April 2015
A A United States
Environmental Protection
tl Agency
Including Energy Efficiency and Renewable
Energy Policies in Electricity Demand
Projections
A resource for state & local air agencies preparing NAAQS
SIPs
INTRODUCTION
In July 2012, the U.S. Environmental Protection Agency (EPA) published the Roadmap for
Incorporating Energy Efficiency/Renewable Energy Policies and Programs Into State and Tribal
Implementation Plans ("the Roadmap")1 in recognition of the potential impact that energy
efficiency and renewable energy (EE/RE) policies and programs can have on reducing air
emissions, and in an effort to make it easier for states, local governments, and tribes to use
EE/RE policies and programs in their state Implementation Plans (SIPs) and Tribal
Implementation Plans (TIPs), particularly those under sections 110, 172, and 175A of the Clean
Air Act.2 The Roadmap lays out four pathways that states, local governments, and tribes can use
in their SIP/TIP to account for the emissions impacts of EE/RE policies and programs:
(1) baseline, (2) control measure, (3) emerging/voluntary measures, and (4) weight of evidence.
This document provides additional information on incorporating EE/RE policies in the baseline
emissions projections pathway.3
The baseline pathway adjusts the electricity demand forecast that state, local, and tribal
agencies use in their SIPs/TIPs to reflect the impacts of on-the-books EE/RE policies and
programs. The baseline pathway can eliminate the need to account for the changes from
individual EE/RE policies and programs "after the fact" using a traditional control strategy
pathway, an approach states have indicated can be onerous. It also can capture the impact of
multiple statewide EE/RE policies, compared to counting individual programs or measures,
which streamlines the process of accounting for EE/RE in a SIP/TIP. Finally, the results of
applying this approach could provide an estimate of future baseline emissions that reflects the
1 For details about the Roadmap, visit http://www.epa.gov/airqualitv/eere/manual.html.
2 For more information on the types of SIPs that this approach is best suited for, see the Roadmap FAQs
at http://www.epa.gov/airquality/eere/resources.html.
3 State, tribal, and local agencies can include EE/RE policies that are currently on the books in a baseline
emissions projection. This is also known as the baseline emissions projection pathway, as detailed in
Appendix E of the Roadmap.
1
-------
impact of EE/RE policies on power plants, which could lower the costs of attaining the National
Ambient Air Quality Standards (NAAQS).
Projecting future emissions from the power sector normally requires information from an
electricity demand forecast as a basis for predicting how future generation requirements will
grow over time. There are a variety of demand forecasts available, including the U.S. Energy
Information Administration's (EIA) Annual Energy Outlook (AEO) used for this illustrative
analysis, and it is important to understand the assumptions, including which EE/RE programs
are already incorporated in the forecast.
EPA has developed a methodology for estimating the energy impacts of key EE/RE on-the-books
policies that are not explicitly reflected in the ElA's AEO 2013 electricity projections and
including them in their baseline projections. These policies include Energy Efficiency Resource
Standards (EERS), dedicated sources of energy efficiency program funding that are adopted in
state law and/or codified in rule or order, and renewable portfolio standards. EPA solicited peer
and public review of this methodology, and comments received have been addressed and
incorporated into this document.
This discussion paper will help agencies understand options for:
• Identifying on-the-books EE/RE policies and estimating the incremental electricity savings of
these policies;
• Developing a methodology for projecting a jurisdiction's energy demand both with and
without the incremental electricity savings; and
• Estimating the change in power sector emissions attributable to the incremental electricity
savings.
This methodology was developed to illustrate how EE/RE policies could be accounted for
in the context of National Ambient Air Quality Standards (NAAQS) State Implementation
Plans (SIPs), as required in Section 110 of the Clean Air Act. While elements of this
analysis may inform issues that are shared between the existing NAAQS SIP
requirements and the proposed Clean Power Plan (Section llld), this analysis is not
intended to offer guidance for complying with any requirements of the Clean Power
Plan.
SECTION I: METHODS AND APPROACH
EPA is providing a methodology that states responsible for developing SIPs for ozone or other
criteria air pollutants could use to estimate their own electricity impacts. Jurisdictions not
currently preparing a SIP/TIP, but interested in better understanding the energy and emissions
impacts of EE/RE policies, can likewise use EPA's methodology to identify strategies for staying
in attainment with the NAAQS.
2
-------
EPA used this methodology to conduct a detailed policy review to produce national numeric
estimates4 of the electricity impacts of state EE/RE policies not accounted for in the AEO 2013
forecast. EPA is not, however, providing state-level estimates of EE/RE impacts, nor estimates
of resulting emissions reductions. The time period covered by this analysis is 2013-2030.
For this analysis, EPA chose to include a core of high-impact, on-the-books policies that states
have generally adopted and illustrate how savings could be estimated for inclusion in NAAQS
SIPs. Because these policies are embodied in state legislation or regulation, there is a
reasonable expectation that they will continue, and thus could be appropriate for inclusion in
the SIPs. EPA recognizes that these assumptions are conservative and acknowledges that there
are other types of programs and measures that could be included in SIPs, such as integrated
resource plans and voluntary programs. EPA also acknowledges that the assumptions limiting
the scope of this illustrative analysis effectively means that not all states with utilities and other
entities reporting their EE/RE program impacts to EIA are reflected. The assumptions used for
this analysis should not be construed as guidance for states to limit the types of on-the-books
policies that may be included in their SIPs. States need to make their own determinations about
what policy instruments to include in their SIPs in consultation with their regional EPA office.
For more information on the steps necessary for applying the results of this methodology to
estimate baseline emissions impacts, see Section II, Applying Results to Estimate Emissions
Reductions.
EPA's approach to estimating the electricity effects of EE/RE policies is described in six steps:
• Step 1: Choose a baseline forecast for electricity demand projections (e.g., AEO 2013 or
newer).
• Step 2: Document EE/RE policies already included in baseline electricity demand
projections.
• Step 3: Identify on-the-books EE/RE policies not included in AEO or the chosen baseline
electricity demand forecast.
• Step 4: Estimate the incremental electricity savings of the on-the-books EE/RE policies
not included in the chosen baseline electricity demand forecast.
• Step 5: Incorporate the incremental electricity impacts of EE/RE policies and lower
projected electricity demand in the model.
• Step 6: Project the change in power sector emissions attributable to the incremental
effects of on-the-books EE/RE policies for future attainment year(s).
Steps 1 through 5 are described in this section, and Step 6 is described in the next section,
Applying Results to Estimate Emissions Reductions. The descriptions for each step include
explanations for how states may conduct their own analysis, as well as the details of the
example analysis that EPA conducted.
4 For more information, go to http://www.epa.gov/statelocalclimate/state/statepolicies.html.
3
-------
Please note that throughout this discussion, EPA refers to savings resulting from EE/RE policies.
EPA recognizes that renewable energy policies result in electricity generation; however, when
compared against an existing forecast, this generation should be viewed as a reduction in the
demand forecast. Therefore, incremental renewable energy generation will be treated the
same as savings from energy efficiency policies and will be referred to jointly as savings for the
purposes of this discussion.
Step 1: Choose a baseline forecast for electricity demand projections.
State, local, and tribal governments can choose from a number of information sources which
provide electricity demand forecasts that can be used in a SIP.5 These include the following:
• U.S. Department of Energy, Energy Information Administration, Annual Energy Outlook
(AEO)
• North American Electric Reliability Corporation (NERC)
• Regional transmission organizations (RTOs)/independent system operators (ISOs) (e.g.,
PJM Interconnection, ISO-New England, NYISO)
• Vertically integrated utilities (e.g., a large power company that operates the electricity
system for a specific region)
• State energy agencies (e.g., State Energy Office or Public Utility Commission [PUC])
• Regional councils that coordinate energy planning (e.g., Northwest Power and
Conservation Council)
Each information source may already reflect different levels of on-the-books EE/RE policies, so
it is important to research and document how each of the policies is incorporated. States can
determine the most appropriate source for their electricity demand forecast (to be used for the
emissions baseline projection) by reviewing the forecast's growth rates, policy assumptions,
and economic conditions to ensure that they are aligned with their own assumptions. Keep in
mind that if a group of states does their air quality modeling on a regional basis and uses
electricity demand forecasts from different information sources, then any inconsistencies
between the approaches and assumptions will need to be reconciled.
Organizations develop demand forecasts for a variety of reasons. The purpose of the specific
demand forecast may influence which informational resource a state, local, or tribal
government chooses. For example, NERC demand forecasts are developed from utility-level
forecasts provided to NERC as part of annual, long-term reliability assessments. Available
regional forecasts may be developed as part of regional transmission planning activities
established by the Federal Energy Regulatory Commission, which may also include separate
scenarios incorporating alternative assumptions about environmental regulations. State, tribal,
and local governments should work closely with their EPA regional office if their demand
5 For more detailed information on how to develop a baseline demand forecast, see EPA's Assessing the
Multiple Benefits of Clean Energy: A Resource for States at
http://epa.gov/statelocalclimate/resources/benefits.html.
4
-------
forecast information comes from one of these organizations to ensure that all environmental
regulations are properly accounted for in the analysis.
Application of Step 1 in EPA's illustrative analysis
EPA and many states use data from the AEO electricity demand forecast when developing a
forecast of emissions from the power sector. EPA relies on ElA's electricity demand forecasts
and underlying EE/RE policy assumptions to project future growth in electricity demand for
the electric power sector. EPA uses the Integrated Planning Model (IPM™)6 electricity market
model and updates its base case application of the platform with the new AEO forecasts as
they become available.7
Step 2: Document EE/RE policies already included in baseline electricity
demand projections.
When choosing a demand forecast, the state, local, and tribal governments should work with
the source of the forecast to understand whether and how the following EE/RE policies are
included in the electricity demand forecast:
• Energy efficiency policies or programs funded by utility ratepayers
• Existing Federal appliance and lighting efficiency standards that are already in effect
• New Federal appliance and lighting standards that are scheduled to take effect over the
forecast period
• State appliance or lighting efficiency standards (if applicable)
• State building energy codes
• Other applicable policies/programs (e.g., codified local policies)
There are at least two ways that these EE/RE policies are captured in an existing demand
forecast:
1. Explicitly modeled policies showing a direct connection between the EE/RE policy and its
impacts on energy demand
2. Indirectly, and either fully or partially, through econometric or other assumptions in the
model
6 IPM was developed and is maintained by ICF Resources. It is used for EPA and other federal, state, and
commercial clients for analysis of power market issues. It has been used by EPA to analyze the potential
impacts of a broad range of air regulations, policies, and legislative initiatives.
7 EPA typically uses AEO electricity demand projections as input assumptions in its IPM base case. IPM
outputs provide estimates of future electric generating unit emissions. It is important to know which
version of AEO that EPA is using as inputs for electricity demand projections in a given IPM run. There is
typically a brief time lag between EPA's modeling platform updates and the most recent AEO forecast
release. For more information, see http://www.eia.gov/forecasts/aeo.
5
-------
Application of Step 2 in EPA's illustrative analysis
To understand the EE/RE policy assumptions included in the AEO 2013 forecast, EPA
reviewed ElA's documentation for the AEO 2013 reference case forecast and consulted with
EIA staff.8 From the review, it is clear that AEO 2013 explicitly includes the impacts of several
existing EE/RE policies,9 including the following:
• Federal appliance and equipment standards for residential and commercial
categories10
• Lighting efficiency standards for various types of lighting technologies11
• Tax credits for energy-efficient appliances and equipment, and investment tax credits
for EE/RE technologies12
• Federal energy efficiency programs and funding:
o American Recovery and Reinvestment Act13
o State Energy Program and Energy Efficiency and Conservation Block Grant
o Weatherization Program
o Green Schools
o Smart Grid expenditures
• Building energy codes for residential and commercial new construction:14
o For example, all states adopt and enforce International Energy Conservation
Code (IECC) 2006 (Residential Building Code) by 2017
• State Renewable Portfolio Standards (RPS):15
o 30 states and Washington, D.C., effective as of October 2012
Step 3: Identify on-the-books EE/RE policies not included in chosen baseline
electricity demand projections.
State, tribal, and local governments should review the EE/RE assumptions within the respective
forecast or talk with the organization providing the demand forecast to determine which on-
the-books state EE/RE policies are incorporated in their forecast, and what on-the-books EE/RE
policies should be evaluated for incremental EE/RE savings. States will need to be aware of
differences between current year (i.e., first year) and cumulative (those that persist and
8 EIA documents their assumptions whenever they produce an updated AEO forecast. It is important to
review the most current assumptions within the respective AEO forecast.
9 This discussion highlights several of the most important policies, but it is not intended to be a
comprehensive review of AEO assumptions. See
http://www.eia.gov/forecasts/aeo/assumptions/pdf/appendix a.pdf for additional information.
10 U.S. EIA (2013c), Appendix A, pp. 187-204.
"Ibid.
12 Ibid.
13 U.S. EIA (2013c), pp. 32.
14 Ibid.
15 U.S. EIA (2013a), pp. 14-17.
6
-------
accumulate over a period of time) impacts associated with any chosen policies, and document
them appropriately for use in calculations.
EE/RE policies represent just one of many assumptions made in electric generation baseline
emissions forecasts. Any EE/RE policies that are explicitly included in an electric generation
baseline projection should document the following:
1. Policy name
2. Whether the policy is codified in state or local rule
3. Year enacted
4. When the policy requirement sunsets
5. Policy requirements (e.g., targets in megawatt hours [MWh] or percentage)
6. Annual energy savings in the base year (MWh)
7. Annual and cumulative energy savings in the future attainment year (MWh)
Application of Step 3 in EPA's illustrative analysis
EPA identified on-the-books EE/RE policies not explicitly included in AEO for the illustrative
example provided in this document.16'17 These policies are as follows:
• Energy Efficiency Resource Standards
• Energy efficiency programs funded by public benefits funds
• Energy efficiency programs funded by the Regional Greenhouse Gas Initiative18
• Energy efficiency programs funded by Forward Capacity Market revenues19
When EPA performed this step to understand the EE/RE policy assumptions included in the
AEO forecast, EPA reviewed ElA's documentation for the AEO reference case forecast and
consulted with EIA staff; EPA recommends that states consider pursuing a similar approach
when using a different forecast, or a newer version of the AEO forecast.
Step 4: Estimate the incremental electricity savings of the on-the-books EE/RE
policies not included in chosen baseline electricity demand projections.
Whether using a version of the AEO forecast or a different forecast entirely, the same
conceptual approach applies:
16 Other EE programs, such as integrated resource plans (IRPs), are also important instruments that
states may choose to include in their own analyses and SIPs; however, EPA chose to focus on these
major policy types that were enforceable and would produce the greatest energy savings impacts across
the states.
17 EPA last checked for changes to state EE/RE policies on August 4, 2014.
18 For more information, see http://www.rggi.org.
19 For example, several states participating in ISO-NE's forward capacity market are using auction
revenues to fund energy efficiency.
7
-------
1. Begin with the baseline electricity demand forecast for the selected jurisdiction. Ensure
that the jurisdictional boundaries for this forecast are consistent with those of the
policies evaluated.
2. Estimate the electricity savings associated with the on-the-books EE/RE policies already
embedded in the baseline electricity demand forecast. This involves evaluating the
historical savings for these policies and establishing an assumption for the share of
savings that would carry forward in subsequent years through the forecast period.20
3. Estimate the electricity savings associated with the on-the-books EE/RE policies not
included in the baseline electricity demand forecast.21 This can be done by using EE/RE
policy targets established in law, where applicable, and/or estimates of savings by EE/RE
program dollars spent.22
4. Where a portion of the embedded electricity savings of the EE/RE policies is indirectly
factored into the baseline forecast, that portion of embedded savings should be
subtracted from the total program savings of these EE/RE policies to calculate the
incremental electricity savings beyond the baseline electricity forecast.
See Appendix A of this document for a detailed examination of how EPA estimated the energy
savings of the following policies not explicitly included in the AEO forecast:
• Energy Efficiency Resource Standards
• Energy efficiency programs funded by public benefits funds (PBFs)
• Energy efficiency programs funded by the Regional Greenhouse Gas Initiative (RGGI)
• Energy efficiency programs funded by Forward Capacity Market (FCM) revenues
Step 5: Incorporate the incremental electricity impacts of EE/RE policies and
lower projected electricity demand in the model.
Using the incremental electricity savings resulting from Step 4, states can apply these savings as
adjustments to the baseline electricity demand projections, thus creating a revised electricity
demand projection. This may include lowering the demand growth rate, integrating savings
from avoided transmission and distribution losses, or accounting for the timing (e.g., peak or
non-peak demand periods) of the electricity savings from specific EE/RE policy types that affect
determination of the marginal generating asset displaced. The form of savings data will depend
on the quantification approach for calculating emissions in Step 6.
National Results
The results of EPA's analysis are presented here to illustrate the expected relationship among
three main components of this analysis: (1) baseline electricity demand projections (AEO
baseline), (2) projections adjusted to reflect a scenario absent embedded savings, and
20 The specifics of this methodology are further detailed in Appendix A.
21 It is important to avoid double counting of policies. For example, if a state has an EERS and a public
benefits fund (PBF), then it may be assumed that the PBF spending goes toward meeting the EERS, so a
state would only consider the EERS.
22 The specifics of this methodology are further detailed in Appendix A.
8
-------
(3) revised electricity demand projections capturing both the embedded and incremental
savings effects.
Figure 1. National Results
4,400,000
4,200,000
§ 4,000,000
~ 3,800,000
3,600,000
3,400,000
Hypothetical Forecast Accounting for Embedded EE Impacts
AEO Baseline Electricity Demand Forecast
Revised Electricity Demand Forecast
Estimated electricity savings
from EERS and EE programs
embedded in the AEO baseline.
Total projected electricity
savings from State EERS and
EE/RE programs (2.2% in 2030).
Incremental electricity
savings from key State
EE/RE policies not
included in the AEO
baseline.
3,200,000
rTJ r\'O n \ r,Y>
Important Sources of Uncertainty
In conducting its analysis, EPA identified important sources of uncertainty that may be relevant
to states' own analyses; these are discussed in detail in Appendix A. Generally, states should
keep in mind the following when employing similar methods:
• The impacts of state EE/RE policies embedded in the baseline electricity demand
projections: It is sometimes the case that sources of demand projections may not know
with absolute certainty what policy impacts are embedded in their projections, often
due to using source data collected from various sources. Therefore, it is important to
recognize that even the best assumptions developed for embedded effects may have
limitations.
• The approval of sufficient EE/RE program budgets necessary to meet the specified
targets: Achievement of program targets is usually dependent upon approval of
adequate energy efficiency program budgets, which may be affected by PUC approvals,
varying levels of funding from legislatures, and other factors. Uncertainty related to
these factors may need to be taken into account.
9
-------
SECTION II: APPLYING RESULTS TO ESTIMATE EMISSIONS REDUCTIONS
Section I focused on the electricity aspects of the EE/RE policies analyzed. In Section II, the
resulting incremental electricity savings are used to estimate the emissions reductions
associated with these electricity savings.
Step 6: Project the change in power sector emissions attributable to the
incremental effects of on-the-books EE/RE policies for future attainment
year(s).
State, tribal, and local governments have several methods available for quantifying the avoided
or displaced emissions from fossil fuel generation as a result of electricity savings from EE/RE
policy/program implementation. These methods range from basic to sophisticated, and vary in
complexity, rigor, resource implications data requirements, and temporal and spatial scales of
outputs.23
Appendix I of EPA's Roadmap suggests which emission quantification approaches would work in
different situations, and include the following:
• Basic approach: eGRID sub-region "non-base load" emission rates
• Basic approach: Capacity factor emission rates
• Mid range approach: Historical hourly emission rates (e.g., EPA's AVERT24)
• Sophisticated approach: Electricity market models (also referred to as energy models)
Figure 2. Emissions Quantification Approaches
4
*
1
eGRID
Capacity
Non-baseload
Factor
Method
Method
M
1
Assumptions are Simpler
Historical
Hourly
Method
Energy
Modeling
Method
Methods are More Sophisticated
23 All are further detailed in the Roadmap, Appendix I.
24 Visit www.epa.gov/AVERT for further details on EPA's AVoided Emissions and geneRation Tool.
10
-------
Appendix I of the Roadmap also suggests that the quantification approaches that work for the
baseline pathway described in this paper would include the following:25
• Energy model approach
• Historical hourly emission rate approach
• Alternative emissions projection tools or analysis
For more sophisticated analyses and/or if the projections for energy efficiency policies and
programs extend out more than 5 years, a state, tribal, or local government should develop
projections of how future generation may change over time. The jurisdiction should examine
each nonattainment area and assign emission rates to new units expected to come online, or
exclude planned retired plants in the jurisdiction's future emission rates. There are several
organizations that project how electric generators will meet future demand and react to new
environmental regulations. EPA recommends obtaining projections of future electric generation
growth from EPA, EIA, electric grid operators, RTOs, ISOs, or NERC, if possible.
SECTION III: CONCLUSION
EPA believes that energy efficiency and renewable energy policies and programs have the
potential to reduce multiple pollutants in a cost-effective manner, and can be one tool that
states, local governments, and tribes can use to help meet the NAAQS. This paper provides
analytical steps that can be used to incorporate EE/RE policies into SIP/TIP baseline electricity
demand projections. EPA welcomes feedback and would like to hear from states, local
governments, and tribes about what additional information might be useful to support their
development of SIP/TIP baselines that incorporate energy efficiency and renewable energy.
25 These quantification methods are only suggestions. Each state should consult with its EPA Regional
Office on the type of alternative emission projection tools or analysis that they plan to use in a SIP/TIP
submission.
11
-------
APPENDIX A: METHODOLOGY AND ANALYTICAL STEPS FOR ESTIMATING
EE/RE POLICY IMPACTS
EPA developed the methodology described herein to conduct a detailed policy review to
produce national numeric estimates of the electricity impacts of state EE/RE policies not
accounted for in AEO 2013 forecast. The time period covered by this analysis is 2013-2030.
EPA applied the following analytical steps to estimate the projected annual energy savings of
energy efficiency policies:
• Step 1: Generate a baseline (i.e., business as usual [BAU]) forecast of state electricity
sales consistent with AEO 2013 regional forecasts.
• Step 2: Estimate the projected impacts of key state on-the-books energy efficiency
policies already embedded in the AEO 2013 forecast of electricity sales.
• Step 3: Estimate projected total energy efficiency savings from key state on-the-books
energy efficiency policies adopted or updated as of June 2013:
o EERS (25 states)
o Funding for energy efficiency programs in non-EERS states (e.g., dedicated
funding from PBFs, RGGI, and FCM revenues)26'27 (five states)
¦ Step 4: Generate a state-adjusted national energy forecast that reflects the energy
savings not captured in (i.e., incremental to) the baseline forecast.
For renewable energy policies, EPA applied the following key analytical steps to estimate the
projected annual energy impacts:
• Step 1: Estimate renewable energy generation from RPS policies adopted or revised
between October 2012, when the AEO 2013 RPS assumptions were formulated, and
June 2013, when this analysis was released for review (two states).
• Step 2: Generate a state-adjusted forecast reflecting policy changes.
Methodology for Generating a Baseline Forecast of State Electricity Sales to
Represent AEO 2013 Regional Forecasts
State-level baseline sales28 data were developed by first using 2012 historical state sales data
from the El A29 and then applying the electricity sales growth rates from AEO 2013. AEO 2013-
based annual average growth rates (AAGR) were calculated for each Electricity Market Module
(EMM) region across the 2012-2040 forecast period. These regional growth rates were then
applied to the 2012 historical sales for each state. For states whose boundaries cross EMM
26 For more information, see http://www.rggi.org.
27 For example, several states participating in ISO-NE's forward capacity market are using auction
revenues to fund energy efficiency.
28 Note that AEO 2013 does not include state-level forecasts, so incremental impacts are calculated
against the BAU electricity sales forecast developed as described in Appendix A.
29 U.S. EIA (2013b).
12
-------
regions, state-specific growth rates were derived by using a historical sales weighted average of
EMM region growth rates.30 The 2012-2040 AAGR was used to forecast sales for 2013-2040.
Table 1 shows the EMM regions and the AAGRs used to forecast sales for each state.
Table 1. Electricity Market Module Regions and AEO 2013 Sales Growth Rates by States
State/Jurisdiction
Electricity Market
Module Region
Average
Annual Growth
Rates
(2012-2040)
Arizona
AZNM, NWPP
1.30%
Arkansas
SPNO, SPSO, SRDA
0.87%
California
AZNM, CAMX, NWPP
0.90%
Colorado
AZNM, NWPP, RMPA,
SPNO, SPSO
1.22%
Connecticut
NEWE
0.22%
Delaware
RFCE
0.51%
District of Columbia
RFCE
0.51%
Florida
FRCC, SRSE
1.18%
Hawaii
HI31
0.78%
Illinois
MROW, RFCW, SRGW
0.43%
Indiana
RFCW
0.41%
Iowa
MROW, SRGW
0.54%
Maine
NEWE
0.22%
Maryland
RFCE, RFCW
0.50%
Massachusetts
NEWE
0.22%
Michigan
MROE, MROW, RFCM,
RFCW
0.33%
Minnesota
MROW
0.54%
Montana
MROW, NWPP, RMPA
0.94%
Nebraska
MROW, RMPA
0.55%
Nevada
AZNM, NWPP
1.20%
New Hampshire
NEWE
0.22%
New Jersey
NYUP, RFCE
0.50%
30 Each state was first mapped to one or more EMM regions, depending on the geographical overlap.
The share of each state's electricity sales (from EIA-861) in a given EMM region was calculated as a
percentage of total sales for that state. These shares represent the contribution of each EMM region's
growth rate to the state's growth rate. The growth rate of each EMM region overlapping a state was
then weighted by the share of each state's sales within that EMM region.
31 Because AEO 2013 includes the contiguous lower 48 states only, the U.S. Average Annual Growth Rate
was applied for Hawaii. Alaska is not included in this analysis because the state has no on-the-books
state EE/RE policies meeting the definition for inclusion in this analysis.
13
-------
State/Jurisdiction
Electricity Market
Module Region
Average
Annual Growth
Rates
(2012-2040)
New Mexico
AZNM, NWPP, RMPA,
SPSO
1.18%
New York
NEWE, NYCW, NYLI,
NYUP, RFCE
0.20%
North Carolina
SRCE, SRVC
1.10%
Ohio
RFCW
0.41%
Oregon
NWPP
0.97%
Pennsylvania
NYUP, RFCE, RFCW
0.48%
Rhode Island
NEWE
0.22%
Texas
AZNM, ERCT, SRDA,
SPSO
0.89%
Vermont
NEWE
0.22%
Washington
NWPP
0.97%
Wisconsin
MROE, MROW, RFCW
0.41%
Methodology for Estimating Energy Savings of State Energy Efficiency Policies
Embedded in AEO 2013
The goal of this analysis was to produce numeric estimates of the energy impacts of state EE/RE
policies not accounted for in AEO 2013 forecast that inform a national estimate. In order to
estimate the impacts not accounted for in the baseline electricity sales forecast, the analysis
necessarily must define the impacts already accounted for in the baseline. Therefore, we define
embedded savings in this analysis as those impacts already accounted for in the AEO 2013
reference case forecast. EPA estimated the embedded savings and subtracted them from
estimates of total state EE/RE policy impacts to yield the incremental savings effects on the
baseline, thus avoiding potential double counting.
AEO 2013 does not explicitly include the impacts of state energy efficiency policies such as EERS
and dedicated sources of energy efficiency program funding. However, we understand the AEO
forecasts to implicitly represent the impacts of those energy efficiency policies addressed in this
analysis. This implicit representation of energy efficiency occurs in two key ways:
1. The AEO forecast incorporates historical data that reflect energy consumption levels and
trends influenced by state-level energy efficiency policies in place at that time. The
effects of these existing policies lower the sales level during the last historic year (e.g., if
2011 is the last historical year of data in AEO 2013, then the 2011 energy demand was
lower than it would have been in the absence of existing energy efficiency policies) and
may also affect AEO's near-term growth rates partially derived from recent historic
demand growth trends (which otherwise would have been expected to be higher in the
absence of existing energy efficiency policies).
14
-------
2. The AEO forecast assumes an ongoing and persistent savings from utility sector energy
efficiency programs as reported in EIA-861 that expire after a defined period of time, or
measure lifetime. Typically, the impacts of energy efficiency programs are estimated in
terms of first-year savings, plus the persistent (cumulative) savings realized from that
program (or energy efficiency measure) over an assumed measure lifetime (a 10-year
lifetime is used for this analysis).32 EPA's assessment of the AEO forecast, however, does
not identify the expected end of these persistent savings (i.e., does not identify a
consequent increase in energy intensity that should accompany the end of an energy
efficiency savings stream), leading EPA to understand that the AEO forecast assumes
that any measure reaching the end of the measure's lifetime will be replaced by an
equal measure, providing an ongoing stream of savings beyond the lifetime of one
efficiency measure.33
Recognizing that AEO 2013 is implicitly affected by these historic and persistent effects of state
energy efficiency policies and programs, EPA concludes that some portion of the total policy-
and program-induced energy efficiency savings are embedded in the AEO 2013 regional
forecast and the AEO 2013-based state-level BAU forecast. EPA therefore developed a
methodology for estimating these embedded savings for each state.34
This methodology involves two steps: estimating national savings from energy efficiency, and
then allocating these national savings to the states covered in the analysis.
For national savings, reported cumulative energy efficiency savings from programs
implemented in prior years (reported as annual effects via EIA-861, and as aggregated in ElA's
Electric Power Annual for residential, commercial, and industrial sectors) are divided by
reported electricity sales (also reported via EIA-861, and as aggregated in ElA's Electric Power
Monthly, Retail Sales of Electricity by State by Sector by Provider). This calculation yields
national average energy efficiency savings as a percentage of sales within the given year.
Because the national average savings is calculated from the most recent year's total cumulative
savings, this value is divided by the average energy efficiency measure lifetime, here assumed
32 In this analysis, first-year energy efficiency savings refer to the savings in the first year that a specific
measure is implemented, and cumulative energy efficiency savings refer to the aggregate stream of
savings resulting from a measure. For example, consider that Program X installs 100 units of energy-
efficient equipment every year for 10 years, and those installations save 100 MWh per year. In Program
X's first year, both first-year savings and cumulative savings are 100 MWh. In the second year, first-year
savings are again 100 MWh because another 100 units of the energy-efficient equipment has been
installed, but the cumulative savings are 200 MWh because both the 100 units from the first year and
the 100 units from the second year are each saving 100 MWh per year.
33 Synapse Energy Economics (2012).
34 During peer review, one reviewer questioned whether it is conceptually possible to calculate
embedded savings from AEO because the level of detail regarding ElA's assumptions in AEO is not fully
transparent, so determining specific embedded savings has inherent uncertainty.
15
-------
to be 10 years. This yields a percentage representing the ongoing annual effects of energy
efficiency that are embedded in the AEO forecast (0.38 percent).35
Allocating the national average embedded savings to the individual states in this analysis uses
state-specific data from the EIA-861. These embedded savings estimates used in the EPA
analysis are presented in Table 2.
Table 2. Energy Efficiency Savings Estimated to be Embedded in AEO 201336
State
Savings Estimated to be Embedded in AEO 2013
(percentage of BAU sales in each year)
Alabama
0.08
Alaska
0.01
Arizona
0.54
Arkansas
0.04
California
1.37
Colorado
0.47
Connecticut
1.34
Delaware
0.00
District of Columbia
0.06
Florida
0.36
Georgia
0.07
Hawaii
0.03
Idaho
0.62
Illinois
0.21
Indiana
0.17
Iowa
0.78
Kansas
0.02
Kentucky
0.10
Louisiana
0.00
Maine
0.54
Maryland
0.25
Massachusetts
0.63
Michigan
0.28
Minnesota
1.31
Mississippi
0.05
Missouri
0.05
35 Synapse Energy Economics (2014).
36 Synapse Energy Economics (2014), Exhibit 3.
16
-------
State
Savings Estimated to be Embedded in AEO 2013
(percentage of BAU sales in each year)
Montana
0.59
Nebraska
0.10
Nevada
0.62
New Hampshire
0.49
New Jersey
0.10
New Mexico
0.19
New York
0.69
North Carolina
0.13
North Dakota
0.02
Ohio
0.32
Oklahoma
0.07
Oregon
0.77
Pennsylvania
0.31
Rhode Island
1.12
South Carolina
0.11
South Dakota
0.03
Tennessee
0.18
Texas
0.15
Utah
0.66
Vermont
1.54
Virginia
0.03
Washington
0.74
West Virginia
0.02
Wisconsin
0.66
Wyoming
0.07
EPA estimates embedded savings for each state by multiplying the percentages shown in
Table 2 by the BAU sales for that state. EPA only estimated embedded savings for the years in
which states achieved savings from energy efficiency policies and, to the extent possible, for
the segments of state electricity load to which the EE/RE policies apply. The next section of this
paper includes discussion of how the cumulative total of the state's embedded savings is
subtracted from the state's total energy efficiency policy savings to yield the impacts that are
incremental to AEO 2013.
Methodology for Estimating Projected Energy Efficiency Savings from Energy
Efficiency Policies
EPA scanned all 50 states to determine which had adopted one or more of the EE/RE policies
included in this analysis as of June 2013. EPA then reviewed the relevant design details for each
17
-------
state policy using publically available information, such as state legislation, state rules and
regulations, PUC orders, summary reports from the American Council for an Energy-Efficient
Economy (ACEEE),37 and the Database of State Incentives for Renewables and Efficiency
(DSIRE).38
EPA estimated state-level energy efficiency savings from EERS policies and dedicated sources of
energy efficiency program funding that are adopted in state law and/or codified in rule or
order. Because these categories are not mutually exclusive, EPA took steps to avoid double
counting of energy savings for states with EERS by treating EERS targets as overall goals that
include savings from individual PBF-funded programs, RGGI-funded programs, and FCM
revenues (in the states that have them). EPA found that qualifying individual programs were
not incremental to the EERS target, so each state with reported savings has either EERS savings
or dedicated sources of energy efficiency program funding.39 EPA did not apply an adjustment
for net versus gross savings; this issue is discussed further in "Important Sources of Uncertainty
in the Analysis" below.
For each policy category, EPA estimated annual first-year electricity savings (i.e., savings
achieved in a given year from programs implemented during that year) and cumulative savings
from energy efficiency measures implemented in the current year and past years. EPA
calculated cumulative savings using state-specific measure lifetimes (see Table 3 below) and
assumed no decay of savings over the life of the measures.40 EPA used a default lifetime of
10 years where state-specific assumptions were not available. EPA did not estimate first-year
savings beyond the requirements of each state's policy period, except for a limited set of states
whose policy indicated a continuation of savings beyond the policy period. For the majority of
states, however, the forecast reverts to the AEO 2013 reference case-based forecast after the
energy efficiency policy period ends.
Table 3. Measure Lifetime by State41
State
Measure Lifetime (years)
Arizona
9.8
California
9.1
Connecticut
9.6
Hawaii
9.2
Massachusetts
11.6
Minnesota
13.8
New Mexico
8.9
Oregon
11.2
37 ACEEE (2012).
38 DSIRE (2013).
39 For more information, see http://www.epa.gov/statelocalclimate/state/statepolicies.html.
40 An alternative, conservative assumption of linear degradation of measure lifetime, is available to
states possessing the appropriate data.
41 ACEEE (2014).
18
-------
State
Measure Lifetime (years)
Rhode Island
11.1
Utah
11.3
Vermont
11.0
Wisconsin
11.4
Default
10.6
Energy Efficiency Resource Standards
An EERS is a policy that sets targets for energy savings over a specified timeframe from end-use
energy efficiency programs operated by utilities or other program administrators. States
typically specify annual first-year or cumulative targets as percentages of electricity sales or as
absolute energy savings. They use different bases for specifying EERS goals: Some states specify
goals based on sales from investor-owned utilities, while others have mandated savings (i.e.,
MWh) targets based on total sales or a subset of total sales.
EPA estimated energy savings for each state using formulas specific to the state's EERS, as
shown below. EPA identified the appropriate sales basis for each state and, if the basis was not
total sales, EPA used 2012 utility-level sales data from EIA-42 and AEO 2013-based growth rates
to develop baseline forecasts of sales of affected utilities (see Table 1). For most states, EPA
assumes full achievement of EERS targets for all years during the compliance period. However,
there are some states for which EPA does not assume full achievement of EERS targets in all
years because of the way that the programs are designed. One example is an EERS policy that
includes cost/rate caps or other design features (e.g., counting savings from building energy
codes or historical energy efficiency programs) that may not lead to incremental energy savings
relative to AEO 2013,43 or are otherwise inconsistent with the EERS targets.44 In addition,
savings were not estimated for purely voluntary EERS.45
The general formulas used to estimate annual first-year and cumulative energy savings for each
year (t) were:
1. EERS With Annual First-Year Energy Efficiency Savings Targets Specified in Percent
Terms
A(t) = r(t) * Z(t-l)
42 U.S. EIA (2014).
43 Building energy codes are already incorporated in the AEO 2013 forecast, so any associated savings
from those existing building codes already assumed in AEO 2013 would not be incremental to the AEO
forecast, and thus are removed from the applicable state's EERS target. However, if the state/utility
program action motivates new code adoption and implementation, those codes are incremental to any
savings already embedded in the AEO forecast.
44 For more information, see the individual state summary sheets at
http://www.epa.gov/statelocalclimate/state/statepolicies.html.
45 EPA acknowledges that it is possible that some savings may result from a voluntary EERS, so this is a
methodological conservatism.
19
-------
c(t) = A( t) + >A(t-l) + ... + A(t-L+1)
/(t) = C(t)-£(t)
Z(t) = B(t)-/(t)
where:
r is the annual first-year percent savings target,
A is the annual first-year energy savings,
L is the measure lifetime,
B is the baseline sales of utilities affected by these specific policies,
C is the cumulative energy savings,
£ is the cumulative savings embedded in the AEO 2013 forecast,
I is the cumulative savings incremental to the AEO 2013 forecast, and
Z is the adjusted sales after application of cumulative incremental savings.
2. EERS With Annual First-Year Energy Efficiency Savings Targets Specified in Absolute
Terms
C(t) = A(t) + ^(t-1) + ... + A(t-L+1)
/(t) = C(t)-£(t)
Z(t) = B(t)-/(t)
where:
A is the annual first-year energy savings target,
L is the measure lifetime,
B is the baseline sales of utilities affected by these specific policies,
C is the cumulative energy savings,
£ is the cumulative savings embedded in the AEO 2013 forecast,
I is the cumulative savings incremental to the AEO 2013 forecast, and
Z is the adjusted sales after application of cumulative incremental savings.
3. EERS With Cumulative Energy Efficiency Savings Targets Specified in Percent Terms
A(t) = C(t) - C(t-l) + A(t-L)
If r(t) is available,
C(t) = r(t) * B(t)
/(t) = C(t)-£(t)
Z(t) = B(t)-/(t)
If r(t) is not available,
Z(t) is calculated by interpolation
/(t) = B(t)-Z(t)
C(t) = /(t) + £(t)
20
-------
where:
r is the cumulative percent savings target,
A is the annual first-year energy savings,
L is the measure lifetime,
B is the baseline sales of utilities affected by these specific policies,
C is the cumulative energy savings,
E is the cumulative savings embedded in the AEO 2013 forecast,
I is the cumulative savings incremental to the AEO 2013 forecast, and
Z is the adjusted sales after application of cumulative incremental savings.
4. EERS With Cumulative Energy Efficiency Savings Targets Specified in Absolute
Terms
A(t) = C(t) - C(t-l) + A(t-L)
If C(t) is available,
/(t) = C(t)-£(t)
Z(t) = B(t)-/(t)
If C(t) is not available,
Z(t) is calculated by interpolation
/(t) = B(t) - Z(t)
C(t) = /(t) + £(t)
where:
C is the cumulative energy savings target,
A is the annual first-year energy savings,
L is the measure lifetime,
B is the baseline sales of utilities affected by these specific policies,
E is the cumulative savings embedded in the AEO 2013 forecast,
I is the cumulative savings incremental to the AEO 2013 forecast, and
Z is the adjusted sales after application of cumulative incremental savings.
Some special considerations that warranted adjustments to the general formulas were as
follows:
1. RPS that define energy efficiency as a qualifying resource: The states of Nevada and
North Carolina have RPS that treat energy efficiency as a qualifying resource, subject to
a quantitative limit. The National Energy Modeling System (NEMS), which is used to
produce the AEO, does not currently have the capability to evaluate tradeoffs between
energy efficiency and renewable energy in cases where both are eligible RPS resources;
so, it relies on renewable energy to meet RPS requirements. For RPS policies explicitly
included in AEO 2013, no incremental energy savings were estimated.
21
-------
2. Compliance type and cost/rate caps: Several states have EERS that use cost-
containment provisions or other design features that may constrain the ability of energy
efficiency program administrators to meet the EERS targets with incremental savings
relative to the AEO. EPA identified five states with such design features—Arizona,
Illinois, Minnesota, Ohio,46 and Texas—and relied on available state-specific academic
reports,47 integrated resource plans,48 and other studies49 to make downward
adjustments to the nominal EERS targets to reflect these design features, as
appropriate.50
3. "All cost-effective energy efficiency" targets: Six states—Connecticut, Maine,
Massachusetts, Rhode Island, Vermont, and Washington—require utilities (or other
energy efficiency program administrators) to implement all cost-effective energy
efficiency. In states with an "all cost-effective energy efficiency" requirement and EERS
targets, EPA used the EERS targets until the policy sunset date and then assumed first-
year savings equivalent to the last policy year, going forward.51 In states with an "all
cost-effective energy efficiency" target without an EERS target through 2020, EPA
estimated savings based on utility plans52 and energy efficiency resource potential
studies.53
4. State legislature or PUC disapproval of energy efficiency program budgets necessary
to meet EERS targets: Two states—Florida and Wisconsin—did not approve requests for
the energy efficiency program budget increases necessary to meet growing EERS
targets, opting instead to maintain current energy efficiency program offerings. In these
states, EPA reduced the EERS nominal targets to the levels achieved with the approved
energy efficiency program budgets.54
Energy Efficiency Program Funding
In states without an EERS policy, but with one or more sources of energy efficiency funding, EPA
developed an approach for estimating the associated savings. The sources of funding evaluated
by EPA include public benefits funds (PBFs), funding from the proceeds of Regional Greenhouse
Gas Initiative (RGGI) allowance auctions, and funding from Forward Capacity Market (FCM)
payments. Data for these energy efficiency programs are mainly available in terms of program
46 EPA recognizes that Ohio's authorizing law was altered in June 2014 by SB 310, which froze cumulative
savings at 4.2 percent (the value expected to be achieved by the end of 2014) for 2015 and 2016. EPA
revised its analysis and forecast to reflect this change.
47 Satchwell (2011).
48 Ameren Illinois (2010), ComEd (2010).
49 Good Company Associates (2010).
50 For more information, see http://www.epa.gov/statelocalclimate/state/statepolicies.html.
51 EPA acknowledges that this is likely a methodological conservatism.
52 Connecticut DEEP-BETP (2012), Mass Save (2012), EMT (2010), EMT (2012), National Grid (2008),
EERMC (2010), VEIC (2009).
53 NWPCC (2010).
54 For more information, see http://www.epa.gov/statelocalclimate/state/statepolicies.html.
22
-------
administrator expenditures (i.e., the cost to the utility of administering energy efficiency
programs, exclusive of customer costs), so EPA calculated savings based on estimates of energy
savings per program dollar spent. For each state with qualifying programs, EPA obtained
information on annual program funding from state55 or utility publications,56 and from state
stakeholder feedback during the public review period, and projected funding for each future
year as equal to the funding for the year for which the latest information is available. The
funding information consists of either actual or committed expenditures, depending on the
data source. Estimates of levelized costs of saved energy (LCSE) were available for some states
from ACEEE (2009a). These are presented in Table 4. The ACEEE report presents costs of saved
energy as reported by programs, except in cases where the methods used by program
administrators to estimate the LCSE were different from ACEEE's standard approach. In such
cases, ACEEE calculates LCSE as:
LCSE = (F*CRF) / A
CRF = (d*(l+d)L) / ((l+d)L-l)
where:
A is the annual first-year energy savings,
F is the annual program funding,
CRF is the Capital Recovery Factor,
L is the measure lifetime, and
d is the discount rate.
ACEEE uses a real discount rate of 5 percent to calculate the Capitol Recovery Factor, and
estimates that the average LCSE across the states included in the report is $0.025/kilowatt hour
(kWh). To apply ACEEE's LCSE estimates in a manner that is consistent with the methodology by
which they were calculated, this analysis also used a discount rate of 5 percent.57 The average
LCSE of $0.025/kWh was used as the default LCSE where state-specific estimates were not
available. In order to adjust for the effects of inflation, EPA converted the dollar values
employed in the ACEEE analysis (reported in 2007$) to 2011$, which is the price metric used
throughout the AEO 2013 analysis. Implicit price deflators for gross domestic product were
assumed as the measure for conversion.58 EPA did not assume a decay of savings during the
measure life, so savings for each year are equal to the lifetime savings averaged over the
measure lifetime.
55 DCSEU (2012), DSEU (2013), NHEU (2012), NJ CEP (2013).
56 MDU (2012), MECA (2011), Northwestern Energy (2012).
57 A 5-percent discount rate is also the average of the two rates (i.e., 3 percent and 7 percent) that EPA
currently uses when performing economic analysis as a part of its rule development; for more
information, see http://yosemite.epa.gov/ee/epa/eed.nsf/webpages/Guidelines.html.
58 U.S. BEA (2013), Table 1.1.9.
23
-------
Table 4. Levelized Cost by State59
Levelized Cost of Saved
Levelized Cost of Saved
State
Energy60
(2007$/kWh)
Energy
(2011$/kWh)
California
$0,029
$0,031
Connecticut
$0,028
$0,030
Iowa
$0,017
$0,018
Massachusetts
$0,031
$0,033
Minnesota
$0,021
$0,022
Nevada
$0,019
$0,020
New Jersey
$0,026
$0,028
New Mexico
$0,033
$0,035
New York
$0,019
$0,020
Oregon
$0,016
$0,017
Rhode Island
$0,030
$0,032
Texas
$0,017
$0,018
Vermont
$0,027
$0,029
Wisconsin
$0,033
$0,035
Default (simple average)
$0,025
$0,027
EPA estimated energy savings from ratepayer-funded programs in each year (t) using the
following formulas:
CRF = {d*{l+d)L) / {{l+d)L-l)
A(t) = (F(t) * CRF) / LCSE(t)
C(t) = A( t) + A( t-1) + ... + A(t-L+1)
where:
CRF is the Capital Recovery Factor,
L is the measure lifetime,
d is the discount rate,
A is the annual first-year energy savings,
F is the annual program funding,
LCSE is the levelized cost of saved energy in 2011$, and
C is the cumulative energy savings.
59 ACEEE (2009a), Table 1.
60 LCSE is based on program administrator costs, not on total resource costs (which include the costs to
participating utility customers).
24
-------
For this analysis, EPA did not estimate the magnitude of savings from energy efficiency
programs funded by dedicated funding sources (i.e., RGGI and FCMs) separately, but instead
incorporated their funds in the Energy Efficiency Program Funding category. This decision was
motivated by the availability of state-level program budget information data, which aggregated
the funding sources.
Methodology for Generating State-Adjusted Forecast That Reflects
Incremental Energy Savings
EPA estimated energy savings that are incremental to the reference case (AEO 2013) by
subtracting cumulative savings embedded in AEO 2013 from total savings from EERS, programs
funded by PBFs, and other program funding sources (e.g., RGGI and FCM):
/(t) = C(t)-£(t)
where:
C is the cumulative energy savings,
E is the cumulative savings embedded in the AEO 2013 forecast, and
I is the cumulative savings incremental to the AEO 2013 forecast.
The state-adjusted electricity sales forecast includes the impact of energy
efficiency savings that are incremental to the BAU reference case. State-level
adjusted sales (Z) are calculated as:
Z(t) = B(t)-/(t)
where:
B is the baseline total sales, and
I is the cumulative savings incremental to the AEO 2013 forecast.
Methodology for Estimating Renewable Energy Sales from Renewable
Portfolio Standards Beyond What Is Captured in AEO 2013
The AEO 2013 reference case incorporates RPS policies or substantively similar laws in place at
the time of forecast development. In general, AEO assumes that utilities will meet the RPS
targets; however, where states have explicitly limited state funding for RPS implementation,
AEO assumes that utilities comply with RPS requirements only to the extent that state funding
allows, as described in the AEO assumptions documents.
This analysis maintains consistency with these limiting assumptions. Because there are only two
states that are not captured in AEO 2013, EPA included the RPS policies for these two states,
Hawaii and Minnesota. The RPS-related energy production in Hawaii is considered incremental
to the AEO forecast because the state is excluded from AEO 2013 modeling.61 Minnesota was
61 U.S. EIA (2013), "NEMS provides electricity market projections for the contiguous lower 48 states
only," p. 13.
25
-------
added because its RPS target was changed62 after the analysis underlying the AEO assumptions
was performed. The expected increase of 1.5 percent for certain utility types after 2020 counts
as incremental to AEO.
EIA did not identify funding limitations for either state, and EPA assumed that their full RPS
targets would be achieved. Table 5 presents final RPS targets used in this analysis for the two
states for which EPA identified updated RPS requirements.
Because the RPS targets for Hawaii were only available for 2015, 2020, and 2030, EPA
estimated sales in intervening years by interpolation. Similarly, the Minnesota targets for 2016,
2020, and 2025 were used to interpolate expected sales levels for all years.
RPS requirements were frozen in percent terms for the years after the RPS policy period.
Table 5. Renewable Portfolio Standard Assumptions Made in This Analysis
State
State
t
RPS Generation (in
housand GWh)
2015
2020
2030
Flawaii63
1.39
2.16
3.43
Minnesota
0.00
0.64
0.65
Important Sources of Uncertainty in the Analysis
In conducting this analysis, EPA used the best available information and adopted assumptions
intended to reduce the likelihood of overstating the impacts of the states' EE/RE policies.
For this analysis, EPA is highlighting four sources of uncertainty to keep in mind when utilizing
these estimates and employing similar methods:
1. The impacts of state energy efficiency policies embedded in the AEO reference case
2. Approval of sufficient energy efficiency program budgets necessary to meet the EERS
targets
3. Variations in state approaches for evaluating and reporting energy efficiency savings
4. Application of regional growth rates to states within that region
62 DSIRE (2013a). Law HF 279 was enacted on May 23, 2013. The 2020 target for Xcel Energy was
increased from 30 percent to 31.5 percent. The target for non-Xcel public utilities was increased from
20 percent to 21.5 percent in 2020 and from 25 percent to 26.5 percent in 2025.
63 NEMS provides electricity market projections for the contiguous lower 48 states only, so the impacts
of Flawaii's RPS are not included in AEO 2013.
26
-------
As discussed earlier, the AEO reference case likely includes the impacts of some programs that
are not explicitly identified in the AEO documentation. Estimating the impacts of existing
energy efficiency policies at the national level, and then applying these national savings to
individual states, requires significant simplifying assumptions about the degree to which these
impacts are embedded in electricity sales projections and the associated magnitude of double
counting. While intrinsic uncertainty exists (and as expressed by one peer reviewer who
questioned whether the concept of calculated embedded savings from the AEO is even
possible), EPA believes that its assumptions are reasonable in light of available data. It also
should be noted that some experts have questioned the accuracy and quality of EIA-861
cumulative or lifetime efficiency data, and that the data has its obvious limitations in that it is
based on data self-reported by utilities.64
Another source of uncertainty relates to the approval of energy efficiency program budgets
necessary to meet the adopted targets. The energy efficiency policy that drives the core results
of this analysis—EERS—depends on the approval of energy efficiency program budgets
necessary to meet the targets, often by PUCs. Several states' EERS legislation includes explicit
cost or rate impact caps that may constrain the ability of energy efficiency program
administrators to meet the nominal EERS targets, and EPA attempts to account for this design
feature in its analysis. However, even in states without specific cost or rate impact caps, PUCs
generally have authority over energy efficiency program budgets and, as the EERS targets
increase in stringency (necessitating larger energy efficiency program budgets), there is
uncertainty over whether PUCs will continue to approve the budgets necessary to achieve the
EERS targets. While recent reports have documented steadily increasing energy efficiency
program budgets,65 and generally good progress with states reporting achievement of EERS
targets,66 this will be an issue for states to track in the future as EERS targets increase. A similar
source of uncertainty resides with state legislatures, which could expand or constrict EERS
across the years covered by this analysis.
A third source of uncertainty in EPA's analysis is the energy savings definitions that states use
when calculating and reporting program impacts. In some states, energy savings are evaluated
and reported to PUCs as gross savings, that is, inclusive of savings attributed to an efficiency
program that would have occurred even in the absence of the program (i.e., program savings
not attributable to a specific program intervention). Other states require the reporting of net
savings, which adjust gross savings by accounting for so-called free-riders, or customers who
receive program rebates even though they would have invested in the efficient equipment
without the program. Net savings also account for free-drivers, or induced market effects, and
other considerations. This difference in how energy savings are defined and measured
complicates efforts to make cross-state comparisons. The degree of uncertainty this conveys to
EPA's analysis is not precisely known, but ACEEE uses a negative adjustment factor of
10 percent applied to gross savings to reconcile the two values for their annual State Scorecard.
64 Per review by Maggie Molina, ACEEE.
65IEE (2012).
66 ACEEE (2011).
27
-------
Such an adjustment is not used in this analysis, although states can make such adjustments in
their own analyses if they so desire. A recent survey indicated that approximately two-thirds of
the states report gross savings and one-third report net savings.67
During the peer review process, another source of uncertainty was raised related to the use of
AEO 2013 average annual growth rates from each EMM region to represent the growth rates of
states within those EMM regions. This is the approach used in this analysis for preparing state-
level baseline sales. While it is recognized that not all states within a given EMM region will
experience similar growth patterns simply because they are grouped geographically, alternative
sources that were identified did not offer a reliable basis for seeking to refine state-specific
growth rates. EPA finds that this assumption is appropriate for this illustrative analysis, but
encourages states to apply their own state-specific assumptions to analyses that they conduct.
67 ACEEE (2012b).
28
-------
References
Ameren Illinois (2010). Electric and Gas Energy Efficiency and Demand-Response Plan,
September 2010.
American Council for an Energy-Efficient Economy (ACEEE) (2009a). Saving Energy Cost-
Effectively: A National Review of the Cost of Energy Saved Through Utility-Sector Energy
Efficiency Programs, September 2009. Available online at
http://www.aceee.org/sites/default/files/publications/researchreports/IJ092.pdf.
ACEEE (2011). Energy Efficiency Resource Standards: A Progress Report on State Experience,
June 2011. Available online at http://www.aceee.org/research-report/ull2.
ACEEE (2012a). The 2012 State Energy Efficiency Scorecard, October 2012. Available online at
http://aceee.org/files/pdf/fact-sheet/el2c-es.pdf.
ACEEE (2012b). A National Survey of State Policies and Practices for the Evaluation of
Ratepayer-Funded Energy Efficiency Programs, February 2012. Available online at
http://www.aceee.org/research-report/ul22.
ACEEE (2014). The Best Value for America's Energy Dollar: A National Review of Cost of Utility
Energy Efficiency Programs, March 2014, Report No. U1402.
Commonwealth Edison Company (ComEd) (2010). Commonwealth Edison Company's 2011-
2013 Energy Efficiency and Demand Response Plan, October 2010. Available online at
http://librarv.ceel.org/content/commonwealth-edison-company%E2%80%99s-2011-
2013-energy-efficiency-and-demand-response-plan.
Connecticut Department of Energy and Environmental Protection Bureau of Energy and
Technology (DEEP-BETP) (2012). 2012 Integrated Resource Plan Final Approval, June
2012. Available online at
http://www.dpuc.state.ct.us/DEEPEnergy.nsf/c6c6d525f7cddll68525797d0047c5bf/cb
827blffa58b2fd85257ald0060c374/$FILE/CTIRP%202012 Appendix%20C%20(Energy%
20Efficiency).pdf.
Database of State Incentives for Renewables and Efficiency (DSIRE) (2013). Rules, Regulations &
Policies for Energy Efficiency. Available online at http://www.dsireusa.org/ (accessed
June 2012).
29
-------
Delaware Sustainable Energy Utility (DSEU) (2013). Monthly Budget Report, February 2013.
Available online at https://imageservll.team-
logic.com/mediaLibrary/191/Feb 2013 Monthly Bond SREC Reports.pdf.
District of Columbia Sustainable Energy Utility (DCSEU) (2012). 2012 Annual Report. Available
online at https://www.dcseu.com/about-dcseu.
Efficiency Maine Trust (EMT) (2010). Triennial Plan of the Efficiency Maine Trust 2011-2013,
April 2010. Available online at www.efficiencymaine.com/docs/EMT Final Tri Plan-
l.pdf.
EMT (2012). Triennial Plan of the Efficiency Maine Trust 2014-2016, October 2012. Available
online at http://www.efficiencymaine.com/docs/TriPlan2-ll-26-2012.pdf.
Good Company Associates (2010). Substantive Changes Made to Rules Governing Texas Utility
Run Energy Efficiency Programs, August 2010. Available online at
http://goodcompanyassociates.com/.
Institute for Electric Efficiency (IEE) (2012). Summary of Ratepayer-Funded Electric Efficiency
Impacts, Budgets, and Expenditures, January 2012. Available online at
http://www.edisonfoundation.net/iee/Documents/IEE CEE2011 FINAL update.pdf.
Mass Save (2012). Massachusetts Joint Statewide Three-Year Electric and Gas Energy Efficiency
Plan, November 2012. Available online at http://www.mass.gov/eea/docs/doer/energy-
efficiency/statewide-electric-and-gas-three-year-plan.pdf.
Montana-Dakota Utilities Co. (MDU) (2012). 2011 Electric Universal System Benefits Report,
February 2012. Available online at
http://leg.mt.gov/content/Committees/lnterim/2011-2012/Energy-and-
Telecommunications/Meeting-Documents/May-2012/2011USBMDU.pdf.
Montana Electric Cooperatives' Association (MECA) (2011). 2011 Final Year-End Universal
System Benefits Pool Report, December 2011. Available online at
http://leg.mt.gov/content/Committees/lnterim/2011-2012/Energy-and-
Telecommunications/Meeting-Documents/May-2012/2011USBCoops.pdf.
National Grid (2008). 2009-2011 Energy Efficiency Procurement Plan, September 2008.
Available online at http://www.ripuc.org/eventsactions/docket/3931-NGrid-
ComplianceProcurePlan(9-3-08).pdf.
30
-------
New Hampshire Electric Utilities (NHEU) (2012). 2013-2014 CORE Energy Efficiency Programs,
September 2012. Available online at
http://www.puc.nh.gov/Regulatory/Docketbk/2012/12-262/TRANSCRIPTS-
OFFICIAL%20EXHIBITS-CLERKS%20REPQRT/12-262%202012-12-
21%20EXH%202%20PSNH%20MERGED%20ATTACHMENT%20A%20AND%20B%20TQ%2
0SETTLEMENT%20AND%20UPDATED%20TO%20INCLUDE%20LATEST%20CORRECTIQNS.
PDF.
New Jersey Clean Energy Program (NJ CEP) (2013). Monthly Report of Progress Toward Goals,
January 2013. Available online at
http://www.nicleanenergv.com/files/file/Librarv/NJCEP%20PTG%20Month%20Report%
20-%20Jan%202013.pdf.
Northwest Power and Conservation Council (NWPCC) (2010). 6th Plan Conservation Target
Calculator, January 2010. Available online at
http://www.nwcouncil.Org/energy/powerplan/6/supplvcurves/l937/UtilitvTargetCalc v
2.0 6thPlan.xls.
Northwestern Energy (2012). Universal System Benefits Activities, 2011 Annual Report, March
2012. Available online at http://www.northwesternenergy.com/.
Rhode Island Energy Efficiency and Resource Management Council (EERMC) (2010). Energy
Efficiency Savings Targets, September 2010. Available online at
http://www.ripuc.org/eventsactions/docket/4202-EERMC-EST-Filing(9-l-10).pdf.
Satchwell, Andrew, et al. (2011). Carrots and Sticks: A Comprehensive Business Model for the
Successful Achievement of Energy Efficiency Resource Standards, Lawrence Berkeley
National Laboratory, LBNL-4399E, March 2011. Available online at
http://emp.lbl.gov/sites/all/files/lbnl-4399e O.pdf.
Synapse Energy Economics, Inc. State Energy Efficiency in the AEO Electricity Forecasts, October
31, 2012. Available online at http://www.syna pse-
energv.eom/Downloads/SvnapseReport.2013-10.0.EE-in-AEQ.12-094.pdf.
Synapse Energy Economics, Inc. State Energy Efficiency Embedded in Annual Energy Outlook
Forecasts, 2013 Update, May 20, 2014.
U.S. Bureau of Economic Analysis (BEA) (2013). Implicit Price Deflators for Gross Domestic
Product, May 2013. Available online at
http://www.bea.gov/iTable/iTable.cfm?ReqlD=13&step=l#reqid=13&step=3&isuri=l&l
305=2012&1306=A&1310=X&1311=0&1301=I&1303=13&1304=2005.
31
-------
U.S. Energy Information Administration (EIA) (2010). Annual Electric Power Industry Database
(Form EIA-861 Database) 2009. Available online at
http://www.eia.doe.gov/cneaf/electricity/page/eia861.html.
U.S. EIA (2011). Annual Electric Power Industry Database (Form EIA-861 Database) 2010.
Available online at http://www.eia.doe.gov/cneaf/electricity/page/eia861.html.
U.S. EIA (2012). Annual Electric Power Industry Database (Form EIA-861 Database) 2011.
Available online at http://www.eia.doe.gov/cneaf/electricity/page/eia861.html.
U.S. EIA (2013a). Annual Energy Outlook 2013, May 2013. Available online at
http://www.eia.gov/forecasts/aeo/pdf/0383(2013).pdf.
U.S. EIA (2013b). Detailed Sales and Revenue Data by State, Monthly Back to 1990 (Form EIA-
826). Available online at
http://www.eia.gov/electricity/data/eia826/xls/sales revenue.xls (downloaded June
2013).
U.S. EIA (2013c). Assumptions to the Annual Energy Outlook 2013, May 2013. Available online at
http://www.eia.gov/forecasts/aeo/assumptions/pdf/0554(2013).pdf.
U.S. EIA (2014). Monthly Electric Utility Sales and Revenue Report With State Distributions (Form
EIA-826). Available online at http://www.eia.gov/electricity/data/eia826 (downloaded
August 2014).
U.S. Environmental Protection Agency (EPA) (2010). Documentation for EPA Base Case v.4.10
Using the Integrated Planning Model, August 2010. Available online at
http://www.epa.gov/airmarkets/programs/ipm/BaseCasev410.html.
Vermont Energy Investment Corporation (VEIC) (2009). Efficiency Vermont: Annual Plan 2010-
2011, November 2009. Available online at https://www.efficiencyvermont.com/.
32
------- |