How to Submit Effective Comments on NEEDS and IPM

Summary

This document is intended to provide guidance on submitting clear, concise, and impactful comments on
NEEDS (National Electric Energy Data System), other inputs to the Integrated Planning Model (IPM), or
outputs from IPM.

While the EPA always welcomes comments on its data, be sure to submit comments to the appropriate
docket by the due date if the comment is regarding modeling in support of a NODA or rulemaking.

When submitting a comment, please remember these following principles:

1) Be as specific as possible regarding what unit, plant, state, regulation, or law you are referring
to. This helps ensure the EPA understands to what you are referring.

a. For comments on specific units, provide the plant name and NEEDS Unique ID

b. For comments on specific plants, provide the plant name and ORIS code

c. For laws and regulations, refer to them by name (not just "laws in [state name]").

Provide citations if possible.

2) Identify if you are referring to a model INPUT or an OUTPUT. When commenting on model
outputs, it is helpful to connect your comments, if possible, about the model's behavior to
inputs and constraints, as inputs and constraints are what EPA can change in the model.

3) Identify the specific data parameter of concern, including the current model input/output
value, and the value you believe it should be. This again helps ensure the EPA understands
what you want changed.

4) Provide strong documentation with clear citations and web links. Strong documentation
includes: operating permits; documentation of firm financial commitments; evidence
(photographs, news articles, press releases) that construction on a new unit or retrofit has
started or that a plant has been retired; copies of state rule, consent decrees, NSR settlements;
designation as a reliability must run (RMR) unit. Weak documentation includes: utility integrated
resource plans (IRPs) or announcements that reflect current intentions but not firm
commitments; and references to historic data/operation without a clearly supported reason
why that historic behavior must continue in the future.

5) Remember that there is a time-lag between when you submit a comment and when it is
reflected in EPA's modeling. Modelling that is released within a few months after your
comment is submitted may not yet reflect any changes related to your comment, due to the
lead time required for conducting such modeling.

6) Remember that some comments may not be "ripe" to support changing model inputs. For
example, a utility might announce its intention to retire a unit in five years, but may not yet
have the necessary approvals. In these cases, EPA welcomes comments about these units and
will monitor the situation and update the model inputs when the retirement is firmer.

7) Check the status of prior comments. EPA will be releasing quarterly a list of the comments it has
received, including their review status and any resulting changes.

Finally, while EPA accepts comments in any format, it is recommended that you use the comment
spreadsheet we've developed. This will help ensure your comments are as clear as possible and remind
you to follow the above principles that facilitate our ability to factor your input into our work going
forward.

-------
Introduction

Commenters play an essential role in our modeling development, providing interpretations of the raw
data and ensuring that the most appropriate data is being used by the Agency. This guidance document
is focused on comments about the EGU sector, and is oriented toward commenters that may be focused
on the behavior of a single unit or a small set of units. This guidance describes information about units
(that may be analyzed directly and/or used as input to models) as well as estimates about potential
emissions from units based on modeling projections and other analytic approaches.

This document is intended to help commenters provide complete information about unit configuration
and behavior that can be understood by EPA.

This document is organized as follows:

•	What data sources are used as inputs to EPA's modeling with IPM?

•	What is contained in the NEEDS database?

•	What are the data outputs of IPM?

•	What information should I include in my comment?

2

-------
V lis.I o .1 . mm . r ,,k " r,I "!ii I:s i 1 i i' ! Imi v ill> II ii ii?

EPA uses several major sources of data to provide inputs to the Integrated Planning Model (IPM). The
inputs include information describing the configuration of each unit in the existing EGU fleet, such as the
size of the unit, type of boiler, and types of fuel it is capable of utilizing. These are distinct from the
outputs, which are modeled projections of how the units could operate in the future including dispatch,
retrofit, retirement, fuel type selected for future-year operation, and construction of new capacity.

It is important to identify if the data that concerns you is an input or an output so that you can comment
effectively. Inputs to IPM are determined by EPA and are based on real, reported data. This data is
constantly changing and, provided with sufficient documentation, the EPA can directly modify the inputs
to the model. Outputs from IPM are based upon solving mathematical equations that use input data.
Comments on outputs (e.g. "This result doesn't match my expectation or other projections") tends to be
less actionable, because directly modifying the outputs would prompt a more complicated consideration
of whether/how such modifications necessitate revisiting other components of the modeled
systemwide solution. It is our preferred approach to adjust model inputs, assumptions, and/or
constraints to produce new outputs that still represent a least-cost (economic) projection; hence our
recommendation that commenters who are concerned with a model output identify justified changes to
relevant input data, assumptions, or constraints likely to drive the output of concern. In any
circumstance, comments on the model outputs can be useful if they flag an unexpected result.
Comments on outputs are most effective, though, when they relate back to the inputs to the model.

Below is a list of the major sources of data the EPA uses to form the inputs to IPM. You can also see the
documentation for IPM for more detailed information about the data inputs and how they are used.

Data inputs:

EPA Air Markets Program Data (AMPD). Continuous emissions monitoring and reporting of sulfur
dioxide (S02), carbon dioxide (C02), and nitrogen oxide (NOx) follow the monitoring regulations in Part
75 of Volume 40 of the Code of Federal Regulations (CFR) for sources included in several programs
administered by EPA. Data for individual units and for groups of units can be found in the Air Markets
Program Database available at https://ampd.epa.gov/ampd/

NEEDS. The National Electric Energy Data System (NEEDS) is the database of existing and planned-
committed units. This includes unit-level information about units that are currently operational and new
units that are firmly anticipated to be operational in the future, and have either broken ground (initiated
construction) or secured financing (these units are termed "planned-committed"). The NEEDS database
is an input for IPM. See the NEEDS User's Guide for additional information.

Coal availability. For units that operate with this fuel, the specific attributes of the fuel are instrumental
in the operation and emissions of the unit. Coal supply, coal quality, the availability of specific types of
coals to individual coal fired generating units, the cost of the coals, the coal transportation network, and
the transportation cost are key components of the endogenous coal market modeling framework within
IPM. Details can be found in chapter 7 "Coal" of the 1PM documentation. For purposes of this document,
we will direct the reader to specific pieces of this extensive documentation.

For units burning coal, we also suggest first reviewing the coal transportation matrix. Table 7-20 in the
IPM documentation provides a crosswalk between plant ORIS code and the coal supply region
(presented both as a code and with the full description). We suggest assessing whether these coal

-------
supply regions are reasonable sources for your unit. That is, if you are purchasing coal from a region that
is not listed, please provide that information to EPA with any information about coal quality, production
and transportation costs that you would like to provide. Alternatively, assess whether source regions
cannot be accessed because of physical transportation limits. Because the cost of delivery is an
important part of the overall cost, assessment of the cost of transportation is important.

Next, we suggest reviewing the coal supply curves. Table 7-6 in the IPM documentation shows the
various coal supply regions and the types/grades of coals that are available in each region. More-
detailed cost curves found in table 7-21 provides information on the cost to extract, the coal quality
including sulfur and heat content, and reserves for mines (i.e., availability) in each coal supply region.

Gas supply matrix and costs IPM Platform 6 uses natural gas supply curves rather than endogenously
modeling the entire US and Canada natural gas system. These curves are derived through a least cost
linear program model solving for the least cost provision of gas to meet total demand. They can be
viewed in Chapter 8 of the IPM platform 6 documentation. These curves provide a cost per production
level for a specific model run year, and are different for each model run year in IPM's time horizon.
These curves were developed using the Gas Market Module (GMM), as described in chapter 8 of the
documentation and they reflect a Henry Hub Spot Price and total gas available to the power sector. The
curves are accompanied by a table of basis differentials which have region specific cost adders reflecting
the transportation costs.

Commenters are welcome to provide comments on the natural gas supply curves and differentials
(including the value and slopes of the curves for any of the years, or how those slopes change among
the years), the cost adders, or underlying GMM analysis.

V i. i ;! J ! ¦ c\is .ii!!i Iiii M i il1 il1 ,1 .ba " «

One of the primary input data sources is EPA's NEEDS database, which can be viewed as an Excel
workbook. NEEDS contains a unit-level description of the EGU fleet that is used by IPM. So, this is an
important data set to verify, and a first source to examine if "output" modeling estimates are not as
expected.

NEEDS contains data on unit characteristics, including:

• Attributes like the name and location of the unit;

• Physical parameters of each unit including unit capacity, heat rate, type of boiler

• Unit online year and retirement year (if it has a firm retirement date)

• Emissions control information, including type of controls and when they were retrofit (or will be
retrofit if firmly known).

• If the unit has been or is firmly scheduled to convert from coal to gas and the year of the
conversion.

• NOx Emissions Rates, reflecting possible combustion control upgrades and operation of post-
combustion controls at units.

o For NEEDS version 6 variations, NOx rates are based on 2017 historical values,
o NOx rates are calculated off historical data and reflect the fuel mix for that year and
burn at the unit. NEEDS represents up to four scenario NOx rates based on historical

-------
data to capture seasonal and existing control variability. These rates are constant and
do not change independent of fuel mix assumed in the model. If the unit undertakes a
post-combustion control retrofit or a coal-to-gas retrofit, then these rates would change
in the model projections,
o NEEDS has prospective NOx rates for units that can upgrade to state of the art

combustion controls and if post-combustion controls are or are not operating (if the unit
has operated the controls seasonally in the recent past),
o The model recalculates the NOx rates if the unit is projected to add a new post-
combustion control.

• S02 Permit Rates. This is the permitted emission level, NOT the emissions rate that the unit will
necessarily operate at in the model. The operational S02 emissions rate is an output of IPM, not
an input, since the modeled S02 rate is highly dependent on fuel choice. Model-projected S02
emission rates for a given unit may not exceed the applicable permit rate shown in NEEDS for
that unit and are often significantly lower than the permit rate.

See the NEEDS User Guide for a list and description of the data fields in NEEDS.

A description of how the values for these various characteristics are determined can be found in the 1PM

Documentation, Chapter 4.

What are the Data Outputs of IPM?

Unexpected unit behavior, state level results, or national results are sometimes the first clue that input
data may need to be updated. EPA has a document called the "Guide to IPM Output Files" that describes
useful output files including:

• System Summary Report (SSR) File: It contains system-wide power sector results for the
lower continental U.S. for each run year. It reports forecasted generation, capacity,
capacity additions, capacity factors, production costs, emissions, fuel consumption &
cost, and allowance prices by model run year. Disaggregation of system-wide data to
plant type data is provided for generation and capacity fields. The plant types are
categorized based on fuel used (e.g., coal, oil/gas, nuclear, hydro), combustion
technology (e.g., turbine, combined cycle gas), control technology (e.g., scrubber, post-
combustion NOxcontrol), and retrofit structure (e.g., coal plant with existing SNCR
retrofit with ACI). In addition to providing the above outputs forecasted for each model
run year, it also gives information on the various regulatory and legal requirements that
were inputted into the model as constraints.

• State Level Emissions File: This file shows EGU emissions at the state level for each run
year. There are two tabs, one showing emissions from all EGUs, and the other showing
emissions only from fossil units greater than 25 MW.

• RPE File: For each model plant, this file shows the projections of fuel consumption,
emissions, capacity, costs (capital, fixed operations and maintenance, and variable
operations and maintenance), and generation

Sometimes, but not always, EPA will parse the model results from a specific year from the model plants
back to "real-world" units (i.e. create parsed file). This parsing is a post-processing step for the IPM

-------
results and is only done to meet specific analytical requirements; it is not a standard model output.
While the parsed files are post-processed, and therefore not a direct output of IPM, they are often
helpful in understanding the choices the model made and why it made them. The parsed file includes
information about how the model chose to dispatch, retrofit, retire, and build EGUs, as mapped to real-
world units. It includes information on unit heat input, fuel choice, generation, emissions, and
retrofit/retirement actions.

V k.i Miirihu.li^h i-MiU i in, In,!.2 in my c

You likely have a lot of information about the unit you are commenting about. EPA is looking at
thousands of units and has access to some information (i.e., through continuous emissions monitoring
system (CEMS) measurements), but may not have the full spectrum of information that you can provide.
Thank you, in advance, for providing this information.

For comments to be effective, the comment needs to be complete, and it needs to be supported by
documentation to the extent possible. Comments like "my unit is a "must-run" unit and is not going to
be retired" need to be supported. For example, describe the particular transmission constraint that is
resulting in the load pocket that the unit is supplying. What documents are available from the facility
operator, local, state, or federal officials to support the claim? Please provide EPA with a copy of that
document, or a link to where the document can be found.

Strong Documentation: Documentation and evidence can support a strong case to change or update
NEEDS when it demonstrates that: a unit must operate within specific parameters; a unit is clearly being
retrofit, retired or built; or that the unit has a choice of options to comply with new state rules consent
decrees, or settlements. It includes, but is not limited to:

• Unit operating permits

• Documentation of firm financial commitments for financing a new unit or retrofit

• Evidence that ground has been broken on a new unit or retrofit (photographs, news articles,
press releases)

• State Rules, Consent Decrees, NSR Settlements, etc.

• Designation as a reliability must run (RMR) unit.

It is extremely helpful to provide a copy or link to the actual documents cited, including a reference for
the page number. This will ensure that we can easily find the information your comment is based on.

Other Documentation: Documentation and evidence that does not demonstrate a firm retrofit,
retirement, or construction of a unit nor shows that a unit must operate in a certain manner can be
helpful, but is generally considered to offer weaker support for making a change to NEEDS and/or IPM
assumptions or constraints and, on its own, may not provide enough certainty to support changing such
data.

For example, long-term utility integrated resource plans (IRP) are not firm commitments to build, retire,
or retrofit units and can and do change from year to year. Similarly, a unit's emissions rate can vary year
to year. While it is helpful to identify instances where there is a large variation in rates between
historical data and the corresponding rates in NEEDS, especially in conjunction with identifying reasons

-------
for the change (e.g., this unit just retrofit an SCR), small variations tend to reflect the choice of base year
for the data, and EPA typically prefers to use a consistent set of base year data for such input
assumptions rather than vary that base year on a unit-by-unit basis. Documentation of this category
includes:

•	Utility IRPs.

•	Historical data on past unit behavior without justification for why that behavior will remain the
same in the future.

•	Press releases about possible, but not firm, fleet changes.

•	Proposed state rules.

Even if the data provided fits in one of these categories, it can still be helpful to submit. Even if EPA
determines the documentation itself does not support an immediate change, it can flag to EPA to watch
for changes at units or in state rules.

7

-------