Response to the Peer Review Report
EPA Base Case Version 5.13 Using IPM
U.S. EPA, Clean Air Markets Division
SECTION 1
INTRODUCTION
In October 2014, the U.S. Environmental Protection Agency (EPA) commissioned a peer review of
the EPA Base Case version 5.13 using the Integrated Planning Model (IPM).1 RTI International, an
independent contractor, facilitated the peer review of the EPA Base Case v. 5.13 in compliance with
EPA's Peer Review Handbook (U.S. EPA, 2006) and produced a report from that peer review. RTI
selected five peer reviewers (Anthony Paul, Meghan McGuinness, Walter Short, Paul Sotkiewicz,
and John Weyant) who have expertise in energy policy, power sector modeling and economics to
review the EPA Base Case v.5.13 and provide feedback. The peer reviewers evaluated the adequacy
of the framework, assumptions, and supporting data used in the EPA Base Case v. 5.13 using IPM,
and they suggested potential improvements.
IPM is a multiregional, dynamic, deterministic model of the U.S. power sector that provides forecasts
of least-cost capacity expansion, electricity dispatch and emission reliability constraints. The EPA
uses the platform to project and evaluate the cost and emissions impacts of various policies to limit
emissions of sulfur dioxide, nitrogen oxides, particulate matter, mercury, hydrogen chloride, and
carbon dioxide (CO2).
The independent peer review panel provided expert feedback on whether the analytical framework,
assumptions and applications of data in the EPA's Base Case v.5.13 using IPM are sufficient for the
EPA's needs in estimating the economic and emissions impacts associated with the power sector due
to emissions policy alternatives.
The panel identified a number of strengths associated with the model. For example, the report stated
that model exceeds other model capabilities in providing a relevant feedback mechanism between the
electric power model and key fuel inputs that drive simulation results.
Other strengths the panel identified include:
The detail with which pollution control technology options and costs are represented
The level of detail at which federal Clean Air Act (CAA) regulations are represented
The ability of the model to allow for the detailed representation of a variety of potential
changes in energy and environmental policies, including complicated features of market-
based programs
1 Documentation and files for EPA Base Case version 5.13 using IPM are located at
http ://www2 .epa. gov/airmarkets/po wer-sector-modeling-platform-v513
1
-------
The accuracy of the emissions control costs and their relationship to retirement decisions
The expansion of model regions from 32 to 64, which allows the model to better represent
current power market operations and existing transmission bottlenecks even within regional
transport organization (RTO) regions
Continuous updates of the representation of domestic coal and natural gas market conditions.
The peer review panel has also provided several areas for investigation and additional
recommendations for the EPA's consideration, including:
Input assumption documentation, which we discuss and respond to in Section 2 of this
document.
Particular elements of the EPA Base Case version 5.13 modeling structure that we discuss and
respond to in section 3 of this document.
Since the peer review was completed, there have been two subsequent versions of the EPA Base
Case's (version 5.14 and 5.15)2 Some of the peer review comments regarding documentation have
already been addressed through these incremental documentations. The two incremental
documentation reports are supplemental to the comprehensive EPA Base Case v.5.13 documentation
report, and are available at the EPA Power Sector Modeling Website.3 The first incremental
documentation report covers changes made in the 2014 update to IPM (v.5.14) base case, and the
second incremental report covers changes made to the 2015 update (v.5.15) base case. The EPA
anticipates when the next version of base case is completed, the EPA will prepare a comprehensive
documentation report analogous to the EPA Base Case v.5.13 documentation.
2 Available at http://www.epa.gov/airmarkets/power-sector-modeling.
3 Available at http://www.epa.gov/airmarkets/power-sector-modeling
2
-------
SECTION 2
INPUT ASSUMPTION DOCUMENTATION
2.1 VOM/FOM Cost Documentation
The Peer Review Report stated that the IPM documentation is a great source of data and information
ranging from capital costs to Variable Operations and Maintenance (VOM) and Fixed Operations and
Maintenance (FOM) costs, heat rates, and the cost and performance characteristics of emissions
controls among other items. However, it was stated that while some information is well documented
(e.g., emissions control from Sargent & Lundy), others are not.
In order to provide more detailed explanation of the VOM and FOM costs, we have made
documentation enhancements in the subsequent EPA Base Case using IPM version 5.14
documentation4 and retained this information in the EPA Base Case using IPM version 5.15.
The panel also suggested increasing FOM costs with age over the modeling time horizon. The FOM
costs for coal steam and oil/gas steam units in the EPA Base Case v.5.13 are age-specific. The cost is
applied to a model plant based on its age at the start of the modeling time horizon, and remains
constant through the modeling time frame. Most existing fossil/steam units also have the option in the
model to incur an assumed representative life extension cost (as an addition to the constant annual
FOM cost), which varies by plant type and is applied at the unit age corresponding to an assumed
representative design life for that plant type. The EPA will consider revising this approach in a future
platform to allow the initially assumed FOM cost to adjust with a unit's age over the modeling time
horizon.
2.2 Financial Assumptions
The panel noted a number of areas in the Financial Assumptions (Section 8) of the EPA Base Case
v.5.13 documentation where assumptions are inadequately documented or their application
insufficiently explained. The peer reviewers suggested that the EPA comb through the full
documentation to identify and address assumptions that do not cite original sources with sufficient
specificity, and to identify areas where more explanation is needed about how the assumption is
applied in the model. In response to these comments, the EPA will take several steps for the
documentation of the next major update of the financial assumptions:
Ensuring all footnotes refer to primary sources and include a current web address, if possible.
Enhancing the documentation by expanding key sections to include relevant background
independent of modeling context, a full explanation for why the EPA selected a certain
parameter or value, and a brief discussion for how that parameter interacts with the modeling
framework.
4 Section 4.2.7 of Base Case version 5.14 Incremental Documentation at http://www2.epa. gov/airmarkets/power-sector-
modeling-platform-v514)
3
-------
Sequence the documentation such that key assumptions are fully detailed when they are first
introduced to prevent the need for readers to search through multiple sections of the
documentation to understand a single concept.
Explaining the relevance of documented financing structures to how the IPM modeling
framework determines solutions.
Further, the panel requested the EPA identify assumptions that differ from generally accepted values
or those used by other well-known entities (especially other government entities), why the choice of a
different assumption was made, and the potential implications of this choice for policy analysis. The
EPA is unaware of any assumptions that differ from generally accepted values and the panel provided
no concrete examples of such an occurrence. Financial assumptions that require a numerical value
are assigned historical values, historical average values, or are based on current values that the EPA
assumes will revert, at least directionally, to longer-term averages that are more suitable to IPM's
intended purpose of projecting fundamentals-based capital investment decisions in the medium- to
long-term. One example of such an assumption is return on equity. The EPA assumes that return on
equity values will trend modestly upwards over time from the current, historically low values. This
assumption is informed by the EPA's desire not to enshrine the current returns on equity for all future
time periods, while recognizing that it is unclear when - or if - a return to higher rates more
consistent with longer-term historical averages will occur. Increasing the return on equity over time
in a modest fashion is intended to reflect a cautious approach appropriate for modeling tools such as
IPM. The approach of incrementally higher returns on equity over time is also adopted in the EIA's
financial assumptions for the Electricity Market Module. The EPA understands that the use of
historical data requires constant evaluation to validate its appropriateness both as a stand-alone value
as well as how that value is integrated with other financial assumptions. The EPA will consider
additional analysis to inform the choice of financial assumptions in the next version of the model and
its documentation.
2.2.1 Social Cost-Minimizing
The peer review panel stated that:
It is not clear that the social cost-minimizing solution supports a competitive
equilibrium in wholesale power markets (with either energy and capacity markets or
energy-only markets) or traditionally regulated regimes. The documentation at
minimum linking the "shadow prices" that come out of the linear programming (LP)
framework with actual prices paid to generation, or available to be collected by
generation, either in competitive wholesale markets or in traditional cost-of-service
regimes would be helpful to understand IPM results. Another way of phrasing this
would be to check that under the shadow prices computed by IPM, entry, exit and
4
-------
retention decisions are "individually rational" such that at these prices, resources would
do exactly the same thing IPM is saying will happen in the social cost minimum. (Peer
Review Report, page 4-5)
The IPM approach of minimizing the cost can be viewed from a social cost perspective, but it can
also be viewed as a competitive equilibrium in power market investment and operation. IPM
assumes that entities offer their supply to the wholesale market at cost, know the cost of other offers
in the markets and act as if other bidders in the market had the same knowledge and behaved in the
same way. Although the IPM solution technique is linear optimization, the modeling framework can
be regarded as a planner's perspective, as it models the market as if a planner were deciding how to
offer supply from a plant into a marketplace, such as might be found in a RTO. Operating in this
way, IPM ensures that there is consistency between the model operations and the resulting marginal
costs ("shadow prices"). IPM operation is consistent with the following view of an efficient planning
process:
The output of each generating plant is assumed to be offered into the market for dispatch at
the short run variable cost, consisting primarily of fuel and variable operating and
maintenance costs.5
The market is cleared based on offers from all plant operators by minimizing the wholesale
delivered cost to meet load in each future time period and region, subject to the constraints
such as transmission and environmental limits.6 The result of the simulated dispatch is a
marginal price for energy and capacity in each future time period, analogous to the short-term
pricing structure of an RTO with markets for energy and capacity, or a competitive wholesale
market outside an RTO.
The expected prices in future time period form the forward price series a planner would use to
determine whether the existing plant should be retained in the fleet, retrofitted to meet
environmental requirements, or retired, and whether new generating capacity will be required
to meet demand in one or more future periods. The model assumes perfect foresight in the
sense that it assumes the future prices are known at the outset of the planning horizon. The
overall optimization determines the price as the shadow price on the appropriate constraints
(demand for the energy price and reserve constraints for the capacity price.)7
IPM solves for the combined effect of the short run dispatch results in each period and the
longer run planning choices simultaneously, assuming that short run markets work efficiently,
and that planners successfully aim at efficient, least cost decisions in the long term using these
prices. Any alterations in short run offers will influence capacity planning decisions, and vice
versa, so that IPM's marginal prices in each time period are consistent with the entry and exit
of generating capacity, as well as with choices about which plants to retrofit.
5 The main components of these steps are noted in the IPM documentation. See IPM documentation Chapter 2.2.2 on
Generation Dispatch Variables.
6 This result is the short term effect of the overall IPM optimization described in Chapter 2 of the IPM documentation.
7 See IPM documentation Chapter 2.2.3 on Demand and Reserve Constraints
5
-------
IPM attempts to reflect the important parameters of the dispatch and planning functions, and multi-
period optimization does so in a way that ensures consistency among prices, dispatch quantities and
entry and exit of generating capacity. As a result, IPM projected prices are consistent with
competitive operation of wholesale markets.
2.2.2 Financial Regions
The peer review panel stated that:
The current financial modeling assumptions seem to apply across the country and may
not be relevant in some organized ISO/RTO wholesale power markets and should be
differentiated by area. For example, in PJM, much of the new entry taking place
across the footprint is merchant only, while in some zones (e.g., Dominion) it is a
vertically integrated entry for the most part (Peer Review Report, page 4-17).
In the development of the EPA Base Case, the EPA considered an approach that would categorize
areas in the United States as one of two types of financial regions - traditional cost-of-service and
competitive. Under this approach, cost-of-service regions would be assigned capital charge rates
based on regulated utility financial assumptions and competitive regions would be assigned capital
charge rates based on merchant financial assumptions.
However, such a modeling approach could result in overbuilding in the cost-of-service region (due to
relatively lower capital charge rates) that does not take into account real-world regulatory
prohibitions of external sales. In practice, there are formal and informal limits constraining new
capacity development that are difficult to capture in modeling simulations. For example, recent
proposals in PJM explicitly limit capacity expansion by some entities to be such that the total
capacity does not exceed internal requirements.8
Even if such new capacity deployment limitations could be implemented in the model, the model
would still face mixed regions in which there are both deregulated and regulated power plants and
participants, which calls into question the categorical assignment of merchant or utility financial
assumptions to the entire region. Recognizing the complexity of interregional electricity trade and
new capacity development under both merchant and utility financing conditions, the EPA Base Case
v.5.13 uses a hybrid approach to capture as many of these variables as possible without introducing
modeling artifacts favoring new capacity development in any particular region.
2.2.3 Financial Assumptions Appropriate for Longer Term Projections
8 https://pjm.eom/~/media/committees-groups/committees/mrc/20121017/20121017-proposed-mopr-summary.ashx
6
-------
The peer review panel stated that:
The financing assumptions based on 2008-2012 require major reworking as this
includes the effect of the worst financial crisis since the Great Depression, and current
financial markets world-wide are showing much lower returns on debt and equity. To
say the current levels are "unsustainable" or "artificial" is irresponsible and does not
reflect experience in Japan, where they are two decades into a similar crisis and still
have lower capital costs as a result (Peer Review Report, page 4-17).
In the EPA Base Case, financing costs are expected to increase over time. The mechanism for
expressing this increase is applying a higher risk free rate for the cost of capital. The five-year
average risk free rate (3.8 percent) is used to inform the ROE rather than the current rate (2.5
percent). It is appropriate to use a five-year average to establish the risk free rate for a number of
reasons:
Current rates are historically low due to the slow pace of recovery from the most recent
recession.
The EPA base case financial assumptions are changed infrequently, and hence, it should not
use assumptions reflective of a single point in time.
Merchant and utility cost of debt, debt-equity ratios, and historical betas are all calculated
based on the last 5 years (2008-2012) of historical data. The same approach to calculate the
risk free rate is used in order to remain consistent in its methodology.
Low interest rates may continue for some time; however, since investments in IPM are based on
expectations for long-term periods of up to 20 years and longer, it is still reasonable to expect
financing costs to return to longer-term trends. This is supported by multiple statements by the
Federal Reserve that they intend to raise rates from recent levels, and that they do not intend to
continue monetary policies adopted in the wake of the financial crisis indefinitely. Both historical
rates and current expectations are reevaluated for each new version of IPM.
7
-------
SECTION 3
Model Structure and Capabilities
We have identified several main topic areas that peer reviewers commented on that are relevant to
modeling structure and capability: 1) model design, 2) demand, 3) transmission, and 4) renewables.
The responses provided in this section explain where it is possible to conduct further research and to
make subsequent improvements in our modeling to address them. We would embark in such an
effort if we assess that these additional capabilities would significantly improve our modeling
capabilities (especially in the context of environmental policies evaluated) without compromising
other aspects of the modeling capabilities (such as model size and run-time) with the available
resources and timelines.9 Therefore, this section intends to clarify the areas where it is possible to
improve or modify our modeling structure and capabilities pursuant to our analytical goals.
3.1 Model Design
The peer review panel recommended "consideration of the addition of one more seasonthe
shoulder seasonto differentiate the spring and fall with respect to load duration curves (LDC)"
(Peer Review Report, page 3-2). In IPM, the seasonal and segmental definition of load duration
curves characterizes the resolution of demand modeling. All else held equal, an increase in the
number of seasons and/or segments could be expected to improve power plant dispatch modeling.
However, in model development, all else cannot be held equal; we must make trade-off decisions in
model design in order to maintain viable model size and acceptable run-time conditions. Since the
model size is directly proportional to number of seasons and segments, base case design must
consider the analytic merits of different seasonal definitions alongside analytic merits of different
numbers of run years, retrofit options and model plants represented in the modeling framework. The
EPA will explore alternate seasonal and segmental definitions, including options that represent a
"shoulder season," in the development of future base cases.
The peer review report also stated:
In constructing the LDCs for the two seasons modeled by IPM, a single year of data
(2011) was employed. The data are hourly load data in each IPM model region. A
more robust approach would take an average over multiple years of data, thereby
generating a more realistic projection than that from a single year of data. The panel
understands that the justification for using a single year instead of an average over
multiple years is to synchronize with the treatment of time in the atmospheric transport
models that EPA sometimes uses in conjunction with IPM. The panel recommends
9 The peer reviewers recognized that the model necessarily includes simplifications, in part because of the need to
maintain reasonable run times so that the model can be used for policy-relevant analysis on a timely basis. Peer
Review Report, page 4-1.
-------
that the multiyear approach be adopted across the board. (Peer Review Report, page 4-
16)
In IPM, the hourly load curves define the load duration curves and hence can affect power plant
dispatch. In general, weather normalized load curves are constructed and used in models instead of
the load curves of a historical year. Past weather data can also be analyzed to pick a year with normal
weather for a region and then to use that year's load curve for that region. The EPA has used the
2011 load curves for the entire country to be consistent with the EPA's treatment of temporal
distribution in air quality models. The EPA will explore this issue further in the future in conjunction
with other considerations informing the EPA's selection of historical data to inform air quality
modeling and benefits analyses.
3.2 Demand
The peer review report stated:
It seems odd to have fixed electricity demands drive the solution of the IPM linear
program base case as changes in demand that occur in response to changes in
economic and noneconomic conditions over time seem like some of the most
important adjustments that would take place in both base case and policy projections.
The fixed demands for the IPM base case seem to also come from the NEMS model,
which raises the question of consistency between the costs and prices in the NEMS
model and in IPM. If the upstream cost/price results from IPM are different than those
from NEMS for the scenario used to generate the NEMS demands that are then fixed
in IPM, this seems like a large potential inconsistency in the baseline, which could be
even more significant in any policy case run from that baseline. One solution would be
to estimate demand curves (demand vs. price functions) with NEMS that are then
incorporated into IPM. This would insure both the internal consistency of the
supply/demand/price picture coming out of IPM and its suitability for considering
policy alternatives. (Peer Review Report, page 4-7)
There are two issues raised here by the peer review panel. The first concerns the appropriateness of
using a fixed demand projection. While we believe the current approach is reasonable, for our next
modeling platform we will explore if and how we should develop elasticities from comparing Annual
Energy Outlook (AEO) reference and alternative cases to support demand representation within IPM
in a more flexible way, along the lines suggested by the peer reviewers. We believe the current
representation is reasonable both because it reflects changes in exogenous economic and
noneconomic conditions over time (a fixed demand approach should not be construed as one that
relies exclusively on historical or current conditions), and because the price fluctuations observed in
the base case (the subject of this peer review) do not have a significant impact on overall demand and
therefore do not warrant such a capability. The second concern is that adopting demand projections
9
-------
from NEMS when there are cost and price differences between NEMS10 and IPM might represent a
significant inconsistency. However, available data even across NEMS scenarios (where costs and
prices may vary substantially, but within the same model, which is also considering electricity
demand endogenously) do not bear this concern out. As an illustrative example, in the 2015 AEO
there is only a 0.1% difference in the average annual growth rate of demand when comparing the
Reference Case with the high and low oil and gas resource cases. This minor difference in demand
relative to a significant change in fuel prices illustrate why demand projections in the EPA Base Case
using IPM do not have to share identical cost and price inputs with a given NEMS scenario to be
internally consistent; rather, the demand projection simply has to contain cost and price inputs that do
not necessitate unreasonable assumptions among the many broader macroeconomic conditions or
trends that are not endogenously adjusted in IPM. Many of the cost and price inputs endogenously
determined by IPM can be shown to have a negligible impact on overall demand levels, especially
when compared with these broader macroeconomic conditions or trends.
The peer review report also stated:
The current model in use by EPA does not allow for the endogenous determination of
energy efficiency options or demand response in dispatch. These two options have
become key parts of wholesale power markets and in integrated resource plans
executed by vertically integrated utilities, but are always evaluated implicitly or
explicitly based on the economics. (Peer Review Report, page 4-6)
EPA Base Case v.5.13 does not model energy efficiency (EE) or demand response (DR) options
endogenously. For the next platform, we will evaluate whether adequate data are available to model
EE or DR options as "negawatt generators" that could capture seasonal load shape impacts and costs
of each EE or DR measure at a regional level.
3.3 Transmission
The peer review report states:
While the IPM model is a pipes-and-bubbles model between the 64 regions, there are
transmission constraints within each of the 64 model regions, and generator
deactivations may create the need to build a great deal of transmission to allow units to
retire in a reliable manner. One suggestion is for EPA to examine the use of internal
regional transmission constraints to get a better sense of the constraints on dispatch.
The second suggestion to account for transmission costs (and/or reliability-must-run
[RMR) generator costs) associated with generator retirements, is to estimate the
"average" cost of new transmission and RMR associated with retirements to account
10 NEMS is the Energy Information Model used for the AEO projections.
10
-------
for these costs in the baseline and in policy cases... Because of the way the grid
works, some non-linearities can be represented in [a pipes-and-bubbles model] and
some transmission studies have informed the pipe and bubbles constraints, but the
model is still limited because of its underlying formulation being pipe and bubbles
which precludes consideration of point to point transmission that would better
represent the actual operation of the current and future electricity grid. Additional
resolution in transmission capacity should be considered. (Peer Review Report, pages
4-16 to 4-17)
The IPM Base Case v.5.13 does include transmission costs for wind units within regions, by
reflecting the cost of transmission in the wind unit cost classes. In the next platform, we will
consider whether/how the number of model regions in the EPA Base Case using IPM could be
increased to capture additional intra-zonal constraints that are not addressed explicitly in the EPA
Base Case v.5.13 (and in the existing subsequent versions so far), by reviewing current market
information for each region and comparing to the existing configuration of zones/regions in IPM to
determine if there are any major intra-zonal constraints, load pockets or generation pockets that are
not modeled explicitly. For any such identified areas, we could define sub-zones in IPM as new
regions and determine that area's interregional transfer capabilities using any available information.
We will also review market information for each region to determine whether we can estimate, and
potentially incorporate, a generic transmission reliability-related retirement cost based on any
identified historical cost of transmission improvements required to enable retirements in the region.
3.4 Renewables
The peer review report stated:
The wind capacity values in Tables 4-21 through 4-23 in the IPM documentation are
reasonable for the initial wind installed into a utility system. However,... the capacity
value of wind declines as more wind is added to a utility system. The decline will be
more rapid if all of the wind capacity is installed at the same location because if the
wind is not blowing at that location when a generator or transmission line fails during
a peak load time, adding more capacity will not increase the contribution to the reserve
margin requirement. However,... were the next increment of additional wind located
at a different site, there is a better chance that wind will be available from at least one
of the sites (i.e., with geographic diversity, the wind sites will have a higher overall
capacity value). Geographically diverse wind sites can also reduce curtailments and
the need for spinning reserves. IPM's wind capacity values do not account for either
the level of wind penetration or the geographic diversity of the wind sites selected.
(Peer Review Report, page 4-9)
11
-------
Solar and wind technologies are intermittent and not dispatchable. Hence, the generation from these
technologies may not be available during system peak. IPM accounts for this issue by reducing the
reserve margin contribution for these technologies based on results from other studies. In the EPA
Base Case v.5.13 (and in the subsequent existing versions so far), these reserve margin contribution
factors remain constant over the time horizon. However, as pointed out by the peer reviewers, these
factors do not remain constant, but reduce with increased renewable penetration in a region. To
address this concern, the EPA chose to impose IPM region-specific deployment constraints that limit
new builds to levels for which it is unnecessary to further reduce the reserve margin contribution of
these potential build options, in order to accurately represent the generation available from wind units
at the system peak. We will continue to evaluate this issue when developing the next modeling
platform.
The peer review report also states:
If wind or solar generation increases unexpectedly, it may not be possible to ramp
down conventional generators fast enough or low enough to accommodate the
renewable energy. In this case, the renewable energy cannot be used. Such
curtailments of wind and solar will not be significant until the fraction of load met by
wind and solar is significantly higher than it is today (today's curtailments of wind and
solar are due mostly to transmission limitations). However, in modeling climate
change or high natural gas prices, penetration of wind and solar may climb to levels
where the unused VRRET11 energy due to ramping constraints significantly impacts
the overall economic viability of these VRRETs. Since IPM makes no estimate of the
curtailments from wind and solar, it again inappropriately advantages VRRETs. This
IPM limitation may be largely removable by adding a step-wise linear function that
accounts for curtailments by reducing the output of VRRETs as more VRRET
capacity is installed in a utility system. (Peer Review Report, page 4-8)
The EPA addresses this concern in the current platform through IPM region-specific deployment
limits. There is no estimate of curtailment in the current base case because new capacity is limited to
levels that have been demonstrated successfully without the need for significant curtailment in the
EPA Base Case v.5.13 (and in the subsequent existing versions so far) or in policy cases related to
these base cases. However, the model may choose not to utilize (i.e., to 'dump') excess energy from
VRRETs during periods of low demand as part of a least-cost solution. This 'dumped' energy is
displayed in the standard IPM outputs. We will continue to evaluate this issue when developing the
next platform.
The peer review report also states:
11 The Peer Review report defines the term VRRET as variable resource renewable energy technology.
12
-------
Spinning reserves are usually conventional generators that are operating, but at less
than their full capacity. As such, they can be quickly brought up to full capacity should
the system require additional power. Because the generation from wind and solar can
drop unexpectedly, generation from these technologies must be backed up by spinning
reserves. Since IPM does not account for spinning reserves, wind and solar are given
an inappropriate advantage in IPM. This advantage may be fairly easy to correct in
IPM by adding a spinning reserve constraint in which the contribution from operating
dispatchable rotating generators contributes positively toward meeting the constraint,
and the addition of VRRETs contributes negatively. (Peer Review Report, page 4-8)
The EPA Base Case using IPM v5.13 includes explicit reserve margin constraints, a partial reserve
margin contribution for intermittent technologies, capacity deployment limits for intermittent
technologies, and various operational constraints related to the operation of fossil fuel-fired capacity;
given this modeling context and the level of intermittent RE generation in the EPA Base Case using
IPM v5.13, a constraint related specifically to establishing a minimum level of spinning reserves
would be redundant. However, in the next platform, we will continue to evaluate if and how the
effect of increasing RE penetration on demand for spinning reserves should be captured more directly
in the model.
The peer review report also states:
Because integration requirements become more difficult as more VRRETs are added
to a utility system, the penetration of VRRETs is naturally limited in the real world.
Since IPM does not capture this increasing integration difficulty, it is forced to use a
surrogate method for limiting the penetration of VRRETs. For wind, this is at least
partially effected by adopting the wind cost classes and their corresponding capital
cost multipliers from EIA's NEMS model. As shown in Table 4-24 of the IPM
documentation, IPM's wind capital cost multipliers range in value from 1.0 to 2.0. The
IPM documentation quotes NEMS documentation to the effect that these multipliers
account for "such factors as distance from existing transmission, terrain variability,
slope and other causes of resource degradation, site accessibility challenges,
population proximity, competing land uses, aesthetics, and environmental factors."
These are real factors that do need to be accounted for. However, as shown in Table 4-
17 in the IPM documentation, over 90% of the wind resource is included in cost class
5, which doubles the capital cost of wind. This is an unrealistically large cost increase
that is not evident in practice. In NEMS, and the panel suspects in IPM, this doubling
of the cost of wind for 90% of the wind sites precludes wind from capturing a
significant share of the electric power market. (Peer Review Report, pages 4-12 to 4-
13)
IPM's cost projections for deployment of wind resources are related to transmission availability and
other factors at the sites where the wind resources are found; IPM cost classes and resources make
13
-------
clear that there are ample opportunities for deployment of cost-effective wind resources. IPM cost
classes reflect the fact that the vast majority of the wind resource in this country is located in areas
that are not amenable to development, but also show that there remain ample opportunities for
deployment of wind projects. Both of these facts are accurately reflected in the EPA Base Case using
IPM v5.13 using wind resources and cost classes. The total onshore potential wind capacity resource
included in the EPA Base Case using IPM v5.13 is vast. For example, the resource available under
cost step 1 far exceeds the amount of wind that is deployed in the EPA Base Case using IPM v5.13,
despite being only a few percentage points of the total wind resource. The resource specified
cumulatively under cost steps 1-3 (cost multiplier of 1 to 1.25 times the base cost) is many times
greater, meaning there are huge amounts of wind resources under the lowest wind classes that are not
being deployed in the model; thus the costs of concern to the reviewers are not constraining the
deployment of wind in the IPM base case. In the next platform we will continue to evaluate the wind
resource assigned to each cost step to ensure an appropriate representation of the availability and
competitiveness of wind technologies.
The peer review report also states:
IPM constrains the annual generation from wind to be no more than 20% of the annual
generation in each region. This artificial limit has already been exceeded in Iowa,
where over 27% of the 2013 electric load was met by wind energy (U.S. EIA, 2014).
The constraint is even more egregious in states/regions like Wyoming and North
Dakota that have substantial wind resources that could be used to generate wind power
for export to nearby regions with larger populations and loads; clearly, this 20% limit
would preclude much of this potential export capability. The peer review panel
recommends that this 20% constraint on wind generation be removed from IPM.
(Peer Review Report, page 4-13)
The annual wind generation constraint by IPM region does not represent a technical limit, but rather a
constraint representing wind penetration levels that could be projected by the model without requiring
significant integration costs not explicitly represented in this platform. For regions that currently
exceed this limit, the limit has been adjusted upward to reflect local conditions. In the next platform
we will continue to evaluate alternatives to this constraint to ensure all appropriate costs are
accounted for under higher penetration scenarios.
The peer review report also states:
Solar thermal power can take advantage of thermal storage of the working fluid to
generate power at a later time. The IPM documentation does not say explicitly
whether it considers this option, but the solar thermal generation profile presented in
Table 4-28 of the IPM documentation suggests that no thermal storage is assumed by
IPM. This needs to be corrected as the primary reason for adopting solar thermal
electric generation capacity is its ability to store energy for dispatch at a later time...
14
-------
Electric power storage technologies can mitigate the variability of wind and solar
power. While the IPM documentation mentions that pumped storage is modeled,
details are not provided. New pumped storage and compressed air energy storage
should be considered in IPM. Because of their ability to ramp up and down quickly,
reservoir-based hydroelectric facilities can also mitigate the variability of wind and
solar power. (Peer Review Report, pages 4-11 to 4-12)
The EPA Base Case v.5.13 (and the existing subsequent versions so far) accounts for new solar
thermal options without storage. In the next platform, we will clarify the operation of pumped storage
in the IPM documentation, consider representation of energy storage technology options in IPM, and
include details about any such options in modeling documentation.
3.5 Heat Rate Improvements
The peer review report stated:
The IPM Base Case v5.13 includes no provision for endogenous investment in heat
rate improvements (HRIs) at coal boilers. However, the version of IPM used in the
analysis that supports the Clean Power Plan does allow for coal boilers to invest in a
6% HRI at a cost. This inconsistent treatment of HRI between the base case and the
Clean Power Plan scenarios is clearly not appropriate. The panel understands that this
resulted from the recent development of the HRI capability that was not available
when the base case was run. Future base cases should include the endogenous HRI
option. (Peer Review Report, page 4-19)
The RIA modeling supporting the Clean Power Plan offers coal steam model plants a heat rate
improvement option that is fully integrated into the IPM modeling framework. This capability
enables IPM to solve for the optimal deployment of heat rate improvement (HRI) technologies on a
plant-by-plant basis in the regulatory scenarios analyzed. The option for heat rate improvement is
only made available in the illustrative plan approaches during the compliance period, in response to
the final rule. The intent of this analysis was to examine the impacts of the regulatory scenarios
analyzed. As such, EPA believes that it is appropriate to limit heat rate improvement opportunities to
the policy case. This approach isolates the impacts of only those improvements that are expected to
occur in response to the policy. Going forward, we will continue to consider what heat rate
performance is reasonable to expect based on the scenario being analyzed, keeping in mind that in
practice, EGU operators have not availed themselves of heat rate improvement opportunities that
already appear economic.
15
-------
RTI Project Number
0213796.001.004
Integrated Planning Model (IPM) Base
Case Version 5.13 Peer Review
Peer Review Report
October 2014
Prepared for
U.S. Environmental Protection Agency
Clean Air Markets Division
1200 Pennsylvania Avenue, NW
Washington, DC 24060
Prepared by
Anthony Paul, Chair
Meghan McGuinness
Walter Short
Paul Sotkiewicz
John Weyant
through
RTI International
3040 E. Cornwallis Road
Research Triangle Park, NC 27709
HRTI
INTERNATIONAL
-------
CONTENTS
Section Page
Section 1 Introduction 1-1
Section 2 Peer Review Process 2-1
Section 3 Summary of findings 3-1
Section 4 Panel findings 4-1
Section 5 References 5-1
Appendixes
Appendix A Conflict of Interest Analysis and Bias Questionnaire A-l
Appendix B Peer Reviewer Resumes B-l
Appendix C Peer Review Panel Charge C-l
Appendix D IPM BaseCase 5.13 Peer Review Panel Meeting D-l
in
-------
LIST OF TABLES
Number Page
Table 2-1. Summary of Peer Review Panel Meetings 2-3
LIST OF FIGURES
Number Page
Figure 4-1. Declining Wind Capacity Value with Increasing Wind Penetration 4-10
iv
-------
LIST OF ACRONYMS
AEO
Annual Energy Outlook
C&I
commercial and industrial
CAA
Clean Air Act
CAMD
Clean Air Markets Division
COI
conflict of interest
CSAPR
Cross-State Air Pollution Rule
CT
Combustion Turbine
EIA
U.S. Energy Information Administration
EPA
U.S. Environmental Protection Agency
ERCOT
Electric Reliability Council of Texas
FOM
fixed operation and maintenance
HRI
heat rate improvement
IGCC
integrated gasification combined cycle
IPM
Integrated Planning Model
ISO
independent system operator
LDC
load duration curves
LP
linear programming
MATS
Mercury and Air Toxics Standards
NEEDS
National Electric Energy Data System
NEMS
National Energy Modeling System
NREL
National Renewable Energy Laboratory
OAR
Office of Air and Radiation
PERI
Princeton Energy Resources International
PV
photovoltaics
ReEDS
Regional Energy Deployment System
RMR
reliability must run
RPS
renewable portfolio standard
RTO
regional transmission organization
VOM
variable operating and maintenance
VRRET
variable resource renewable energy technology
WACC
weighted average cost of capital
v
-------
SECTION 1
INTRODUCTION
The Clean Air Markets Division (CAMD) within the U.S. Environmental
Protection Agency's (EPA's) Office of Air and Radiation (OAR) has requested a peer
review of Base Case v5.13 of the Integrated Planning Model (IPM), a multiregional,
dynamic, deterministic model of the U.S. power sector that provides forecasts of least-
cost capacity expansion, electricity dispatch and emission reliability constraints. CAMD
uses IPM Base Case v5.13 to project and evaluate the cost and emissions impacts of
alternative policies to limit emissions of sulfur dioxide, nitrogen oxides particulate
matter, mercury, hydrogen chloride and other toxic air pollutants as well as emissions of
carbon dioxide (CO2) and other greenhouse gases over a modeling time horizon of 2016-
2050 to carry out its mission and to support periodic policy and regulatory analyses of the
electric power sector.
RTI International (RTI), an independent contractor, supported CAMD by
facilitating a peer review of Base Case v5.13 in compliance with EPA's Peer Review
Handbook (U.S. EPA, 2006). The peer review was conducted to ensure that the power
sector model meets EPA's goals for transparency of EPA technical analyses by making it
easy for stakeholders and expert reviewers to examine specific estimated impacts of
potential new policies, to evaluate the technical credibility of EPA's projections, and to
comment on the consequences of modeled policies. RTI selected four peer reviewers and
one peer review panel chair who are experts in energy policy, power sector modeling and
economics to review Base Case v5.13 and provide feedback. Results of the peer review
are intended to evaluate the adequacy of the framework, evaluate assumptions and
supporting data used in Base Case v5.13, and identify potential modifications to improve
CAMD's forecasting ability.
This report includes a description of the peer review process, the peer review
panel report and EPA's responses to the peer review panel report. In addition, all
materials provided to the peer reviewers to support the review, such as the panel charge
and the technical work product, as well as peer reviewer resumes and a conflict-of-
interest (COI) disclosure form, are provided in the appendices.
1-1
-------
SECTION 2
PEER REVIEW PROCESS
In February 2014, EPA requested that RTI facilitate a peer review of Base Case
v5.13, used by CAMD within EPA's OAR to support policy and regulatory analyses of
the electric power sector. RTI managed the peer review independently and according to
guidelines in EPA's Peer Review Handbook (U.S. EPA, 2006). RTI initiated the process
of identifying and selecting five peer reviewers in April 2014 and completed the process
in June 2014.
RTI surveyed the literature and on-line resources and gathered recommendations
from individuals knowledgeable of the subject matter in order to identify qualified
candidates for consideration. Qualified candidates were those with knowledge of the
workings of Base Case v5.13 and expertise in power sector forecasting. Of the 21
identified candidates, 4 were excluded from consideration because they or their
significant other were involved in the development of ICF International's IPM, EPA's
Base Case, or direct competitors of the model, and there was considered to be an inherent
COI.
Per instructions from EPA, RTI aimed to select four or five reviewers from the
candidate pool based on all of the following criteria:
ฆ expertise, knowledge and experience of each individual
ฆ adherence to the COI guidance in the EPA's Peer Review Handbook (2006)
ฆ panel balance with respect to the expertise required to conduct the review and
the diversity of relevant scientific and technical perspectives
Five candidates were then highlighted based on recommendations from subject
matter experts and those with relevant expertise. In addition, a panel composed of these
five candidates was identified as balanced, independent and expert with various
backgrounds from academia, nongovernmental organizations, and private consulting.
The five highlighted candidates were then contacted to ascertain their availability
and potential COI. Each candidate completed a COI disclosure form to identify any and
all real or perceived COI or bias, including funding sources, employment, public
statements and other areas of potential conflict in accordance with EPA's Peer Review
Handbook (U.S. EPA, 2006). A template of the COI form completed by the candidates is
2-1
-------
included in Appendix A. RTI staff supporting the peer review also underwent a COI
investigation to corroborate the independence and a lack of bias across all components of
the peer review.
Based on the candidates' availability and qualifications, the information provided
in the completed COI disclosure forms, and an independent COI investigation conducted
by RTI staff, RTI selected the following five candidates:
ฆ Meghan McGuinness, Bipartisan Policy Center
ฆ Anthony Paul, Resources for the Future
ฆ Walter Short, retired National Renewable Energy Laboratory (NREL)
ฆ Paul Sotkiewicz, PhD, PJM Interconnection
ฆ John Weyant, PhD, Stanford University
Two of the selected peer reviewers, Mr. Paul and Mr. Short, reported no COI on
the disclosure form. Dr. Sotkiewicz reported that he attended a CAMD workshop in May
2013 reviewing IPM modeling efforts conducted for the Mercury and Air Toxics
Standards (MATS) and Cross-State Air Pollution Rule (CSAPR), that he has participated
in panel discussions about the transmission and generation constraints of IPM, and that he
works for a bulk power entity. Dr. Sotkiewicz further stated that although he will be
leading PJM's efforts to comment on the Clean Power Plan, he has no financial interests
that would be affected by IPM Base Case v5.13 modeling results. Ms. McGuinness
reported that she worked at EPA CAMD from 2003 to 2006, and currently works with a
private-sector version of ICF's IPM. Dr. Weyant reported that he maintains an active
grant with the Climate Economics division of EPA conducting model comparisons, and
that he participated in a three-person review of IPM for ICF in 2008.
None of the reported COI was deemed inconsistent with EPA's Peer Review
Handbook (2006). EPA reviewed and approved the list of candidates selected by RTI as
appropriate choices from the candidate pool to form an independent and balanced panel.
Based on expertise, availability and further recommendations, Mr. Paul was invited to
serve as the chair of the peer review panel. Copies of the selected candidate resumes are
included in Appendix B of this report.
2-2
-------
RTI staff provided the peer reviewers with the following materials to guide the
evaluations:
ฆ EPA-developed Peer Review Charge (see Appendix C)
ฆ Documentation for EPA Base Case v. 5.13: Using the Integrated Planning
Model (U.S. EPA, 2013)
The peer review panel met several times by conference call and attended one in-
person meeting over the following months, submitting the final peer review panel report
in October 2014. A summary of the peer review panel meetings is provided in Table 2-1.
Table 2-1. Summary of Peer Review Panel Meetings
Date
Meeting Type
Meeting Topic(s)
Outside Attendees
July 21, 2014
Conference call
Introductions
Leland Deck (EPA), Jennifer
Richkus (RTI)
August 6, 2014
Conference call
Review of charge
Leland Deck (EPA), William
Meroney (EPA), Jennifer
Richkus (RTI)
August 15, 2014
Conference call
Discussion of Base Case
v5.13 structure
Leland Deck (EPA), William
Meroney (EPA), Jennifer
Richkus (RTI)
August 21, 2014
Conference call
Renewable portfolio
standards, heat rate
improvement (HRI), life
extension costs
Leland Deck (EPA), William
Meroney (EPA), Jennifer
Richkus (RTI)
August 29, 2014
Conference call
Turndown constraints, load
duration curves
Leland Deck (EPA), William
Meroney (EPA), Jennifer
Richkus (RTI)
September 3, 2014
Conference call
In-person meeting agenda,
transparency and recognizing
tradeoffs
Leland Deck (EPA), William
Meroney (EPA), Jennifer
Richkus (RTI)
September 9, 2014
In-person meeting
Review of panel materials;
see Appendix D for the
meeting agenda
Leland Deck (EPA), William
Meroney (EPA), Ryan Sims
(EPA), Venkatesh Boddu
(ICF), Jennifer Richkus (RTI)
September 22, 2014
Conference call
Emission control technology,
fixed operation and
maintenance costs
Leland Deck (EPA), William
Meroney (EPA), Jennifer
Richkus (RTI)
September 30, 2014
Conference call
Report development
Jennifer Richkus (RTI)
Peer reviewers were provided with an honorarium for approximately 50 hours of
effort and travel costs to attend the in-person meeting. The panel chair was provided with
a larger honorarium to oversee the review and guide the technical and substantive aspects
of the peer review. The following sections provide the findings of the peer review panel.
2-3
-------
SECTION 3
SUMMARY OF FINDINGS
The purpose of this independent peer review panel was to provide expert feedback
on whether the analytical framework, assumptions and applications of data in IPM Base
Case v5.13 are sufficient to EPA's needs for estimating the economic and emissions
impacts associated with the power sector and emissions policy alternatives. While the
panel identified a number of strengths associated with the model, the panel has also
provided several areas for investigation and additional recommendations for EPA's
consideration:
ฆ Pollution control technology options and costs are well represented in the
model.
ฆ The model exceeds other model capabilities in providing a relevant feedback
mechanism between the electric power model and key fuel inputs that drive
simulation results.
ฆ Although simplifying assumptions is necessary to conduct analyses of
complex systems such as the power sector, the model contains
oversimplifications in the reference case that may have adverse impacts on the
model results (e.g., load duration curves and treatment of renewables).
ฆ An analytical version of the linear program would greatly enhance
transparency and the ability of the public to understand how the model works.
ฆ Source(s) of underlying assumptions are referenced in an incomplete manner
in several sections of the documentation such as variable operation and
maintenance (VOM) costs, fixed operation and maintenance (FOM) costs, life
extension costs and financial assumptions.
ฆ A mixed integer programming framework would do more to capture the entry
and exit reality and would then obviate the need to use model plants.
ฆ Features of the current model that do not exist and would be desirable include:
endogenous determination of end-use energy efficiency and demand response
in dispatch, endogenous improvement in boiler heat rates, endogenous
construction of new interregional transmission capability.
ฆ There is a current absence of consideration of renewable integration factors in
IPM. IPM modelers should review the detailed approach used in NREL's
Regional Energy Deployment System (ReEDS) model (Short et al., 2011).
3-1
-------
ฆ The panel recommends consideration of the addition of one more seasonthe
shoulder seasonto differentiate the spring and fall with respect to load
duration curves (LDC).
3-2
-------
SECTION 4
PANEL FINDINGS
From the perspective of the peer review panel, market models are by necessity
just simplified representations of complex, real-world markets designed to provide
insights about how policy or market fundamentals will change outcomes. The panel
understands that it is in this spirit that EPA has employed IPM by ICF for examining the
effect of various environmental policies on the electric power sector.
The review panel also notes that real-world complexities only grow when
considering markets for electric power that exist at the intersection of economics, markets
and power system engineering. Some of the simplifying assumptions that distinguish a
power market model from a power market itself are more costly than others in causing
the model to deviate from the real world. Improving on some simplifying assumptions to
better reflect market realities may yield important insight that may be crucial to
understanding the effects of environmental policy on the electric power system and
associated costs. Doing so, however, may be costly from a model design, maintenance
and production standpoint.
It is in this context of balancing model complexity with the need to extract useful
and transparent insights that the review panel offers this document of observations for
further improvements and enhancements to IPM used by EPA to evaluate the effects of
environmental policy on the electric power sector. Because each panel member has
examined IPM from differing experiences and perspectives, the recommendations are not
listed in any particular order, but rather are to be interpreted as committee consensus on
issues that should be addressed for the next iteration of IPM.
By necessity, many simplifying assumptions have been made in the development
of the reference case in order to allow for the model to solve in a reasonable amount of
time. However, the panel has raised a number of instances where oversimplifications in
the reference case may be problematic (e.g., load duration curves and treatment of
renewables). For the sake of transparency, as well as for internal EPA decision making, it
may make sense for EPA to develop a hierarchy of the market characteristics or other
details that are most important to represent in greater detail at the expense of others. This
hierarchy may change over time between reference cases as EPA's policy agenda
changes. From the current documentation, it is difficult to discern how EPA has
4-1
-------
determined the appropriate level of detail with which to represent various aspects of the
system and markets.
While EPA has emphasized the importance of maintaining a computing time
(approximately 7 hours) that allows model runs to be turned around quickly in order to
meet the demands of policy makers, EPA should investigate the implications for both
computing time and model results of adding complexity to some of the runs in some of
the areas the panel identifies.
Ultimately, the value of EPA's base case depends on the extent to which EPA can
use it to develop robust analyses of the policies it will be proposing and finalizing over
the life of the base case. New policy approaches will require new capabilities or features
in the base case, and the base case should be developed with some insight into what those
features should be. In practice, this may be somewhat challenging because the base case
is likely to be finalized before key policy design decisions are made. That said, there is
generally some sense of what is coming, and it would be useful for the documentation to
include some discussion of relevant features that could be valuable, even if they are
ultimately not included due to technical or budget/time constraints. The role of IPM in
analyzing regulations that affect power sector CO2 emissions, for example, emphasize the
importance of this point. For example, the current lack of endogenous demand-side
energy efficiency in the model means that when EPA analyzes policy scenarios, a key
compliance option is not actually competing with alternatives, but must be forced in
through exogenous assumptions.1 This limits the value of EPA's analysis for
stakeholders. Similarly, the base case's currently crude representation of state renewable
portfolio standards (RPSs) may not matter much for an initial national-level analysis of
the proposal, but may impact the value of state- or regional-level results depending on the
role that existing RPSs play in implementation plans.
The remainder of this section addresses the strengths of IPM identified by the
panel, followed by a discussion of the nine key issues that the panel views as most
significant and worthy of further development for the next base case.
1 As is discussed later in this report, the reference case also does not include the option to endogenously
invest in heat rate improvements, though endogenous heat rate improvement options in policy scenarios
run in support of the Clean Power Plan. This inconsistency is also problematic for determining the
impact of policy scenarios.
4-2
-------
Strengths
The panel would like to recognize some of the strengths of IPM and highlight
some of the many improvements that EPA has made in updating the model for Base Case
v5.13.
The detail with which pollution control technology options and costs are
represented in the model and refined by EPA with each version of the base case remains a
real strength of IPM. These inputs, as well as the level of detail at which existing federal
Clean Air Act (CAA) regulations are represented in the base case, provide a strong
foundation for analyzing the implications of the costs (and, after exporting results to air
quality models, the benefits) of regulations affecting conventional pollutants and
mercury. The ability of the model to allow for the detailed representation of a variety of
energy and environmental policies, including complicated features of market-based
programs, is also an ongoing strength. The accuracy of the emissions control costs and
their relationship to retirement decisions were borne out in the 2011 PJM study of the
effect of MATS and CSAPR, which used EPA cost data and PJM revenue and going-
forward cost data to come up with what has turned out to be a reasonably accurate
assessment of generator retirements in PJM (PJM Interconnection, 2011).
The expansion of model regions from 32 to 64 is also an important feature of
v5.13. This significant improvement allows the model to better represent current power
market operations, represent existing transmission bottlenecks even with RTO regions,
and now accurately models transmissions zones with regional transmission organizations
(RTOs). As we note later in this report, however, more could be done to improve the
model's representation of transmission.
We also commend EPA for continuing to update the representation of domestic
coal and natural gas market conditions. The coal and gas modules are crucial to providing
a relevant feedback mechanism between the electric power model and key fuel inputs that
will drive simulation results. This is a strength of IPM that other commercially available
software used by utilities and RTOs does not always have where fuel prices are taken as
given, and scenario analyses are necessary to get a sense of the differential effects of fuel
price dynamics.
The volume of detailed data on FOM and VOM costs for all unit types, including
some coal steam by age and nuclear units, is very helpful in independently verifying
4-3
-------
some results that come out of model runs, but, as we mention, better documentation of
sources would be helpful.
Transparency and Documentation
The IPM documentation is a great source of data and information ranging from
capital costs to VOM costs to FOM costs, heat rates, and the cost and performance
characteristics of emissions controls among other items. However, while some
information is well documented (e.g., emissions control from Sargent & Lundy), others
are not. Explicit equations defining the model itself would be very helpful and make the
results from IPM more transparent. Specifically, publishing an analytical version of the
linear program would greatly enhance transparency and the ability of the public to
understand how the model works.
Throughout the model documentation, there are many places where the source(s)
of underlying assumptions are referenced in an incomplete manner. For example, the
documentation contains very limited information on the approaches or information used
to develop assumptions for VOM costs, FOM costs and life extension costs, as well as
how they are actually applied in the model.2 Similarly, the panel noted a number of areas
in the Financial Assumptions section (Section 8) where assumptions are inadequately
documented or their application insufficiently explained. EPA and ICF should comb
through the full documentation to identify and address assumptions that do not cite
original sources with sufficient specificity, and to identify areas where more explanation
is needed about how the assumption is applied in the model. It would also be helpful for
EPA and ICF to identify assumptions that differ from generally accepted values or those
used by other well-known entities (especially other government entities), why the choice
of a different assumption was made, and the potential implications of this choice for
policy analysis.
2 For example, it is not clear from the documentation that FOM costs are assigned to a unit based on its
age at the start of the model, and do not escalate as the unit ages over the course of the model. Similarly,
it is not clear from Table 4-10 in the IPM documentation that life extension costs are applied annually in
the model. It is also not clear how the cost is applied to nuclear units, which are not able to double their
lifetimes in the base case.
4-4
-------
Implications of Linear Program Structure
It is not clear that the social cost-minimizing solution supports a competitive
equilibrium in wholesale power markets (with both energy and capacity markets or
energy-only markets) or traditionally regulated regimes. The documentation at minimum
linking the "shadow prices" that come out of the linear programming (LP) framework
with actual prices paid to generation, or available to be collected by generation, either in
competitive wholesale markets or in traditional cost-of-service regimes would be helpful
to understand IPM results. Another way of phrasing this would be to check that under the
shadow prices computed by IPM, entry, exit and retention decisions are "individually
rational" such that at these prices, resources would do exactly the same thing IPM is
saying will happen in the social cost minimum.
The current framework uses a dynamic linear program, but, in reality, entry and
exit decisions are lumpy, all-or-nothing decisions. Units cannot be divided up as if they
were indivisible. A mixed integer programming framework would do more to capture this
reality and would then obviate the need to use model plants, and each individual unit
could be modeled with all of its details as known in the National Electric Energy Data
System (NEEDS) (U.S. EPA, 2014).
One big advantage of an LP formulation is ease of computation. In addition,
under very general (nondegeneracy) assumptions, linear programs generate unique global
optima, avoiding the multiple local optima outcome that might result from nonconvex
programs and/or less rigid simulation-type alternatives where iterative convergence
algorithms are used to find approximate equilibria.
However, a least-cost solution from a national LP model is equivalent to optimal
solutions at all of the state/regional levels only if strong efficient market assumptions are
made. Put differently, real-world characteristics of economic agents may lead to a
situation in which the national solution that is computed by the LP is not incentive-
compatible at the subnational levels. With companies, power pools, state regulatory
bodies and state legislatures all in play here, there are likely to be market and political
power constraints on the global solution that operated at the regional or state levels. Put
differently still, this means that a fairly large number of what economists call side
payments and possible alliance formations would be necessary in order for the incentives
from the national LP to be completely compatible at the subnational level, which seems
very unlikely. This is where the documentation of the shadow prices and linkages to
4-5
-------
payments available in competitive markets or in a cost-of-service regime would be
advantageous. One standard test for this could be comparing results between the model
and the real world even if only for a base year or two. However, the current National
Energy Modeling System (NEMS) calibration precludes this, but it would be a good
diagnostic to consider anyway.
Demand Response and Energy Efficiency
The current model in use by EPA does not allow for the endogenous
determination of energy efficiency options or demand response in dispatch. These two
options have become key parts of wholesale power markets and in integrated resource
plans executed by vertically integrated utilities, but are always evaluated implicitly or
explicitly based on the economics. The base case would be enhanced by seeing how
much efficiency and demand response would enter the model prior to examining any
policy cases.
A key challenge for IPM in its ability to evaluate the impacts of policy scenarios
is the lack of price-responsive demand. While the documentation reflects some capability
to implement price elasticities, EPA noted that this feature has been used under only
limited circumstances. EPA also noted that the lack of retail price characterization in the
model makes it challenging to effectively implement price response in demand.
The panel discussed that the flattening of demand projections has been largely due
to income, rather than price effects. Further, because of the extent to which EPA
currently simplifies load duration curves, price effects are likely to be smaller than might
otherwise be estimated. EPA should consider whether there is opportunity within IPM for
capturing income effects.
In order to effectively model regulations that affect CO2 emissions from the
power sector, EPA should consider options for allowing energy efficiency investments to
compete with other potential compliance options within the model. While the simplest
from the perspective of model constructionwould likely be to allow energy efficiency
to displace generation on the supply side (via model plants that are "negawatt"
generators), even this approach requires the development of location-specific energy
efficiency supply curves, a challenging undertaking that has the disadvantage of putting
the demand on the supply side of the equation. Another alternative would be to develop
representations of distribution households and commercial and industrial (C&I)
4-6
-------
customers by size and income and allow "representative households or businesses" to
make efficiency decisions under a budget constraint (i.e., how much can they save vs. the
investment) to keep the demand on the demand side of the market. Such an approach is
not easy in any way, but given the trends in overall energy consumption, it may prove
fruitful. In either case, the panel recommends that EPA work with ICF to evaluate the full
range of options for including energy efficiency, and include this discussion in the model
documentation.
It seems odd to have fixed electricity demands drive the solution of the IPM linear
program base case as changes in demand that occur in response to changes in economic
and noneconomic conditions over time seem like some of the most important adjustments
that would take place in both base case and policy projections. The fixed demands for the
IPM base case seem to also come from the NEMS model, which raises the question of
consistency between the costs and prices in the NEMS model and in IPM. If the upstream
cost/price results from IPM are different than those from NEMS for the scenario used to
generate the NEMS demands that are then fixed in IPM, this seems like a large potential
inconsistency in the baseline, which could be even more significant in any policy case
run from that baseline. One solution would be to estimate demand curves (demand vs.
price functions) with NEMS that are then incorporated into IPM. This would insure both
the internal consistency of the supply/demand/price picture coming out of IPM and its
suitability for considering policy alternatives. The following section discusses other
issues regarding the use of NEMS electricity demands.
Renewables
The IPM Base Case v5.13 has a very simplified representation of variable
resource renewable energy technology (VRRET) such as wind and solar. IPM benefits
wind and solar by ignoring some of the integration issues associated with them (e.g.,
induced reserve requirements and curtailments) and yet penalizes the same technology by
adopting capital-cost penalty functions and by failing to consider the possibilities for new
transmission, demand response and geographic diversity of wind siting. The only
integration issue addressed directly by IPM is the capacity value of wind and solar, but
again the representation is simplified and nondynamic.
VRRETs like wind and solar have unique characteristics that require unique
modeling approaches. The most significant difference between VRRETs and
conventional generators like coal and gas-fired power plants is the variable and uncertain
4-7
-------
nature of the resources VRRETs depend on, which limits their dispatch ability. This
variability and uncertainty precludes system operators from dispatching VRRETs at will.
As a result, the full capacity of a VRRET generator cannot be counted on to be available
when peak loads need to be met; additional reserves, primarily spinning reserves, are
needed to ensure that loads are met when generation from VRRETs declines
unexpectedly, and when VRRET generation is higher than expected, it may not be
possible to use it all because conventional generators may not be able to ramp down fast
enough or low enough to accommodate the VRRET generation.
Spinning reserves are usually conventional generators that are operating, but at
less than their full capacity. As such, they can be quickly brought up to full capacity
should the system require additional power. Because the generation from wind and solar
can drop unexpectedly, generation from these technologies must be backed up by
spinning reserves. Since IPM does not account for spinning reserves, wind and solar are
given an inappropriate advantage in IPM. This advantage may be fairly easy to correct in
IPM by adding a spinning reserve constraint in which the contribution from operating
dispatchable rotating generators contributes positively toward meeting the constraint, and
the addition of VRRETs contributes negatively.
If wind or solar generation increases unexpectedly, it may not be possible to ramp
down conventional generators fast enough or low enough to accommodate the renewable
energy. In this case, the renewable energy cannot be used. Such curtailments of wind and
solar will not be significant until the fraction of load met by wind and solar is
significantly higher than it is today (today's curtailments of wind and solar are due
mostly to transmission limitations). However, in modeling climate change or high natural
gas prices, penetration of wind and solar may climb to levels where the unused VRRET
energy due to ramping constraints significantly impacts the overall economic viability of
these VRRETs. Since IPM makes no estimate of the curtailments from wind and solar, it
again inappropriately advantages VRRETs. This IPM limitation may be largely
removable by adding a step-wise linear function that accounts for curtailments by
reducing the output of VRRETs as more VRRET capacity is installed in a utility system.
IPM does consider contingency reserves that are used to continue to meet loads
when there is a system failure (e.g., an operating generator malfunctions or a transmission
line is damaged). IPM uses a reserve margin constraint to ensure there is enough capacity
in the system that loads can continue to be met even when there is such a failure. The full
capacity of dispatchable generators like coal and gas contributes to this system reserve
4-8
-------
margin. However, since wind and solar resources are variable, electric system planners
do not count their full capacity toward the reserve margin. But most planners do count a
fraction of wind and solar capacity toward the reserve margin requirement because
statistically there is a reasonable probability that some fraction of the wind or solar
capacity will be generating at the time of peak load and a system outage. Similarly, IPM
counts a fraction of wind and solar capacity toward the reserve margin as specified in
Tables 4-21 through 4-23 for wind and Table 4-29 for photovoltaics (PV) and solar
thermal.
The wind capacity values in Tables 4-21 through 4-23 in the IPM documentation
are reasonable for the initial wind installed into a utility system. However, as shown in
Figure 4-1 below, the capacity value of wind declines as more wind is added to a utility
system. The decline will be more rapid if all of the wind capacity is installed at the same
location because if the wind is not blowing at that location when a generator or
transmission line fails during a peak load time, adding more capacity will not increase the
contribution to the reserve margin requirement. However, were the next increment of
additional wind located at a different site, there is a better chance that wind will be
available from at least one of the sites (i.e., with geographic diversity, the wind sites will
have a higher overall capacity value). Geographically diverse wind sites can also reduce
curtailments and the need for spinning reserves. IPM's wind capacity values do not
account for either the level of wind penetration or the geographic diversity of the wind
sites selected. Once again, the constant IPM values for wind capacity values
inappropriately advantages wind. However, it should be possible to approximate a
declining capacity value for wind in IPM using a piece-wise linear curve to reduce the
capacity value of wind as a function of the penetration level. The use of only 64 regions
in the IPM Base Case v5.13 would limit the ability to directly capture geographic
diversity in wind sites, but an approximate method such as that used in NEMS could be
applicable.
4-9
-------
Figure 4-1. Declining Wind Capacity Value with Increasing Wind Penetration
0.35 -
03
Q_
O
T3
0.25
CO
d
r o.i5
ฆa
m
Q.
tI5
O
0.05
2000 3000 4000
Installed capacity (MW)
Source: Milligan. 2011
IPM's treatment of the capacity value of solar has the same shortcomings as that
described above for wind. However, unlike wind, the constant solar capacity values
shown in Table 4-29 of the IPM documentation are not appropriate for even the initial
solar installations in a utility system. Because peak loads generally occur at the time of
summer air conditioning loads, solar generation is much more likely to be available at
that time. Thus, the capacity value of the initial solar installations in a utility system may
be as high as 40 to 80% of the solar capacity. On the other hand, the capacity value of PV
generally declines even faster than that of wind as PV penetration increases. Once enough
solar has been installed to decrease the summer afternoon peak below the summer
evening peak, the capacity value of PV falls off precipitously since the sun has set or is at
least low on the horizon by that time. As with wind, it should be possible to approximate
a declining capacity value for PV in IPM using a piece-wise linear curve to reduce the
capacity value of PV as a function of the penetration level. Geographic diversity of solar
does not yield as large a benefit as it does for wind because the solar resource is
geographically more homogenous than that of wind.
4-10
-------
The integration issues with respect to solar thermal power are the same as those of
PV with one major exception. Solar thermal power can take advantage of thermal storage
of the working fluid to generate power at a later time. The IPM documentation does not
say explicitly whether it considers this option, but the solar thermal generation profile
presented in Table 4-28 of the IPM documentation suggests that no thermal storage is
assumed by IPM. This needs to be corrected as the primary reason for adopting solar
thermal electric generation capacity is its ability to store energy for dispatch at a later
time. Ideally, IPM would endogenously determine the optimum level of thermal storage
associated with each solar thermal plant, but once again the relationships are highly
nonlinear and difficult to capture in a single LP optimization model. NREL has
conducted considerable work in this area and should be consulted, probably with the
ultimate goal of exogenously selecting a thermal storage system with a fixed number of
hours of storage3 to be considered for all possible solar thermal plants in IPM. Since the
marginal cost of another increment of thermal storage is relatively inexpensive, the
optimum storage size is normally quite large, which would allow IPM to assume that the
solar thermal plant is essentially dispatchable, alleviating the need to deal with
curtailments, additional reserves and nonunity capacity factors.
IPM modelers should review the detailed approach used in NREL's ReEDS
model and the more approximate methods of the U.S. Energy Information
Administration's (EIA's) NEMS model for reserves, curtailments and capacity credit.
Because these models are recursive optimizations,4 they can and do use highly nonlinear
representations (calculated between the annual optimizations) to compute integration
3 An hour of thermal storage for a solar thermal system is defined as the amount of thermal storage
required to run the solar thermal generator at full output for an hour.
4 In a recursive optimization model, an optimization is conducted for each model period. For example, the
model optimizes for year n. then moves on to optimize for year n + 1. The optimization for year n may
include multiple future periods (e.g., n + 1, n + 2, ...), but only the year n results are used from that
optimization. The next optimization will start with year n + 1, and may optimize over years n + 1, n + 2,
etc., but again, only the results from year n + 1 are retained. In NEMS, this approach is used for electric-
sector capacity expansion. It allows the NEMS electric-sector capacity expansion to make decisions
with some foresight about the future, while at the same time allowing the period n electric capacity
expansion results to be part of the larger economy-wide equilibrium methodology within NEMS for
period n. It also has the advantage that nonlinear relationships can be introduced into the NEMS electric
capacity expansion if some inputs for period n + 1 depend nonlinearly on the results from period n
and/or earlier periods. NREL's ReEDS model uses a similar recursive approach with the simplification
that each optimization is just for period n (i.e., there are no decision variables for periods n + 1, n + 2,
etc.). This reduces the size of each optimization and allows the introduction of nonlinearities with
respect to the variability of wind and solar, but limits the foresight in the model; the only foresight
afforded the investor modeled by ReEDS is an exogenously specified trajectory of expected future
conventional fuel prices.
4-11
-------
factors. The single linear optimization approach of IPM will preclude the direct
application of the ReEDS and NEMS methods, but as briefly mentioned above there may
be approximations that would be considerably better than the current absence of
consideration of these factors in IPM. There is also always the option of reconstructing
IPM to make it solve recursively as ReEDS and NEMS.
For wind and solar, the IPM documentation cites NREL as the source of the
capacity factors, but does not provide a reference or a contact. The IPM documentation
states that the capacity values for wind and solar were derived from Annual Energy
Outlook (AEO) 2012 results using the ratio of the NEMS capacity values to the NREL
capacity factors. This is not a reasonable approach for two reasons: (1) the NEMS model
used in AEO has limitations with respect to its calculation of capacity values for wind
and solar, and (2) the AEO base case is a very low wind and solar penetration scenario
that should yield higher capacity values than would be possible in a high wind/solar
penetration scenario. At a minimum, the IPM documentation should include information
on how NEMS makes its capacity value calculations.
Electric power storage technologies can mitigate the variability of wind and solar
power. While the IPM documentation mentions that pumped storage is modeled, details
are not provided. New pumped storage and compressed air energy storage should be
considered in IPM. Because of their ability to ramp up and down quickly, reservoir-based
hydroelectric facilities can also mitigate the variability of wind and solar power. It is not
clear how IPM handles water releases at hydro plants for nonpower purposes (e.g.,
irrigation).
IPM should also consider demand response options that can mitigate the
variability of wind and solar power. For example, on most days, the Electric Reliability
Council of Texas (ERCOT) receives more offers for "responsive reserves" from the
demand side than the 50% (1150 MW) it is legally allowed to accept (Tweed, 2011).
Because integration requirements become more difficult as more VRRETs are
added to a utility system, the penetration of VRRETs is naturally limited in the real
world. Since IPM does not capture this increasing integration difficulty, it is forced to use
a surrogate method for limiting the penetration of VRRETs. For wind, this is at least
partially effected by adopting the wind cost classes and their corresponding capital cost
multipliers from EIA's NEMS model. As shown in Table 4-24 of the IPM
documentation, IPM's wind capital cost multipliers range in value from 1.0 to 2.0. The
4-12
-------
IPM documentation quotes NEMS documentation to the effect that these multipliers
account for "such factors as distance from existing transmission, terrain variability, slope
and other causes of resource degradation, site accessibility challenges, population
proximity, competing land uses, aesthetics, and environmental factors." These are real
factors that do need to be accounted for. However, as shown in Table 4-17 in the IPM
documentation, over 90% of the wind resource is included in cost class 5, which doubles
the capital cost of wind. This is an unrealistically large cost increase that is not evident in
practice. In NEMS, and the panel suspects in IPM, this doubling of the cost of wind for
90% of the wind sites precludes wind from capturing a significant share of the electric
power market. The August 2014 NEMS documentation of the NEMS "renewable fuels
module" cites a report by Princeton Energy Resources International (PERI) as the source
for these numbers. One of the peer reviewers for this review was the NREL point of
contact for this PERI work. NREL never endorsed the report as the fraction (90% or
more) of wind resources included in the highest cost bin (2 x cost) are regarded as too
high. NREL's ReEDS model includes some of these costs explicitly; for example, for
each wind resource site in ReEDS, an additional cost is added to the wind capital cost
based on the slope of the terrain at that site. The peer review panel recommends that these
cost multipliers be reviewed independently using actual field cost data and modified
appropriately. No such cost multipliers are used in IPM for solar.
Although not mentioned in the IPM Base Case v5.13 documentation, IPM
apparently uses an additional method for limiting the penetration of wind. IPM constrains
the annual generation from wind to be no more than 20% of the annual generation in each
region. This artificial limit has already been exceeded in Iowa, where over 27% of the
2013 electric load was met by wind energy (U.S. EIA, 2014). The constraint is even more
egregious in states/regions like Wyoming and North Dakota that have substantial wind
resources that could be used to generate wind power for export to nearby regions with
larger populations and loads; clearly, this 20% limit would preclude much of this
potential export capability. The peer review panel recommends that this 20% constraint
on wind generation be removed from IPM.
Were EPA/ICF only to remove the IPM 20% limit on wind and reduce the cost
multipliers in accordance with actual data, in some scenarios, wind penetration might
increase beyond reasonable levels. Thus, it is imperative that these IPM modifications be
made in concert with the implementation of the integration factors described above (e.g.,
curtailments, additional reserve requirements and declining capacity values).
4-13
-------
To capture the variability of wind and solar generation, IPM uses the 2011 hourly
loads in each region and the regional wind and solar generation profiles illustrated in
Tables 4-20 and 4-28, respectively, to identify which load duration curve section the
wind and solar generation contributes to. The illustrative PV summer profile has some
(very small) PV generation even in the middle of the night. If this is generation from
moonlight, then the winter profile should have the same, but it does not. This is a minor
point, but the panel suggests using a different PV summer profile for illustrative
purposes.
VRRETs also differ from most conventional generators in that the bulk of the cost
of their power is the capital cost of the wind and solar plants themselves since they have
no fuel costs. Thus, it is more critical for VRRETs that financing of capital costs be
accurately modeled. Currently, IPM does not distinguish between regulated and
deregulated regions; IPM uses a cost of capital that is a weighted average between
regulated and market financing. As discussed in the Financial Assumptions section of this
report, the use of an average cost of capital can disadvantage high capital cost
technologies like renewables and nuclear because the low cost of capital of a pure
regulated utility is not included.
Finally, wind and solar thermal power also differ from conventional generators in
that their resources are often best in regions where there is little population and
consequently little load (people do not generally want to live in exceptionally windy
areas or desert areas with high direct insolation). These areas of limited population/load
generally also do not have significant transmission links to other regions. Therefore, to
access some of the best wind and solar thermal resources will require the construction of
new transmission lines. This process is currently underway in Texas to move power from
the strong winds of west Texas to population centers like Dallas and San Antonio. IPM
Base Case v5.13 does not allow for the building of new transmission lines. As discussed
under Interregional Transmission of this peer review report, modeling the addition of new
transmission lines is a difficult undertaking due to the highly nonlinear nature of AC
power flow and the nonproportional cost of power lines. However, the review panel
recommends that transmission expansion using DC power flow approximations be
considered for IPM.
Load Duration Curves and Seasons
4-14
-------
The 8,760 hours that comprise a single year are grouped by IPM into 2 seasons
(May 1-September 30 and October 1-April 30). Within each season, IPM uses a six-step
piecewise linear representation of LDC. The 2 seasons and 6 time segments per season
account for the entire year in 12 total segments. The motivation for the LDC approach to
dealing with time is computational constraints. Without such constraints, a more detailed
modeling approach would be to drop LDCs all together and model the year as 8,760
chronologic hours with day-ahead unit commitment and real-time economic dispatch by
known RTO or vertically integrated utility region. This would open the door for even
more detailed modeling of system dispatch (e.g., unit commitment and ramping) that
would better represent how the system actually operates and would likely influence the
capacity factors at which various different types of units operate. However, the
computational constraint that the model must solve in 7 hours is real, and the application
of seasons and LDCs is a popular approach to the computational problem.
The choice of two seasons and six segments per season is an important one that
the panel feels has not been investigated sufficiently by EPA. EPA points out the virtue
of seasonal LDCs instead of a single annual LDC in Section 2.3.5 of the IPM
documentation. EPA does not state it this way, but the LDC approach captures issues
related to daily dispatch and does not capture issues related to weekly or seasonal
dispatch; therefore, the application of LDCs to seasons is better than an annual LDC.
However, LDCs really only apply to single days, maybe a week, and so a more robust
application of the seasons/LDC approach would be 365 seasons (i.e., a single LDC for
each day of the year where unit commitment and economic dispatch could be brought to
the forefront to better account for how systems are actually operated).
Recognizing the computational infeasibility of such an approach, the panel
recommends at least serious consideration of the addition of one more seasonthe
shoulder seasonto differentiate spring and fall. A three-season approach would still
treat the summer as 5 months (allowing for a representation of the summer ozone season
for Clean Air Interstate Rule compliance), but would treat winter as only 3 months with
the other 4 months grouped together in a spring/fall season. This would allow IPM to
better capture the operation of intermediate load coal and oil steam plants that, in reality,
are often totally shut down in the shoulder seasons. Recognizing computational
constraints, a reduction in the number of segments in the LDCs from 6 to 4 would
amount to an identical number of 12 annual segments (3x4 instead of 2 x 6). Another
approach that would preserve the 12 time segments would be 4 seasons and 3 time
4-15
-------
segments per season. There are clearly tradeoffs between more seasons and more time
segments. It is the understanding of the panel that these tradeoffs are not well understood
by EPA. If the two-season, six-segment approach is retained, it should be justified in
documentation.
In constructing the LDCs for the two seasons modeled by IPM, a single year of
data (2011) was employed. The data are hourly load data in each IPM model region. A
more robust approach would take an average over multiple years of data, thereby
generating a more realistic projection than that from a single year of data. The panel
understands that the justification for using a single year instead of an average over
multiple years is to synchronize with the treatment of time in the atmospheric transport
models that EPA sometimes uses in conjunction with IPM. The panel recommends that
the multiyear approach be adopted across the board.
NEMS for Business-as-Usual
There are many benefits to and issues in using NEMS to calibrate the IPM Base
Case. The biggest benefits are that it is fairly detailed, openly available and well known.
However, we recommend that EPA give consideration to some of the issues with this
approach. One issue is that it is fairly constrained around the current state of the energy
system because it is strongly calibrated to the current state of the energy system; factors
adjusted to match those base year conditions are not unique, such that projections
probably become less reliable as time moves on past a decade or two, and the model may
be less responsive to large policy changes by the end of that time frame than would be
realistic. In addition, nonlinear regulatory rate setting rules, market power and other
departures from competitive market assumptions are difficult or impossible to implement
within this model structure. Finally, both NEMS and IPM appear to be fairly conservative
regarding the amount of technological and institutional change that might be expected by
the end of the modeled time horizon. What are the justification for and implications of
these assumptions?
Interregional Transmission
While the IPM model is a pipes-and-bubbles model between the 64 regions, there
are transmission constraints within each of the 64 model regions, and generator
deactivations may create the need to build a great deal of transmission to allow units to
retire in a reliable manner. One suggestion is for EPA to examine the use of internal
4-16
-------
regional transmission constraints to get a better sense of the constraints on dispatch. The
second suggestion to account for transmission costs (and/or reliability-must-run [RMR)
generator costs) associated with generator retirements, is to estimate the "average" cost of
new transmission and RMR associated with retirements to account for these costs in the
baseline and in policy cases.
The construction of the IPM baseline scenario precludes the building of new
electric transmission capacity which seems overly restrictive, especially for certain types
of scenarios (e.g. technological breakthroughs in renewables electricity generation that
work best in areas that are remote, etc.). In addition, the treatment of transmission
capacity from region to region in IPM is a pipe and bubble formulation which aggregates
both loads and generating units.
Computational and data limitations probably necessitate some aggregation
although parallel processing methods may help make the computational limits less
restrictive in the near future.
Because of the way the grid works, some non-linearities can be represented in this
way and some transmission studies have informed the pipe and bubbles constraints, but
the model is still limited because of its underlying formulation being pipe and bubbles
which precludes consideration of point to point transmission that would better represent
the actual operation of the current and future electricity grid. Additional resolution in
transmission capacity should be considered.
Financial Assumptions
The current financial modeling assumptions seem to apply across the country and
may not be relevant in some organized ISO/RTO wholesale power markets and should be
differentiated by area. For example, in PJM, much of the new entry taking place across
the footprint is merchant only, while in some zones (e.g., Dominion) it is a vertically
integrated entry for the most part. Additionally, the financing assumptions based on
2008-2012 require major reworking as this includes the effect of the worst financial
crisis since the Great Depression, and current financial markets world-wide are showing
much lower returns on debt and equity. To say the current levels are "unsustainable" or
"artificial" is irresponsible and does not reflect experience in Japan, where they are two
decades into a similar crisis and still have lower capital costs as a result. One example of
an assumption that does not make sense is a weighted average cost of capital (WACC)
4-17
-------
for merchants that is higher for a gas combustion turbine (CT) than a coal or nuclear base
load resource or for environmental retrofits on coal units. Reality would indicate that coal
and nuclear are far riskier than a CT, and biasing the cost of capital in favor of retrofits
over some new resources may explain why the MATS analysis found so few coal units
retiring and opting to retrofit as opposed to seeing retirements that have actually taken
place and the prevalence of new build gas units.
IPM Base Case v5.13 includes a number of costs associated with the continued
operation, maintenance and refurbishment of power plants. These include FOM costs,
environmental retrofit costs and life extension costs. The peer review panel finds the
environmental retrofit costs of IPM to be representative of real-world costs and a useful
data set for all modelers. The FOM costs and the increase in those costs for coal and
oil/gas plants with the age of the plant as shown in Table 4-9 of the Base Case v5.13
documentation also appear reasonable. Although not stated in the IPM documentation,
the increase in FOM cost is apparently applied only once to those plants that exist at the
start of the model run (i.e., 2016). The peer review panel recommends that these increases
in FOM costs be also applied to new plants when they reach the ages shown in Table 4-9
and to plants that existed in 2016 as they continue to age into their next decade. The peer
review panel further recommends that increasing FOM costs with age be considered for
combustion turbines, combined cycle, integrated gasification combined cycle (IGCC),
renewables, storage, hydro and fuel cell plants. Nuclear should also be added to Table 4-
9, and increasing FOM costs should be considered for nuclear as well.
Table 4-10 of the Base Case v5.13 documentation presents the costs to double the
lifetime of generation plants in IPM. The panel understands that the life extension cost is
based on FERC Form 1 data and is in addition to environmental retrofit and FOM costs.
After extensive discussion with ICF and EPA experts, it became clear to the panel that
these life extension costs are relatively low because they cover only the replacement of a
limited number of essential plant components that may fail despite regular maintenance
efforts. The panel also understands that these costs have been reviewed extensively by
ICF and EPA. Thus, the panel recommends that the Base Case v5.13 documentation be
revised to convey these facts about the life extension costs. In particular, the
documentation should be expanded to state that these life extension costs
ฆ are in addition to FOM costs,
ฆ cover only the replacement of a limited number of essential plant components
that may fail in spite of regular maintenance efforts, and
4-18
-------
ฆ are applied once for each power plant in only the year during which that plant
would have reached its original life term.
Heat Rate Improvements
The IPM Base Case v5.13 includes no provision for endogenous investment in
heat rate improvements (HRIs) at coal boilers. However, the version of IPM used in the
analysis that supports the Clean Power Plan does allow for coal boilers to invest in a 6%
HRI at a cost. This inconsistent treatment of HRI between the base case and the Clean
Power Plan scenarios is clearly not appropriate. The panel understands that this resulted
from the recent development of the HRI capability that was not available when the base
case was run. Future base cases should include the endogenous HRI option.
One important aspect of the potential for CO2 emissions reductions from
investments in HRI are changes in utilization at boilers that do and do not make
improvements. The modeling framework employed by IPM does capture these effects
and is a strength of the approach. A problem with the IPM approach is that HRI is treated
as binary rather than continuous. It is modeled as an all-or-nothing proposition when in
fact it could be modeled incrementally per the EPA commissioned work by Sargent &
Lundy in 2011 that examined the costs and different opportunities for HRI. Finally, in
wholesale competitive markets, there are academic papers showing HRIs of the fossil
fleet over time, and perhaps these improvements have already been exhausted.
4-19
-------
SECTION 5
REFERENCES
Milligan, M. (2011). Capacity value of wind plants and overview of U.S. experience.
National Renewable Energy Laboratory (NREL). Retrieved from
http://www.nrel.gov/docs/fyllosti/52856.pdf
PJM Interconnection. (2011). Coal capacity at risk for retirement in PJM:
Potential impacts of the finalized EPA cross state air pollution rule and proposed
National Emissions Standards for Hazardous Air Pollutants. Retrieved from
http://pim.eom/~/media/documents/reports/20110826-coal-capacitv-at-risk-for-
retirement.ashx
Short, W., Sullivan, P., Trieu, M., Mowers, M., Uriarte, C., Blair, N., ...Martinez, A.
(2011). Regional Energy Deployment System (ReEDS) technical report NREL/TP-6A20-
46534. Retrieved from http://www.nrel.gov/analvsis/reeds/pdfs/reeds documentation.pdf
Tweed, K. (2011). Demand response and renewables: Too good to be true? Greentech
Media, Inc. Retrieved from http://www.greentechmedia.com/articles/read/demand-
response-and-renewables-too-good-to-be-true
U.S. Energy Information Administration (EIA). (2012). Annual energy outlook 2012 with
projections to 2035. DOE/EIA-0383. Retrieved from
http://www.eia.gov/forecasts/aeo/pdf/0383(2012).pdf
U.S. EPA (2006). Peer Review Handbook, 3rd Edition, EPA Number
EPA/100/B-06/002. Prepared for US EPA. Retrieved from
http://www.epa.gov/peerreview/
U.S. EIA (2014). State Profile and Energy Estimates: Iowa. Retrieved from
http ://www. eia. gov/ state/?sid=IA
U.S. EPA. (2014). National Electric Energy Data System (NEEDS) v.5.13. Retrieved
from http://www.epa.gov/powersectormodeling/BaseCasev513.html
U.S. EPA. (2013). Documentation for EPA Base Case v.5.13: Using the Integrated
Planning Model. Retrieved from
http://www.epa.gov/powersectormodeling/BaseCasev513.html
5-1
-------
APPENDIX A
CONFLICT OF INTEREST ANALYSIS AND BIAS QUESTIONNAIRE
Instructions:
This disclosure form has been developed in accordance with EPA's Peer Review
rd
Handbook, 3 Edition (2006). The questions help identify any conflicts of interest and
other concerns regarding each candidate reviewer's ability to independently evaluate
Integrated Planning Model's (IPMฎ) Base Case v5.13. The Peer Review of the Base Case
(as used in IPM for EPA's power sector analyses) will provide EPA an independent
evaluation of whether IPM meets EPA's goals for analytical transparency, and
recommendations on ways to improve the Base Case and related documentation.
Analytical transparency is a critical component to make it possible for stakeholders and
expert reviewers to examine specific estimated impacts of potential new policies, to
evaluate the technical credibility of EPA's projections and to comment on the
consequences of modeled policies.
Please answer Yes, No or Unsure in response to each question to the best of your
knowledge and belief. If you answer Yes or Unsure to any of the questions, please
provide a detailed explanation on a separate sheet of paper.
Answering Yes or Unsure to any of the questions will not result in disqualification for
serving as a peer reviewer. The responses to the questionnaire will only be used to help
RTI International ensure a balanced, unbiased group of peer reviewers. Responses will
not be publicly released without consent of the candidate. However, if you are selected
to serve on the peer review panel, RTI International will include the signature page as
part of the published peer review record.
It is expected that the candidate make a reasonable effort to obtain the answers to each
question. For example, if you are unsure whether you or a relevant associated party (e.g.,
spouse, dependent, significant other) has a relevant connection to the peer review subject,
a reasonable effort such as calling or emailing to obtain the necessary information should
be made.
A-l
-------
Conflict of Interest Questions
1. Have you had previous involvement with the development of IPM or Base Case
5.13, which is under review? Yes/No/Unsure
2. Is there any connection between IPM and any of your and/or your spouse's (or
other relevant associated party's):
a. Compensated or non-compensated employment, including government
service, during the past 24 months? Yes/No/Unsure
b. Sources of research support and project funding, including from any
government, during the past 24 months? Yes/No/Unsure
c. Consulting activities during the past 24 months? Yes/No/Unsure
d. Expert witness activity during the past 24 months? Yes/No/Unsure
e. Other Financial Connections to IPM holding to be reworked as we
discussed Yes/No/Unsure
3. To the best of your knowledge and belief, is there any direct or significant
financial benefit that might be gained by you or your spouse (or other relevant
associated party) as a result of the outcome of peer review of IPM 5.13?
Yes/No/Unsure
4. Have you made any public statements (written or oral) or taken positions that
would indicate to an observer that you have taken a position on IPM or a closely
related topic under review? Yes/No/Unsure
5. Have you served on previous advisory panels, committees or subcommittees that
have addressed IPM under review or addressed a closely related topic?
Yes/No/Unsure
6. Do you know of any reason that you might be unable to provide impartial advice
on the matter under review or any reason that your impartiality in the matter
might be questioned? Yes/No/Unsure
7. To the best of your knowledge and belief, is there any other information that
might reasonably raise a question about whether you have an actual or potential
personal conflict of interest or bias regarding the matter under review?
Yes/No/Unsure
A-2
-------
Conflict of Interest Analysis and Bias Disclosure Form Signature Page
Please sign below to certify that:
1. You have fully and to the best of your ability completed this disclosure form,
2. You will update your disclosure form promptly by contacting the RTI
International peer review facilitator if relevant circumstances change,
3. You are not currently negotiating new professional relationships with, or
obtaining new financial holdings in, an entity (related to the peer review subject)
which you have not reported, and
4. This signature page, based on information you have provided, and your CV may
be made public for review and comment.
Signature
Date
(Print name)
A-3
-------
APPENDIX B
PEER REVIEWER RESUMES
B-l
-------
Phone: (202)328-5148
Professional Preparation
Carnegie Mellon University
University of Wisconsin,
Madison
Johns Hopkins University
ANTHONY PAUL
Resources for the Future
1616 P Street, N.W., Washington, DC 20036
Email: Paul@rff.org
Civil and Environmental Engineering
Engineering and Public Policy
Economics
Geography and Environmental
Engineering
Appointments
Center Fellow, Resources for the Future
Independent Contractor
Project Assistant, University of Wisconsin, Madison
Independent Contractor
Research Assistant, Resources for the Future
Environmental Engineer, Corning, Inc.
Selected Recent Publications
Fax: (202) 939-3460
B.S., 1997
M.S., 2006
(Expected) Ph.D., 2015
2007 - present
2006 - 2007
2005 - 2006
2000 - 20004
1997-2000
1996
"The Costs and Consequences of Clean Air Act Regulation of C02 from Power Plants" 2014.
Annual Economic Review: Papers and Proceedings. 104(5), 557-562. with D. Burtraw, J. Linn, K.
Palmer.
"Designing by Degrees: Flexibility and Cost-Effectiveness in Climate Policy" 2014, February. RFF
Discussion Paper 14-05.
"Reliability in the Electricity Industry under New Environmental Regulations" 2013, Energy Policy,
w. D. Burtraw, K. Palmer, B. Beaseley, M. Woerman.
"Modeling a Clean Energy Standard for Electricity: Policy Design Implications for Emissions,
Supply, Prices, and Regions," 2013. Energy Economics. Vol 36. with K. Palmer, M. Woerman
"Taxing Electricity Sector Carbon Emissions are Social Cost" 2013, November. RFF Discussion
Paper 13-23-REV. with K. Palmer, B. Beasley.
"Analysis of the Bingaman Clean Energy Standard Proposal," 2014. Review of Energy Economics.
36(1): 113-129. withK. Palmer, M. Woerman.
-------
"Cost-effectiveness and Economic Incidence of a Clean Energy Standard," 2012, September.
Economics of Energy and Environmental Policy. Vol. 1, No. 3, pp. 59-86. withB. Mignone, T.
Alfstad, A. Bergman, K. Dubin, R. Duke, P. Friley, A. Martinez, M. Mowers, K. Palmer, S.
Showalter, D. Steinberg, M. Woerman, F. Wood.
"Secular Trends, Environmental Regulations and Electricity Markets," 2012, July. The Electricity
Journal. Vol. 25, No. 6, pp. 35-47. with D. Burtraw, K. Palmer, M. Woerman.
"Retail Electricity Price Savings from Compliance Flexibility in GHG Standards for Stationary
Sources," 2012. Energy Policy. Vol. 42, pp. 67-77. with D. Burtraw, M. Woerman.
"The Variability of Potential Revenue from a Tax on Carbon," 2012, May. RFF Issue Brief 12-03.
with K. Palmer, M. Woerman.
"The Role of Energy Efficiency Spending in Maryland's Implementation of the Regional
Greenhouse Gas Initiative," (with Matthias Ruth,, Benjamin F. Hobbs, Daraius Irani, Jeffrey
Michaels, Yihsu Chen, Kimberly Ross, Karen Palmer, Erica Myers), Energy Policy. Vol. 38. No. 11,
pp. 6820-6829, 2010.
"Federal Policies for Renewable Electricity: Impacts and Interactions," 2010. Energy Policy. Vol.
39, No. 7, pp. 3975-3991. with K. Palmer, M. Woerman, D. Steinberg.
"From Regions to Stacks: Spatial and Temporal Downscaling of Power Pollution Scenarios," 2010,
May. IEEE Transaction on Power Systems. Vol. 25, No. 2, pp. 1179-1189. with B. Hobbs, M. Hu, Y.
Chen, J. Ellis, D. Burtraw, K. Palmer
"A New Look at Residential Electricity Demand Using Household Expenditure Data," 2010,
November. RFF Discussion Paper 10-57. with H. Fell, S. Li.
Awards and Fellowships
Outstanding Civil Engineering Student, Carnegie Mellon University, 1997
Andrew Carnegie Scholar, Carnegie Mellon University, 1997
-------
MEGHAN COLLEEN MCGUINNESS
202-285-6805
mmcguinness@bipartisanpolicy.org
EDUCATION
S.M., Technology and Policy, 2008. Massachusetts Institute of Technology.
Focus area in Environmental Policy and Economics
Thesis title: "The Effects of Interactions between Federal and State Climate Policies:
Implications for Federal Climate Policy Design"
Completed coursework at Harvard Law School (Environmental Law), and Kennedy
School of Government (Analytic Frameworks for Policy)
B.A., summa cum laude, Economics and Environmental Studies, 2000. Middlebury College.
PROFESSIONAL EXPERIENCE
Associate Director for Energy and Environment
Bipartisan Policy Center Energy Project (formerly National Commission on Energy Policy)
Washington, DC. October 2009 - present.
Led BPC's Electric Grid Initiatives, including engagement and negotiations with task
force members, preparation of recommendations and final reports, and outreach to
stakeholders and policymakers
Led modeling of power sector/energy system impacts of a variety of energy and climate
policies, including EPA regulations, clean energy standards, cap-and-trade programs, and
carbon taxes
Provided legislative and analytical support to Senate Energy and Natural Resources
Committee staff and individual congressional offices
Conducted outreach to DOE, FERC, EPA congressional offices, and industry
associations in support of BPC policy recommendations
Consultant, Environment Group
NERA Economic Consulting. Boston, MA. September 2008 - September 2009.
Advised private and public sector clients on climate policy design and the economics of
environmental regulations
Researcher
MIT Center for Energy and Environmental Policy Research. Cambridge, MA. June 2007 - August 2008.
Estimated C02 abatement in the EU Emissions Trading Scheme (with Denny Ellerman)
for book evaluating the impact of the first phase of the program
Analyzed UK and Spain power sector response to the EU ETS
Analyzed the economic and environmental implications of interactions between U.S.
state and federal climate policies
Policy Analyst
United States Environmental Protection Agency. Clean Air Markets Division, Program Development
Branch. Washington, DC. December 2003 - August 2006.
Advised EPA officials and management on issues related to design and analysis of cap-
and-trade programs
Conducted economic analysis and developed provisions for major regulatory and
legislative cap-and-trade programs, including the Clean Air Interstate Rule, Clean Air
Mercury Rule, and proposed multipollutant legislation
Conducted regulatory outreach to industry stakeholders and state environmental agencies
1 of 4
-------
Drafted guidance on developing new regional and national strategies for air emissions
reductions for Clean Air Act Advisory Committee workgroup
Research Assistant and Senior Research Assistant
Resources for the Future, Washington, DC. September 2000 - November 2003.
Led economic analysis of health benefits from cap-and-trade programs for the power
sector
Analyzed industrial sector response to carbon tax scenarios
Constructed model and led economic benefits analysis for report for Maryland
Department of Natural Resources on the costs and benefits of fish consumption
advisories for mercury
Drafted portions of RFF report on the use of cost-benefit and cost-effectiveness analysis
for valuing health outcomes in policy analysis
Research Assistant
Middlebury College Department of Economics, Middlebury, VT. Spring 2000.
Led econometric analysis for research (with Jon Isham) on the co-production of
community-based water projects
Tutor and Grader
Middlebury College Department of Economics, Middlebury, VT. 1999-2000.
Ran weekly drop-in help sessions for introductory macroeconomics students
Graded weekly homework assignments for introductory microeconomics students
HONORS AND AWARDS
Gold Medal for Exceptional Service, U.S. EPA, 2006.
Bronze Medal for Commendable Service, U.S. EPA, 2005.
Scott A. Margolin '99 Environmental Studies Award for best demonstrating integrated study of the
physical and human environments, Middlebury College, 2000.
Department of Economics Thesis Prize, Middlebury College, 2000.
Phi Beta Kappa, Middlebury College, 2000.
New England Small College Athletic Conference All-Academic, 1999, 2000.
RESEARCH AND PUBLICATIONS
"Power Sector Transition: GHG Policy and Other Key Drivers." With Jennifer Macedonia, Blair Beasley
and Stuart Her. Bipartisan Policy Center. May, 2014.
"A New Organization for Cybersecurity Across the Electric Grid." Bulletin of the Atomic Scientists. April
13,2014.
"The Administration's Clean Energy Standard Proposal: An Initial Analysis." Bipartisan Policy Center
Staff Paper. April, 2011.
"Environmental Regulation and Electric System Reliability." With Jennifer Macedonia, Joe Kruger, and
Lourdes Long. Bipartisan Policy Center Staff Report. April 2010.
2 of 4
-------
"After Summit, We Need the Senate to Act." The Environmental Forum 27 (2). March/April 2010.
"Overlapping State and Federal Climate Programs: Economic and Policy Considerations." Climate
Policy Economic Insights, Issue 2. NERA Economic Consulting. April 2009.
"C02 Abatement in the UK Power Sector: Evidence from the EU ETS Trial Period." With Denny
Ellerman. MIT CEEPR Working Paper WP-2008-010. September.
"The Effects of Interactions between Federal and State Climate Policies." With Denny Ellerman. In
Cap-and Trade: Contributions to the Design of a U.S. Greenhouse Gas Program. MIT CEEPR. Also
released as MIT CEEPR Working Paper WP-2008-004. May.
"Technical Memorandum on Analysis of the EU ETS Using the Community Independent Transaction
Log." With Raphael Trotignon. MIT CEEPR Working Paper WP-2007-013. December.
"Overview: Importance of Sources, Chemistry, Technology and Environmental Factors in Setting U.S.
Mercury Control Policies." With Sam Napolitano. Paper prepared for China Conference on Mercury,
Beijing, China. October, 2005.
"Uncertainty and the Net Benefits of NOx Emissions Reductions from Electricity Generation." 2003.
With Dallas Burtraw and Ranjit Bharvirkar. Land Economics 79 (3). August.
"The Benefits and Costs of Fish Consumption Advisories for Mercury." 2002. With Paul Jakus and Alan
Krupnick. Washington, DC. Resources for the Future Discussion Paper 02-55. October. Also condensed
into white paper for National Oceanic and Atmospheric Administration.
CONSULTING REPORTS
Accounting for Differences in the Timing of Emissions in Calculating Carbon Intensity for the California
Low Carbon Fuels Standard (with D. Harrison and N. Nichols), prepared for the Renewable Fuels
Association, April 2009.
Economic Comments on Nuclear Regulatory Commission DSEIS for Indian Point Energy Center (with D.
Harrison and others), prepared for Entergy Corporation, March 2009.
An Economic Evaluation of Borrowing as a Method to Contain Costs in a Greenhouse Gas Emissions
Cap-and-Trade Program (with D. Harrison and others), prepared for the Electric Power Research
Institute. November 2008.
PRESENTATIONS
"Cybersecurity and the North American Electric Grid: New Policy Approaches to Address an Evolving
Threat." Presentation to Washington State Cybersecurity Summit 2. Tacoma, WA. April 2014.
"Capitalizing on an Evolving Power Sector: Policies for a Modern and Reliable U.S. Grid." Presentation
to the NASEO Energy Policy Outlook Conference. Washington, DC. February 2014.
"The Regional Impacts of a Clean Energy Standard." Presentation to the RFF/EPA workshop A Federal
Clean Energy Standard: Understanding the Important Policy Elements. Washington, DC. July 2011.
"Emissions Trading: Background, Prior Programs, and Implications for a U.S. Carbon Cap-and-Trade
Program." Presentation to the ALI-ABA course Clean Air Law, Policy, and Practice, Washington, DC.
December 2008.
3 of 4
-------
"Implications of the DC Circuit's Decision Vacating the Clean Air Interstate Rule." Presentation to MIT
Joint Program on the Science and Policy of Global Change. Cambridge, MA. August, 2008.
"The Clean Air Interstate, Mercury, and Visibility Rules: Implications for Emissions Markets."
Presentation to the Environmental Markets Association 9th Annual Fall Meeting. San Francisco, CA.
November, 2005.
"Summary of Economic Analyses of the Clean Air Mercury Rule." Presentation to the Center for Clean
Air Policy, Climate and Air Quality Dialogue, Warrenton, VA. February, 2005.
"The Benefits and Costs of Fish Consumption Advisories for Mercury." Paper presented at Camp
Resources X, Wilmington, NC, August, 2002.
4 of 4
-------
Walter D. Short
walterdshort@gmail.com
Professional Experience
2012 - 2014 Retired; consulting on a limited basis on energy modeling issues.
1980 -2011 National Renewable Energy Laboratory (NREL), Golden, CO, Group Manager, Principal
Researcher.
Leadership and management of R&D groups and projects. Technical analysis and
modeling of energy technologies, markets and policies. Broad knowledge of energy
markets, renewable and energy efficiency technologies and climate change issues.
Engineering experience with solar thermal and buildings technologies.
Sample of Accomplishments at NREL:
ฆ Initiated, built up, and led the NREL Energy Forecasting and Modeling Group (EF&M). The
group of 20 professionals contributes to analysis for the wind, solar, biomass, geothermal,
hydrogen, and vehicles programs as well as cross-cutting analysis for the U.S. Department of
Energy (DOE) Office of Energy Efficiency and Renewable Energy.
ฆ Initiator and principal developer of the Regional Energy Deployment System model (ReEDS)
of long-term capacity expansion in the U.S. electric sector. ReEDS was the tool used to
develop the forecast and costs for the 2008 DOE/AWEA study 20% Wind Energy by 2030,
the 2012 DOE Sunshot Vision study, the 2012 DOE Renewable Electricity Futures study, the
revised NREL Western Wind and Solar Integration Study and the DOE Wind Vision 2014
study.
ฆ Conceptualized and led the development of the Renewable Energy Load Matcher model at
NREL to estimate the upper bound on the contribution that wind and photovoltaics can make
to U.S. power generation.
ฆ Conceptualized and contributed to the implementation of the Stochastic Energy Deployment
System model for the market assessment of energy technologies in all economic sectors under
uncertainty. Led team of 6 laboratories and 3 contractors in the model development.
ฆ Identified need, garnered support, hired critical analyst and brought in-house a commercially-
available, electricity production cost model, PROSYM, for analysis of the integration of
renewables and Plug-in Hybrid Electric Vehicles within specific electric utility systems. This
led to the eventual use within the EF&M group of the commercially-available GridView and
PLEXOS models.
ฆ Co-chair and co-author of the Scenarios for a Clean Energy Future. "This 2000 report for the
Interlaboratory Working Group was the subject of two Senate hearings, has been cited in
proposed federal legislation, was the subject of an entire issue of Energy Policy, and has
played a significant role in international climate change debates" (2003 quote from co-chair
Marilyn Brown of ORNL).
ฆ Initiated the Renewable Energy Modeling Forum for comparison of the treatment of
renewable energy in various market models. This forum morphed into REMAP (Renewable
Energy Modeling and Analysis Partnership) which compared model results for various
scenarios. Participants included: EIA, RFF, ICF, EPA, NREL, BNL, OnLocation, etc.
1974 - 1980 Senior energy analyst Stanford Research Institute, Menlo Park, CA.
1971 - 1973 U.S. Army, Lieutenant.
Education and Training
M.S., Operations Research, Stanford University, (1974)
B.S., Mathematics, University of Georgia, (1971)
-------
Curriculum Vitae
Sotkiewicz Page 1
PAUL MICHAEL SOTKIEWICZ
Current Position and Contact Information
Chief Economist, Market Services Division
PJM Interconnection, L.L.C.
955 Jefferson Avenue
Valley Forge Corporate Center
Norristown, Pennsylvania 19403
E-mail: sotkip@pjm.com
Phone: (610) 666-4351
Mobile: (610) 955-2411
Areas of Expertise
Electricity Market Design, Energy Economics, Environmental Economics, Power System Economics, Regulatory
Economics, and Industrial Organization
Education
PhD, Economics, University of Minnesota, 2003
Dissertation: "The Impact of State-Level Public Utility Commission Regulation on the Market for
Sulfur Dioxide Allowances, Compliance Costs, and the Distribution of Emissions"
M.A., Economics, University of Minnesota, 1995
B.A. (High Honors), History/Economics, University of Florida, 1991
Academic and Professional Experience
Chief Economist, 2010 - present, and Senior Economist, 2008-2010, Market Services Division, PJM
Interconnection, LLC
Lead and support market design initiatives related to Demand Response Compensation in the Energy
Market, Shortage Pricing, the RPM Capacity Market, Price Responsive Demand, and any other market
design related issues as they arise
Lead the PJM whitepaper research initiatives related to the impact of environmental regulations on PJM's
markets including potential climate change policy and EPA rulemakings, transmission cost allocation, and
other timely relevant topics
Report to the PJM Board Competitive Markets Committee on market performance and special topics that
highlight recent events or future impacts reading PJM's Energy, Capacity, and Ancillary Services Market
Support State Government Policy, Federal Policy, and Member Services outreach on topical issues facing
PJM, PJM members, state and federal regulators, and state legislatures and Congress.
Lead and support strategic analysis of relevant commodity market, macroeconomic, and environmental
policy trends that affect PJM's Energy and Capacity Markets
Director, 2010 - present, Board of PJM EnviroTrade, LLC
Developed the analytical auction model to perform a single auction for solar renewable energy certificates
(SRECs) with differing state eligibility requirements across multiple jurisdictions
Guided strategic direction of PJM EnviroTrade
Independent Consultant, 2006 - 2007
Provided energy, environmental, and regulatory and government agencies in the United States and globally.
Supported the Florida Department of Environmental Protection implementation of the EPA Clean Air
Interstate Rule in 2006, as well as litigation defense of the rule implementation through emissions
compliance cost modeling and direct testimony as an expert witness during litigation.
-------
Curriculum Vitae
Sotkiewicz Page 2
Independent expert reviewer for the Belize Electricity Commission in its 2007 initial decision on the Annual
Review Proceeding for Belize Electricity Limited examining the logic and robustness of the existing tariff
methodology ranging from rate accounts to rate design, cost allocation, and prudence decision regarding
purchased power contracts.
Wholesale electricity market design training and advice for the Electricity Regulatory Authority of Vietnam in
2007 covering energy, ancillary service, and capacity market design fundamentals as well as market design
experience from the United States, Europe, and Latin America.
Director of Energy Studies, Public Utility Research Center, University of Florida, 2000-2008
Responsible for conducting regulatory training seminars in electricity for the PURC/World Bank Training
course conducted twice yearly.
Develop and deliver training and advising for energy regulators in Latin America and the Caribbean, Africa,
and Southeast Asia as well as in the United States.
Community and Industry outreach activities throughout Florida on Energy related issues.
Research, analysis, and monitoring of energy markets and energy industry regulation and restructuring.
Teaching undergraduate courses in the Government Regulation of Business and Managerial Economics.
Advising doctoral students on their Ph.D. dissertations at the University of Florida and at other universities
Economist, Office of Economic Policy, 1998-1999 and Office of Markets, Tariffs, and Rates, 1999-2000
Federal Energy Regulatory Commission.
Performed market design analysis for filings from Independent System Operators (ISOs), with an emphasis
on the New York Independent System Operator (NYISO) and California Independent System Operator
(CAISO) markets
Briefed Commission staff on aspects of ISO market designs, implementation, and current issues.
Wrote technical inserts for Commission orders.
Monitored electricity markets including centralized ISO markets and bilateral markets.
Conducted applied research on electricity market design for energy and ancillary service markets.
Performed Merger analysis
Instructor, Department of Economics, University of Minnesota, 1992-1998
Taught large lectures of Principles of Microeconomics and Principles of Macroeconomics. In addition to
preparing lectures, exams, and assignments, managed a staff of teaching assistants and graders.
Taught small classes of Principles of Microeconomics, Intermediate Microeconomics, Environmental
Economics and Public Economics. Advised undergraduate students' Senior Projects.
Member of the department committee that helped redesign the Principles of Economics courses to satisfy
the University's Council on Liberal Education International Perspectives requirement.
In 1998 authored Principles of Microeconomics Study Guide for Independent and Distance Learning
Program, University of Minnesota.
Teaching Assistant Coordinator (1995-96) responsible for creating and implementing training seminars for
new teaching assistants and graduate student instructors. This included conducting site visits for
observation of teaching and coordinating the videotaping of teaching in the classroom.
Instructor, Department of Economics, Augsburg College, Minneapolis, Minnesota, 1995-1997
Taught classes in Money and Banking, Labor Economics and Principles of Microeconomics.
Teaching Assistant, Department of Economics, University of Minnesota, 1991-1992
Led recitation sections for Principles of Microeconomics.
Publications and Book Chapters
Smith, J.C.; Beuning, S.; Durrwachter, H.; Ela, E.; Hawkins, B.; Kirby, B.; Lasher, W.; Lowell, J.; Porter, K.;
Schuyler, K.; Sotkiewicz, P.; "The Wind at Our Backs", IEEE Power and Energy Magazine, Volume 8, Issue 5,
September-October 2010.
Holt, Lynne; Sotkiewicz, Paul M.; and Berg, Sanford V.; "Nuclear Power Expansion: Thinking About
Uncertainty", Electricity Journal, Volume 23, Issue 5, July 2010.
-------
Curriculum Vitae
Sotkiewicz Page 3
Helman, Udi; Singh, Harry, and Sotkiewicz, Paul; "RTOs, Regional Electricity Markets, and Climate Policy", in
Generating Electricity in a Carbon Constrained World, Fereidoon P. Sioshansi, Editor, Academic Press,
September 2009.
Holt, Lynne; Sotkiewicz, Paul; and Berg, Sanford; "(When) to Build or Not to Build?: The Role of Uncertainty in
Nuclear Power Expansion", Texas Journal of Oil, Gas, and Energy Law, (3)2,2008.
Sotkiewicz, Paul M. and Vignolo, J. Mario, "Towards a Cost Causation Based Tariff for Distribution Networks
with DG." IEEE Transaction on Power Systems, Vol. 22, No. 3, August 2007, pp. 1051-1060.
Sotkiewicz, Paul and Vignolo, Jesus Mario. "Distributed Generation." The Encyclopedia of Energy Engineering
and Technology, Vol. 1, pp 296-302. Ed. Barney Capehart. New York: CRC Press, Taylor and Francis Group,
2007.
Sotkiewicz, Paul. "Emissions Trading." The Encyclopedia of Energy Engineering and Technology, Vol. 1, pp.
430-437. Ed. Barney Capehart. New York: CRC Press, Taylor and Francis Group, 2007.
Vignolo, Jesus Mario and Sotkiewicz, Paul M., "Towards Efficient Tariffs for Distribution Networks with
Distributed Generation", Cogeneration and On-site Power Production, November-December 2006, pp. 67-75.
Jamison, Mark A. and Sotkiewicz, Paul M., "Defining the New Policy Conflicts," Public Utilities Fortnightlyฑ July
2006, pp. 36-40, 50.
Sotkiewicz, Paul M. and Vignolo, Jesus Mario "Nodal Pricing for Distribution Networks: Efficient Pricing for
Efficiency Enhancing DG." IEEE Transaction on Power Systems, Vol. 21, No. 2, May 2006, pp. 639-652.
Sotkiewicz, Paul M. and Vignolo, Jesus Mario "Allocation of Fixed Costs in Distribution Networks with
Distributed Generation," IEEE Transaction on Power Systems, Vol. 21, No. 2, May 2006, pp. 1013-1014.
Sotkiewicz, Paul M., and Lynne Holt, "Public Utility Commission Regulation and Cost Effectiveness of Title IV:
Lessons for CAIR." Electricity Journal 18(8): 68-80, October 2005.
O'Neill, Richard P., Sotkiewicz, Paul M., Hobbs, Benjamin F., Rothkopf, Michael H., and Stewart, William R.
Jr., "Efficient Market Clearing Prices in Markets with Non-Convexities." European Journal of Operational
Research, Volume 164, Issue 1, 1 July 2005, Pages 269-285.
Vignolo, J. Mario and Sotkiewicz, Paul M., "Distribution Network Loss Allocations with Distributed Generation
Using Nodal Prices." (Proceedings of the Seventh IASTED International Conference on Power and Energy
Systems, December 1, 2004 Clearwater Beach, FL).
O'Neill, Richard P., Helman, Udi, Sotkiewicz, Paul M., Rothkopf, Michael H., and Stewart, William R. Jr.,
"Regulatory Evolution, Market Design, and the Unit Commitment Problem" The Next Generation of Unit
Commitment Models, B. Hobbs, M. Rothkopf, R. O'Neill, and H.P. Chao editors. 2001.
Sotkiewicz, Paul M. "Opening the Lines", Forum for Applied Research and Public Policy, Special Issue on the
Role of Public Power in Utility Restructuring, Summer 2000, pp. 61-64.
Selected Working Papers and Unpublished Manuscripts
O'Neill, Richard P., Sotkiewicz, Paul and Rothkopf, Michael. "Equilibrium Prices in Exchanges with Non-convex
Bids." PURC Working Paper, January 2006, updated September 2007.
Sotkiewicz, Paul M. and Vignolo, Jesus Mario "The Value of Intermittent Wind DG under Nodal Prices and
Amp-mile Tariffs, PURC Working Paper, December 2006.
Sotkiewicz, Paul M. "Cross-Subsidies That Minimize Electricity Consumption Distortions," Working Paper,
Public Utility Research Center, University of Florida, September 2003, revised February 2005.
"Price Tests for Entry into Markets in the Presence of Non-Convexities", with R.P. O'Neill, B. Hobbs, W. Stewart,
-------
and M. Rothkopf., mimeo December 2003.
Curriculum Vitae
Sotkiewicz Page 4
Sotkiewicz, Paul M. "Considerations for the Design of Restructured Electricity Markets and Institutions,"
Working Paper, Public Utility Research Center, University of Florida, November 2002, revised September 2003.
Helman, Udi and Sotkiewicz, Paul M. "Market Design in the Northeastern U.S. Wholesale Electricity Markets:
Events and Issues, 1999-Mid 2000". Mimeo, Office of Markets, Tariffs, and Rates, Federal Energy Regulatory
Commission, June 2000.
"Design Issues for Short-Term Electricity and Ancillary Service Markets" Joint with R.P. O'Neill, U. Helman, J.
Cardell, B. Hobbs, W. Stewart, and M. Rothkopf. Mimeo, Office of Economic Policy, Federal Energy Regulatory
Commission, January 2000.
Grant Awards
2007 Fulbright Senior Specialist Grant in Economics with a specific request for expertise in electricity
markets, electricity regulation, and distribution tariff design, Universidad de la Republica,
Montevideo, Uruguay. Grant award $2,800 plus travel expenses.
2007 Principal Investigator, PPIAF/World Bank Grant to conduct two on-site training courses on the
regulation of the electric power sector and on independent power producers and power purchase
agreements for the Electricity Authority of Cambodia. Grant award $59,900.
Honors and Awards
2006 "Efficient Market Clearing Prices in Markets with Non-Convexities" published in European Journal
of Operational Research received New Jersey Policy Research Organization Bright Idea Research
Award in Decision Sciences.
2003 Transportation and Public Utilities Group, Ph.D. Utilities Dissertation Award for "The Impact of
State-Level Public Utility Commission Regulation on the Market for Sulfur Dioxide Allowances,
Compliance Costs, and the Distribution of Emissions"
1992-97 Distinguished Instructor, Department of Economics, University of Minnesota
1995-96
1994-95 Walter Heller Award for Outstanding Teaching of Economic Principles, Department of Economics,
1993-94 University of Minnesota
1992-93
1991-92 Distinguished Teaching Assistant, Department of Economics, University of Minnesota
1991 Phi Beta Kappa, University of Florida
Ph.D. Dissertations Supervised and Dissertation Committees
2008 Committee Member (unofficial). Joshua David Kneifel, "Essays in Renewable Energy and
Emissions Trading", Department of Economics, University of Florida.
2007 Thesis Director. Jesus Mario Vignolo, "Cost-causality Based Tariffs for Distribution Networks with
Distributed Generation," Universidad de la Republica Instituto de Ingenieria Electrica.
2005 External Examiner. Juan Daniel Oviedo, "Regulation of Regional Monopolies in Natural Gas
Markets", Midi Pyrenees School of Economics, Universite de Toulouse I.
Referee and Review Experience
-------
IEEE Transactions on Power Systems
Curriculum Vitae
Sotkiewicz Page 5
Ecological Economics
Environmental Science and Technology
Determining the Economic Value of Coastal Preservation and Restoration on Critical Energy Infrastructure,
prepared for The Economic and Market Impacts of Coastal Restoration: America's Wetland Economic Forum II,
September 28, 2006 Washington, DC
National Research Council of the National Academy of Sciences report entitled "Changes in New Source
Review Programs for Stationary Sources of Air Pollutants", February 2006
California Energy Commission (CEC) Energy Innovations Small Grant (EISG) Program
Energy Journal
Journal of Environmental Economics and Management
IEEE PES Letters
IASTED International Journal of Power and Energy Systems
The Next Generation of Unit Commitment Models B. Hobbs, M. Rothkopf, R. O'Neill, and H.P. Chao editors
2001.
Professional Affiliations
American Economic Association
International Association for Energy Economics
Association of Environmental and Resource Economists
-------
CURRICULUM VITAE
(as of January 1 ,2014)
John P. Weyant
BACKGROUND
ADDRESSES
Room 260 Huang Engineering Center
Stanford University
Stanford, CA 94305-4121
(650) 723-3506
861 Allardice Way
Stanford, CA 94305
(650) 494-3570
PROFESSIONAL INTERESTS
Application of quantitative methods to policy development and strategic planning.
ACADEMIC APPOINTMENTS
STANFORD UNIVERSITY
Professor - Dept. of Management Science and Engineering (1/2000 - present)
Professor- Engineering-Economic Systems and Operations Research (9/96-12/99)
Professor - Department of Engineering-Economic Systems (9/89-8/96)
Associate Professor - Department of Engineering-Economic Systems (9/84-8/89)
Senior Research Associate - Department of Operations Research (9/80-8/84)
Research Associate - Department of Operations Research (6/77-8/80)
EDUCATION
HARVARD UNIVERSITY:
Postdoctoral Fellow - John F. Kennedy School of Government (1976-7)
Research Topic: Quantitative Models in Energy Policy
UNIVERSITY OF CALIFORNIA, BERKELEY:
Doctor of Philosophy - Management Science (1976)
Minor Fields: Economics, Operations Research, Organization Theory
RENSSELAER POLYTECHNIC INSTITUTE:
Master of Science in Operations Research and Statistics (1971)
Master of Science in Management (1970)
B.S./M.S. in Aeronautical Engineering and Astronautics (1969/70)
1
-------
RESEARCH EXPERIENCE
STANFORD UNIVERSITY
PROGRAM ON INTEGRATED ASSESSMENT MODELING DEVELOPMENT
DIAGNOSTICS AND INTERCOMPARISONS: Cutting edge research on uncertainty,
technology dynamics, and fine scale climate impacts, model diagnostics development, and
scenario ensemble construction for IAM community- Lead PI (20 Pis at four institutions) 8/10-
ENERGY MODELING FORUM: The EMF conducts systematic comparative studies of
energy-economic models applied to policy problems of current interest.
Director (9/84-present)
Executive Director (J.L. Sweeney, Director): 1/83 - 8/84
Deputy Director (J.L. Sweeney, Director): 12/79 - 12/82
Associate Director (J.L. Sweeney, Director): 8/78 - 11/79
Research Staff (W.W. Hogan, Director): 6/77 - 11/79
SNOWMASS SUMMER WORKSHOPS ON INTEGRATED ASSESSMENT OF CLIMATE
CHANGE: Inter-disciplinary workshops on critical issues for integrated assessment - 6/95-
PRECOURT INSTITUTE FOR ENERGY EFFICIENCY:
Deputy Director (J.L. Sweeney, Director): 9/07-present
GLOBAL CLIMATE AJND ENERGY PROJECT
Staff (Lynn Orr, Director): 9/02-9/06
GENERAL MOTORS COLLABORATIVE LABORATORY ON WORK SYSTEMS
Co-Director (With Arthur Veinott): 9/00-9/03
INTERNATIONAL ENERGY PROGRAM: The IEP conducted studies on international energy
policy issues of current interest.
Research Staff (A.S. Manne, Director). 7/81 - 8/85
Research Staff (H.S. Rowen, Director): 9/77 - 6/81
COMBINING ENERGY MODELS PROJECT: The CEM project attempted to develop a set of
rules for the combination of energy models of different types.
Research Staff (W.W. Hogan & L.J. Lau, Principal Investigators): 8/77 - 4/81
HARVARD UNIVERSITY
ENERGY AND ENVIRONMENTAL POLICY CENTER: 4/76 - 6/77
RAND CORPORATION
Air Force Energy Problems (for J.R. Gebman): 6/75 - 3/76
Air Force R&D Planning (for G.K. Smith): 6/74 - 9/74
Air Quality Modeling (for J.R. Gebman): 6/72 - 9/72
Aerodynamics Computer Module (for G.K. Smith): 6/70 - 9/70
U.C.. BERKELEY
ENERGY & RESOURCES PROGRAM:
Post Graduate Researcher (J.P. Holdren, Director): 9/74-6/75
2
-------
OPERATIONS RESEARCH CENTER:
Postgraduate Researcher (R.C. Grinold, Director): 10/72-12/73
RENSSELAER POLYTECHNIC INSTITUTE
URBAN & ENVIRONMENTAL STUDIES PROGRAM:
Graduate Assistant (W.A. Wallace, Director): 1/71 - 9/71
TEACHING EXPERIENCE
STANFORD UNIVERSITY
COURSES TAUGHT:
Energy Policy and Strategy Modeling (EES 283 & EES&OR 483 & MS&E 473)
1979-2001
International Environmental Policy (MS&E 92Q): 2001-present
Policy and Strategy Analysis (MS&E 190): 2006-
Public Policy Analysis (MS&E 290): 2000-2006 (with William Perry)
Climate Policy Analysis (MS&E 294): 2004-present
Energy Policy Analysis (MS&E 295): 2005-present
Department of Management Science & Engineering
Seminar on Business & Technology
Department of Engineering-Economic Systems 1987 - 1993
Models and Applications of Operations Research in Society (O.R. 50/150)
Department of Operations Research 1979-1980
Contingency Planning - The World Oil Market
(O.R. 348 A, B, and C, with A. S. Manne)
Department of Operations Research 1981 - 1982
Seminar in Energy/Economic Modeling
(Econ 360ABC, with L.J. Lau and H.G. Huntington)
Fall 1983, Winter & Spring 1984
Financial Decisions, IE 235
Department of Industrial Engineering
Winter 1990, Fall 1990, Winters 1993, 1994 &5
DISSERTATION COMMITTEES (Total - 168; Principal Adviser - 39):
Principal Advisees (Completed Only)
Gregory Hamm (Engineering-Economic Systems - 1986)
Xia Shi (Engineering-Economic Systems - 1989)
Douglas Robinson (Engineering-Economic Systems - 1990)
Peter Lilienthal (Engineering-Economic Systems- 1991)
3
-------
Hean-Lee Poh (Engineering-Economic Systems -1991)
Sylvia Kwan (Engineering-Economic Systems - 1994)
Eric Johnson (Engineering-Economic Systems - 1994)
Ming-Fai Sit (Engineering-Economic Systems - 1994)
Elisabeth Browne (Engineering-Economic Systems - 1995)
Thomas Hoff (Engineering-Economic Systems and Operations Research - 1996)
Robert Earle (Engineering-Economic Systems and Operations Research - 1996)
Chi-Peng Chu (Engineering-Economic Systems and Operations Research - 1996)
Shu-Cheng Liu (Engineering-Economic Systems and Operations Research - 1997)
Enrique Garza-Escalante (Eng.-Economic Systems and Operations Research- 1998)
Quingxuan Meng (Engineering-Economic Systems and Operations Research - 1998)
Michael Hsu (Engineering-Economic Systems and Operations Research - 1999)
Akira Maeda (Engineering-Economic Systems and Operations Research - 1999)
Karen (Cushing) Sepucha (Eng.-Economic Systems and Operations Research - 1999)
Karl Knapp (Eng.-Economic Systems and Operations Research - 1999)
Kevin Zhu (Eng.-Economic Systems and Operations Research - 1999)
Antje Kann (Eng.-Economic Systems and Operations Research - 2000)
Wenlong Weng (Eng.-Economic Systems and Operations Research - 2001)
Michelle Freed (Eng.-Economic Systems and Operations Research - 2001)
John McConnell (Eng.-Economic Systems and Operations Research - 2001)
Erin Baker (Eng.-Economic Systems and Operations Research - 2002)
Jochen Kleinknecht (Eng.-Economic Systems and Operations Research - 2002)
Fehmi Ashaboglu (Eng.-Economic Systems and Operations Research - 2002)
Kazuhiro Ninomiya (Management Science and Engineering - 2003)
Tao Yao (Management Science and Engineering - 2005)
Albert Whangbo (Management Science and Engineering - 2005)
Geoff Blanford (Management Science and Engineering - 2006)
Oytun Eskiyenenturk ((Management Science and Engineering - 2006)
Leslie Holmes Hummel (Interdisciplinary Program on Environment and Resources-2006)
Katherine Calvin (Management Science and Engineering-2007)
Dhruv Sharma (Management Science and Engineering -2010)
Nikit Abhyankar (Interdisciplinary Program on Environment and Resources -2013)
Danny Cullenward (E-IPER & JD Stanford Law School- 2013)
John Bistline (Management Science and Engineering - 2013)
Jordan Wilkerson (Management Science and Engineering - 2014)
ENGINEERS THESIS ADVISEES (3):
Teodoro Myslabodski (Operations Research)
Anousheh Alamzad (Engineering-Economic Systems)
Vincent Lui (Engineering-Economic Systems and Operations Research)
4
-------
OTHER ACTIVITIES
PROFESSIONAL SOCIETY MEMBERSHIPS (Year JoinecT)
The Operations Research Society of America (1974)
The Institute of Management Sciences (1976)
The American Economic Association (1976)
The International Association of Energy Economists (1978)
The Association for Public Policy Analysis and Management (1979)
The Econometric Society (1981)
The Academy of Political Science (1981)
The Mathematical Programming Society (1982)
The American Statistical Society (1983)
The Association of Energy Engineers (1984)
JOURNALS REFEREED
Energy Economics (Editor in Chief 2002-)
Interfaces (Area Editor: Public Policy 1988 - 1991)
Petroleum Management (Editorial Advisory Board 1984 - 1991)
Operations Research (Associate Editor 1983-1987)
The Energy Journal (Editorial Board 1985 - present)
Environmental Modeling and Assessment (Editorial Board 1995-present)
Management Science
Resources and Energy
International Studies Quarterly
Resources and Energy
Economic Inquiry
The Journal of Economic Dynamics and Control
The European Journal of Operational Research
The Journal of Policy Modeling
Energy Policy
Policy Studies
CONFERENCE ORGANIZATION
Program Chairman: International Association of Energy Economists,
Sixth Annual North American Meeting, San Francisco, November 1984 (36 Sessions & 5
Plenary Addresses)
Co-General Chairman: International Association for Energy Economics,
Eleventh Annual North American Meeting, Los Angeles, October 1989
Program Advisory Committee: International Association of Energy Economists
Seventh Annual North American Meeting, Philadelphia, December 1985
Eighth Annual North American Meeting, Cambridge, MA, November 1986
Ninth Annual International Meeting, Luxembourg, July 1988
Tenth Annual International Meeting, Caracas, June 1989
Individual Sessions Organized at Professional Society Meetings:
5
-------
Institute for Operations Research and Management Science-17
American Economics Association - 3
International Association for Energy Economics - 8
MAJOR ADVISORY BOARDS AND COMMITTEE MEMBERSHIPS
-N.R.C. Committee on Nuclear and Alternative Energy Systems (1975-77)
-The Institute for the Future - Advisory Board on NSF Interactive Modeling (1978-82)
-Scientists' Institute for Public Information - Oil Emergency Task Force (1980)
-Energy Research Commission of Sweden-Research Program Evaluation Comm. (1982-84)
-Office of Technology Assessment - Advisory Board on U.S. Gas Supply (1983-84)
-Chairman: E.P.A. Peer Review Panel on Acid Deposition Research (1983-87)
-Advisory Board - Electric Power Research Institute Visibility Valuation Project (1983-87)
-National Academy of Sciences: Committee on The Gas Research Institute (1985-87)
-California Public Utilities Comm.-Advisory Board on Utility Model Reviews (1986-88)
-Advisory Board on Utility Model Reviews, California Public Utilities Commission (1986-89)
-Peer Reviewer Final Assessment Report-National Acid Precipitation Assessment Prog. (1990)
-Secretary of Energy Advisory Board (1992-1993)
-Nat. Renewable Energy Lab., Analytic Studies Division Advisory Board (1993-1996)
-Convening Lead Author: Intergovernmental Panel on Climate Change Second Assessment
Report - Working Group m on Economic and Social Dimensions (1993-1995)
-Lead Author: The Contribution of the Social Sciences to Global Climate Change Policy:
A State of the Art Report (1993-1996)
-Collaborating Faculty Member: International Institute for Applied Systems Analysis (1994-)
-Review Panel, Program on Climate Change, Division of Energy Research,
U.S. Department of Energy (1994-)
-Director: Snowmass Workshops on Climate Change Impacts and Integrated Assessment
(1995-)
-Advisory Board, Consortium on International Earth Sciences Information Network (1995-2002)
-Participant, Forum on Global Change, Joint Program on the Science and Economics of Global
Change, Massachusetts Institute of Technology (1995-).
-Lead Author, Intergovernmental Panel on Climate Change, Special Report on Emissions
Scenarios (1996-1998).
-Advisory Board: Yale/National Bureau of Economic Research Program on Economics and
Policy Issues in Global Warming National Science Foundation Center (1996-)
-Adviser: National Institute for Environmental Studies, Japan (1997-)
-Chairman: External Review Panel- Electric Power Research Institute: Environment Division
-Coordinating Lead Author, Intergovernmental Panel on Climate Change, Third
Assessment Report, Working Group II on Climate Impacts and Working Group in on
Climate Change Mitigation (1998-2001).
-Independent Expert Review Panel, Energy Information Administration Report on Likely Costs
and Energy Sector Impacts of the Kyoto Protocol on Climate Change Policy. Report Prepared
for the U.S. House of Representatives Committee on Science (1998).
-Chairman, Lawrence Berkeley Laboratory, Director's Review Panel, Environmental Energy
Technologies Division (1998).
-Co-founder, The Boathouse Group of Climate Negotiators fro the twelve largest carbon emitting
6
-------
countries (2003-).
-Review Editor, Intergovernmental Panel on Climate Change, Assessment Report Number Four,
(2004-2007).
- American Statistical Association Advisory Board for the Energy Information Administration,
U.S. Department of Energy (2006-)
-California Air Resources Board, (ARB) -Economic and Technology Advancement Advisory
Committee (ETAAC) (2007-).
- World Bank - Advisory Board for Latin American and Carribean Region, (2007-)
- World Bank - Academic Advisory Board for World Development Report - 2010.
-National Academy of Sciences America's Climate Choices Study
- Review Editor, Inter-Governmental Panel on Climate Change.
- Chairman, Scientific Advisory Board, European Commission's AMPERE Project on Integrated
Assessment Model Diagnostics.
-Steering Committee, Latin American Modeling Project
-Member: Economic and Technology Advancement Advisory Committee to the California Air
Resources Board on Implementation of AB32 the Climate Solutions Act of 2006
- American Statistical Association Review Committee, the Energy Information Administration -
U.S. Department of Energy (2006-2012).
-Review Panel, Program on Integrated Assessment of Climate Change, Division of Energy
Research, U.S. Department of Energy.
-Co-Editor in Chief, Energy Economics
-Editorial Board, The Energy Journal
-Chairman: Steering Committee of the Integrated Assessment Modeling Consortium (IAMC)-55
International Member Institutions.
-Collaborating Faculty Member: International Institute for Applied Systems Analysis.
-Participant, Forums on Global Change, Joint Program on the Science and Economics of Global
Change, Massachusetts Institute of Technology.
-Adviser: National Institute for Environmental Studies, Japan.
-Member, Steering Committee, The International Project on "Developing a Technology Strategy
for Dealing With Climate Change," Lead by Pacific Northwest National Laboratory for An
International Consortium of Government and Industry Sponsors.
-Chairman, Scientific Advisory Board, European Commission's ADVANCE Project on Cutting
Edge Research for Integrated Assessment Modeling.
- Biological and Environmental Research Advisory Board, Office of Science, U.S. Department of
Energy (2014-2016).
7
-------
OTHER UNIVERSITY ACTIVITIES
Freeman Spogli Institute for International Studies
SENIOR FELLOW (1998-)
Woods Institute for the Environment
SENIOR FELLOW (2007-)
Precourt Institute for Energy
SENIOR FELLOW (2010-)
Northeast Asia - United States Forum on International Policy
FELLOW (1982 - 1985)
Engineering Library Committee
MEMBER 1986-1995
CHAIRMAN 1988-1991
University Committee on Libraries
MEMBER 1988-1991
OFFICES IN PROFESSIONAL SOCIETIES
President, Northern California Chapter-International Association of Energy Economists
Vice President, U.S. Institute of Energy Economics
Vice President, International Association for Energy Economics
Chairman, IAEE Nominating Committee for 1989
Selection Committee - Best Paper in the Energy Journal for 1997 - 1998
CONSULTING
Rand Corporation
Electric Power Research Institute
U. S. Department of Energy
Environmental Protection Agency
Pan Heuristics
Applied Decision Analysis
Science Applications, Inc.
Charles River Associates
U.S. Arms Control and Disarmament Agency
United States Synthetic Fuels Corporation
National Acid Precipitation Assessment Program
California Energy Commission
Federal Trade Commission
8
-------
HONORS and AWARDS
STANFORD UNIVERSITY
Adelman Frankel Award for 2008, U.S. Association for Energy Economics, for Unique and
Enduring Contributions to the Field (Ninth Individual Award in International Competition)
Nobel Peace Prize, 2007, Significant Contributions to the Intergovernmental Panel on Climate
Change Award.
HARVARD UNIVERSITY:
POSTDOCTORAL FELLOW, National Science Foundation, Harvard University (1976-1977)
UNIVERSITY OF CALIFORNIA. BERKELEY:
REGENTS FELLOW (1972-1974)
DISTINCTION Plus, Ph.D. Qualifying Examination in Economic Theory (June 1973)
DISTINCTION Plus, Ph.D. Qualifying Examination in Management Science (Dec. 1973)
PUBLICATIONS
JOURNAL ARTICLES (last five years only)
Allen A. Fawcett, Katherine V. Calvin, Francisco C. de la Chesnaye, John M. Reilly, John P.
Weyant, " Overview of EMF 22 U.S. Transition Scenarios," Energy Economics, Vol. 31,
Supplement 1, Pages S198-S211, Available online 28 October 2009.
Leon Clarke, John Weyant, Energy Economics, Introduction to the EMF 22 Special Issue on
Climate Change Control Scenarios, Vol. 31, Supplement 1, Page S63,Available online 26
October 2009.
Richard H. Moss, .., J.P. Weyant.., The Next Generation of Scenarios for Climate Chnage
Research and Assesment, Nature, Vol. 463, No. 7282, 11 February, 2010, pp 747-756.
D. P. van Vuuren,, James A. Edmonds, Mikiko Kainuma, Kevwan Riahi and John P. Weyant
(2011) Special Issue: The Representative Concentration Pathways in Climatic Change, Climatic
Change: 109:1-2, 241 pp. Introduction and Overview paper plus editor of whole volume.
John P. Weyant, Accelerating the development and diffusion of new energy technologies:
Beyond the "Valley of Death", Energy Economics, Volume 33, Issue 4,
July 2011, pp. 674-682.
P. A. Matson, T. Dietz..., J. P. Weyant, et al., America's Climate Choices; Advancing the
Science of Climate Change, Board on Atmospheric Sciences and Climate, National Academy of
Sciences, 2010.
9
-------
Competing or Coordinating: IT R&D Investment Decision Making Subject to Information Time
Lag (Tao Yao, John Weyant, Baichun Feng*), Information Technology and Management, Vol.
12 (3), 213-228, 2011
L.E. Clarke, L.E., V. Krey, J.P. Weyant, Regional energy system variation in global
models: results from the Asian Modeling Exercise scenarios. Energy Econ. 34 Supplement 3,
S293-S305, 2012.
Bistline, J. E. and J. P.Weyant (2013). Electric Sector Investments under Technological and
Policy-Related Uncertainties: A Stochastic Programming Approach. Climatic Change.
J. Weyant, B. Knopf, E. De Cian, I. Keppo, and D. van Vuuren, Introduction to the Emf 28 Study
On Scenarios For Transforming The European Energy System, Climate Change Economics, Vol.
4, Suppl. 1, 2013. (also, editors of whole special issue volume).
Kriegler E, J.P. Weyant, et al. (2013). The role of technology for achieving climate policy
objectives: overview of the EMF 27 study on global technology and climate policy. Special Issue
of Climatic Change, (forthcoming) (also editors of two volume special issue).
Sugiyama, M., O. Akashi, K. Wada, A. Kanudia, J. Li, and J. Weyant, 2013. Energy Efficiency
Potentials in Global Climate Change Mitigation; A comparison of modeling approaches, in
Special Issue of Climatic Change, forthcoming.
Rose, S., E. Kriegler, A. Popp, and J. Weyant, 2013. Bioenergy in energy transformation
and climate management, Special Issue of Climatic Change, forthcoming.
Weyant, J.P., Fawcett, A.A., Clarke, L.C. eds. Forthcoming. Energy Modeling Forum 24.
Energy Journal Special Issue.
Allen A. Fawcett, Leon C. Clarke*, Sebastian Rausch, John P. Weyant; Overview of EMF 24
Policy Scenarios, The Energy Journal, in press.
Clarke, L.C., Fawcett, A.A., McFarland, J., Weyant, Overview of the EMF 24 Technology
Scenarios, Energy Journal, in press.
10
-------
APPENDIX C
PEER REVIEW PANEL CHARGE
C-l
-------
Charge to the Peer Reviewers:
Integrated Planning Model (IPM) Base Case Version 5.13 Peer Review
Peer Review of Integrated Planning Model Formulation
Background
The U.S. Environmental Protection Agency uses the Integrated Planning Model (IPMฎ)
developed by ICF International (ICF) to project the impact of potential emissions policies
on the U.S. electric power sector in the 48 contiguous states and the District of Columbia
over the 2016-2050 time horizon. IPM is a multi-regional, dynamic, deterministic linear
programming model of the U.S. electric power sector. It provides forecasts of least cost
capacity expansion, electricity dispatch, and emission control strategies while meeting
energy demand and environmental, transmission, dispatch, and reliability constraints.
EPA uses IPM to evaluate the cost and emissions impacts of alternative policies to limit
emissions of sulfur dioxide (SO2), nitrogen oxides (NOx), carbon dioxide (CO2), and air
toxics including mercury (Hg) and hydrochloric acid (HC1) from the electric power
sector's operations. IPM's deterministic, linear programming formulation not only
supports a large scale model with the required level of detail, but it also allows model
runs to be performed, quality assured, and delivered within turnaround times (2-3 days)
required by EPA and the many decision makers who use IPM results for policy analysis.
The level of detail of IPM outputs at the state, regional and national adds levels
transparency to EPA technical analyses by making it easy for stakeholders and expert
reviewers to examine the specific estimated impacts of potential new policies, to evaluate
the technical credibility of EPA's projections, and to comment on the consequences of
modeled policies.
EPA's Needs for a Power Sector Model
To support periodic policy and regulatory analyses of the electric power sector, EPA
needs to routinely access a model of the electric power sector capable of analyzing the
projected impact of environmental policies in the 48 contiguous states and the District of
Columbia. The model must be able to evaluate the costs and impacts of proposed
environmental programs affecting the power sector, such as programs limiting emissions
of sulfur dioxide (SO2), nitrogen oxides (NOx), carbon dioxide (CO2), hydrogen chloride
(HC1), and mercury (Hg). The model must be able to provide forecasts of the future
impacts of a wide variety of potential environmental policies affecting generation
capacity expansion and retirements, electricity dispatch, fuel use, and emission control
strategies expected to be adopted in order to meet potential changes in energy demand
and environmental, transmission, dispatch, and reliability constraints.
The model must incorporate sufficient engineering, financial and geographical detail, as
well as the current status of the power sector, in order to provide EPA the ability to
analyze emission control options encompassing a broad array of emission control
technologies along with emission reductions through fuel switching, changes in capacity
C-2
-------
mix and electricity dispatch strategies. The model must be able to capture the complex
interactions among the electric power, fuel, and environmental markets.
The power sector model that EPA uses must meet EPA's goals for transparency,
scientific integrity, technical accuracy, peer review, and public participation in regulatory
development proceedings. One critical component of achieving all these goals is periodic
peer review of the power sector model used by EPA. The peer review of the model must
follow the procedures and standards of EPA's current policies and guidance on peer
review as well as the Office of Management and Budget's Final Information Quality
Bulletin for Peer Review (2004).5 Other necessary components of achieving EPA's
transparency and public participation goals include the model documentation, input data
used by the model, and sufficient model results information (i.e., output data) to support
communication and review objectives.
Purpose of Peer Review
In November 2013 EPA released a new base case (designated EPA Base Case
v.5.13). This new EPA base case incorporates important structural improvements and
data updates with respect to EPA's previous base case version (v.4.10).6
Base cases serve as the starting point against which EPA compares potential policy
scenarios. Base Case v.5.13 is a "business-as-usual" projection of electricity sector
activity that takes into account only those Federal and state air emission laws and
regulations whose provisions were either in effect or enacted and clearly delineated at the
time the EPA base case was finalized in August 2013.
This peer review will focus on the use of EPA Base Case v.5.13 in IPM with the intent of
obtaining expert feedback on the adequacy of using the base case and IPM to meet EPA's
needs for estimating the economic and operational behavior of the power sector under
alternative emissions policies over a modeling time horizon of 2016-2050.
Among the goals of the peer review are:
(a) given the scope and intended purposes of EPA's base case v.5.13, evaluate
whether the IPM analytical framework, assumptions and applications of data are
appropriate for meeting EPA's needs for a policy base case and power sector
model to be used to estimate the impacts of emissions policy alternatives;
5 EPA's current peer review guidance contained in the EPA Peer Review Handbook, 3rd Edition (2006), and
the Addendum to the EPA Peer Review Handbook, 3rd Edition (2009). Both EPA documents, as well as
the OMB Bulletin are available at http ://www. epa. gov/peerreview/. EPA is currently completing a new
version of the EPA Peer Review Handbook; the 4th edition PRH is scheduled for release in 2014.
6 The version number (4 or 5) indicates a major update affecting fundamental projection
inputs or outcomes (such as fuel resource assumptions or regions). The portion of the
version name after the 'dot' (10 or 13) indicates the use of Energy Information Agency's
(EIA) Annual Energy Outlook (AEO) information (most importantly the AEO demand
projections), in this case AEO 2013.
C-3
-------
(b) identifying specific strengths, weaknesses, limitations, and errors in the base case
formulation;
(c) proposing specific options for correcting errors and fixing or mitigating
weaknesses and limitations in the base case formulation;
(d) evaluate whether the parameters of the EPA base case v. 5.13 and the IPM model
have an appropriate basis (e.g., econometric estimation, published financial and
power sector data) that is supported from existing literature.
Topics to be Addressed
1. Identify strengths and weaknesses in the structure of the base case and the model
formulation with respect to the following:
a. Objective function. Are all elements necessary to meet EPA's analytical needs
included in IPM's least cost objective function? Are any extraneous or
inappropriate elements included in the objective function? Are the different terms
within the objective function consistent with each other? Does the objective
function account correctly for time and geographical factors?
b. Constraints. Are all the constraints necessary to meet EPA's analytical needs
included? Are there any extraneous or confounding constraints? Are the
constraints correctly formulated? Within individual constraints are the terms
consistent with each other? Is the sense of the constraint (i.e., <>,=) correct?
Could any constraints be simplified?
c. Decision variables and their indexes. Are all the decision variables necessary to
meet EPA's analytical needs included? Are there any extraneous or confounding
decision variables? Are decision variables unambiguously defined? Could any be
simplified? Do they correctly represent time and geographical factors?
2. Identify strengths, weaknesses, limitations, and errors in the base case and the IPM
formulation representation of:
a. Power sector operation. Are the base case and IPM's representations of model
regions, electricity demand and growth, electricity peak demand, transmission,
generation, dispatch, capacity additions, capacity factors, reserve margins
adequate to meet EPA's analytical needs?
b. Generating resources. Are the base case and IPM's representations of existing
generating units and new (potential) capacity that the optimization model
"builds," differentiations based on plant types (e.g., coal steam, combustion
turbine, combined cycle, integrated gasification combined cycle, nuclear,
biomass, wind, fuel cells, solar photovoltaic and thermal, geothermal, landfill gas,
and repowerings), and capability of reflecting locational variations adequate to
meet EPA's analytical needs?
c. Emission factors and control alternatives. Are the base case and IPM's
representations of emission factors, existing controls and available control
alternatives available within the model adequate to meet EPA's analytical needs?
Are the least-cost retrofit combinations, control differentiations based on pollutant
(e.g., sulfur dioxide, nitrogen oxides, mercury, carbon dioxide and hydrochloric
acid), and constraints from impacts retrofit sequencing incorporated in the model
adequate to meet EPA's analytical needs?
C-4
-------
d. Emission policies. Are the base case and IPM's ability to represent emission
policies that are differentiated geographically (e.g., nationally, by state, by region,
by specific plants), by policy mechanism (e.g., straight caps, cap and trade with
and without banking, seasonal limits, rate limits, tonnage limits), by allowance
allocation and processing approach (e.g., with and without flow control, output vs.
input allocation), and by pollutant adequate to meet EPA's analytical needs?
e. Power system finance and economics. Are the base case and IPM's handling of
general financial assumptions, including costs affecting dispatch, capacity
additions and retirements, retrofits, repowerings, differentiations based on
investment risk, and externalities adequate to meet EPA's analytical needs?
f. Fuels. Are the base case and IPM's representation of fuel supply, demand, and
cost information for multiple fuel types and subtypes, different supply and
demand regions, and different representations of supply dynamics (e.g.,
endogenous treatment vs. exogenous fixed price streams), competing fuel demand
from non-electricity sectors, and fuel transportation or transmission adequate to
meet EPA's analytical needs?
g. Regional resolution. Is the new regional representation in the base case and IPM
(64 US regions plus 11 Canadian regions) appropriate for both representing the
power sector and meeting EPA's needs for emission estimation?
Topics Not to Be Addressed
1. This peer review is not intended to obtain comments on the choice of a deterministic,
least-cost linear programming approach for representing the power sector. Given
EPA's choice of this methodology, reviewers are asked to provide expert input on the
strengths and weaknesses of the base case and IPM's linear programming
formulation. For example, reviewers are not being asked for comments on discrete,
non-linear, stochastic, or other approaches that might have advantages to the chosen
deterministic linear programming approach in representing certain aspects of power
system behavior. On the other hand, reviewers are being asked for expert evaluation
on elements in the base case and model formulation that can be improved within the
framework imposed by a deterministic linear programming approach.
2. This is a peer review of the base case and model's mathematical formulation and
structure, not the specific data that is currently used to populate the base case and
model. The quality and adequacy of that data will have a major bearing on many of
the same questions that apply to the model formulation. However, for purposes of this
peer review, panel members are asked to evaluate the formulation assuming that the
model will be populated with the best and most comprehensive data available.
C-5
-------
APPENDIX D
IPM BASECASE 5.13 PEER REVIEW PANEL MEETING
RTI International - 701 13th St NW Ste 750
September 9, 2014
9:00 am Q&A with EPA/ICF - Modeling Framework 30 minutes
Load Duration Curves
Demand Response Elasticity
Renewables
10:30 am Q&A with EPA/ICF - Power System Operation 30 minutes
Transmission Constraints
11:30 pm Q&A with EPA/ICF - Generating Resources 60 minutes
Life Extension Capital Costs
12:30 pm Lunch 60 minutes
1:30 pm Q&A with EPA/ICF -Financial Assumptions 45 minutes
Debt Equity Ratios
2:00 pm Q&A with EPA/ICF - Coal and Natural Gas 30 minutes
Shadow Pricing
2:30 pm Panel Discussion 60 minutes
3:30 pm Consensus Report Development 60 minutes
Report Assignments
4:30 pm Wrap up
D-l
------- |