (MW)
900 1,000
or greater
Little Gypsy 1 I
Little Gypsy 2 |
Little Gypsy 3 I
Ninemile Point 1 [
Ninemile Point 2 I
Ninemile Point 3 I
Ninemile Point 4 I
Ninemile Point 5 I
Sterlington 7AB I
Sterlington 7C |
3,000
II 11 ¦Hill
IIIIIIIIIII I I III I III
Oil
urn
i M |i ii
¦
IF
3,400 3,500 3,600
Hour of Year
Colors represent the level of EGU generation.
The reference EGU records file (a structural array in MATLAB®) holds the name, ORISPL number
(a value assigned to each plant site by DOE), EPA unit ID, a lookup table pointer to the two-
dimensional matrices, iocational information, and fuel information for each EGU. Table 4 shows an
example record for the Handley Generation Station, Unit 5. The "Unique ID" shown in Table 4 is a
unique identifier created within AVERT, consisting of the ORISPL number concatenated with the
EPA unit identification number as a string.71 In addition, each record stores a lookup value (not
shown in Table 4), which codes for the location of the plant in the two-dimensional data files.72
71 The pipe character ("|") is used to separate the ORISPL and Unit ID for legibility.
72 Due to coding limitations in Excel and MATLAB, a small number of units have modified UnitlDs relative to the
UnitID that appears in CAMD data. For example, units with a UnitID of "1 -1" in CAMD data may instead use a
UnitID of "N1" in AVERT.
6 &EPA
State and Local Climate
and Energy Program
-------
AVERT User Manual Version 4.2
Table 4. Example record in the reference EGU records file.
Name
Handley Generating Station
UnitID
5
NERCSub
ERCT
NERCSubJx
26
State
TX
State_lx
44
LUTValue
4022
ORISPL
3491
Lat
32.7278
Lon
-97.2186
County
Tarrant
FuelPrimary
Pipeline Natural Gas
FuelSecondary
Diesel Oil
PrimeFuelType
Gas
UniquelD
349115
CSIRegion
TX
CSIRegionIX
9
EGU records also include expected CO2 emissions data for units that do not report CO2 to CAMD
on an hourly basis. Expected CO2 emissions for these units were calculated as the product of an
assumed fuel-specific CO2 content factor and the unit's heat input for each hour. CO2 factors in
tons of CO2 emitted per MMBTU of fuel consumed were calculated using the "unspecified coal,"
"natural gas," and "distillate fuel oil" carbon content values codified for EPA's Greenhouse Gas
Reporting Program in 40 CFR Part 98, Subpart C. Units with a fuel type other than coal, oil, or gas
were assumed to have the same carbon emissions factor as oil-fired units.
All AVERT versions after AVERT v3.1.1 (including v4.2) divide the contiguous United States into 14
distinct regions.73 These regions are aggregates of one or more balancing authorities. Each
balancing authority is an entity tasked with the actual operation of the electric grid and ensures that
the demand for electricity in every minute of every day is met by adequate supply from the grid's
power plants. In effect, these entities are the smallest discrete component of the grid's operation.
There are about 75 balancing authorities active in the United States today, with each of the nation's
4,400 emitting power plants assigned to one of these entities.
Within AVERT, these 75 balancing authorities—and their constituent power plants—are assigned
to one of 14 regions. These assignments were developed according to delineations based on
geography (e.g., all balancing authorities in California are assigned to the "California" region) or
electrical transmission (e.g., balancing authorities in Florida's panhandle). In many situations, an
AVERT region is based around a "core" balancing authority (e.g., PJM, MISO, CAISO) with other
smaller balancing authorities grouped with that larger entity for convenience or because there may
be substantial transfers of electricity between regions. In most situations, AVERT's regional
assignment is closely based on the regional assignments from ElA's 930 dataset.74 Using ElA's
73 These regions include the 48 contiguous states plus Washington, D.C. Power plants in Alaska and Hawaii are not
required to report hourly data to EPA's Continuous Emissions Modeling System (CEMS) dataset used by AVERT
and are thus excluded from analysis in the tool.
74 See https://www.eia.qov/beta/electricitv/qridmonitor/.
dS ®EPA
State and Local Climate
and Energy Program
-------
AVERT User Manual Version 4.2
930 dataset and ElA's 861 dataset for 2018, we match each electric utility with a balancing
authority, and each balancing authority with an AVERT region.75 Each electric utility is assigned to
one and only one balancing authority, and each balancing authority is assigned to one and only
one AVERT region. Retail sales from each utility are grouped by state and AVERT region to
determine how each state's electricity sales are split up across the 14 AVERT regions. This is done
in order to inform users how they may wish to allocate electricity impacts across different AVERT
regions, in situations where a state spans more than one region (see Appendix G for more
information). Finally, using data from ElA's 860 dataset for each analysis year, we match each
EGU with a balancing authority and an AVERT region for purposes of creating RDFs.76
Analysis based on smaller regions, such as utility service territories, risks missing important
interdependencies between the EGUs in a larger region (e.g., the impact of New Jersey load
reductions on Ohio EGUs). Using still larger regions, such as the Eastern Interconnect, spreads the
influence of load changes too widely, making it difficult to ascribe load changes at a particular
location to a reasonable cohort of EGUs.
Data from the National Emissions Inventory
To determine emission rates for three pollutants—PM2 5, VOCs, and NH3—we rely on data from the
full, triennial NEI and interim year NEI point source data.77 Emissions data for these pollutants is
not available from the CAMD dataset on an hourly basis, but is available from the NEI on an annual
basis.78 Using a methodology developed by EPA for eGRID, we match EGUs in the CAMD dataset
with facilities in the NEI.79 Using total emissions for each EGU in either the NEI or point source NEI,
we divide by the EGU's heat input (as reported in the CAMD dataset) to calculate Ib-per-MMBtu
emission rates. In some cases where there is missing or known anomalous data in the NEI, EGUs
are assigned the average rate of similar plants (i.e., the same prime mover and fuel type). These
cases include situations where emissions data for one or more pollutant is missing in the NEI for an
EGU, situations where an EGU match is unable to be made between the CAMD dataset and the
NEI dataset, and situations in which emissions published in the NEI are known to be sourced from
outdated data.80 Emission rates for these three pollutants for each year are stored within the Main
Module and are applied to calculate unit-specific emissions for each unit during each hour. When
users upload an RDF, NEI emission rate data for the appropriate year is automatically selected in
AVERT.
The NEI datasets used in AVERT vary depending on which AVERT dataset the user has selected.
Table 5 describes this assignment. Note that the full NEI is published triennially, with the last official
release in 2021 for the 2017 data year. Datasets for other years (e.g., 2018, 2019, and 2020) are
75 See https://www.eia.gov/electricitv/data/eia861/.
76 See https://www.eia.gov/electricitv/data/eia860/.
77 See https://www.epa.gov/air-emissions-inventories/national-emissions-inventorv-nei.
78 For more information on the NEI, see the 2017 National Emissions Inventory: January 2021 Updated Release,
Technical Support Document, available at https://www.epa.gov/sites/default/files/2021 -
02/documents/nei2017 tsd full ian2021.pdf. More information on the most recent NEI data used in AVERT
(released January 2023 for the 2020 data year) is available at https://www.epa.gov/svstem/files/documents/2023-
01/NEI2020 TSD Section3 Point.pdf.
79 See https://www.epa.gov/egrid/egrid-pm25-methodology.
80 Users should note that there are different reporting requirements for EGUs between the triennial NEI and interim
year point source NEIs. More EGUs are likely to be assigned an average rate in the interim years when reporting
requirements capture a smaller universe of EGUs.
SEPA
State and Local Climate
and Energy Program
-------
JtVert
AVoid«d Emissions and q»rwRatlon Tool^^
wwwepagovTavSt'^W
AVERT User Manual Version 4.2
based on data compiled annually in the NEI point source inventories and can be accessed through
the Emission Inventory System (EIS)81 for certain users. More recent years (including 2021 and
2022) do not currently have corresponding NEI datasets released. For 2021 and 2022 analyses,
AVERT uses information from the most recent NEI data year (2020). For power plants that were
newly constructed in 2021 or 2022, an average emission rate is used, based on existing power
plants that are similar to the newly constructed plant in terms of fuel type and prime mover type.
Table 5. NEI data used for each AVERT data year in AVERT v4.2.
AVERT data year
NEIyear used
2017
2017 (full NEI)
2018
2018
2019
2019
2020
2020
2021
2020
2022
2020
81 See https://www.epa.qov/air-emissions-inventories/emissions-inventorv-svstem-eis-qatewav.
SEPA
State and Local Climate
and Energy Program
-------
AVERT User Manual Version 4.2
Appendix C: Renewable Energy Hourly Profiles
AVERT's Main Module provides regional estimates of hourly capacity factors for generic solar and
wind projects. These capacity factors are meant to provide quickly accessible options to review
example renewable project portfolios in each of the regions discussed here. The user is
encouraged to develop site-, state-, or region-specific RE load profiles. Where such information is
not available or for the purposes of exploration, the proxy capacity factors in the Main Module
provide a reasonable basis for expected wind and solar hourly profiles. Hourly capacity factors can
also be scaled to reflect an annual capacity factor specified by the user. EPA routinely revisits the
default capacity factors and methodologies described below, and updates them to be consistent
with the latest available reported and modeled data.
Rooftop and Utility-Scale Photovoltaic
Annual hourly capacity factors for rooftop PV and utility PV were obtained from the National
Renewable Energy Laboratory's PVWatts v.1 tool.82 Each hourly capacity factor assembled for
each AVERT region is based on the average PV capacity factor for one to 16 cities in the region.
The number and location of the sampled cities were chosen to provide a representative distribution
of the AVERT region's insolation (energy from sunlight) at the largest load centers.
Onshore Wind
Annual hourly capacity factors for onshore wind were developed based on EPA's Power Sector
Modeling Platform v6 - January 2020 Reference Case data, which are primarily used as inputs to
the Integrated Planning Model (IPM) modeling platform.83 Table 4-38 from the January 2020
Reference Case data shows onshore regional potential wind capacity (MW) by techno-resource
group (TRG) and cost class for 63 electric regions in the contiguous United States.84 Table 4-39
shows onshore wind generation profiles (kWh of generation per MW of capacity). These data were
downloaded in Excel format.
Each of the 63 electric regions are matched to one of the 14 AVERT regions. Annual hourly
capacity factors were calculated for each of the 63 regions, then averaged for each of the 14
AVERT regions, weighted by the total potential for onshore wind capacity in each IPM region from
all TRGs and cost classes.
Finally, to better approximate observed hourly capacity factors, the annual capacity factors derived
from the January 2020 Reference Case data were compared with annual historical (2017-2020)
capacity factors for the same regions.85 The annual historical capacity factors were used to scale
82 National Renewable Energy Laboratory, n.d. PVWatts: A Performance Calculator for Grid-Connected PV
Systems. Accessed December 14, 2012. https://pvwatts.nrel.gov/.
83 EPA. 2019. Documentation for EPA's Power Sector Modeling Platform v6 - January 2020 Reference Case.
https://www.epa.aov/airmarkets/documentation-epas-power-sector-modelina-platform-v6-ianuarv-2020-
reference-case.
84 Each of the 63 regions are subdivided into smaller regions that are specific to individual states (e.g., PJM
Dominion in Virginia, PJM Dominion in North Carolina). As a result, there are 120 different IPM regions in the
contiguous United States, each with wind potential data further broken down by TRG and cost class.
85 Annual historical wind capacity factors were calculated using data from EIA 860 and EIA 923. See
https://www.eia.gov/electricitv/data/eia860/ and https://www.eia.gov/electricitv/data/eia923/.
SEPA
State and Local Climate
and Energy Program
-------
JtVert
AVoid«d Emissions and q»rwRatlon Tool^^
wwwepagovTavSt'^W
AVERT User Manual Version 4.2
the hourly capacity factors from the January 2020 Reference Case data using the following
approach:
annual historical
January 2020 Reference Case ^ capacity factors final hourly
hourly capacity factors January 2020 Reference Case capacity factors
annual capacity factors
Note that several regions (Carolinas, Florida, and Southeast) do not have reported historical wind
generation in the EIA datasets. As a result, these regions exclusively rely on capacity factor data
from the IPM dataset.
Offshore Wind
Offshore wind speed data were assembled using information from the Bureau of Ocean Energy
Management (BOEM) 2019 modeled hourly offshore wind dataset.86 These data were downloaded
in GIS point format. Figure 37 shows the coverage of the wind speed data in gray.
Figure 37. Map of offshore wind data and lease regions coded by AVERT region.
Next, wind speed data were filtered using a dataset of actual and proposed wind lease areas.87
Each screened BOEM lease area contains thousands of wind speed data points. Each one of
these points contains 24-hour wind speed data for each month, which represents a typical hourly
wind speed for a representative day for any given month.
BOEM lease areas were then assigned to each AVERT region—some AVERT regions comprise
one single BOEM lease area, while other AVERT regions have many BOEM lease areas. Next, the
average hourly wind speed (for each 24-hour interval) was calculated from all data points across
86 Bureau of Ocean Energy Management. 2019. Renewable Energy GIS Data: Hourly Wind Speeds.
https://www.boem.gov/renewable-enerqv/mappinq-and-data/renewable-enerqv-qis-data.
87 Bureau of Ocean Energy Management. 2019. Renewable Energy GIS Data: Wind Planning Areas, Wnd Energy
Areas and Renewable Energy Leases. https://www.boem.gov/Renewable-Enerqv-GIS-Data/. This dataset
describes the areas that are most likely to be developed with offshore wind in the next several years. This aligns
with the time horizon modeled in AVERT. In other words, it is unlikely that areas outside the current designated
and proposed BOEM lease zones would be developed with offshore wind in the near future (e.g., more than 5
years from the present day).
SEPA
State and Local Climate
and Energy Program
-------
JtVert
AVoid«d Emissions and q»rwRatlon Tool^^
wwwepagovTavSt'^W
AVERT User Manual Version 4.2
the lease areas within each AVERT region. Each 24-hour period was then replicated over the
course of the entire month to develop a single hourly wind speed dataset for each AVERT region,
containing one data point for each hour of the year.
Using data from NREL's 2016 report Offshore Wind Energy Resource Assessment for the United
States, the developers of AVERT applied a gross power curve and estimated losses to each hourly
wind speed data point to estimate the net power output for each windspeed.88 The team then
divided the net power output by the nameplate capacity of the representative power curve to
determine a scalable net hourly capacity factor for each of the hourly data points for each of the
AVERT regions.
Only AVERT regions with proximity to actual or proposed offshore wind lease areas can model the
addition of offshore wind generators. For example, the Gulf Coast and the Great Lakes do not have
actual or proposed offshore wind lease areas, so they are not included in this analysis. AVERT
provides offshore wind capacity factors for the New England, New York, Mid-Atlantic, Carolinas,
California, and Northwest AVERT regions. AVERT does not provide capacity factor data for the
Texas, Midwest, Central, Florida, Southeast, Southwest, Tennessee, or Rocky Mountains AVERT
regions. When users enter a non-zero capacity for offshore wind in AVERT regions that do not
have hourly offshore wind capacity factors, the model will simply display a change of 0 MW for
each hour. Note that currently, BOEM's wind planning areas are located only in certain areas of the
United States coastline. For example, there are no BOEM wind planning areas in Florida or Texas,
which means that the AVERT regions that largely encompass these states do not have offshore
wind profiles.
Users are encouraged to develop site-specific capacity factor profiles for RE options whenever the
data are available. It is important to note that AVERT is not a tool for formal greenhouse gas
accounting or establishing who may take claim credit for emission reductions of RE programs or
projects. It is recommended that companies follow the protocols from the World Resources
Institute's GHG Protocol and the Federal Trade Commission's Guides for the Use of Environmental
Marketing Claims for the purposes of greenhouse gas and carbon footprint accounting.89
88 National Renewable Energy Laboratory. 2016. 2016 Offshore Wind Energy Resource Assessment for the United
States. Section 7.3.1, Figure 9. https://www.nrel.qov/docs/fv16osti/66599. pdf.
89 Federal Trade Commission. 2012. Guides for the Use of Environmental Marketing Claims.
https://www.ftc.gov/sites/default/files/documents/federal register notices/guides-use-environmental-marketing-
claims-green-guides/greenguidesfrn.pdf.
SEPA
State and Local Climate
and Energy Program
-------
AVERT User Manual Version 4.2
Appendix D: Overview of AVERT's Statistical Module
For each region, the MATLAB®-based Statistical Module provides the model's core statistical
analysis. (For installation instructions, see Appendix A.)
Data analysis within the Statistical Module is conducted in five steps, described briefly in the sub-
sections below:
1. Parsing the base year into "bins" of hours with similar levels of total regional fossil-fuel
load.
2. Collecting statistical information (probability distributions for generation, heat input, and
emissions) on how each fossil-fuel EGU has responded to regional load requirements in
each hour of the base year.
3. Extrapolating this statistical information to extend to potential lower and higher fossil-fuel
loads not experienced in the base year.
4. Estimating the ranges of generation, heat input, and emissions likely to be experienced by
each EGU for each fossil-fuel load bin (or approximate regional load).
5. Preparing outputs for export to AVERT's Main Module.
Appendix E provides step-by-step instructions to using the Statistical Module.
The Statistical Module is also equipped to estimate how fossil-fuel EGUs respond to regional load
requirements given changes in the fleet of EGUs available in future years. The module inputs
information from AVERT's Future Year Scenario Template to identify existing EGUs to be retired,
the expected impact of pollution-control retrofits on existing EGUs' emission rates, and new EGUs
coming on line. AVERT then re-estimates all statistical information based on each region's
projected fleet of fossil-fuel EGUs for a particular future year scenario. (See Appendix F for a more
detailed description of the Future Year Scenario Template.)
Parsing Generation Demand into Fossil-Fuel Load Bins
In its first step, the Statistical Module sums up all fossil-fuel generation in each hour under analysis
to arrive at a total regional fossil-fuel load by hour (see Figure 38, which includes a detail of hours
3,000 to 4,000).90
90
Hour 3,000 = May 5. Hour 4,000 = June 15.
68
SEPA
State and Local Climate
and Energy Program
-------
Bsd^SRT
missions and qenaRaiion Tool^^
www.epa.gov/avan
AVERT User Manual Version 4.2
AVoidad Emissions.
Figure 38. 2011 hourly sum of fossil-fuel generation in the Texas region.
Hour of the Year
60.000
| 50,000
C
| 40,000
0)
c
S 30,000
"a3
=3
± 20,000
!/>
o
"" 10,000
3300 3400 3500 3600
Hour of Year
3700 3800
3900 4000
These hourly sums of fossil-fuel generation are sorted from lowest to highest generation level and
grouped into 41 "fossil-fuel load bins" for the purpose of collecting statistics for each EGU at each
approximate load level (see Figure 39).91 Thirty-seven of the bins contain 224 or 225 hours; the
second lowest and second highest bins contain 204 or 205 hours; and the bins for the lowest and
highest fossil-fuel loads contain just 20 hours each to best represent these extreme load levels.92
Bin thresholds (the fossil-fuel load levels dividing the bins) and bin medians vary by region.93
Figure 39 illustrates how the bins are formed relative to total system fossil-fuel load. The figure
shows a typical "load duration curve" in dark red for fossil generation in Texas in 2011. This curve
represents all fossil-fuel load levels of the year (8,760 data points) in declining order. The horizontal
axis represents the fraction of time that fossil generation is at or above a certain level (e.g., fossil
generation only exceeds the 20 percent marker in 20 percent, or 1,752, hours). The vertical axis
shows the fossil-fuel load for each point on the curve.
91 "Load" always refers to regional, system-wide demand, and never to individual unit generation. The fossil-fuel
load bins group together hours that have similar generation levels ignoring their chronological order.
92 The ranges of the historical fossil-fuel load bins are determined is as follows: A region's 8,760 one-hour loads are
sorted from low to high, and then divided into 39 bins each containing 224 or 225 hours, depending on rounding.
For each bin (excluding the highest and lowest of the 39 bins, described below), the maximum threshold is the
MW load of the highest-load hour in the bin, and the minimum threshold is the MW load of the highest-load hour
the next lower bin. Bin "widths" are the high bin threshold in MW minus the low bin threshold. Bin medians are the
load (in MW) of the median hour of the bin. The highest and lowest of the 39 bins are each divided into two parts,
such that there are 41 fossil-fuel load bins from historical data in every region. The lowest of the 39 original bins
is split into the 20 hours with the lowest loads and the remaining 204 or 205 hours; the highest bin is split into the
20 hours with the highest loads and the remainder.
93 AVERT results include additional fossil-fuel load bins designed to capture regional load levels that did not occur
in the base year (see the "Extrapolation to Higher and Lower Fossil-Fuel Loads" sub-section below).
&EPA
State and Local Climate
and Energy Program
60,000
50,000
40,000
30,000
20,000
10,000
1,000
2,000 3,000 4,000 5,000 6
,000 7,000 8,000 9,000
-------
JtVert
AVoid«d Emissions and q»rwRatlon Tool^^
wwwepagovTavSt'^W
AVERT User Manual Version 4.2
Figure 39. 2011 fossil fuel load duration curve for the Texas region, indicating load bins.
60,000
55,000
50,000
S
45,000
l/l
2
O
"5 40,000
0J
I—
-C
I-
c 35,000
CO
T3
ro
5 30,000
"qj
3
= 25,000
V)
V)
O
Ll_
20,000
15,000
10,000
The horizontal axis has 42 light blue lines on it, representing the outside thresholds of the 41 fossil-
fuel load bins. Thirty-eight of these lines are evenly spaced from zero percent to 100 percent,94
capturing 224 or 225 hours each. In other words, each of these bins represents slightly over 2.5
percent of the hours in the year (again, grouped according to total fossil load in that hour rather
than by chronology). At the extreme ends, there are two additional light blue lines very near to the
zero and 100 percent markers. These additional lines fall 20 hours from the extremes; therefore
there are two bins at the extremes with 20 hours each, and two bins just prior to the extremes with
204 or 205 hours each.95
Wherever a percentage threshold crosses the fossil-fuel load duration curve, it creates a horizontal
line, representing a fossil-fuel load bin threshold. These are the horizontal grey lines shown on the
chart above, closely spaced in the lower middle of the graph and spreading out toward the highest
and lowest loads. This is because the majority of hours experience total fossil load that is neither
extremely high nor extremely low. In other words, there are more hours represented in the middle
of the fossil load range (for example, the regime from ~20,000 to ~32,000 MW in Figure 39) than at
very high or very low fossil loads. In order to capture an approximately equal number of hours in
each bin, each bin in the middle of the fossil load range captures a narrower range of MW. This is
shown in Figure 39 based on the spacing between the grey horizontal lines, where the points along
the load duration curve that fall between two horizontal threshold lines are the points in the
corresponding fossil-fuel load bin, and the points in each bin (apart from the end binds, as
discussed above) represent roughly 2.5 percent of hours in the year.
94 Each line represents a demarcation of 2.56%.
95 20 hours is represented by 0.23% and 99.77% on this axis.
State and Local Climate
and Energy Program
-------
id Emissions and q«rwRatlon Tool
AVERT User Manual Version 4.2
Collecting Statistical Information
Next AVERT gathers statistics about how each EGU responds to the generation requirements of
each fossil-fuel load bin. Three types of probability distributions are constructed: frequency of
operation by fossil-fuel load bin, generation level by fossil-fuel load bin, and heat input and
emissions by generation level.
Frequency of Operation by Fossil-Fuel Load Bin
In this first set of probability distributions, AVERT calculates the share of hours within each fossil-
fuel load bin for which a particular unit is turned on (i.e., has generation greater than zero). Figure
40 shows the frequency of operation for three EGUs in the Texas region in 2011.
Figure 40. 2011 frequency of operation by fossil-fuel load bin for three indicative EGUs in the Texas
region.
W A Parish WAP6
701 MW
Victoria Power Station 9
286 MW
T H Wharton
73 MW
100%
90%
J 80%
£ 70%
Q.
? 60%
g 50%
£ 40%
.2 30%
tJ
2 20%
10%
•
•
• •• • • *
•
•
••
*
•
•
•
•
•
•
20,000 30,000 40,000 50,000
Fossil-Fuel Load Bins (MWh)
20,000 30,000 40,000 50,000
Fossil-Fuel Load Bins (MWh)
20,000 30,000 40,000 50,000 60,000
Fossil-Fuel Load Bins (MWh)
In the figure above, the 701 MW coal-fired EGU shown on the left operates in nearly every hour of
the year, with its probability of operation dropping below 90 percent only at the lowest fossil-fuel
load requirements. This pattern is typical of a baseload EGU that operates continually with the
exception of maintenance outages scheduled to occur at low load requirement levels. The middle
EGU, a 286 MW gas-fired station, operates only rarely at low load requirements, but its frequency
of operation increases steadily with regional demand. At fossil-fuel load levels above 40,000 MW,
this EGU operates in nearly every hour. This pattern is typical of an intermediate-load EGU such as
a combined-cycle EGU. The 73 MWgas turbine on the right is a peak-load EGU, operating only at
the highest load requirements.
Generation Level by Fossil-Fuel Load Bin
The second set of probability distributions calculated by AVERT describes generation output for
each EGU in operation in each fossil-fuel load bin.96 AVERT divides each EGU's generation into 19
evenly spaced "unit generation bins."97 Figure 41 depicts the intersection of these two types of bins.
Smaller fossil-fuel load bins (where the vertical lines are closer together) indicate a higher
concentration of hours at those load levels.
96 For each fossil-fuel load bin, AVERT filters out the units which did not generate, and reviews only the operational
units.
97 The thresholds between unit generation bins are unit-specific.
State and Local Climate
and Energy Program
-------
JtVert
AVoid«d Emissions and q»rwRatlon Tool^^
wvvw.epa.gov/avert
AVERT User Manual Version 4.2
Figure 41. Schematic of unit generation bins and fossil-fuel load bins.
Unit Generation Bins (MW)
(n = 19, width = maximum unit
capacity (MW) / 19)
J
Low Extrapolation
(n = varies by
region, width =
median historical
bin width)
Historical
(n = 41, width =
regional historical
fossil-fuel load
(MW), divided
into bins by hours)
High Extrapolation
(n = varies by
region, width =
median historical
bin width)
Fossil-fuel Load Bins (MW)
For each of the 41 fossil-fuel load bins, AVERT determines the number of hours in which the unit
generated at an amount within each of the 19 unit generation bins. In this way, the model creates a
discrete probability distribution of generation for each fossil-fuel load bin during all hours in which
the EGU is in operation.
Figure 42. shows the probability distributions of generation at two EGUs in the Texas region. The
axis to the bottom right of each plot represents the region's fossil-fuel load bins. The axis to the
bottom left represents unit generation bins, from zero to the EGU's maximum generation in the
base year. The vertical axis is the probability that the EGU is operating at the given unit generation
level in each fossil-fuel load bin.
72
SEPA
State and Local Climate
and Energy Program
-------
AVERT User Manual Version 4.2
Figure 42. 2011 generation level by fossil-fuel load bin and unit generation bin for two
indicative EGUs in the Texas region.
Lake Hubbard 2
B 533 MW
State and Local Climate
and Energy Program
73
SEPA
-------
bJ^ERT
AVoidod Einissiortt and qenaRailori Tool^^
www.epa.gov/avarl
AVERT User Manual Version 4.2
Heat Input and Emissions by Generation Level
The final set of probability distributions relate EGU heat input and S02, NOx, and CO2 emissions to
unit generation. For heat input and emissions of SO2, NOx, and CO2, statistics for the ozone
season and non-ozone seasons are gathered and stored.98 AVERT creates eight discrete
probability distributions—ozone season SO2, NOx, and CO2 emissions and heat input and non-
ozone-season SO2, NOx, and CO2 emissions and heat input—for each EGU at each of the 19 unit
generation bins. Probability distributions are not a function of regional fossil-fuel load. Figure 43
displays a single EGU's emissions of SO2 and NOx relative to its generation level.
Figure 43. 2011 ozone-season emissions of SO2 (right graph) and NOx (left graph) by generation level
at an indicative EGU in the Texas region.
Oklaunion Power Station 1
Extrapolation to Higher and Lower Fossil-Fuel Loads
The end purpose of AVERT is to allow users to estimate the emissions changes resulting from
recent historical or expected/proposed near-future energy policies. In either case, the range of
fossil-fuel load requirements in the base year may be insufficient to represent all scenarios of
interest. For example, a scenario might require the user to examine regional load requirements that
are lower than the range represented by the base year. In contrast, a user can choose to estimate
the emission changes from policies already in place today, which could entail examining a scenario
with fossil loads higher than the range represented by the base year.
In the third step of AVERT, load requirements outside the base-year range are estimated by
extrapolating each EGU's statistics below and above base-year regional load requirements. Two
sets of probability distributions are subject to extrapolation: probability of operation and generation
level for each fossil-fuel load bin . A flexible number of fossil-fuel load bins are constructed below
the regional minimum load (with a lowest bound of zero), and above the regional maximum, such
that the new maximum bin threshold is the coincident maximum generation of all of the fossil-fuel
EGU on the system—that is, the level of load that could be reached if every fossil-fuel EGU were
98 Where "ozone season" is considered to be May through September, inclusive, for most states (states with
different ozone season designations are not recognized in this model). Ozone season distinctions are used to
capture differences in emissions output where generators are required to reduce emissions output during
selected times of the year to reduce ozone formation. Heat rate (heat input divided by generation) and CO2 rates
are not considered to change considerably from season to season, but are recorded in these categories for
computational convenience.
&EPA
State and Local Climate
and Energy Program
-------
JtVert
AVoid«d Emissions and q»rwRatlon Tool^^
wwwepagovTavSt'^W
AVERT User Manual Version 4.2
operating at its maximum output." Bin thresholds and medians vary by region. Theoretically, the
regional extrapolated maximum can be reached by the simultaneous use of every EGU in the
system, but in practice load curves that reach this maximum are unlikely.
Extrapolating the Probability of Operation
The minimum amount of potential generation is zero, a level that would require zero generation
from all EGUs being modeled in the RDF. If an EGU is already at a zero probability of operation at
a low load requirement, the zero value is maintained into the lower potential fossil-fuel load bins.
Any probability of operation above zero is extrapolated linearly down to zero from the probability at
the lowest recorded load level.
The potential maximum generation is the combined simultaneous maximum output of all EGUs in a
region; to reach that maximum point, all EGUs in the region need to be operating at their full
capacity. EGUs that have a 100 percent probability of operation at the base-year's highest fossil-
fuel load bin maintain that probability of output. Any probability of operation lower than 100 percent
is extrapolated linearly up to the potential maximum from the probability at the highest recorded
load level.
Figure 44 displays extrapolated values for the probability of operation for three EGUs in the Texas
region. In this figure, black points represent the probability of operation at the base-year fossil-fuel
load, and gray points represent the probability of operation at potential high and low fossil-fuel
loads beyond the base-year range. Extrapolation is simple in the figure to the left (Tenaska), as the
EGU does not operate at all during the lowest loads and operates continually during the highest
load periods. The middle figure (Bayou) requires downward extrapolation to a zero probability of
generation at a zero load, and the figure to the right (Permian Basin) requires upward extrapolation
to meet the highest load requirements.
Figure 44. 2011 base year and extrapolated probabilities of operation for three indicative EGUs in the
Texas region.
100%
90%
o
80%
70%
O
c
60%
Ol
E
50%
o
40%
Mm
—
2
u_
20%
10%
Tenaska Kiamichi CTGDB1
349 MW
Bayou Cogeneration CG804
81 MW
1 . , , , • ~
/
\
1
1
1
. 1
Permian Basin CT3
89 MW
10,000 20,000 30,000 40,000 50,000 60,000
10,000 20,000 30,000 40,000 50,000 60,000
•A
10.000 20.000 30,000 40,000 50,000 60,000 70,000
Fossil-Fuel Load Bins (MWh) Fossil-Fuel Load Bins (MWh) Fossil-Fuel Load Bins (MWh)
Black points represent the probability of operation during base-year load periods. Gray points are the
probability of operation at potential low and high loads beyond the base-year range.
99 The number of fossil-fuel load bins outside the base-year range is determined as follows: For each region, the
median of the fossil-fuel load threshold times four sets the MW size of the extrapolated bins. Bins of this size are
extended below the base-year minimum until zero is exceeded and above the base-year maximum until the
coincident maximum peak load is exceeded. The lowest and highest bins are truncated to begin at zero and end
at the coincident maximum, respectively.
State and Local Climate
and Energy Program
-------
AVERT User Manual Version 4.2
Extrapolating the Generation Level
EGUs not only run more often at higher load requirements, but also need to generate higher levels
of output to meet the requirements of the higher fossil-fuel load bins. Within the base-year range,
EGU generation is described as a series of discrete probability distributions for each fossil-fuel load
bin.
The extrapolated space encompasses fossil-fuel load bins that extend to the highest possible
coincident peak generation (i.e., all EGUs in a region operating at maximum capacity
simultaneously) and down to zero. New fossil-fired load bins are created at intervals equal to four
times the size of the median load bin.
To extrapolate to potential higher fossil-fuel load bins, AVERT assumes that each unit will have a
100 percent probability of generating at its highest output if fossil load is at its theoretical maximum,
and 0 percent of generating at any other level of output given maximum fossil load. For each level
of generation (i.e., within each unit generation bin), AVERT determines the slope of the line
connecting the unit's probability of generating at that level at the highest historical fossil load bin to
either 100 percent probability (for the highest unit generation bin) orO percent probability (for all
other generation bins). The unit's probability of generating at that level is then extrapolated
accordingly. Once all levels of generation have been extrapolated, AVERT normalizes the height of
the extrapolated load bins such that the total value of all points in each load bin sums to one. A
similar process is repeated for lower extrapolated fossil-fuel load bins, except that AVERT assumes
that the unit's output will be zero if total fossil load is zero.
The example in Figure 45 (below) shows an extrapolation of EGU generation to potential lower and
higher fossil-fuel load bins. The graphs show base-year unit generation bins (on the left-hand
horizontal axis) for any fossil-fuel load bin (on the right-hand horizontal axis). The height of the
surface represents the probability of operating at a given generation level in a particular fossil-fuel
load bin. Below about 15,000 MW (the lowest fossil-fuel load bin median in 2011) and above about
56,000 MW (the highest fossil-fuel load bin median in 2011), the surface in the figure showing data
for the historical year (A) is blank, indicating that no hours fell into those bin combinations in the
base year.
To extrapolate to higher and lower fossil-fuel load levels, a line is extended linearly toward the
corner constraints described above—100 percent probability of generating at full output given
maximum fossil load, and no unit output at a zero fossil load. The bottom graph (B) shows the
results of this extrapolation. From 60,000 MWto the peak load, this method returns generation
exclusively at this EGU's maximum, 701 MW. When the probability of operation is zero, the
generation output is automatically set to zero as well.
SEPA
State and Local Climate
and Energy Program
-------
J/VeW
AVoidod Einissiortt and qenaRailori Tool^^
www.epa.gov/avarl
AVERT User Manual Version 4.2
Figure 45. 2011 base year (A) and extrapolated (B) probabilities of generation levels for
an indicative EGU in the Texas region.
W A Parish WAP6
653 MW
O
'-t—
O
LL
,40,000
50,000
60,000
.30,000
10,000 ,KjWsn
j 20,000"
fos
,40,000
50,000
60,000
.nn30.000
10'000'U,I "
FOSs«F^LOa<,B
77
SEPA
State and Local Climate
and Energy Program
-------
AVERT User Manual Version 4.2
Statistical Analysis
The fourth step in the Statistical Module is to estimate the predicted range and expected (average)
generation, heat input, and SO2, NOx, and CO2 emissions for each EGU at each of the fossil-fuel
load requirement bins, from zero MW up to the coincident maximum generation of all of fossil-fuel
EGUs in a region.
AVERT's Monte Carlo analysis (contained within the Statistical Module) uses discrete probability
distributions to estimate key variables' range and expected value for each EGU in each fossil-fuel
load bin. For each EGU and fossil-fuel load bin, AVERT draws three random numbers between
zero and one:
• The first number drawn is compared to the EGU's probability of operation at the selected
fossil-fuel load bin. If the number drawn is greater than the probability of operation, the
EGU is turned off and draws two and three are not conducted. If the number drawn is less
than the probability of operation, the EGU is turned on. Figure 46 illustrates this first draw.
Starting in fossil-fuel load bin number 15 (representing a particular system-wide fossil load
level), the simulator randomly draws a value of 0.25. This value is slightly lower than the
probability of operation in bin 15 (approximately 0.40), and this EGU is "turned on."
Figure 46. EGU frequency of operation and example of random draw selection.
1.00
0.90
0.80
= 0.70
I0'60
§ 0.50
£0.40
"¦ 0.30
0.20
0.10
0.00
• The second number drawn is compared to the discrete cumulative distribution function of
EGU generation for each fossil-fuel load bin. The model selects the EGU's unit generation
bin as the next highest EGU output greater than the cumulative probability value indicated
by the number drawn. Figure 47 illustrates this step: the random draw is 0.33, which results
in the selection of unit generation bin 12.
State and Local Climate
and Energy Program
Fossil-fuel Load Bin
-------
JtVert
missions and q«neRaiion Tuol^^
wvvw.epa.gov/avert
AVERT User Manual Version 4.2
AVoid«d Emissions.
Figure 47. EGU generation histogram, cumulative probability distribution, and example of
random draw selection.
2,000
1,800
1,600
!2
= 1,400
o
J 1,200
0
> 1,000
1 800
£ 600
LL
400
200
0
.
—i
* Random
=n
Draw #2
33
1
1
/
11
1
|
I I
| .
¦
y
m-trii
1
1
1
ll
1.00
0.90
0.80
0.70
c
T3
i_
o
a
a!
o
>
a
c
a>
3
a-
a)
10
9
8
7
6
5
4
3
2
1
0
Randc
)m Draw #3
=0.80
/
/
/
/
/
/
L
I-J
H
r
1
1.00
0.90
0.80
w
0.70 £ £
-------
AVERT User Manual Version 4.2
Generation, heat input, and emissions output from each EGU at each fossil-fuel load bin is
recorded for thousands Monte Carlo runs.100 Each of these runs repeats the process described
above of drawing and applying three random numbers in sequence. Examples of the projected
generation and NOx emissions at a hypothetical fossil-fuel load of 30,000 MW for the 270 fossil-fuel
EGUs in the Texas region are shown in Figure 49 and Figure 50.
Figure 49. Generation (MW) for 1,000 Monte Carlo runs at 270 EGUs in the Texas region at a fossil-fuel
load of 30,000 MW(2011),
100
200
300
400
500 600
Monte Carlo Run
700
800
900
1,000
Figure 50. NOx ozone season emissions (lb) for 1,000 Monte Carlo runs at 270 EGUs in the Texas
region at a fossil-fuel load of 30,000 MW (2011).
100
200
300
400
500 600
Monte Carlo Run
700
800
900
1,000
AVERT takes the average (expected value) generation; heat input; and SO2, NOx, and CO2
emissions for each EGU within a region across thousands of Monte Carlo runs and records these
values in a new structural array.
Statistical Output
The final step is to generate an Excel output file to store the expected values of generation and
emissions of each EGU at every level of base-year and potential load for each region, and a time
series of the total base-year fossil-fuel load in each hour of the year; this file becomes an input file
100 The base dataset provided by EPA, and the default for users of the Statistical Module, is 1,000 Monte Carlo runs
and 500 generation-only Monte Carlo runs.
&EPA
State and Local Climate
and Energy Program
-------
JtVert
AVoid«d Emissions and q»rwRatlon Tool^^
wwwepagovTavSt'^W
AVERT User Manual Version 4.2
to the Main Module. Eleven sections of the output file each are composed of a matrix of the names
and identifiers of each EGU and the expected value at each fossil-fuel load bin, with one section
devoted to each of the following:
• Generation (MW)
• Heat input (MMBtu, ozone season)
• Heat input (MMBtu, non-ozone season)
• SO2 emissions (lb, ozone season)
• SO2 emissions (lb, non-ozone season)
• NOx emissions (lb, ozone season)
• NOx emissions (lb, non-ozone season)
• CO2 emissions (tons, ozone season)101
• CO2 emissions (tons, non-ozone season)
101 CO2 emissions are divided into ozone and non-ozone seasons to maintain algorithmic consistency with SO2 and
NO* emissions. In AVERT results, changes in CO2 emissions are presented in terms of annual totals.
State and Local Climate
and Energy Program
-------
AVERT User Manual Version 4.2
Appendix E: AVERT's Statistical Module: Step-by-Step
Instructions
This section provides step-by-step instructions for using AVERT's Statistical Module to prepare
inputs for AVERT's Main Module.
Step 1: Determine Your Windows Operating Environment
The Statistical Module is designed to work in a 64-bit operating system environment, so you will
first need to determine if your Windows system operates in a 32-bit or 64-bit environment.
Generally, this information is displayed among the "Properties" of "My Computer" in Windows XP,
or "Computer" in Wndows Vista, Windows 7, or Wndows 8. In Windows 10, press the Start button,
type "Settings," select "System," select "About," and check under "Device specifications."
Instructions for determining your Wndows environment can be found at
https://support.microsoft.com/en-us/help/15056/windows-32-64-bit-faq..
Step 2: Download the Statistical Module Executable
Download the following two files and save them to your computer:
1) AVERT's Statistical Module executable package:
"AVERT StatMod [Year] 64bit_package." Download this MATLAB executable at
https://www.epa.gov/avert.
2) The MATLAB Compiler Runtime (MCR). Download the Wndows 64-bit version of the
MCR for R2012b from the Mathworks website at
https://www.mathworks.com/products/compiler/matlab-runtime.html.
Do not download a newer version of the MCR even if one is available from the
Mathworks website. It is important to download version R2012b, also known as version 8.0.
,The AVERT executable is packaged in a form that can only be compiled by an MCR of the
same vintage.
Once the AVERT package is downloaded, we
recommend creating a folder on your computer
titled "AVERT Statistical Module." Place the AVE
executable package in the folder and run the file
The package will decompress to three files and
three subfolders. These folders must stay in the
same folder as the Statistical Module for the
program to operate successfully.
The folders are:
• AVERT Future Year Scenarios: This
folder contains the future year scenario
template. Other versions of the future year scenario template must be saved in this folder
in order for the AVERT Statistical Module to see them.
• AVERT Output: This folder will hold RDFs generated by the Statistical Module.
State and Local Climate
and Energy Program
AVERT F utu re Yea r Sc en a ri o s
AVERT Output
CAMD Input Files
^ AVERT_StatMod_2021_v1_&4bit
fi^l avert_statm o d_2021 _v1 _64b it_p kg
I 1 readme
-------
AVERT User Manual Version 4.2
• CAMD Input Files: This folder contains MATLAB-formatted flat data files with hourly
generation and emissions from each fossil EGU in the United States. The most recent year
of data is packaged by default with the Statistical Module. Other years of data can be
obtained from EPA, and must be put in this folder to be accessed by the Statistical Module.
The three files are:
• avert_StatMod_[Year]_v1_64bit_pkg: The downloaded file containing all of the other files
and folders.
• AVERT_StatMod_[Year]_v1_64bit: The executable that will run the Statistical Module
once the MATLAB compiler is installed.
• readme.txt: Basic instructions on the folders and instructions for obtaining the MATLAB
compiler.
Step 3: Download the CAMD Database
The Statistical Module package contains, by default, the most recent data year of data. If another
year is desired, additional CAMD Power Sector Emissions Data compatible with AVERT are
available at https://www.epa.qov/avert. For most purposes, users will want to obtain the most
recent data year. Download the file and save it in the subfolder "CAMD Input Files."
Step 4: Install the MATLAB Compiler Runtime
The Statistical Module executable requires additional, free MATLAB software in order to function.
For additional instruction on how to verify that the MCR is installed properly on your computer,
consult the readme.txt file. As noted on the previous page, it is critical to download and install the
correct version of the MCR. Using a different version will give you an error message when you try
to run the executable file.
Step 5: If Desired, Complete a Future Year Scenario Template
The Statistical Module can create either a base-year or a future year scenario (i.e., a scenario in
which some EGUs are retired, new EGUs are
brought online, and other EGUs change emission
rates). The process of creating a Future Year
Scenario Template is described in Appendix F. A
template for the Future Year Scenario Template is
available in the folder "AVERT Future Year
Scenarios."
Step 6: Launch the AVERT
Executable
Click on the AVERT executable to launch the
Statistical Module. A window labeled "Input for
AVERT Model" will open. In this window, select:
• The number of Monte Carlo runs (default
value is 1,000).
State and Local Climate
and Energy Program
R3 Input for AVERT Model
l-=-l
Avoided Emissions and Generation Tool (AVERT) Statistical Module
Synapse Energy Economics, March 2013
Enter number of Monte Carlo runs:
I I
Enter number of generation-only Monte Carlo runs:
500
I
Minimum annual generation to participate (MWh):
1000
I
Write output file?
Y
I
Please name this run.
I
I
OK
Cancel
-------
AVERT User Manual Version 4.2
• The number of generation-only Monte Carlo runs (default value is 500).
• The minimum annual generation for an EGU to be considered in the model (default value is
1,000 MWh).
• Whether or not an output file should be written. Choose "Y" to create an input file for
AVERT's Excel-based Main Module; choose "N" to skip writing an output file (typically used
for test runs only).
• Designate a run name. This name will be part of the
output file name.
Tip: The number of Monte Carlo runs directly influences
how long a run takes to execute. To ensure that inputs and
outputs are correctly read, perform a test run with a small
number of Monte Carlo runs and generation-only Monte
Carlo runs (10 each). For final runs where the output will be
used in the Main Module, use at least 1,000 Monte Carlo
runs and 500 generation-only Monte Carlo runs. These are
the same parameters used in the base dataset supplied by
EPA.
Step 7: Choose a Base Year for
Analysis
A window labeled "Choose CAMD Dataset" will open. In this
window, choose the data file for your desired base year. A second window showing a progress bar
will also be visible.
Step 8: Choose a Base- or
Future Year Scenario
A window labeled "Choose Future Year
Scenario" will open. In this window, choose a
base- or future year scenario for this analysis.
A second window showing a progress bar will
also be visible.
Note that each historical baseline year has a
unique Future Year Scenario Template. Use
only the one associated with the historical
baseline year of interest. In other words,
having chosen a 2012 base year and using a
future year scenario, ensure that the base year
of the template is also 2012. In the example
here, three scenarios have been created for
the 2012 base year, titled "10PctRetire,"
"MidwestCTs," and "S02RateRed."
&EPA
State and Local Climate
and Energy Program
H MENU 1 1=1
Choose CAMD Dataset
AVERTCAM DArray_20 OS. mat
AVERTCAM DArray_20 0 9. mat
AVERT_CAM DArray_2010. mat
AVERTJCAM DArray_2011. mat
AVERT_CAM DArray_2012. mat
0 MENU
Choose Future Year Scenario
-------
cndVEi
AVERT User Manual Version 4.2
Step 9: Choose Region(s) of interest
A window labeled "Choose one or more regions:" will open.
Choose the region for analysis and click "OK," or choose
"Select All" to launch an analysis of each of the 14 AVERT
regions in turn. Two additional windows showing progress bars
will be visible.
The selection of a region launches the full Monte Carlo
analysis, which can take up to several hours to complete
depended on computer processing speed, region selected, and
number of Monte Carlo runs selected.
A progress bar will inform the user of how far the program has
progressed through various analysis stages, with the Monte
Carlo runs usually taking the longest time. After all Monte Carlo
runs are complete, an output file will be written if selected at the
start of the program.
Each region is complete when the status bar reads "Finished
with [Region]." If more than one region is chosen, the program
automatically proceeds to the next region.
0
x
Choose one or more regions:
California
H
Carolines
Central
Florida
Mid-Atlantic
Midwest
New England
New York
Northwest
Rocky Mountains
Southeast
Southwest
Tennessee
Texas
V
Select all
OK
Cancel
85
State and Local Climate
and Energy Program
SEPA
-------
AVERT User Manual Version 4.2
Appendix F: AVERT's Future Year Scenario Template
AVERT is equipped to estimate change in emissions in scenarios projecting a future year by
making adjustments to the regional fleet of fossil-fuel EGUs before calculating the probability
distributions and expected values discussed in Appendix D.
The user implements these changes before running AVERT's Statistical Module using a
spreadsheet called "AVERT Future Year Scenario Template [Year] v.4.1where year is the data
year associated with the file. (Note that the AVERT v4.2 update did not involve any changes to the
Statistical Module or Future Year Scenario Template, so the latest versions of those resources
continue to be those labeled "v4.1.") The most recent historical baseline year template is stored in
a subfolder of the Statistical Module called "AVERT Future Year Scenarios." To access and use the
Future Year Scenario Template, download the Statistical Module, following instructions in Appendix
E. To access other historical baseline year datasets and future year scenario templates, visit
https://www.epa.gov/avert.
For each user-defined scenario, the user is strongly
recommended to save this file (in the same
subfolder) using the following naming convention:
"AVERT Future Year Scenario [Year] v.4.1
[Scenario X]," where "Scenario X" is a user-defined
name and "Year" is the historical baseline year.
The Future Year Scenario Template workbook
stores ancillary information about each EGU in the
system, and also includes worksheets that the user modifies directly and then inputs into the
Statistical Module. This section describes AVERT's process for projecting three types of user-
specified adjustments to the fossil-fuel generation fleet:
• Retiring existing EGUs
• Adding additional "proxy" EGUs
• Changing emission rates for existing EGUs to represent pollution-control retrofits
Users can make adjustments for multiple regions simultaneously; AVERT will correctly associate
each EGU with its appropriate region.
Please note that each historical baseline year has a unique Future Year Scenario Template.
Use the template associated with the historical baseline year of interest.
Retirement of Existing EGUs
EGUs can be retired from the analysis using the "Retires_Modifications" worksheet. The user finds
the EGU of interest and selects "yes" in the dropdown menu under "Retire?"
Addition of Proxy EGUs
New EGUs can be added to the fossil-fuel fleet in the "Additions" worksheet. This process is more
complex than that for retirements and retrofits, and requires some knowledge of the types of EGUs
expected to be added into the system. To add a new EGU, the user finds a "proxy" EGU for which
statistics are already recorded in AVERT and modifies this proxy to meet the requirements of the
oEPA
State and Local Climate
and Energy Program
AVERT Future Year Scenarios
AVERT Output
CAMD Input Files
IS1 AVERT StatMod 2012 vl 64bit_package
* AVERT_RegionNames
AVERT_StatM o d_2012_vl_64 b it
_| readme
-------
bJ^ERT
AVoidod Einissiortt and qenaRailori Tool^^
www.epa.gov/avarl
AVERT User Manual Version 4.2
user-defined scenario. In most cases and in most regions, the diversity of EGUs is sufficient to
provide a proxy for most traditional fossil-fuel generating resources. If completely new types of
resources are to be added (i.e., advanced combined cycle, integrated gasification combined cycle,
or fossil-fuel backup for wind plants), the proxies available for selection may be insufficient.
In the "Additions" worksheet, the user copies existing rows to create one row for each new EGU
required (see Figure 51)
Figure 51. 2011 Screenshot of example EGUs in the "Additions" worksheet.
Either select a county from the
dropdown, or enter manually
Fuel Unit . Description Capacity
# Region Unit ORSPL UNHID p , ..J™, State
Type Type (Note that "0 MW" units did not run in 2011.) (MW)
Lat - Lon -
County
County County
1
TX
Gas
ec
Bayou Cogeneration Plant
CG802
10298
CG802
This is a 84 MW unit. It is located in Harris
County, TX. In 2011, it ran for 516 GWh at a
capacity factor of 70%.
TX
Bastrop
30.126 -97.296
2
TX
Gas
cc
Bayou Cogeneration Plant
CG802
10298
CG802
This is a 84 MW unit. It is located in Harris
County, TX. In 2011, it ran for 516 GWh at a
capacity factor of 70%.
TX
Bastrop
30.126 -97.296
3
TX
Gas
cc
Bayou Cogeneration Plant
CG802
10298
CG802
This is a 84 MW unit It is located in Harris
County, TX. In 2011, it ran for 516 GWh at a
capacity factor of 70%.
TX
Bastrop
30.126 -97.296
12
TX
Gas
CC
Bayou Cogeneration Plant
CG802
10298
CG802
This is a 84 MW unit. It is located in Harris
County, TX. In 2011, it ran for 516 GWh at a
capaci ty factor of 70%.
TX
Bastrop
30.126 -97.296
13
TX
Gas
CC
Bayou Cogeneration Plant
10298
CG802
This is a 84 MW unit. It is located in Harris
County, TX. In 2011, it ran for 516 GWh at a
TX
Bastrop
30.126 -97.296
For each proxy EGU, the user selects from dropdown menus in the relevant cells:
• The region in which the EGU is to be placed
• The fuel type (gas, oil, coal, or other)
• The unit type (combined cycle, combustion turbine, steam, or other)
The user then chooses an appropriate proxy from a list of available EGUs that meet these selected
criteria, and the worksheet automatically fills in the ORISPL code and Unit ID of the proxy and
creates a brief description including the base-year capacity, output, and capacity factor of the EGU.
Next, the user adapts the proxy to more closely meet requirements by selecting:
• The desired capacity of the new EGU (i.e., it does not have to be the same size as the
historical EGU)
• The state and county in which the EGU will be located
The worksheet automatically fills in the latitude and longitude center of that county as the new
EGU's default location. If a more precise location is known, the user can override this
latitude/longitude selection by manually entering the correct coordinates. The only purpose of this
location selection is to map EGUs in AVERT's Main Module. The latitude and longitude serve no
function in model calculations.
For units added in the Future Year Scenario Template, PM, VOCs, and NH3 emission rates for are
automatically assigned in the Main Module. These emission rates are identical to the emission
rates of the proxy unit.
&EPA
State and Local Climate
and Energy Program
-------
mt^/EF
1 AVERT
I Ft! w:l»-. :"i: 1 y -IK-R.ll Ir.r
wwwepagovTavSt'^W
Pollution-Control Retrofits
AVERT User Manual Version 4.2
Expected changes in emission rates due to pollution-control retrofits are also made in the
"Retires_Modifications" worksheet. The user finds the EGU of interest; selects "Yes" in the
dropdown menu under "Revise Emissions Rates?"; and inputs new rates in Ib/MWh for SO2 and
NOx, and in tons/MWh for CO2, in columns I, J, or K, respectively. To leave the rate for a particular
pollutant unchanged, the user leaves the relevant cell blank. New rates entered by the user must
be greater than zero. These adjusted emission rates will be employed in AVERT as single point
estimates of the mean rate; no probability distribution for adjusted emissions is developed for
retrofit EG Us.
Running Future Year Scenarios in AVERT
When running AVERT's Statistical Module, the user is presented a menu of future year scenario
files saved in the "FutureYearScenario" subfolder. The user can choose one of these files or select
"Present Year Analysis (no modifications)." If the user selects a "Present Year Analysis," the model
does not read or use any changes to the dataset, including retirements, additions, or changed
emission rates. If the user selects a particular future year scenario, the retirements, additions, and
emissions modifications from that scenario's workbook are read into the Statistical Module. Once a
region has been selected for analysis, the Statistical Module reports the individual EGUs that have
been removed from or added to the region.
Note that each historical baseline year has a unique Future Year Scenario Template. Use the
template associated with the historical baseline year of interest.
Future year scenarios require an additional level of calculations before the five steps described in
Appendix D can be carried out. For each fossil-fuel load bin, the average generation level of each
EGU (retired and active, and including new units) must be determined in a separate Monte Carlo
analysis. The results of this analysis change the system fossil-fuel load perceived by all of the
remaining EGUs to generate the correct output. Because the generation of each EGU is
independently derived, each unit's generation is not affected by the generation levels of other
EGUs.
If the net change to generating capacity of retiring old and adding new EGUs results in an increase
in total generation, the region will incorrectly generate an amount greater than the required fossil
load. AVERT determines how much to back down the "perceived" fossil-fuel load in each bin to
output the appropriate amount of generation for that bin.102 For net reductions in generation, the
algorithm is simply reversed: perceived system fossil-fuel load is increased, allowing each EGU to
generate more than it otherwise would and make up the gap left by retired EGUs.
102
In this case, "perceived" load is the fossil-fuel load bin for which the model assigns generation and emissions.
88
SEPA
State and Local Climate
and Energy Program
-------
AVERT User Manual Version 4.2
Appendix G: AVERT Regions and Instructions for States
that Cross Regional Boundaries
AVERT 3.1.1 divides the contiguous United States (including Washington D.C., but not Alaska,
Hawaii, or other territories) into 14 distinct regions. These regions are aggregates of one or more
balancing authorities. Refer to Appendix B for additional detail on how the regions were developed.
Each EGU is assigned to exactly one AVERT region. These assignments may change overtime as
a result of changes in the dynamics of supply and demand. The current version of AVERT assigns
EGUs to regions based on the assignments in the 2018 edition of ElA's 861 database. Electrical
boundaries do not necessarily represent political boundaries. As such, only 24 states and the
District of Columbia are encompassed entirely within one AVERT region. The remaining states
straddle two or more AVERT regions. Refer to Table 1 (on page 17) for the AVERT regions and the
states they contain, either in whole or in part.
This section provides instructions for states that are split across more than one AVERT region.
AVERT results represent the emission changes of the programs only on generators that are
contained within that AVERT region. AVERT regions are defined not by state geography but by the
electric sales that occur within their borders, as defined by utility-reported sales to EIA. To capture
the emission changes of a state-wide energy policy across two or more AVERT regions, the energy
policy must be parsed between the two (or more) AVERT regions straddled by the state.
Figure 52 shows a schematic of such an example. In this case, State A crosses AVERT regions 1
and 2, and thus is only partially represented in each. However, the vast majority of State A's sales
are located in and serve AVERT region 2. With some exceptions, as an approximation, the effects
of the energy policy should be split into the two AVERT regions ratably, such that 90 percent of the
program (represented by 90 percent of the sales) is run within AVERT region 2, and 10 percent of
the program is run within AVERT region 1.
Figure 52. Schematic of recommendation to states that cross AVERT regions.
The exception to this rule is if the user has explicit knowledge of the location of new energy
projects, programs, or policies and can readily identify the region into which they will fall using the
map in Figure 3.
Table 6 indicates, by state, the approximate fraction of electricity sales found in each AVERT
region. This table was constructed by reviewing how much fossil electricity was sold in 2018 in
each AVERT region. For example, an air quality planner in Arkansas reviewing the displaced
emissions benefit of 2,000 MW of wind would run AVERT twice—once with 1,475 MWin the
Midwest region, and once with 525 MW in the Central region—and then aggregate the results of
these runs. An Arizona air quality planner, conversely, would run AVERT only once, with all the
energy changes attributed to the Southwest region.
State and Local Climate
and Energy Program
-------
JtVert
AVoid«d Emissions and q»rwRatlon Tool^^
wvvw.epa.gov/avert
AVERT User Manual Version 4.2
Table 6. State apportionment in AVERT regions, based on electricity sales in 2018.
State
Region
California
Southeast
Carolinas
Florida
Tennessee
Mid-Atlantic
New England
New York
Northwest
Rocky
Mountains
Southwest
Texas
Midwest
Central
Alabama
74%
26%
Arkansas
74%
26%
Arizona
100%
California
100%
Colorado
100%
Connecticut
100%
District of Columbia
100%
Delaware
100%
Florida
6%
94%
Georgia
98%
2%
Iowa
94%
6%
Idaho
100%
Illinois
65%
35%
Indiana
21%
79%
Kansas
100%
Kentucky
15%
30%
55%
Louisiana
93%
7%
Massachusetts
100%
Maryland
100%
Maine
100%
Michigan
4%
96%
Minnesota
99%
1%
Missouri
65%
35%
Mississippi
23%
32%
44%
Montana
91%
2%
7%
North Carolina
96%
4%
North Dakota
53%
47%
Nebraska
4%
96%
New Hampshire
100%
New Jersey
100%
New Mexico
5%
60%
35%
Nevada
100%
New York
100%
Ohio
100%
SERA
State and Local Climate
and Energy Program
-------
AVERT User Manual Version 4.2
State
Region
California
Southeast
Carolinas
Florida
Tennessee
Mid-Atlantic
New England
New York
Northwest
Rocky
Mountains
Southwest
Texas
Midwest
Central
Oklahoma
5%
95%
Oregon
100%
Pennsylvania
100%
Rhode Island
100%
South Carolina
100%
South Dakota
25%
25%
50%
Tennessee
98%
2%
Texas
1%
86%
5%
7%
Utah
97%
3%
Virginia
100%
Vermont
100%
Washington
100%
Wisconsin
100%
West Virginia
100%
Wyoming
62%
38%
91
SEPA
State and Local Climate
and Energy Program
-------
AVERT User Manual Version 4.2
Appendix H: Frequently Asked Questions
AVERT Inputs
Are users restricted to the energy profiles created within AVERT's Main Module?
No. The Main Module maintains simple wind and solar profiles for various regions of the United
States for the convenience of users, but does not restrict users to these profiles. Users are
encouraged to create energy profiles that reflect their regions and assumptions. Such profiles can
be copied into the manual entry page of the Main Module.
Can I review RE options other than wind and solar generation?
Yes. New non-intermittent, must-take renewable generation, such as hydroelectric generation or
geothermal generation, can be approximated using either the manual hourly data entry or the
preset EE sections. For example, if you want to model the expected changes resulting from a new
hydroelectric generator, you could click on "Enter hourly data manually" in AVERT's Step 2 and
enter the expected hourly generation curve. If you assume the non-intermittent resource functions
as a purely baseload resource, you could use the "annual GWh" setting as a proxy.
Is there a way for baseload renewables to be included?
You can model changes in non-emitting, must-take baseload renewables like geothermal or
hydroelectricity in AVERT using two different methods.
1. You can use the "annual GWh" setting in Step 2. This means you are essentially modeling
an EE program as a proxy for baseload renewables. In order to account for transmission
and distribution losses, you must first correct your total annual GWh value by reducing its
value according to the value of the T&D losses for that year. You can find annual T&D
losses in Table 2 of the "Library" tab in the Main Module.
2. If you have an 8,760 hour profile of your EGU, you can enter the profile directly into the
"Manual Energy Profile Entry" page of the Main Module. Note that in this case
adjustments to T&D losses are automatically included.
How do you handle biomass, waste combustion generators, or CHP generators in AVERT?
If biomass, waste combustion, or CHP generators are emitting and have capacities greater than 25
MW, they are included in EPA's Power Sector Emissions Data.
AVERT is not currently equipped to estimate the emissions of emitting generators that do not report
to CAMD. However, if you know the expected generation and emissions from a new biomass,
waste, or CHP generator, you could review the estimated displaced emissions and generation from
the inclusion of that generator using AVERT (assuming an hourly energy profile is known for the
new EGU) and then add in that generator's emissions post hoc. To do so, follow the steps below:
1. Determine the estimated energy profile for the CHP generator and associated stack
emissions.
2. Input the energy profile for CHP generator into AVERT under "manual energy profile
entry" in Step 2.
3. Run the Main Module to determine emissions offset due to new CHP generator.
4. Subtract the CHP stack emissions from emissions offsets to determine the total change in
emissions.
SERA
State and Local Climate
and Energy Program
-------
AVERT User Manual Version 4.2
net emissions reduction from CHP generator = AVERT displaced emissions + CHP stack
emissions
There is no current option to review emissions displaced from new biomass, waste, or CHP
generators if they do not already report to CAMD.
Are there any plans to incorporate electricity production from biogas facilities into AVERT?
If the facility is an emitting generator and has a capacity greater than 25 MW, it is currently included
within the AMP EGU dataset. Otherwise, there are no current plans to incorporate electricity
production from these types of facilities. Often, these facilities may generate electricity according to
their onsite needs and fuel supply and may not be affected by regional changes in load or dispatch.
Can storage technology be modeled in AVERT?
Yes. Energy storage, such as batteries, pumped hydroelectric generation, and compressed air
storage, is used to capture excess energy at low-cost (often low-demand) periods, and release that
energy at high-cost (often high-demand) periods. AVERT draws statistics from, and dispatches
against, a base year time-series of demand. You can create an energy storage assumption (how
much energy is drawn off a renewable resource, when it is released, and what the associated
losses are) and create a manual energy profile that reflects this energy storage assumption. As
with all other assumptions in AVERT, users are encouraged to create a time-series of energy
changes that fits their region and assumptions.
DOE has an emissions calculator that uses AVERT as its engine. This web-based tool is called
Grid Impact Emissions Quantification (GRIDPIQ). It specifically models storage and other
distributed energy resources scenarios; however, the outputs are only regional. For more
information, visit https://qridpiq.pnnl.qov/app/#!/.
How does AVERT account for the dispatch of new RE into the existing system?
RE generation sources typically have very low variable operating costs; in other words, they are
very inexpensive to operate once they are constructed. Typically, low-operating-cost resources are
dispatched first, and increasingly expensive resources are dispatched thereafter. RE sources are
assumed to dispatch first (a common assumption across many economic dispatch models), and
thus can be modeled as an equivalent reduction in demand. AVERT simply compares the
generation and emissions of all fossil resources before the new RE resource (i.e., at the equivalent
of full demand in each hour) and after the new RE resource (i.e., at the equivalent of a reduced
demand in each hour). The difference in generation and emissions between the before and after
scenarios represents the emissions displaced by RE.
How does AVERT account for the dispatch of new EE into the existing system?
EE usually results in a reduction in demand (in some cases and for some types of programs, it may
result in a shifting of demand to off-peak hours). AVERT simply compares the generation and
emissions of all fossil resources before the new EE resource (i.e., at the equivalent of full demand
in each hour) and after the new EE resource (i.e., at the equivalent of a reduced demand in each
hour). The difference in generation and emissions between the before and after scenarios are the
emissions that are displaced by EE.
Is there a bound on the smallest change that is appropriate to model?
No, there is no specified lower bound, but users may find the following guidance useful when
analyzing small changes. Users can review the output chart titled "Hourly Results by Week" for an
State and Local Climate
and Energy Program
-------
JtVert
AVoid«d Emissions and q»rwRatlon Tool^^
wwwepagovTavSt'^W
AVERT User Manual Version 4.2
indication of how closely their expected energy changes are captured in hour-to-hour unit changes
for one week. For very small inputs, this graphical interface will indicate a rougher hour-to-hour
energy profile—i.e., the resulting change in generation will look less like the amount of energy
change expected. Note that all numerical results are shown rounded to the nearest 10 unit.103
Dashes indicate that AVERT reported a value greater than zero, but lower than the level of
reportable significance. In some cases, no reasonably sized energy policy will result in reportable
changes. For example, the review of monthly results for a single small county in a low load month
may often result in low significance results. However, the user can use discretion to determine if an
energy policy has resulted in an acceptable level of significance based on the signal-to-noise
diagnostic and the degree to which critical results are below the level of acceptable significance
(i.e., are obscured by dashes in numerical results).
For some smaller inputs, users may find increased "noise" in a given model run. Although the
resulting annual changes in emissions or annual average emission rate may resemble what one
might expect for results, based on the inputs, results for individual EGUs may exhibit more volatile
behavior. For example, in a modeling run that is intended to reduce fossil load (e.g., the user has
input some quantity of EE or RE), some individual EGUs may exhibit decreases in generation (as
one might expect), while others may exhibit unexpected increases in generation. For a check of
these results in aggregate, the user should view the "Signal-to-noise diagnostic," found on the
"Display Results" page of AVERT's Main Module. As described in this manual (p. 47), this scatter
plot shows the changes in generation calculated by AVERT (on the y-axis) against the energy
change input by the user. More reasonable results (from a program size perspective) will appear
closer to 1:1 lines. Smaller load changes have more noise (i.e., scatter) in this plot, while larger
load changes have a straighter line relationship. The R2 value in the title of the chart indicates how
much of the change in generation can be explained by the user-input energy change. For
examples, an R2 value of 0.9 indicates that AVERT has captured 90 percent of the change in
generation required by the user, while a value of 0.7 indicates that AVERT has only correctly
captured 70 percent of the energy change input by the user (i.e., noise accounts for 30 percent of
the observed variability).
Figure 53 shows two different energy profiles with very different R2 values from the same region,
and designed similarly. The graph on the top is a 1.5 percent load reduction during the peak 20
percent of hours. The reduction is sufficiently sized such that the generation reduction is able to
match the requirement very closely—over 99 percent of the reduction in generation is a direct
result of the energy change input. The graph on the bottom is a 0.25 percent load reduction during
all hours of the year. The reduction is insufficiently sized in this case and results in a wide range of
uncertain results. By comparison, 92 percent of the generation reduction can be attributed to the
energy policy—the rest is noise.
In general, modeling runs that produce a high level of noise (i.e., a low R2) may be useful for
describing high-level results (such as annual and regional changes in emissions) but may be less
useful for describing changes in generation or emissions on an hourly basis or at any one individual
103 The Power Sector Emissions Data are reported in integer units of MWh (generation), lb (NO* and SO2), tons
(CO2), and MMBtu (heat input). Results in AVERT are rounded to the closest 10 MWh, lb NO* and SO2, tons
CO2, and MMBtu fuel input.
State and Local Climate
and Energy Program
-------
cndVEi
AVERT User Manual Version 4.2
EGU. Modeling runs with comparatively less noise (and larger R2 values) are better suited for these
purposes.
Figure 53. Examples of two different load reductions with different R2 values in the signal-to-noise
diagnostic. Top: 1.5 percent load reduction in peak 20 percent of hours. Bottom: 0.25 percent load
reduction in all hours.
Change in Total Unit Generation Relative to Energy Impact
(MW) for All Hours, R2=1.00
Expected Change in Generation (MW)
-1,800 -1,600 -1,400 -1,200 -1,000 -800 -600 -400 -200 0
-400 $
-800 &
Change in Total Unit Generation Relative to Energy Impact
(MW) for All Hours, R*=0.92
Expected Change in Generation (MW)
-200 -150 -100
-300
Is there a hound on the largest change that is appropriate to model?
There is not a formal bound on the largest project, program, or policy that should be modeled in
AVERT, In general, users should note that AVERT is designed to review marginal operational
changes in load, rather than large-scale changes that may change fundamental dynamics. As a
guideline, EPA suggests that modeled scenarios generally not deviate 15 percent from baseline
fossil generation in any given hour.
vvEPA
State and Local Climate
and Energy Program
-------
AVERT User Manual Version 4.2
With this 15 percent guideline, analysts should use their judgment in deciding whether the results
are appropriate for their uses. To assess appropriateness of results, analysts can consider the
number of hours out of 8,760 (the number of hours in a year) that exceed the 15 percent threshold
and how much greater than 15 percent the resultant fossil generation values are. Analysts should
also consider their specific interest in using AVERT. For example, an analyst interested in only
annual results may likely be less sensitive to the 15 percent guideline than an analyst interested in
the hourly results that span the hours where the threshold is exceeded.
In addition, users may encounter situations in which their selected change to load produces a new
hourly load that is outside the range calculable by AVERT. This may occur in situations where load
is reduced by 50 percent or more relative to the hour with the lowest load (i.e., well outside the
recommended threshold of 15 percent). In situations where load is increased, users may encounter
this issue at load increases as low as 10 percent for some regions. In these situations, users
should refine their load changes such that an error is not produced on the "Manual User Input"
page.
For reference, Table 7 provides each AVERT region's total annual load, in GWh. Knowing the total
annual load in each region may be helpful when using AVERT's "reduce generation by X% in all
hours" option to model the impacts of a broad-based EE program. Figure 54 identifies the
distribution of hourly loads in each of the 14 AVERT regions, using data from 2022. Table 7 also
shows the maximum and minimum possible load levels able to be modeled in AVERT. These
values provide helpful context if one knows the absolute size of a project, policy, or program in
units such as MW and one wants a sense of the percentage of regional load that it represents.
Note that fossil loads in Table 7 and Figure 54 have not been adjusted to reflect the transmission
and distribution losses inherent to each region. Demand-side measures (such as EE or distributed
solar) avoid not only electricity demand, but also the electricity associated with transmission and
distribution losses. As a result, demand-side programs increase the avoided fossil load by an
additional 5-9 percent, depending on the region and the year being analyzed.
Table 7. Total regional fossil loads in AVERT regions, 2022.
AVERT region
Total annual fossil
load (GWh)
Maximum possible
hourly load (MW)
Minimum possible
hourly load (MW)
California
85,194
27,936
1,565
Carolinas
102,593
24,715
4,836
Central
148,702
39,403
4,119
Florida
184,555
37,014
10,370
Mid-Atlantic
476,645
99,333
27,783
Midwest
490,318
101,141
28,286
New England
49,809
15,220
1,564
New York
66,357
19,023
2,457
Northwest
118,165
20,466
3,782
Rocky Mountains
52,308
11,518
3,102
Southeast
177,108
38,650
11,197
Southwest
81,971
17,645
3,617
Tennessee
78,855
16,475
2,658
Texas
261,628
57,977
7,043
SERA
State and Local Climate
and Energy Program
-------
JtVert
AVoid«d Emissions and q»rwRatlon Tool^^
wvvw.epa.gov/avert
AVERT User Manual Version 4.2
Figure 54. Characteristics of regional hourly fossil loads, 2022.
125,000
TJ
ro
o
3
O
5 S
o
QJ
bJ3
E
100,000
75.000
50,000
25,000
j—I
—I
30:
qBBb :
B B t
AVERT Region
Maximum
Average
Minimum
Why might actual planned offshore wind projects trigger a warning in AVERT?
The Excel Main Module gives the user a warning when they have entered an energy profile that
collectively exceeds 15 percent of load in at least one hour of the year. This warning could appear
when modeling certain large policies, programs, or projects, including ambitious offshore wind
projects that may be implemented in the near future. However, this warning will not prevent the
user from modeling large quantities of offshore wind—the user can simply ignore the warning
message if desired.
Are energy changes applied over the whole region? Is there a way to apply them at the
state, county, or municipal level only?
Because AVERT does not model transmission constraints within a region, energy changes are
assumed to have change emissions throughout the selected AVERT region. A limitation of AVERT
is that it is insensitive to the physical location within a region of new projects, programs, or policies,
despite the fact that real-world dispatch decisions may be quite sensitive to specific locations of
resources resulting from energy policies as well as EGUs. AVERT assumes that energy changes
are spread across the modeled region. It cannot currently identify the differential effects of local
versus regional energy changes. Such differentiation requires the use of a production cost model.
For more information, see "Limitations and Caveats" in Section 2 of this manual. Detail on changes
at the state and county level are available on the output sheet "Annual Results by County."
AVERT Results
Does AVERT account for losses?
Yes, AVERT accounts for three types of losses. First, gross generation as collected in the Power
Sector Emissions Data is corrected to account for parasitic consumption of energy onsite at fossil-
fired EGUs. AVERT applies a parasitic loss factor to each EGU based on unit and fuel
characteristics and subsequently calculates emissions based on each unit's "net" output of energy
State and Local Climate
and Energy Program
-------
JtVert
AVoid«d Emissions and q»rwRatlon Tool^^
wwwepagovTavSt'^W
AVERT User Manual Version 4.2
exported to the grid. Second, reductions in fossil load due to EE and distributed PV (and increases
in fossil load from increased demand) are corrected to account for avoided grid (transmission and
distribution) losses, using region-specific, year-specific grid loss factors. Wind and utility-scale PV
profiles are not corrected for these losses, as it is assumed they are located at a similar distance
from load centers as fossil-fired EGUs. Finally, additional loss factors are assumed for offshore
wind due to the fact that these resources are commonly located far from load centers, and because
associated transmission lines may be underground or underwater. Using loss estimates from
NREL, these factors act to reduce the hourly capacity factors of offshore wind resources.104 See
Appendix C for more detail on how offshore wind capacity factors have been developed in AVERT.
How can users assess the accuracy of the results returned by AVERT's Main Module?
The current version of the Main Module is not equipped to return information on the accuracy or
uncertainty of the model results.
While the Monte Carlo analysis run by AVERT creates information useful for some forms of
uncertainty analysis, using this information to assess the accuracy of the results returned by the
Main Module would require simultaneously performing a Monte Carlo analysis on the baseline
scenario and on the modified scenario, and returning uncertainty metrics associated with the
difference between these two scenarios. The current version of AVERT does not contain this
information. EPA is exploring a future version of AVERT that could perform explicit uncertainty
analyses and allow users to assess the accuracy of results returned by the model.
What is the accuracy of the map chart in AVERT's Excel-based Main Module?
The map is a visual cue only, and should not be used as a precise rendering of the location or
influence of change in generation or emissions from any EGU or cohort of EGUs. Maps could be
used for visual presentations to show the general location of emissions.
Why do some EGUs show positive increases in generation with decreases in system load?
Some EGUs show positive increases in generation with decreases in load because the EGU
statistics indicate either a slight increase in the probability of operation at very low loads or an
increase in generation at very low loads. Spot checks indicate that most of the EGUs that show
generation increases with decreases in system load are due to baseload EGUs that show a lower
probability of generation at mid-range loads than at either very high or very low loads. In other
words, these EGUs counterintuitively increase the probability of operation as system load levels
become very low. Further inquiry into these EGUs suggests that they have prolonged maintenance
outages during spring or autumn—i.e., during periods of generally low load, but possibly not the
lowest in the year. Therefore, the EGU will register as non-operational through a wide swath of
medium-low loads, but may operate during the very lowest loads of the year. Therefore, the
statistics capture this behavior and increase expected generation by a small margin when system
load is reduced from very low load periods. This pattern is almost always observed in trough
periods.
104 These loss factors include wake losses, electric losses, availability losses, and other loss categories, as defined
by NREL in National Renewable Energy Laboratory. 2016. 2016 Offshore Wind Energy Resource Assessment
for the United States. Section 7.3.1, Figure 9. https://www.nrel.gov/docs/fv16osti/66599.pdf.
SEPA
State and Local Climate
and Energy Program
-------
AVERT User Manual Version 4.2
What kinds of emission rates does AVERT produce?
AVERT's Annual Regional Results table (see Figure 21) displays information on modeled
emissions, generation, and emission rates. Two types of emission rates are shown on the lower
box of this page: "Average Fossil" and "Marginal Fossil."
The first column labeled "Average Fossil" is an emission rate calculated by dividing the mass (tons
or pound) of each pollutant in the baseline ("Original") by the level of generation (MWh) in the
baseline ("Original"). These values are derived from the power plants in the AVERT dataset prior to
any user-defined change. This is an annual average emission rate of those EGUs in EPA's Power
Sector Emissions database.105 It is specifically a "fossil" rate because it does not include generation
or emissions from other power plants, including nuclear, hydro, wind, solar, or other plant types.
See U.S. EPA's eGRID tool for information on average emission rates that are inclusive of these
other EGU types.106
The second column, labeled "Marginal Fossil," is calculated by dividing AVERT's estimated change
in emissions by AVERT's estimated change in generation. This predicted change in emissions and
generation is the impact on the power sector as calculated by AVERT due to the user-defined
scenario. This value is called a marginal rate because it describes a change in emissions per unit
change in generation. As AVERT only models fossil-fired EGUs in the Power Sector Emissions
database—not other types of power plants—we refer to this as a "Marginal Fossil" rate.
Although the Annual Regional Results focus on annual emission rates, it is also possible to
calculate more temporally detailed "Average Fossil" and "Marginal Fossil" emission rates using
data in the advanced outputs (see page 54). AVERT reports both original and changes in
generation and emissions for every hour. Users can divide an emission quantity by the
corresponding hourly generation and calculate either "Average Fossil" and "Marginal Fossil"
emission rates for a single hour. Users can also use these data to create weighted average values
for weeks, months, or other time periods.
Does AVERT account for lifecycle emissions?
No. At this time, AVERT users are only able to analyze emissions related to direct combustion
(EGU and ICE vehicle) and select emissions relating to fueling and the volatilization of fuel in ICE
vehicles. Users interested in exploring lifecycle emissions may wish to utilize the Greenhouse
gases, Regulated Emissions, and Energy use in Technologies (GREET) Model developed by
Argonne National Laboratory.107
Can AVERT results be used for spatial analyses?
AVERT contains two types of data with geospatial attributes. The first type of data is point source
(latitude and longitude) emission data from EGUs. The annual-aggregated EGU emission data with
latitude/longitude attributes can be accessed by viewing the tab titled "Summary" in the Excel file.
AVERT also presents this annual EGU data aggregated at the county-, state-, and regional-level in
some outputs. (Note that AVERT users will have to click the button labeled "Click here to restore
default Excel data" on the welcome page of the model to access this "Summary" page.) Users can
105 The AVERT Average Fossil rate will be nearly, but not exactly, equal to an average emission rate derived directly
from EPA's Power Sector Emissions database.
106 See https://www.epa.gov/earid.
107 See https://qreet.es.anl.gov/for more information on GREET.
SEPA
State and Local Climate
and Energy Program
-------
AVERT User Manual Version 4.2
utilize the latitude and longitude locations to visualize the annual data displayed on this page, or
apply this latitude and longitude data to hourly, EGU-specific data for all six pollutants, which can
be accessed by viewing the tabs titled "SO2 (lb)," "NOx (lb)," "CO2 (short tons)," "PM2 5 (lb)," "VOCs
(lb)," and "NH3 (lb)."
The second type of geospatial data is avoided ICE vehicle emissions aggregated at the county-
level (AVERT also presents these data at the state and regional level). These data are a
generalization of the emissions not occurring on roadways in counties where EVs are being
deployed according to the user's scenario. These data are most easily accessible for geospatial
analysis (via FIPS codes) through the page titled "Output: Annual Results by County, Including
Vehicles."
Users may be interested in utilizing these spatial data to understand how policies and programs
modeled in AVERT affect overburdened communities or other communities of interest. To support
these equity analyses, users could employ a geographic information system (such as ArcMap,
QGIS, Google Earth, or another program) to analyze and visualize AVERT data alongside
EJScreen108 or another spatial dataset of interest.
Can AVERT results be used for mobile source regulatory analyses?
No. AVERT may not be used for mobile source regulatory analysis, including SIP and
transportation conformity analyses. Consult the most recent EPA guidance document for applying
EPA's MOtor Vehicle Emission Simulator (MOVES) model at: https://www.epa.gov/moves/latest-
version-motor-vehicle-emission-simulator-moves.
AVERT Statistical Module
Why is the model driven by system fossil generation instead of by demand or total load?
AVERT is a statistically based model that tracks and reproduces EGU behaviors. EGUs are
forecast to operate in the near future much as they operate today. In electrical system dispatch,
determining how an EGU operates is largely driven by two factors—total demand on the system
and the cost of operation for any given EGU. In economic dispatch, more expensive EGUs are
dispatched at higher levels of demand.
In AVERT, the degree to which an EGU will be dispatched in the future is assumed to be the same
as the degree to which it has been dispatched at a historical level of demand .This assumption
incorporates the relative operational cost of different EGUs. In other words, EGUs that dispatched
at high demand were likely more expensive to operate and thus are likely to be on the margin at
the same level of demand.
However, if AVERT were to observe the behavior of EGUs against total system demand, it would
miss an important factor: the large number of non-fossil EGUs that are dispatched at very low
operating costs—such as solar, wind, hydroelectric, and nuclear operations. These non-fossil EGU
have the same effect as lowering system demand, or specifically, system demand that needs to be
met by fossil generation. Therefore, AVERT dispatches against demand for fossil resources, rather
than total system demand. New EE or RE policies are assumed to reduce the demand for fossil
resources. Figure 55 below illustrates the difference between total system demand and demand for
fossil resources.
108 See https://www.epa.gov/eiscreen.
State and Local Climate
and Energy Program
-------
AVERT User Manual Version 4.2
Figure 55. Diagram schematic of system demand over two days, divided into
fossil and non-fossil components illustrating system and fossil demand.
E
OJ
~
E
OJ
ts
Does the sum of all unit generation in any given Monte Carlo run add up to the size of the
load bin (i.e., the expected fossil generation)?
Not necessarily. The generation from each EGU is calculated independently in each Monte Carlo
run, meaning that there is no constraint that forces the output of all EGUs to equal the exact size of
the load bin. The total sum of all unit generation from any given Monte Carlo run may be slightly
larger or smaller than the load bin. However, over large numbers of Monte Carlo runs, the average
output of each EGU will sum quite closely to the expected fossil generation, or the size of the load
bin.
Is it possible to displace baseload EGUs in AVERT?
Yes. AVERT treats baseload EGUs, or units that run during most hours, including during baseload
hours of the year, the same as all other EGUs. There is no distinguishing characteristic that either
promotes or prevents an EGU from running during any given hour, except for how it has operated
in the past. If an EGU has experienced little downtime in the past and operates continuously even a
low levels of load, the model will replicate this behavior going forward. This type of EGU is unlikely
to be displaced by EE or RE programs in AVERT. However, an EGU that ramps from high output in
the daytime to low output on off-peak hours may show a displacement if system demand declines
due to RE or EE.
Can AVERT capture or replicate ramping behavior?
No. AVERT performs a separate calculation for each hour of the year and does not evaluate the
rate at which an EGU increases or decreases generation. Capturing that behavior requires a
chronological dispatch model.
Can AVERT capture or replicate spinning reserves behavior?
Yes. AVERT captures historical generation patterns. EGUs that maintain a spinning reserve (i.e.,
maintain a minimum level of generation in most operating hours) will reflect this pattern in the
statistics gathered by AVERT. The Lake Hubbard EGU shown in Figure 42. is a classic example of
an EGU that appears to maintain a spinning reserve. It maintains an output of about 150 MWper
hour for most hours of the year. However, when system demand climbs above 45,000 MW, it
quickly climbs towards an output of 500 MW.
101
SEPA
State and Local Climate
and Energy Program
-------
AVERT User Manual Version 4.2
Can AVERT capture transmission constraints or changes in transmission?
Generally, no. AVERT operates on the simplifying assumptions that there are no transmission
constraints between load centers and EGUs within a region and that regions are independent of
each other. Therefore, AVERT is insensitive to the location where new EE or RE resources are
placed within a region, and thus does not capture transmission constraints. However, the behavior
of some EGUs may be influenced by historical transmission constraints, and this behavior is
captured by AVERT. For example, in "load pockets," or areas of constrained inbound transmission,
reliability EGUs may run at lower regional load levels than would otherwise be dictated by
economic dispatch. Because AVERT is not an economic model, it simply replicates the behavior of
these EGUs, which may capture some elements of current transmission constraints.
Due to the same simplifying assumptions that prevent AVERT from operating as a transmission-
constrained dispatch model, AVERT cannot capture future changes in transmission, which typically
change which future EGUs can compete to provide the lowest-cost energy in a particular area.
Are recent fuel prices reflected in AVERT?
Yes. To the extent that fuel prices have influenced dispatch during the base data year you choose,
AVERT will reflect those dispatch decisions. AVERT cannot, however, change dispatch based on
future economic or regulatory conditions, such as expected fuel prices, emissions prices, or
specific emissions limits. AVERT should not be used for this type of analysis, as such changes
require an economic dispatch model.
Are emissions control technologies reflected in AVERT?
Yes. To the extent that emissions controls were in operation at the time that data were collected in
the base data year, emissions will reflect operational (and operating) control technologies. To the
extent that a user requires a review of dispatch with different emission rates, they can override
observed emission rates using the "Future Year Scenario Template" as described in Appendix F.
Modeling emissions prices, specific emissions limits, or fuel switching requires an economic
dispatch model.
Are predicted changes in fuel or emissions prices reflected in AVERT?
No. AVERT should not be used for this type of analysis; capturing this behavior requires an
economic dispatch model.
What other tools are available to me to estimate changes in emissions aside from AVERT?
You can model generation and emissions changes caused by new energy policies in a production-
cost dispatch model. These programs simulate real dispatch decisions based on explicit costs and
operational constraints, optimizing generator use to minimize costs. Some of these models are
sensitive to EGU ramp-rates, transmission constraints, and outage schedules.
For the most part, these models are highly detailed, proprietary, and require specialized labor and
licensure to operate and use, as well as some degree of proprietary knowledge for fuel costs and
operational constraints.
AVERT provides an alternative, publicly available tool to estimate changes in emissions in near-
term years. Users who wish to conduct analyses more than 5 years from the baseline must use
AVERT's statistical module and future year scenario template. This type of analysis requires
access to future year hourly, unit-specific generation and emissions data (e.g., from an electric-
State and Local Climate
and Energy Program
-------
AVERT User Manual Version 4.2
sector dispatch model designed to forecast future generation) that can be entered in place of
AVERT's historical data.
Future Year Scenarios
Why is there a different future year template for each historical baseline year?
AVERT is sensitive to the composition of the electric fossil fuel fleet. Every year, the composition of
the fleet changes slightly as new EGUs are added or retired. To accommodate this changing fleet,
AVERT creates a new future year scenario template for each historical baseline year. Using a
mismatched pair in AVERT's Statistical Module (e.g., a historical baseline year of 2017 but a future
year template of 2019) risks accidentally using proxy "new" EGUs that did not exist in 2017, and
thus will not be incorporated into a 2017 analysis.
Why are some generators excluded from AVERT's Future Year Scenario Template?
AVERT considers EGUs that report to EPA's CAMD only. This may exclude generators with less
than 25 MW of capacity or generators that did not operate in a particular year.
Why do some generators appear in the Future Year Scenario Template but not in the RDF?
AVERT's Statistical Module allows users to exclude small, low-generation units from consideration
in the emissions analysis. Small peakers have statistics that may be non-representative of
expected generation patterns (i.e., they cannot be readily extrapolated or interpreted outside of
specific events). By default, AVERT excludes units that have generated less than 1,000 MWh per
year. For a 25 MW unit (the smallest reporting unit), this would be the equivalent of 40 hours of
generation over the year, or less than 0.5 percent of all possible operational hours.
In the future scenario demo, do the total avoided emissions include the impact of retired
units, or just the energy impacts adjusted for retirements?
Results from AVERT runs using the Future Year Scenario Template do not include changes in
emissions at user-specified retired units. These units are assumed to be retired in both the "before"
and "after" cases.
The purpose of the retirements category is to exclude from consideration any units that are likely to
be non-operational in the future year, regardless of the energy change modeled.
Does EPA provide projections for use in AVERT's Future Year Scenarios?
At this time, EPA is not providing EPA projections for use in AVERT. Future scenarios are meant to
be developed by users. You can use AVERT's future year scenario template to make known
changes in the regional dataset. Users who wish to conduct analyses more than 5 years from the
baseline must use AVERT's statistical module and future year scenario template. This type of
analysis requires access to future year hourly, unit-specific generation and emissions data (e.g.,
from an electric-sector dispatch model designed to forecast future generation) that can be entered
in place of AVERT's historical data.
Along these lines, EPA has partnered with the Eastern Regional Technical Advisory Committee
(ERTAC), a group of state environmental agency senior staff and multi-jurisdictional organizations
(e.g., LADCO, MARAMA, WESTSTAR, SESARM, NESCAUM), to provide AVERT-compatible
RDFs for ERTAC-specified custom future years.
State and Local Climate
and Energy Program
-------
AVERT User Manual Version 4.2
EPA will issue periodic updates to the historical data files available for download, but will not
release stand-alone future scenarios. At this time, EPA anticipates releasing new RDFs in the
second quarter of each year.
Electric Vehicles
If a new electric car is expected to be charged for the next 12 to 15 years, how should we
consider the results generated by AVERT?
As described in the above FAQ ("Does EPA provide projections for use in AVERT's Future Year
Scenarios?"), AVERT is best used to conduct short-term analyses. In practice, we recommend that
users restrict their analyses to a five-year time horizon from the year of the RDF. Analyses that
extend past this window may intersect with years when the operation and composition of the grid is
substantially different from the baseline year and may not include the structural change that may
have been caused by the intervention modeled.
The medium- to long-term impact of EVs on the grid is also of interest to many users as the power
sector's emission rates for many pollutants are expected to decline overtime. While AVERT is not
intended to answer these questions, EPA has built a reference output to help put the AVERT
results in context: the output page titled "Reference: Modeled Emission Rates Over Time." On this
page, users can compare the CO2 emission rate as modeled in AVERT with a set of emission rates
modeled by NREL in its Cambium data set (for more information on this feature, see page 52). CO2
SRMER and LRMER are provided for different approaches and different years, allowing users to
estimate how grid impacts of charging vehicles are likely to change in the future.
Which power plants does AVERT assume charge EVs?
AVERT sums the load change from all entered resources (EERE and EVs) in each hour and
applies the net change to the fossil power plants in the selected grid region (as defined by the
RDF). When modeling EV adoption in the near future, users are recommended to model the
amount of EE and RE resources that are estimated to come online during the same period to
understand the joint effect of both changes together. For example, if a user is modeling the number
of EVs added to the grid over a three-year period, they should also model the amount of EE and
RE expected to be added over the same three years.
Should I use AVERT's definition of "new" vehicle or "existing" vehicle?
Wth AVERT, users can select an emissions profile for a new or existing ICE vehicle. In many
situations, users are likely to want to select "new" for this vehicle type. This allows a user to
compare the impact of some number of new EVs relative to the same number of new ICE vehicles.
In some cases, users may wish to compare the impacts of replacing an existing ICE vehicle with a
new EV. In these situations, users should select "existing" vehicle. In both options, the grid impacts
associated with charging EVs will remain the same; changing the option between "new" and
"existing" only modifies the emission rate associated with ICE vehicle emissions.
What other sources for charging profiles are there? How would I use them in AVERT?
One potential source for EV charging profiles is the Electric Vehicle Infrastructure Projection Tool
(EVI-Pro) Lite.109 EVI Pro-Lite is the source for AVERT's "light-duty vehicle" charging profile, but
109 U.S. Department of Energy. Electric Vehicle Infrastructure Projection Tool (EVI-Pro Lite). Available at:
https://afdc.enerqv.qov/evi-pro-lite/load-profile.
State and Local Climate
and Energy Program
-------
AVERT User Manual Version 4.2
users can utilize EVI-Pro Lite to develop their own custom charging profiles. Charging profiles
derived from this tool may be a better representation than the defaults available in AVERT, as EVI-
Pro Lite allows users to modify many different options currently not available in AVERT. Users
should reference EVI-Pro Lite materials at https://afdc.enerqv.qov/evi-pro-lite/load-profile to learn
how to use the tool and generate load profiles.
Users should note that they can export the results from EVI-Pro Lite into a CSV file. Some post-
processing may be needed before importing the results into AVERT. As of January 2023, the load
profiles will produce results in 15-minute intervals, which must be averaged to produce hourly
values usable in AVERT.
Can the vehicle emissions be exported to COBRA?
Yes. When a user generates a COBRA text file, emission changes will be generated for the power
sector as well as the vehicle emissions from the transportation sector (see page 53 for more
information). This text file will contain a row of vehicle emission changes for every county in the
region currently selected for modeling. These county-specific changes are developed by allocating
the total regional emissions changes for each pollutant to each county based on the share of VMT
in that county relative to the regional total. When COBRA reads the AVERT-generated text file, the
emissions will be automatically classified to the appropriate sector (i.e., electricity generation or
"transportation," as the sector is called in COBRA). Because vehicle emissions are emitted close to
the ground (unlike pollutants emitted from EGUs, which are emitted from a stack high in the air),
these pollutants tend to not travel far distances and, as a result, tend to be more impactful to local
communities.
105
SEPA
State and Local Climate
and Energy Program
-------
JtVert
AVoid«d Emissions and q»rwRatlon Tool^^
wwwepagovTavSt'^W
AVERT User Manual Version 4.2
Appendix I: Web-Based AVERT
In 2018, EPA released a web-based version of the AVERT Main Module. The Web Edition provides
a streamlined interface with much of the same functionality as the downloadable Excel-based Main
Module, without the need to use Excel software or upload separate RDFs. The Web Edition relies
on the most recent year of input data and has a more limited range of input and output options. It
uses the same underlying methods, calculation algorithms, and regional data inputs, and it reflects
the same assumptions as the Excel-based Main Module.
Differences Between the Web Edition and the Excel Main Module
The web-based Main Module has the following limitations:
• The Web Edition relies on a single year of input data, whereas the Excel versions can
incorporate data from any year that is posted on the AVERT website.
• The Web Edition does not allow the user to manually input a custom load profile with 8,760
hourly values, model increases in generation or load, or scale RE capacity factors, like the
Excel version does.
• For EVs, the Web Edition uses default values for vehicle composition, the
weekend:weekday ratio (97 percent), and the share of ICE miles replaced by an EV (100
percent). It does not allow the user to input a custom charging profile.
• The Web Edition uses built-in default RDFs, so it does not support future year scenario
planning.
• The Web Edition provides a more streamlined range of result formats than the Excel
version, consisting of two data tables, three graphs, downloadable CSV data, and a
COBRA CSV file.
However, the Web Edition also offers several advantages over the Excel-based Main Module:
• The Web Edition provides some accessibility advantages because it can run in any major
web browser, without the need for Excel software or the need to download, save, and
upload separate RDFs.
• The graphs in the Web Edition have some dynamic capabilities that allow the user to
customize the geographic area displayed and save a variety of formats (e.g., JPG, PDF) to
display in presentations or reports.
• The Web Edition can run an analysis for a single state, in addition to running an analysis
for an AVERT region (see below for more detail). The Excel version analyzes one region at
a time, thus requiring manual post-processing for aggregation if users want to model state-
level energy changes in states that cross regional boundaries (see Appendix G for more
information on these analyses).
• The Web Edition provides a direct connection to import data into the web version of EPA's
COBRA model.
State and Local Climate
and Energy Program
-------
AVERT User Manual Version 4.2
Ultimately, some users will find that the Web Edition meets their needs, while others who wish to
use different data years, custom load profiles, custom EV parameters, future RDFs, or additional
result formats should use the Excel version.
Web AVERT State Analysis
For states that span more than one region, the Web Edition allocates energy changes across
regions and performs the multiple regional analyses simultaneously. This section describes how
the Web Edition performs analyses for states that cross regional boundaries. The methodology for
analyses in states that are entirely in one AVERT region is essentially the same as the
methodology for analyses of a region.
For states that span more than one region, AVERT apportions the energy changes from the
applied scenario proportionately to the two or more regions. For all input options except a
percentage reduction (EE by percent), the user-input amounts of EE, RE, and/or EVs are prorated
to the relevant AVERT regions based on the state's electricity sales in each region. Table 6 shows
state apportionment in AVERT regions based on electricity sales in 2018. The Web Edition then
performs one or more regional analyses as normal using these prorated inputs.
For percentage generation reduction ("Percentage reductions in some or all hours"), AVERT uses
2018 state and regional sales data to calculate the proportion of each region's sales that originated
in the selected state. The user-specified percentage is scaled by this amount. This means that a
percentage reduction in a state that crosses regional boundaries typically corresponds to a smaller
percentage reduction when considered at a regional level. For example, Alabama is in both the
Southeast and Tennessee AVERT regions. Alabama's 2018 fossil sales constituted 66.6 terawatt-
hours (TWh) of the 228.1 TWh of fossil sales in the Southeast region. A 10 percent reduction in
electricity use in Alabama therefore corresponds to a 6.66 TWh reduction in the Southeast region,
which represents a 2.9 percent reduction in total Southeast region sales. AVERT will therefore
model 2.9 percent for the Southeast.
Wthin "Percentage reductions in some or all hours," the Web Edition applies a simplifying
assumption when modeling a targeted program (i.e., percentage reduction during a peak
percentage of hours). To streamline the analysis, AVERT applies the reduction to the top X percent
of hours in each individual region. These may or may not be the exact same hours that constitute
the top X percent of hours in terms of state-specific loads.
Some states that cross regional boundaries are in at least one region that has capacity factors
available for offshore wind and at least one region that does not. In these situations, the entire
offshore wind capacity entered by the user is allocated to the region that supports offshore wind.
This is the case regardless of whether the state has coastline adjacent to potential offshore wind
sites or even has any coastline at all. The rationale for allowing landlocked states to model offshore
wind is that such states may still invest in offshore wind capacity elsewhere, and their electric
regions may still import electricity from offshore wind generation.
Input Validation
To ensure the user's inputs are realistic and within the calculable range, web AVERT implements
two rounds of input validation:
SEPA
State and Local Climate
and Energy Program
-------
AVERT User Manual Version 4.2
• First pass: Check each individual input to make sure it is a valid positive number. Users
who want to model reverse EERE scenarios (negative inputs) should use the Excel Main
Module.
• Second pass: Check the aggregate load shape after all inputs have been combined to
ensure that it is within AVERT's recommended range. There are four possible outcomes:
o If any individual hour has a load reduction that amounts to 15 percent or more
of that hour's regional generation, AVERT gives a warning but allows the user
to proceed.
o If any individual hour has a load reduction that amounts to 30 percent or more
of that hour's regional generation, AVERT gives an error message and
prevents the user from proceeding until they revise their inputs to get the
resulting load shape below the 30 percent threshold,
o If any individual hour has an increase in load that would be too large for
AVERT to model, AVERT gives an error message and prevents the user from
proceeding until they revise their inputs to create an aggregate load shape with
an increase that AVERT can model. This is analogous to the error message
that the Excel version provides. Each region's maximum calculable load varies
in percentage terms, but it tends to be on the order of a 15 percent increase,
o Otherwise, the user can proceed without any warning.
For state-specific AVERT runs that involve multiple AVERT regions, the program creates a
separate new energy impact profile for each affected AVERT region. These regional energy impact
profiles are aggregated to create a single graph onscreen, but the region-specific profiles are what
actually feed into the AVERT displacement calculations, which are performed by region. The
second pass validation step independently tests each new region-specific energy impact profile
against the corresponding regional hourly loads in the RDF. AVERT returns a warning or error
message, respectively, if any region in the analysis sees an exceedance of 15 percent or 30
percent. It also returns an error message for any exceedance of a region's maximum calculable
load.
108
SEPA
State and Local Climate
and Energy Program
-------
AVERT User Manual Version 4.2
Appendix J: Electric Vehicles in AVERT
Inputs and Assumptions
This appendix describes the user-modifiable inputs and background assumptions found in AVERT
related to modeling EVs. At the end of this appendix, there are also practical examples
demonstrating how analysts can use AVERT to answer questions about EV emission impacts. EVs
are motor vehicles that obtain some or all of their power supply from batteries, which are ultimately
charged by power plants on the electric grid. Analysts can use AVERT to estimate the emissions
impacts of EV deployment or EV policies and programs. Like EE and RE, EVs are treated as an
energy resource in AVERT, so analysts can include EERE and EV resources together in the same
scenario. Along with power sector impacts, AVERT estimates the emissions avoided when EVs
displace ICE vehicles.
AVERT includes default assumptions for several parameters to help users complete EV analyses
more easily. These assumptions are easy to edit if scenario-specific information is available.
Some users may choose only to interact with the Primary Inputs, described in Step 2: Set Energy
Scenario (see page 27). More advanced users may wish to modify the default settings in the
Detailed Inputs, described here. AVERT also utilizes a set of background assumptions located in
the "Library" tab of the Main Module that EPA generally advises against modifying. For information
on limitations related to modeling the emission impacts of EVs in AVERT, see Appendix K.
EV Detailed Inputs - Excel Main Module Only
In Step 2: Set Energy Scenario of the Main Module, users may click the "Enter detailed EV data"
button. This will bring users to a page called "EV Detailed Inputs." On this page, users may modify
the default settings of more advanced inputs. This page is separated into three different parts: Part
I. Charging Profiles, Part II. Vehicle Composition, and Part III: Model Year.
Part I: Charging Profiles
Part I is separated into several tables to assist the user in creating a 8,760-hour, user-defined
charging profile. In Table A, users can select from a list of pre-defined charging profiles.
AVERT provides three charging patterns to choose from:
• Light-duty vehicle: Charging from a composite of likely chargers, according to data in
NREL's EVI-Pro Lite tool.110
• Bus: Little charging throughout the day, with higher levels of charging overnight.111
• Manual: User-defined.
Each charging pattern is defined on a 24-hour basis, where each hour's load is defined by a
percentage of the day's total charging requirement. For example, if the hour ending in 8 had a
value of 20 percent, it would mean that 20 percent of charging occurs between 7:01AM and
110 Charging pattern data are based on results of EVI-Pro Lite for St. Louis, Missouri. Available at
https://afdc.enerqv.qov/evi-pro-lite/load-profile/assumptions. More information on how this model can be used for
developing charging patterns can be found in Appendix H. St. Louis was chosen as a "baseline" city due to its
geographical centrality in the United States and due to its relatively moderate climate.
111 This charging profile has been created by EPA.
SEPA
State and Local Climate
and Energy Program
-------
AVERT User Manual Version 4.2
8:00AM.112 Separate charging patterns are provided for weekdays and weekends; AVERT applies
the correct charging pattern to each day of the week automatically. See Figure 56 and Figure 57 for
a comparison of weekday and weekend charging patterns for the two default options.
Figure 56. Default "light-duty vehicle" weekday and weekend charging patterns found in AVERT.
Figure 57. Default "bus" weekday and weekend charging patterns found in AVERT.
Users also have the option to input their own hourly charging profile for weekdays and weekends.
The sum of all 24 hourly values must equal 100 percent.
In Table C, users specify the ratio of weekend to weekday energy consumption. On weekends,
EVs are assumed to drive less and therefore consume less electricity.113
AVERT automatically estimates the total annual and monthly charging that will need to take place
given the number of vehicles specified (see "Calculations" section below). It will then spread these
total quantities over each hour based on the charging percentage information provided in this step.
112 For information on how charging requirements are calculated, see the "Calculations" section on page 114.
113 Results from EVI-Pro Lite for St. Louis, Missouri, show that weekends use 97 percent of the energy that is
consumed by EVs on weekdays.
^ &EPA
State and Local Climate
and Energy Program
-------
AVERT User Manual Version 4.2
Part II: Vehicle Composition
On the second part of this page, users can make further adjustments to the types of vehicles being
modeled. Here, users can define the share of BEVs and PHEVs that are cars or trucks. This
selection modifies the modeled EV efficiencies (trucks are generally larger and less efficient than
cars) and also modifies the emission rates modeled for the displaced ICE vehicles. Default data in
this section are based on recent LDV annual sales data from the Federal Reserve.114
For scenarios considering transit buses, users can also modify the fuel type used by displaced
fossil fuel-powered transit buses. Based on VMT data from the MOVES runs (described further in
Background Assumptions below), AVERT features default values for what share of transit buses
have conventionally been powered by gasoline, diesel, or compressed natural gas (CNG)
nationally. Unlike the BEV and PHEV adjustments, these bus adjustments do not modify the
efficiency of EV buses; they only modify the avoided emissions from ICE vehicles. For scenarios
considering school buses and LDVs, only a single fossil fuel type is available for each vehicle
class: diesel and gasoline, respectively.
Part III: Model Year and ICE Replacement
On the third part of this page, users select a model year to analyze. In the first selectable
parameter, EV model year, users can choose a year from 2023 to 2028. This parameter
determines two things:
• The modeled EVs' efficiencies. EVs are expected to become more efficient each year.115
• The ICE vehicles are subject to different emissions standards and thus, model year is an
important input in determining the impact of EVs on different pollutants.
The default value is 2023.
The second selectable parameter, ICE vehicle being displaced, affects the emission rates of fossil
fuel-powered vehicles only. There are two options:
• A selection of "New" will mean that the emission rates of displaced vehicles will be based
on new vehicles from the specified EV model year. This selection suggests that new EVs
will displace the same number of new fossil fuel-powered vehicles that would have
otherwise been added to the vehicle fleet.
• A selection of "Existing" will mean that emission rates of displaced vehicles will be based
on a weighted average of all vehicles that are on the road. This selection suggests that
new EVs will replace the average existing vehicle.
Users should carefully consider which setting makes the most sense for their scenario. For most
analyses, users will probably be best served by selecting "New," as it allows users to perform a
prospective analysis wherein some number of new EVs are purchased in lieu of some number of
new fossil fuel-powered vehicles. The "Existing" setting is likely most useful for users who are
interested in performing a comparison of their existing vehicle fleet with a future alternative
114 These data are available from the Federal Reserve website at
https://fred.stlouisfed.ora/araph/?id=LTOTALNSA.LTRUCKNSA#Q. Types of EVs sold may not necessarily match
the types of all LDVs sold, due to vehicle availability, customer choice, or some other reason. Users can update
these percentages as they see fit.
115 Data on EV efficiency overtime are drawn from National Renewable Energy Laboratory (NREL). 2017.
Electrification Futures Study, https://www.nrel.gov/docs/fv18osti/70485.pdf.
State and Local Climate
and Energy Program
-------
JtVert
AVoid«d Emissions and q»rwRatlon Tool^^
wwwepagovTavSt'^W
AVERT User Manual Version 4.2
featuring EVs. Users can easily toggle between these two vehicle parameters after a model run as
they do not influence power sector load. The default value is "New."
Users can also specify an ICE replacement rate. This parameter addresses the concept that some
drivers of EVs may not drive as many miles with their EV as they might with an ICE vehicle. For
example, if the EV has less range than a driver's ICE vehicle, they might continue to use an ICE
vehicle for longer trips. AVERT assumes a default rate of 100 percent, meaning that 100 percent of
miles driven by an ICE vehicle are replaced with an EV. Users can modify this parameter if they
have relevant data for their analyses.
AVERT automatically applies the associated emission rates for the selected state or region, vehicle
type, fuel type, modeled year, vintage, and replacement rate.
Background Assumptions
AVERT uses a set of background assumptions to facilitate the calculation of EV emission impacts.
MOVES Modeling
EPA used MOVES to generate emission factors for fossil fuel-powered vehicles.116 MOVES is an
emission modeling system developed by EPA that estimates emissions for mobile sources at the
national, county, and project level for criteria air pollutants, greenhouse gases, and air toxics.
MOVES serves as EPA's repository of vehicle emission factors, drawing from five decades' worth
of emission measurement on hundreds of thousands of vehicles. The model can combine these
emission factors with vehicle activity (e.g., VMT, vehicle starts) and fleet characteristics (e.g., age
distribution, speed distribution) to produce emissions estimates.
Within the AVERT context, EPA used MOVES to produce metrics related to VMT and total
emissions across the following variables:
• States (48 contiguous states plus Washington, D.C.)117
• Vehicle type (passenger car, passenger truck, transit bus, and school bus)
• Fuel type (varies by vehicle type)
• Vehicle model year (the year the vehicle was made: 2023-2028)
• Vehicle age (new or fleet average)
• Modeling year (the year in which the analysis was conducted: 2023-2028)
• Modeling month (the month in which the analysis was conducted)
Reported metrics for each data item include:
• VMT by vehicle type
116 All analysis was conducted using MOVES4.0.0 with model database version movesdb20230615, published by
EPA's Office of Transportation and Air Quality. MOVES and more information about it (including technical
documentation, policy guidance, and user tools) can be obtained from https://www.epa.gov/moves/latest-version-
motor-vehicle-emission-simulator-moves.
117 Because Alaska, Hawaii, and Puerto Rico are not currently supported in AVERT, vehicle emission rates for these
regions were not generated.
State and Local Climate
and Energy Program
-------
AVERT User Manual Version 4.2
• Total emissions for CO2, NOx, SO2, PM25,118 VOCs,119 NH3
For each data item and each pollutant, emissions are divided by VMT to calculate emission rates
measured in lb/mile.
MOVES4 was used to produce estimates of VMT by county.120 These data are used to allocate
displaced vehicle emissions to each county. See "Step 2: Set Energy Scenario" on page 27 for
more information about how these locations are specified, and see the Calculations section below
for how these county-level data are applied.
Other Modeling
AVERT incorporates a number of other assumptions to facilitate the calculations behind charging
impacts and displaced vehicle emissions. All of these assumptions are found on AVERT's Library
tab. EPA generally recommends that users not modify these assumptions.
• Typical VMT per year: AVERT assumes that nationwide, passenger cars and passenger
trucks travel about 11,543 miles in each year.121 Meanwhile, transit buses are assumed to
travel about 43,647 miles per year, while school buses are assumed to travel about 12,000
miles per year.122 Depending on the region selected, AVERT modifies these VMT values to
reflect the typical miles driven in specific states and regions.123
• EV efficiency: AVERT describes EV efficiency in terms of kWh per VMT (i.e., the number of
kWh required to travel one mile). AVERT assumes that EV efficiency improves each year
as technology improves. AVERT uses different efficiencies for different types of vehicles,
including cars, trucks, BEVs, PHEVs, transit buses, and school buses.124125
• Percentage of PHEV miles driven on electricity: PHEVs can be driven using electricity
stored in batteries or using a conventional fossil fuel-powered engine. AVERT assumes
118 MOVES reports three different types of PM2.5 emissions: those created from vehicle exhaust, brake wear, and tire
wear. While EVs still create PM2.5 emissions related to brake wear and tire wear, AVERT currently assumes that
brake wear and tire wear for ICE vehicles and EVs are the same, which would yield no net change for PM2.5
emissions. Thus, these two PM2.5 sources are not included in AVERT's displaced emissions calculations. This is
acknowledged to be a simplifying assumption in light of some evidence that EVs may have lower brake wear
PM2.5 emissions than ICE vehicles due to regenerative braking.
119 MOVES reports three different types of VOC emissions: those created from vehicle exhaust, evaporation, and
refueling. EVs are able to avoid all three types of VOC emissions. As a result, the VOC emission rate in AVERT
is the sum of all three emission types.
120 VMT comes from MOVES4.0.0 with model database version movesdb20230615, published by EPA's Office of
Transportation and Air Quality. MOVES and more information about it (including technical documentation, policy
guidance, and user tools) can be obtained from https://www.epa.gov/moves/latest-version-motor-vehicle-
emission-simulator-moves.
121 Federal Highway Administration (FHWA). Highway Statistics 2021. February 2023. Table VM-1.
https://www.fhwa.dot.qov/policvinformation/statistics/2021/vm1.cfm.
U.S. Department of Energy Alternative Fuels Data Center (DOE AFDC). Average Annual Vehicle Miles Traveled
by Major Vehicle Category. Updated February 2020. https://afdc.enerqv.gov/data/10309.
Federal Highway Administration (FHWA). Highway Statistics 2021. February 2023. Tables VM-1, VM-2, VM-4,
MV-1, and MV-10. https://www.fhwa.dot.gov/policvinformation/statistics/2021/.
124 Islam, E.S., R. Vijayagopal, and A. Rousseau. 2022. A Comprehensive Simulation Study to Evaluate Future
Vehicle Energy and Cost Reduction Potential, Report to the U.S. Department of Energy, Contract ANL/ESD-22/6.
https://vms.taps.anl.qov/research-hiqhliqhts/u-s-doe-vto-hfto-r-d-benefits/.
125 Islam, E.S., R. Vijayagopal, A. Moawad, N. Kim, B. Dupont, D.N. Prada, and A. Rousseau. 2021. A Detailed
Vehicle Modeling and Simulation Study Quantifying Energy Consumption and Cost Reduction of Advanced
Vehicle Technologies Through 2050. Report to the U.S. Department of Energy, Contract ANL/ESD-21/10.
https://vms.taps.anl.qov/research-hiqhliqhts/u-s-doe-vto-hfto-r-d-benefits/.
State and Local Climate
and Energy Program
122
123
-------
AVERT User Manual Version 4.2
that 54 percent of PHEV miles are driven using electricity.126 Different vehicles with
different specifications related to range and utilization may see a different share of miles
driven on electricity and gasoline.
• Climate adjustments: AVERT includes a single annual adjustment factor (from 0 to 8
percent) to account for average regional temperature differences relating to charging
requirements. EVs driven in regions that are substantially warmer or colder than AVERT's
moderate-temperature baseline are expected to require more electricity.127 Using climate
data from EVI-Pro Lite, AVERT assigns each region a climate adjustment factor that
increases the electricity required to drive a single mile (see Table 8).128 The MOVES
monthly emission rates reflect temperature impacts on ICE vehicle emission rates as well.
Table 8. Climate adjustments to charging requirements in AVERT regions.
AVERT region
Local average
temperature (°F)
Climate adjustment to
charging requirements
California
68
100%
Carolinas
68
100%
Central
50/68
104%
Florida
68/86
104%
Mid-Atlantic
50
108%
Midwest
50/68
104%
New England
50
108%
New York
50
108%
Northwest
50/68
104%
Rocky Mountains
50
108%
Southeast
68
100%
Southwest
68
100%
Tennessee
68
100%
Texas
68
100%
Calculations
This section describes how AVERT combines the above inputs and assumptions to estimate
emission impacts in the power sector and from vehicles.
126 Plotz, P., M. Cornelius, Y. Li, G. Bieker, and P. Mock. 2020. Real-world Usage of Plug-in Hybrid Electric
Vehicles: Fuel Consumption, Electric Driving and CO2 Emissions. The International Council on Clean
Transportation, https://theicct.ora/publications/phev-real-world-usaae-sept2020.
127 EVs being driven in hot or cold regions may require more electricity due to increased HVAC use, battery
inefficiencies, and charging inefficiencies.
128 Given that St. Louis, Missouri, is AVERT's baseline city for establishing an EV charging profile, it is defined as
AVERT's "baseline" climate. St. Louis, Missouri, has an average temperature of 68°F, which is typical of many
cities across the United States. Per EVI-Pro Lite, an EV driving in a city that is 18°F warmer than a city with the
68°F baseline (like St. Louis) uses about the same additional amount of electricity as an EV driving in a city that
is 18°F colder. Some regions contain a mix of cities with different temperatures. (At this time, AVERT does not
apply a climate adjustment factor that reflects how climate in any one region changes month to month.) Each
region receives a climate adjustment as appropriate relative to this baseline. See Table 9 of the AVERT Main
Module's "Library" tab for region-specific adjustment factors and additional explanation.
State and Local Climate
and Energy Program
-------
JtVert
AVoid«d Emissions and q»rwRatlon Tool^^
wwwepagovTavSt'^W
AVERT User Manual Version 4.2
Power Sector
Power sector impacts (measured in kWh) are first calculated on a monthly basis for each type of
vehicle. Monthly impacts are then converted into hourly kWh impacts. AVERT combines these kWh
impacts with other energy resources in the scenario and then calculates the generation and
emission impacts from specific power plants.
Monthly kWh impact values are calculated as follows:
A=BxCxDxExF
Where
A is monthly power sector impacts. These are the resultant kWh changes in
demand due to EV charging in a given month.
B is the number of vehicles input by the user.
C is monthly VMT. This number is estimated by multiplying the typical VMT per
year for that region by MOVES' estimate of the percentage of annual miles
traveled in a month.129
D is vehicle efficiency—the amount of electricity required to travel one mile
(measured in kWh per mile traveled), adjusted for a region's climate.
E is the fraction of miles driven on electricity. This fraction is 100 percent for BEVs
and 54 percent for PHEVs.
F is the fraction of miles replaced with an EV. In AVERT, the default is 100
percent.
Next, these monthly kWh impacts are converted into daily impacts by calculating the number of
weekday days and weekend days in each month. These data are combined with the ratio of
charging that occurs on weekdays versus weekends to estimate the total kWh consumed for any
one weekday and any one weekend day in each month.
Third, daily kWh impacts are converted to hourly impacts by allocating the daily kWh across a 24-
hour period using the charging pattern specified by the user. This hourly kWh profile is then
repeated for each weekday of the month, with a separate kWh profile applied to each weekend.
Different profiles are repeated for each of the 12 months.
Finally, hourly impacts for all vehicle types are summed and combined with the other energy
resources in the scenario (e.g., EE, RE). As a result, when users model the effect of EV
deployments along with EE or RE measures, they may find that the aggregate impact on the power
sector is an energy decrease in some or all hours (see Figure 58). The resultant profile is divided
by one minus the T&D loss factor to estimate the total vehicle impacts on wholesale power sector
generation.
129 According to the data from MOVES, VMT varies by month, with summer months having more miles traveled than
winter months. Per the MOVES data, these monthly shares do not change by state or by vehicle type.
State and Local Climate
and Energy Program
-------
AVERT User Manual Version 4.2
Figure 58. Illustrative example of the combined impacts on marginal generation from EV charging and
solar resources.
600
Hour Ending
Vehicles
Emission impacts from vehicles are calculated on a monthly basis for each type of vehicle. For the
purposes of AVERT, "emission impacts from vehicles" refer to the impacts associated with
emissions from vehicle tailpipes and other emissions closely related to the driving and fueling of
vehicles. Unlike power sector impacts, they are not calculated at a daily or hourly level.
Monthly vehicle impact values are calculated for as follows:
A=BxCxDxExF
Where
A is monthly vehicle impacts. These are the resultant emission changes (in lb) due
to avoided ICE vehicle usage (i.e., due to EV charging) in a given month.
B is the number of vehicles input by the user.
C is the monthly VMT. This number is estimated by multiplying the typical VMT per
year for that region by MOVES' estimate of the percentage of annual miles
traveled in a month.
D is the ICE vehicle emission rate. This is the emission rate derived from MOVES
for the vehicle and fuel type being evaluated. This value changes depending on
the state or region that has been selected.
E is the fraction of miles driven on electricity. This fraction is 100 percent for BEVs
and 54 percent for PHEVs.
F is the fraction of miles replaced with an EV. In AVERT, the default is 100
percent.
AVERT repeats the above calculation for each vehicle and fuel type being modeled in a particular
run. AVERT then repeats the calculation for each pollutant.
State and Local Climate
and Energy Program
-------
AVERT User Manual Version 4.2
Each pollutant's change in emissions is then summed and applied to each county. Each county
receives a share of emission impacts commensurate with that county's VMT relative to the
aggregate VMT for the selected region. For example, if a user chooses to model emission impacts
over the "Entire Region," emission impacts will be distributed over the entire region, based on the
VMT in all counties in that region (see Figure 59). If a user chooses to model emission impacts
over just one state in the region, emission impacts will be distributed over the state's counties that
are within the selected AVERT region, based on the total VMT in all of the counties in that state in
the selected AVERT region. This process is applied separately for each vehicle type (as VMT are
distributed non-uniformly by type and county) and reported in AVERT's outputs.
Figure 59. Formula for estimating VMT for a county in a selected region.
, v-, County A VMTjupi
County A (Annual VMT) in Region Y=ZRegion Y VMTmoves ——
2-iRegion Y VMlNEI
How to: Analyzing Emissions Impacts of Electric Vehicles
The AVERT EV module can help answer questions such as the following:
• What is the net emissions impact of adding a certain number of EVs in one year and
displacing an equivalent number of ICE vehicles?
• What is the net emissions impact of adding EVs cumulatively over multiple years and
displacing an equivalent number of ICE vehicles?
• How much EERE offsets the generation requirement of a certain number of EVs?
• How do vehicle charging profiles affect emissions changes?
• Where, at the county-level, are emissions changing due to vehicle charging and avoided
ICE vehicle emissions?
• In which months do NOx emissions or other pollutant emissions vary?
• What are the ozone season implications of adding EVs?
• What are the emissions impacts of adding electric school buses or electric transit buses?
Other common EV questions are associated with AVERT, but require either data or tools external
to AVERT:
• What are the health and associated economic impacts of increases in vehicle
electrification?
• How can existing environmental justice tools (such as EJScreen) be used in coordination
with AVERT results?
When considering the following examples, users should bear in mind the caveats and limitations as
described in Appendix K. particularly those discussed in the EV sub-section.
State and Local Climate
and Energy Program
-------
AVERT User Manual Version 4.2
Example 1: What is the impact of deploying 39,000 new battery-powered electric
vehicles in 2022 in North Carolina?
a) What are the expected emissions changes from these BEVs?
b) In which months do NOx emissions vary?
c) How are emission changes distributed within each county?
Methodology
1) Open AVERT.
2) Load your region and baseline year of interest (in Example 1, Carolinas, 2021).
3) Choose the location of the EV deployment. For Example 1, choose North Carolina.
4) Enter the number of BEVs you would like to model. For Example 1, enter 39,000 BEVs (about
10 percent of annual vehicle sales in North Carolina in 2019). Use Table 1 in AVERT to assist
if necessary (shown below as this document's Table 9).
Table 9. AVERT Step 2: Set Energy Scenario, Table 1.
Table 1. Sales and
stock comparison
Percent of annual
vehicle sales in North
Carolina
Percent of registered
vehicles in North
Carolina
Light-duty vehicles
10.0%
0.5%
Transit buses
0.0%
0.0%
School buses
0.0%
0.0%
5) Enter the estimated amount of EERE expected to be added by 2022.
a) In this example, because you are starting with 2021 as a baseline year, add the amount of
EE, onshore wind capacity, and utility solar PV capacity that is expected to be added in
2022. You can use Table 2 in AVERT (shown below as this document's Table 10) to help
develop the EERE portion of a scenario. Table 2 reports the average annual addition of
EERE; it will adjust to the location of EV deployment.
i) EE: Reduce each hour by constant MW: 138 MW
ii) Onshore wind capacity: 0 MW
iii) Utility solar PV capacity: 531 MW
Table 10. AVERT Step 2: Set Energy Scenario, Table 2.
Table 2. EERE EV
comparison for
North Carolina
Historical additions
(Annualavg.2018-2020)
EERE required to offset
EV demand
EERE required divided by
historical additions
MW
GWh
MW
GWh
MW
GWh
EE (retail)
138
1,211
12
103
8%
8%
Onshore wind
0
0
-
-
-
-
Utility solar
531
1,084
42
85
8%
8%
Total
681
2,393
53
188
-
-
State and Local Climate
and Energy Program
-------
bJ^ERT
AVoidod Einissiortt and qenaRailori Tool^^
www.epa.gov/avarl
AVERT User Manual Version 4.2
6) Click "Next"; then, in Step 3, click "Click here to calculate changes to generation and
emissions."
7) In Step 4, analysts can select from a variety of pre-made summary tables, charts, and figures.
To answer question 1a ("What are the expected emissions changes from these BEVs?"),
click the green "Annual regional results—with vehicle" button. The resulting table is shown
below as this document's Table 11.
Table 11. Results for Example 1: Annua! regional results, including vehicles.
From fossil generation
From vehicle
Net changes
Total emission changes (lb)
SO2
-1,000,690
-2,550
-1,003,240
NOx
-1,350,980
-36,750
-1,387,730
CO2
-3,068,879,630
-384,206,390
-3,453,086,020
PM2.5
-200,570
-2,410
-202,980
VOCs
-83,920
-73,970
-157,890
NHs
-82,730
-21,970
-104,700
8) To answer question 1b ("In which months do NOx emissions vary?"), click the "Results by
month" button in Step 4 (see button circled in red in Figure 60). Note that these results show
both regional (AVERT Carolinas) power sector emissions changes and state-level vehicle
emissions changes.
Figure 60. How to view results by month, including avoided vehicle transportation emissions.
Step 4: Display Results
Summary tables - Power sector only
Annual regional results
Results for top ten
peak days
Daily NOx results by county
Summary tables, charts, and figures -
Power sector and avoided vehicle emissions data
Annual regional results - 1
with vehicle
Annual county results - 1
with vehicle
Results by selected 1
geography
Results by month
1
Emission rates over time
Charts and figures - Power sector only
Map of generation and 1
emissions changes 1
Hourly results by week 1
Monthly results by
selected geography 1
Signal-to-noise
diagnostic
]
COBRA text file generation
Enter a filepath, then click
the button to save a
COBRA text file.
Please be patient.
This calculation may take
up to twenty minutes to
run on older machines.
SMOKE text file generation
Enter a filepath, then click
the button to save
SMOKE text files.
Please be patient.
This calculation may take
up to twenty minutes to
run on older machines.
1. Regional Data
File
3. Run Scenario
4. Display
Results
Start new
scenario
State and Local Climate
and Energy Program
119
SEPA
-------
AVERT User Manual Version 4.2
9) To answer question 1c ("How are the emission changes distributed in each county?"),
click the "Annual county results—with vehicle" button in Step 4. Analysts can use the filtering
function to choose counties and pollutants of interest. For example, Table 12 shows the results
for Mecklenburg County, North Carolina.
Table 12. Results for Example 1: Annual results by county, including vehicles.
State
County
FIPS code
Pollutant
From fossil
generation
From vehicles
Net
changes
NC
Mecklenburg County
37119
S02 (lb)
0
-270
-270
NC
Mecklenburg County
37119
NOx (lb)
0
-3,930
-3,930
NC
Mecklenburg County
37119
CO2 (tons)
0
-20,210
-20,210
NC
Mecklenburg County
37119
PM2.5 (lb)
0
-260
-260
NC
Mecklenburg County
37119
VOCs (lb)
0
-7,910
-7,910
NC
Mecklenburg County
37119
NHs(lb)
0
-2,340
-2,340
Note that in this example, Mecklenburg County shows no emission changes from fossil
generation, but an amount of emission changes from vehicles. In every AVERT region there
are counties, like Mecklenburg County, that do not have any emitting EGUs (this means that in
these counties there are no EGUs per EPA's Power Sector Emissions Database). This means
that scenario results will not show increases or decreases in emissions from fossil generation
in those counties. However, all counties will receive some, if small, amount emissions benefits
from vehicles when EV scenarios are run. Users will observe that counties with more of the
region or state's VMT will see the largest benefits. In AVERT, changes to vehicle emissions are
first allocated for the selected state or region, and then allocated to each county, proportional
to VMT.130
Example 2: What if Florida were evaluating the impacts of a proposed policy that
increased the sales of electric vehicles by 5 percent each year from 2022 to 2024?
(EV sales are 5 percent of total vehicle sales in 2022, 10 percent in 2023, and 15 percent in 2024.)
a) What are the cumulative emissions changes in 2024?
b) What are the health impacts of Florida's EVs in 2024?
Methodology
1) Open AVERT.
2) Load your region and baseline year of interest (in Example 2, Florida, 2021).
130 Users can view the county-level VMT assumptions by clicking the green "Welcome" button on "Step 4: Display
Results," then clicking the grey button labeled "Click here to restore default Excel." Users can then navigate to
the tab titled "CountyFIPS." Columns G through J describe the county-level VMT for different vehicle types. Users
can filter the data using columns D or O (states and AVERT regions, respectively) to show only the counties that
are in their state or region. For our example, we see that Mecklenburg County has about 10 percent of North
Carolina's passenger car and truck VMT, which means that it has 10 percent of North Carolina's emission
changes.
State and Local Climate
and Energy Program
-------
AVERT User Manual Version 4.2
3) Develop three single-year scenarios following the methodology described in Example 1. (EPA
suggests creating three separate files and saving them individually.) Table 13 describes the
inputs for the three scenarios in this example.
Table 13. Inputs for Example 2.
Scenario
A
B
C
Modeled year
2022
2023
2024
EV sales percentages
5%
10%
15%
BEVs
58,000 (EVs deployed in
2022)
174,000 (58,000 EVs
deployed in 2022 plus
116,000 EVs deployed in
2023)
348,000 (174,000 EVs
deployed in 2022 and
2023 plus 174,000 EVs
deployed in 2024)
EE
34
68
102
Onshore wind
0
0
0
Utility PV solar
1,062
2,124
3,186
In Scenario C, in certain hours, the load change is greater than the 15 percent guideline suggested by A VERT. For this
example, these warnings can be ignored because hours exceeding this 15 percent guideline make up fewer than 2 percent
of hours, and no single displacement is larger than 21 percent of hourly load. Including these hours is unlikely to
substantially impact annual-level results.
4) Run the three single-year scenarios, saving each one as a separate file. Each file represents
the emissions changes in that year, relative to 2021.
5) To answer question 2a ("What are the cumulative emissions changes in 2024?"), click the
"Annual regional results—with vehicle" button and sum the "net changes" from each of the
three files.
6) To answer question 2b ("What are the health and associated economic impacts of Florida's
EVs in 2024?"), perform a series of steps using EPA's COBRA model.
a) First, during Step 4 in AVERT, generate a COBRA CSV file for each of the three AVERT
scenarios. Each one will be automatically named "COBRA.csv," so you may want to
rename the files to something more helpful in File Explorer or Finder (e.g.,
"COBRA_Flo rida_Sce narioAJ n puts. csv").
b) Download and install the COBRA model, if you have not already.
c) Open COBRA. Perform an analysis for each CSV file, using 2023 as a baseline year and a
3 percent discount rate.131
d) Export the COBRA results from each scenario. Open a separate document in Excel, import
the results of the three scenarios, and add the results of the three datasets together. Table
14 shows the results of each of the scenarios, as well as the combined impacts, with a
focus on total monetized health benefits (low). You can view many other monetized and
non-monetized benefits in the COBRA results.
131 For more information about COBRA and developing COBRA scenarios, see www.epa.gov/cobra.
State and Local Climate
and Energy Program
-------
JtVert
AVoid«d Emissions and q»rwRatlon Tool^^
wwwepagovTavSt'^W
AVERT User Manual Version 4.2
Table 14. COBRA results for Example 2.
Scenario
A
B
C
A, B, and C
Modeled year
2022
2023
2024
Three-year
cumulative benefits
Total monetized
health benefits, low
(2017 $ million)
$24
$46
$65
$136
Caveats
• Certain health impacts and monetized benefits reflect future benefits (beyond 2024) from
avoided deaths.
• Health impacts and monetized benefits reflect only reduced PM2 5. They do not include the
impact of ground-level ozone or a price on carbon.
Example 3: How might different charging profiles for transit buses affect emissions
from additional electric vehicles in New York?
a) How do total annual benefits change when a charging profile changes?
b) How do emissions change when transit buses of different fuel types are displaced?
Methodology
1) Open AVERT.
2) Load your region and baseline year of interest (in Example 3, New York, 2021).
3) Develop two scenarios following the methodology described in Example 1. (EPA suggests
creating two separate files and saving them individually.) Table 15 describes the inputs for the
two scenarios.
Table 15. Inputs for Example 3.
Scenario
A
B
Modeled year
2022
2022
Electric transit buses
600
600
Charging profile
Bus
Manual
EE
245
245
Onshore wind
53
53
Utility PV solar
142
142
4) Next, select a charging profile for your two scenarios. Click on the green button labeled "View
detailed EV data." Modify the parameters in Table A and Table B:
SEPA
State and Local Climate
and Energy Program
-------
JtVert
AVoid«d Emissions and q»rwRatlon Tool^^
wwwepagovTavSt'^W
AVERT User Manual Version 4.2
a) For Scenario A, select the "Bus" charging profile in Table A. This is the default bus
charging pattern, which models buses being charged primarily overnight.
b) For Scenario B, select "Manual" in Table A, and then enter a charging profile in Table B. In
this example, you can enter the charging profile described in Table 16, which describes a
situation where half of bus charging occurs overnight, and the other half in mid-day,
between the morning and evening rush hours.132 Copy and paste this same charging
profile for both weekdays and weekends (for transit buses, the default "bus" profile used is
identical for both weekdays and weekends).
Table 16. Manual charging inputs for Example 3B.
Hour ending
Percent charging
1
11%
2
8%
3
5%
4
3%
5
2%
6
1%
7
1%
8
1%
9
2%
10
3%
11
5%
12
8%
13
11%
14
8%
15
5%
16
3%
17
2%
18
1%
19
1%
20
1%
21
2%
22
3%
23
5%
24
8%
5) Run the two single-year scenarios, saving each one as a separate file.
6) To answer 3a ("How do total annual benefits change when a charging profile changes?"),
click the "Annual regional results—with vehicle" button and compare the "Net Changes" from
each of the two files. Error! Reference source not found, compares a subset of the results, f
ocused on NOx emissions. Between the two scenarios, Scenario B, which includes larger
amounts of daytime charging, produces 1,070 fewer short tons of power sector NOx emissions.
132 Users may also wish to consult other sources (e.g., NREL's EVI-Pro Lite) for more information regarding
charging profiles, particularly for scenarios relating to LDVs.
State and Local Climate
and Energy Program
-------
AVERT User Manual Version 4.2
Table 17. Outputs for Example 3, annual regional results, including vehicles, NOX emissions (lb).
Scenario
From
fossil generation
From
vehicles
Net changes
Scenario A
-972,350
-58,320
-1,030,670
Scenario B
-971,280
-58,320
-1,029,600
Scenario A less Scenario B
-1,070
0
-1,070
7) The vehicle emissions reductions results shown above in Table 17 are calculated assuming
that the vehicles that EV buses displace resemble the current fleet of transit buses in New
York—in this case, 79 percent diesel, 7 percent CNG, and 15 percent gasoline. Users may
also be interested in evaluating scenarios targeting different fuel types.
8) To answer 3b ("How do emissions change when transit buses of different fuel types are
displaced?"), users do not need to rerun AVERT, and can modify bus fuel types by returning
to the EV Detailed Inputs page by either navigating by tabs or returning to Step 2 and clicking
"View detailed EV data."
9) To evaluate a scenario where all displaced buses are diesel buses, users would first navigate
to the "View detailed EV data" page and then to "Part II. Vehicle Composition." Here, users
should set the percentage associated with diesel transit buses to 100 percent, and gasoline
and CNG to 0 percent. Users should review "Part III. Model and Year and ICE Replacement"
within the EV Detailed Inputs page to ensure vehicle replacement type aligns with the
intended model year and age. At this point, you may wish to save this file as "Scenario C".
a) To view results, navigate to the page titled, "Annual Regional Results, Including Vehicles"
either by clicking the tab named "10_Vehicle"133 or by clicking through Step 2 and Step 3 in
the Main Module.134
Table 18 shows that Scenario C, which assumes that EV buses only replace diesel buses,
displaces 11,290 more lb of NOx emitted from buses compared with Scenario B, which assumes
that EV buses displace a mix of different bus types.
Table 18. Outputs for Example 3, annual regional results, including vehicles, NOx emissions (lb).
Scenario
From fossil generation
From vehicles
Net changes
Scenario B
-971,280
-58,320
-1,029,600
Scenario C
-971,280
-69,610
-1,040,890
Scenario B less Scenario C
0
11,290
11,290
133 See Step 9 under Example 1 above for information on how to enter default Excel mode.
134 Tip: AVERT does not need to be rerun to evaluate changes to inputs that only affect vehicle emissions—users
can navigate by using Excel's tabs to save time. However, if users make changes to inputs that change electric-
sector impacts, or if they are not sure if their changes result in different electric sector impacts, they should re-run
AVERT using the button found on Step 3.
State and Local Climate
and Energy Program
-------
AVERT User Manual Version 4.2
Appendix K: Caveats and Limitations
Caveats and Limitations: Power Sector
• Snapshot analysis: AVERT provides a representation of the dynamics of electricity
dispatch (i.e., which EGUs are put into operation in which hours) in a historical base year.
However, it does not model changes in dispatch due to transmission resources, fuel prices,
emissions allowances, demand for electricity, or the variable running cost of individual
EGUs.135 The use of AVERT to estimate forward-looking dispatch decisions is made more
difficult when there are changes to the electrical grid (e.g., new transmission resources,
EGU retirements, pollution control retrofits, or new EGUs), commodity prices (such as fuel
or emissions allowances), or operational restrictions (e.g., "reliability must run"
designations, curtailment due to new emissions caps). AVERT characterizes EGU
retirements, pollution control retrofits, and new EGUs in its Future Year Scenario Template,
but the scenarios created are only as good as the user's predictions of future conditions.
• No explicit ramping or cycling: AVERT does not model changes in ramping (periods
when EGUs are changing to a new generation level) and cycling (fluctuating generation
levels) behavior resulting from energy policies, retirements, environmental controls, or new
EGUs.136 AVERT does not capture the changes in the frequency of ramping and cycling of
fossil-fuel EGUs that can result from variability in wind- and sun-powered generation. In
addition, it does not capture the ability of slow- or fast-cycling plants to respond to hour-to-
hour changes in demand.
• Average outcome: AVERT generates an average outcome for each EGU, rather than a
specific and precise trajectory. The default RDFs produce generation and emissions levels
that are averaged across thousands of hypothetical scenarios of a recent past year. These
levels are the statistically expected outcome, and should not be mistaken for an assertion
of what did happen in a past year or what will happen in a future year.
• Limited resolution for generation: AVERT estimates regional changes in emissions. To
do so it predicts the most likely generation levels for individual EGUs given a particular
regional fossil-fuel load level and the most likely emission rates for individual EGUs given a
particular generation level. Results at the individual EGU level (and for counties containing
small numbers of EGUs) have very limited "resolution"; the accuracy of the results is
limited at small spatial and temporal scales.
• Limited ability to capture impacts related to for energy policies with small MW
inputs: Due to the precision limitations within AVERT, when analyzing smaller-scale
energy programs, AVERT may return a higher level of "noise" in the changes in emissions.
With small inputs, users may notice a divergence between desired changes in generation
and modeled changes in generation. Small changes may be overwhelmed by random
effects, such as historical non-economic forced outages and weather events. Users are
135 For example, new emissions controls may entail additional variable costs incurred by specific units. These
additional costs could affect dispatch, but are not captured by AVERT.
136 Models that do not capture ramping or cycling dynamics are generally referred to as having non-chronological
dispatch—i.e., there is no explicit sense of time or timing built into the model.
SEPA
State and Local Climate
and Energy Program
-------
JtVert
AVoid«d Emissions and q»rwRatlon Tool^^
wwwepagovTavSt'^W
AVERT User Manual Version 4.2
encouraged to use emission rates pre-generated from AVERT for small-scale projects.137
There is no specific limit on the smallest project that can or should be reviewed in AVERT,
but users should be aware that modeling very small energy polices may produce answers
that are within the rounding errors of the tool.138 For additional guidance on modeling small
energy projects or policies, see Appendix H. For guidance on how to interpret the noise in
the results (particularly for small programs) see the text related to "Signal-to-noise
diagnostic" on page 47.
• Limited ability to capture dispatch implications of very large energy policies: AVERT
is designed to model marginal changes in system demand. Very large-scale energy
projects or programs may fundamentally change the way in which dispatchers operate a
system. In particular, there is little precedent in the United States for understanding how
high penetrations of variable renewable resources (such as wind and solar) impact other
EGUs in a system. In some cases, very high penetrations of these resources may result in
patterns that are not often observed in the historical dataset, such as the curtailment of
slower-cycling coal plants or an increase in the dispatch of fast-cycling peaking plants to
smooth irregularities.139 Appendix H discusses reasonable maximum levels of load change
for the purposes of obtaining useful results from AVERT.
• Precision of results: AVERT reports results rounded to the nearest 10 units (i.e., MWh, lb
of PM2 5, SO2, NOx, or tons140 of CO2). In general terms, users should consider the number
of significant figures in their specified MW load change, and limit their use of AVERT
results to that number of significant figures.
• Non-communicating regions: AVERT models one region at a time, assuming that each
region generates sufficient electricity to meet its own requirements; imports and exports of
electricity between regions are assumed to stay constant with changing load requirements.
Similarly, changes to emissions are restricted to the confines of the AVERT region selected
for the analysis. The basis of this assumption is that analyses on smaller-sized regions
would risk missing important interdependencies between EGUs across larger, well-
integrated regions. Using yet larger regions than those in AVERT, however, would spread
the influence of load changes too widely, making it difficult to ascribe load changes to
particular EGUs.
137 Pre-generated emission rates for recent and historical years are available at https://www.epa.qov/avert/avoided-
emission-rates-qenerated-avert.
138 The smallest size that AVERT can resolve appropriately will vary by region and program, depending on how the
program is distributed across time and the number of units that reduce output in response. AVERT allows users
to review the noise in expected results via a post-run diagnostic (discussed under "Signal-to-noise diagnostic" on
page 46). Ultimately, the user must use these resources and their best judgement to determine if the results for
small projects return adequate information and appear reliable. For more information, see Appendix H.
139 See Brown, P. 2012. U.S. Renewable Electricity: How Does Wind Generation Impact Competitive Power
Markets? Congressional Research Service. R42818. https://www.fas.org/sqp/crs/misc/R42818.pdf. See also
National Renewable Energy Laboratory. 2016. Eastern Renewable Generation Integration Study.
https://www.nrel. qov/qrid/erqis. html.
140 In AVERT, all references to tons are short tons (2,000 lb.), not metric tons.
State and Local Climate
and Energy Program
-------
AVERT User Manual Version 4.2
• Unconstrained transmission: AVERT looks at the dynamics of each region as a whole
regardless of transmission constraints.141 The model represents how the regional electricity
system actually operated in the base year given the existing transmission infrastructure,
but is completely insensitive to the physical location within a region of new resources or
demand change resulting from energy policies, as well as to the location within a region of
retirements, environmental retrofits, and new EGUs modeled in the Future Year Scenario
Template. In contrast, actual electricity dispatch decisions may be quite sensitive to the
specific locations of resources, including (but not limited to) whether renewable resources
are located close to consumers (at "load center") or in sparsely populated areas.
• Limited capture of individual EGU dynamics: Fossil-fuel EGUs, especially those using
steam cycles, tend to operate at higher efficiencies and with lower emission rates while in
steady-state operation at or near their maximum output (although NOx emissions in
particular may be exacerbated by high-output operations). The AVERT approach does
account for emission rates appropriate to different levels of generation (which may be
associated with periods of fast-ramping or cycling by fossil-fuel EGUs), but does not
account for inefficiencies that may be associated with rapid cycling.
• Infrequent emission events for SO2: In some limited circumstances, infrequent extreme
emission events may be over-represented in the AVERT dataset. For example, instances
during which an EGU switches from one fuel to another (e.g., from natural gas to oil), or
EGU equipment experiences malfunctions, may cause SO2 emission rates for one or more
hours to be hundreds or thousands of times higher than emission rates in other hours with
similar levels of generation.142 Under these conditions, SO2 emission rates produced by
AVERT may appear different than they might otherwise be expected to, given the low
prevalence of these anomalous-emission hours. These unusual emission rates appear only
in certain years, at certain EGUs within a limited number of AVERT regions, and only for
SO2 (one of the four pollutants reported in AVERT). Depending on the region and year,
these high emission rates produce annual total SO2 emissions -13 percent to 141 percent
different from actual observed emissions. Known instances of this issue include:
¦ 1 EGU (of 225) in the 2022 New York RDF
¦ 3 EGUs (of 215) in the 2022 Southeast RDF
¦ 5 EGUs (of 192) in the 2022 Florida RDF
¦ 1 EGU (of 127) in the 2022 Southwest RDF
¦ 1 EGU (of 189) in the 2021 Florida RDF
¦ 4 EGUs (of 228) in the 2021 New York RDF
¦ 1 EGU (of 260) in the 2020 California RDF
¦ 1 EGU (of 152) in the 2020 Carolinas RDF
¦ 1 EGU (of 190) in the 2020 Florida RDF
¦ 8 EGUs (of 223) in the 2020 New York RDF
¦ 1 EGU (of 200) in the 2020 Southeast RDF
¦ 9 EGUs (of 203) in the 2019 New York RDF
141 Transmission is the infrastructure to transport electricity from generators to load centers (i.e., from the source of
generation to electricity consumers). It can be "constrained" when the thermal (or other) limits prevent as much
energy as is needed from moving across wires. When transmission is "bound" under these circumstances,
dispatch begins to reflect local requirements, rather than regional requirements. In other words, generators may
dispatch in a non-economic fashion when transmission is constrained.
142 The data for these circumstances are reported as measured, which contrasts with the maximum potential
concentration substitute data discussed in Appendix B.
SEPA
State and Local Climate
and Energy Program
-------
AVERT User Manual Version 4.2
¦ 1 EGU (of 213) in the 2019 Southeast RDF
¦ 1 EGU (of 132) in the 2018 New England RDF
¦ 1 EGU (of 222) in the 2018 New York RDF
¦ 5 EGUs (of 227) in the 2018 Southeast RDF
¦ 6 EGUs (of 205) in the 2017 New York RDF
¦ 4 EGUs (of 207) in the 2017 Southeast RDF
To account for these infrequent emission events, AVERT outputs are modified to not report
the marginal SO2 emissions for those EGUs affected by infrequent emission events. This
conservative modification ensures that AVERT does not overstate the SO2 emissions
impacts from EERE or EVs. Additionally, for the regions and years above, regional SO2
total emissions are based on actual reported emissions (CAMD data) rather than AVERT's
modeled data. EPA is currently evaluating approaches to improve the modeling of units
with infrequent events for future versions of AVERT.
Caveats and Limitations: Modeling Electric Vehicles
The following section describes several key limitations and caveats to using AVERT to estimate
emissions associated with EVs.
• Time horizon: AVERT provides estimates of changes in emissions in a single, near-term
year that is assumed to be taking place within five years of the selected RDF. It does not
provide estimates of emissions over the lifetime an EV may operate. Over a longer period
of time, electric sector emissions associated with EV charging may decrease as the grid
continues to become cleaner. For more information, see the FAQ section titled "Electric
Vehicles" on page 104.
• Lifecycle: At this time, AVERT only addresses vehicle emissions associated with
combustion and evaporation of volatile chemicals from vehicles during refueling and
nonuse. AVERT does not account for lifecycle emissions (e.g., those related to upstream
fuel production and transportation, upstream manufacturing, or downstream reclamation).
For more information, see the FAQ section titled "Electric Vehicles" on page 104.
• Distribution of vehicle emission changes: AVERT assumes that changes in ICE vehicle
emissions (e.g., from tailpipes and associated vehicle emission sources) are allocated
across counties in the modeled AVERT region in line with historical VMT in that county. For
example, if a county represents 1 percent of historical VMT relative to total VMT in the
entire AVERT region, that county will be allocated 1 percent of avoided emissions from ICE
vehicles. In reality, this allocation may be different as some counties may see near-term
penetration of EVs that does not match this proportional VMT assignment (e.g., for reasons
related to demographics or charging station accessibility). However, because counties are
likely to be similar geographically to the typical area an EV drives in the course of a day,
we expect this allocation to be a reasonable representation.
• Climate adjustments: AVERT assumes that EV efficiencies decrease when vehicles are
driven in warmer- or colder-than-average temperature conditions. This assumption is
applied at a regional and annual level. In other words, a single factor changes the charging
efficiency for a given region based on its annual difference in temperature from AVERT's
baseline temperature conditions. Future versions of AVERT may modify vehicle
State and Local Climate
and Energy Program
-------
AVERT User Manual Version 4.2
efficiencies based on weather using more specific regional data, or may modify vehicle
efficiencies on a month-to-month basis rather than using a single annual value.
Future marginal emission rates: The page "Reference: Modeled Marginal Emission
Rates Over Time" displays region-specific projections for two different kinds of marginal
emission rates, both from NREL's Cambium data set. Users interested in projections of
marginal emission rates may wish to consult with other regional estimates published by
Independent System Operators (ISOs) or Regional Transmission Organizations (RTOs),
utilities, regulators, state energy agencies, or other organizations.
Mobile source regulatory analyses: AVERT may not be used for mobile source
regulatory analyses, including SIP and transportation conformity analyses. Consult the
most recent EPA guidance document for applying EPA's MOVES model at:
https://www.epa.qov/moves/latest-version-motor-vehicle-emission-simulator-moves.
MOVES: MOVES predicts average emissions by vehicle class in a given model year,
based on average operation and activity levels. It does not estimate vehicle or
manufacturer-specific emissions, or consider driving patterns of individual drivers.
Emissions for displaced "existing" vehicles are based on average emissions of vehicles
aged 0-30, weighted by vehicle age. These emissions may vary substantially from specific
model years that some programs might target for displacement. For more information
about MOVES, see: https://www.epa.gov/moves.
129
SEPA
State and Local Climate
and Energy Program
-------
AVERT User Manual Version 4.2
Appendix L: Version History
EPA has added several enhancements to AVERT since the first release in 2014. Table 19 catalogs
the version history of AVERT in reverse chronological order and notes key changes. For a detailed
description of what is new in the most recent version of AVERT, see "What's New in AVERT 4.2?"
in the beginning of this user manual.
Table 19. AVERT version history.
Version #
Release date
Updates, bugfixes, and notes
4.2
October 31,
2023
Updates:
• Updated vehicle emission rates using results from MOVES4. As part of this
update, users can now analyze vehicle model years 2023-2028 (updated
from 2020-2025).
• Updated the source for data on VMT by county from NEI to MOVES4.
4.1
April 25, 2023
Updates:
• Released 2022 Regional Data Files (RDFs) based on data downloaded
from https://campd.epa.qov/data/bulk-data-files on March 30. 2023, and
updated transmission and distribution losses for 2022.
• Estimated emission rates for fine particulate matter (PM2.5), volatile organic
compounds (VOCs), and ammonia (NH3) for 2022 and updated them for
2020 and 2021. For all three years, power plant emission rates for these
three pollutants have been updated to rely on the 2020 National Emissions
Inventory (NEI) point source file. For power plants that were newly
constructed in 2021 or 2022 and do not yet exist in the NEI, a typical
emission rate is used, based on power plants that are similar to the newly
constructed plant in terms of fuel type and prime mover type, and were
operating in 2020. This means that AVERT runs performed with 2020 and
2021 RDFs in v4.1 will produce different emission impacts for PM2.5, VOCs,
and NH3 compared with runs using these same RDFs in earlier versions.
• Updated the vehicle sales and stock and the historical energy efficiency and
renewable energy additions tables to years 2019-2021. Updated the long-
run marginal emission rate reference to Cambium 2022.
• Updated Web Edition to use AVERT v4.1 with 2022 RDFs.
4.0
January 31,
2023
Updates:
• Added capability to model the impact of electric vehicles on electric power
sector emissions and displaced emissions from internal combustion engine
vehicles.
• Added new summary outputs that include vehicle-related emissions
changes.
• Added a new output page that references long-run marginal emission rates
to allow for comparison with AVERT results.
3.2
March 29,
2022
Updates:
• Released 2021 RDFs and updated transmission and distribution losses for
2021.
• Estimated emission rates for PM2.5, VOCs, and NH3 for 2021 and updated
them for 2020. For both years, AVERT relies on emission rate data from the
2019 NEI point source file, so EGUs built in 2020 and 2021 are not yet
listed in the NEI For these EGUs, a typical emission rate is calculated
based on new EGUs operating in 2019 that have the same fuel type and
prime mover.
3.1.1
December 9,
2021
Bugfixes:
• Addressed issues with analyzing the 2020 Midwest RDF.
• Addressed visual glitches that occurred when loading RDFs under certain
conditions.
State and Local Climate
and Energy Program
-------
AVERT User Manual Version 4.2
Version #
Release date
Updates, bugfixes, and notes
3.1
October 5,
2021
Updates:
• Updated onshore wind power profiles and capacity factors in the Main
Module.
• Incorporated PM2.5, VOC, and NH3 data from the NEI.
• The AVERT Web Edition still uses 2019 RDFs but reflects the updated
Main Module v3.1 with respect to onshore wind capacity factors and new
pollutants.
Updates:
• Revised AVERT regions to reflect the modern electric grid. The 14 new
AVERT regions are based on aggregations of one or more balancing
authority(ies).
3.0
September
15, 2020
Note: Prior to AVERT 3.0, there were 10 AVERT reqions based on aqareqations
of the 26 eGRID regions (also in use in the Energy Information Administration's
(ElA's) Annual Energy Outlook from 2011 to 2019). The switch to a new regional
topology in AVERT 3.0 was driven by the fact that these regions are in some
cases out of date as the electric grid has evolved and because certain data on
electricity demand are not readily available for these regions.
• Added offshore wind.
• Added the ability to scale renewable energy capacity factors (Excel-based
AVERT only).
• Added statewide analysis functionality (web AVERT only) (see Appendix I).
Bugfixes:
• Removed "worst-case" substitute emissions data points from the underlying
EPA Clean Air Markets Division (CAMD) input files.
Updates:
• Incorporated line loss factors from EIA, which provides unique values for
each year.
2.3
May 30, 2019
Note: AVERT 2.3 is a deprecated version of AVERT with 10 reqions and fewer
features than the most recent version of AVERT. EPA is no longer supporting
data updates, enhancements, or bugfixes to this version of AVERT. However,
for users who want to use this previous version of AVERT, the Main Module,
RDFs for years 2007-2018, Statistical Module packages for years 2007-2018,
and user manual are available for download at www.epa.qov/avert.
2.2
March 4,
2019
Updates:
• Users can now output AVERT calculations to CO-Benefits Risk Assessment
Health Impacts Screening and Mapping Tool (COBRA) and Sparse Matrix
Operator Kernel Emissions Model (SMOKE) formats even if the modeled
changes in load exceed AVERT's recommended limit of 15 percent of
regional load in any hour.
2.1
October 19,
2018
Updates:
• Added new columns on "Manual Energy Profile Entry" page that tell the
user when the entered generation change exceeds both the recommended
and calculable ranges of AVERT in each hour.
• Added new pop-up box to "Step 3: Run Impacts" that explains how a user
can remedy entered generation change that exceeds both the
recommended and calculable ranges of AVERT.
2.0
May 31, 2018
Updates:
• Added output files compatible with EPA's COBRA Health Impact Screening
Tool.
Bugfixes:
• Corrected code in the Statistical Module to ensure that AVERT will work
with the newest version of MATLAB.
131
SEPA
State and Local Climate
and Energy Program
-------
AVERT User Manual Version 4.2
Version #
Release date
Updates, bugfixes, and notes
Updates:
• Added PM2.5.
Note: RDFs produced Driorto summer 2017 do not contain PM?s emission data,
and they include generation data in "gross" rather than "net" (corrected for
parasitic losses) terms. If you load an RDF produced in 2017 or earlier, another
pop-up box will alert you to these considerations and suggest that you download
a newer RDF from EPA's website.
RDF import pop-up example for data files produced prior to summer 2017.
AVERT
X
1.6
July 31, 2017
Note that this regional data file does not include PM2.5 data and quantifies
emission impacts based on gross generation. To obtain inputs with PM2.5 data
and net generation, click on the hyperlink under the AVERT map,
OK
•
•
Adjusted the Statistical Module and RDFs to account for additional
generation impacts associated with parasitic loads at the point of
generation.
Improved the way data are extrapolated for peak hours.
Bugfixes:
• Corrected summation of annual nitrogen oxides (NOx) values.
• Removed mismatches in CAMD data-to-AVERT data import pipeline.
1.5
March 6,
2017
Updates:
• Added adjustment factor to account for avoided line losses associated with
energy efficiency and distributed renewable energy profiles.
• Added daily avoided NOx by county results.
• Improved data display on map figure.
• Modified rounding of results to tens rather than hundreds place.
• Added caution message for larger-than-recommended energy profiles.
• Updated compatibility to Excel for Mac 2016.
Bugfixes:
• Corrected unit labeling of NOx and sulfur dioxide (SO2) data blocks in
RDFs.
• Corrected peak-day-finding formula in post-processing sheets.
1.4
April 25, 2016
Updates:
• Added compatibility with Excel for Mac 2011.
1.3
April 28, 2015
Updates:
• New pop-up box depicts percent generation in each state within an AVERT
region. Instructions added for states that reside in multiple AVERT regions.
Bugfixes:
• Corrected SMOKE output function bug.
132
SEPA
State and Local Climate
and Energy Program
-------
AVERT User Manual Version 4.2
Version #
Release date
Updates, bugfixes, and notes
1.2
November 21,
2014
Updates:
• Modified default wind capacity factor data to more closely represent
measured wind speeds.
Bugfixes:
• Corrected transposition of NOx and SO2 columns in the Monthly Impact
Data by County table in Step 4.
1
February 18,
2014
Original public version of AVERT.
State and Local Climate
and Energy Program
133
SEPA
------- |