20180529Admitted Exhibit 76.pdf

RMP Cross Exhibit
BRADLEY G.MULLINS
Consultant,Energy &Utilities
333 SW Taylor Street,Suite 400 January 24,2017
Portland,Oregon 97204
Phone [phone]
[e-mail] o
Oregon Public Utility Commission
201 High St SE,Suite 100
Salem,Oregon 97301
Re:LC 66 -Initial Comments on behalf of ICNU on the 2016 Integrated Resource Plan of
Portland General Electric Company
Dear Commissioners,
I appreciate the opportunity to provide initial comments on behalfof the Industrial
Customers of Northwest Utilities ("ICNU")on the 2016 Integrated Resource Plan ("IRP")of
Portland General Electric Company ("PGE"or the "Company").ICNU is a non-profit trade
association representing large electric utilitycustomers located throughoutthe Northwest,
including customers of the Company.In addition to these comments,TylerPepple of Davison
Van Cleve will also be filing comments on behalf of ICNU in this matter.
In summary,I recommend the Commission Ig acknowledge two Supply-side actions in
the Company's 2016 IRP Action Plan:
1)Supply-side action "a.Renewable Resources"and,
2)Supply-side action "b.Capacity Resources."
With respect to Supply-side action "a.",my analysis demonstrates that a Just-in-Time
("JIT")strategy is a more prudent,and less risky,way to plan for renewable portfolio standards
("RPS")compliance.Such a strategy would postpone the need for a physical RPS compliance
until 2030,or beyond.
With respect to Supply-side action "b.",the Company has not demonstrated a near-term
need to acquire a supply-side capacity resource.My analysis shows the Company is surplus in
capacity until the winter of 2021.After accounting for potential market purchases,the 2021
deficit is onlyapproximately243MW.Rather than issuing an immediate request for proposal
("RFP")for this potential capacity need,I recommend the Company pursue a flexible approach.
Specifically,I recommend that the future capacity need be further monitored and studied by the
O
Page 1 of 16
Initial Comments of Bradley G.Mullins RMP Cross Exhibit
On Behalf of the Industrial Customers of Northwest Utilities
Page 2 of 16
Company over the next two years,considering changes in loads,availability of market
purchases,and other demand-side alternatives.This will provide the Company with greaterflexibilityinplanningforthisneed,if conditions-such as loads,contract extensions,resource
availability,demand response opportunities-change.
I.COMMENTS
In general,ICNU is concerned with the prospect of substantial capacity and renewable
resource additions in the IRP action plan.The Company has justcompleted a series of major
capital projects,including the construction of three large utility-ownedgenerating facilities:
Port Westward II,Tucannon River Wind,and Carty Generating Station.These three generating
facilities collectively consisted of over $1.2 billion in capital and have placed material upward
pressure on the Company's rates in recent years.This recent rate pressure,however,would
seem small relative to the potential rate increases associated with the proposal in the action plan
for 850 MW of thermal capacity and 515 MW of Pacific Northwest wind capacity.
Put simply,the analysis in the 2016 IRP is inadequate to justify such significant
resource actions.The significant amount of work and effort undertaken by the Company to
prepare the IRP is appreciated.The Company's analysis,however,fails to answer some basic
questions-such as the amount of capacity availablethrough front office transactions-and is
based on certain methodologies with which ICNU fundamentallydisagrees.OMyconcernwiththeCompany's approach can be categorized into three general areas.
First,I disagree with the proposed resource adequacy assessment,based on the black-box,
Renewable Energy Capacity PlanningModel ("RECAP")model.Second,I disagree with
several aspects of the methodologyemployed to conduct portfolio analysis.Third,I disagree
with the methodologies used by the Company to evaluate various RPS compliance strategies.
Each of these areas will be discussed in the sections that follow.
II.RESOURCE ADEQUACY
In the 2016 IRP,the Company proposes a new methodology for evaluatingresource
adequacy,based on the RECAP model.RECAP is an unlicensed,freely available computer
program developed by the California-based consulting firm Energy +Environmental
Economics ("E3").Given its black-box nature,ICNU does not necessarily believe it is
appropriate to use the RECAP model in Oregon to establish resource adequacy requirements for
the Company.Rather,the use of a traditional,PlanningReserve Margin ("PRM")is a more
straightforward,proven way for the Company to evaluate resource adequacy.As it has been
deployed,the RECAP model would result in an effective increase to the PRM from 12.0%to
approximately 19.4%.E Yet,the existing PRM has produced reasonable reliability in the
See 2016 IRP,Volume II,Appendix P,Table Pl.The 19.4%figure was calculated by taking the averageofthevaluesonthelineTRM%over the period 2017 through 2021.
O
LC 66 -2016 IntegratedResource Plan of Portland General Electric Company
Initial Comments of Bradley G.Mullins RMP Cross Exhibit
On Behalf of the Industrial Customers of NorthwestUtilities
Page 3 of 16
Company's service territory for many years.Accordingly,the Company's proposal to increase
the PRM is inappropriate.Given the peaking capability of the Northwest hydroelectric system
and other unique characteristics of the region,the Northwest Power and Conservation Council
calculated a winter Adequacy Reserve Margin-a functional equivalentof a regional PRM-of
0.1%in 2026,2 a value which is directionally and conceptually inconsistent with the
Company's proposal to increase its effective PRMs by 61.6%.
1.The RECAP Model is Not Suited to Model Resource Adequacy in the Northwest
The RECAP is a compilation of python computer programs and excel spreadsheets,
which rely on a "[n]eural network model [...]to estimate complex relationships between inputs
and outputs using hidden nodes that weigh and transform input data and optimize fit to output
data."*The model was developed for use in California and is highlycomplicated.Due to its
black-box nature,the results it produces are also not necessarily transparent or easy to
understand.In fact,when benchmarked against the PRM used in the 2013 IRP,it produces
materially different results that cannot be reconciled in a straightforward manner.
While the RECAP model was developed in Califomia,it is not actually used by
regulators in Califomia for resource adequacy.Resource adequacy requirements for Califomia
utilities have historically been based on a deterministic PRM,similar to that which PGE has
historically used.*There have been efforts in Califomia to move towards a more probabilistic
O approach.My understanding,however,is that the modelingtools used to develop those studies
are much more comprehensive than the RECAP model,relying more on historical data and less
on data extrapolated through use of neural network models.
The RECAP model is not suitable to evaluate resource adequacy for utilities in the
Northwest.Unlike many other regions,hydroelectric conditions are the principal driver of
reliability in the Northwest.Accordingly,regional hydrology should be a principal
consideration in evaluating resource adequacy for any utility located in the Northwest.Yet,the
RECAP model only considers a limited time series of historical hydro data and does so only for
Company resources.
Markets in the Northwest also are different than in California.With a substantial
amount of surplus market capacity available through the Bonneville Power Administration,the
Mid-Columbia publics,*British Columbia's Powerex,and approximately3,000 MW of in-
region capability from independent power producers,it is common for utilities in the Northwest
to rely more heavily on capacity available through bilateral markets for purposes of resource
27 Northwest Power and Conservation Council,Seventh Power Plan,at 11-23,Table 11-7 (represents the
average of Q4 and Ql)(Feb.2016).
1/E3,Renewable Enerav Capacity Planning Model,User Manual at 7 (June 17,2015)
I See e.e.CAISO,April 21"Public Workshop,Regional Resource Adequacy Revised Straw Proposal at
26-34.Available at http://www.caiso.com/Documents/Acenda-
Presentation RegionalResourceAdequacy-RevisedStrawProposal.pdf
17 The Mid-Columbia Publics include:Grant PUD (Wanapum and Priest Rapids),Chelan PUD (Rock Island
and Rocky Reach),Douglas PUD (Wells).
O
LC 66 -2016 IntegratedResource Plan of Portland General Electric Company
Initial Comments of Bradley G.Mullins RMP Cross Exhibit
On Behalf of the Industrial Customers of Northwest Utilities
Page 4 of 16
adequacy.A critical question in evaluatingthe Company's resources adequacy is the extent to
which the Company should rely on regional markets.The RECAP model,however,does not
attempt to determine the depth of markets in the Northwest.
In fact,it is a flaw of the 2016 IRP that it did not analyze the degree to which the
Company can rely on market purchases to satisfy load requirements in a reliable manner.
Given the depth of the markets in the Northwest,as well as availability of a substantial amount
of winter import capability from California,market availability should be a fundamental
question addressedby the IRP of any utility in the Northwest prior to building a new resource.
Notwithstanding,the Company assumed little-to-no market capability in the RECAP
model,and the market capability that was included was modeled in a way that is inconsistent
with how the RECAP model is designed.The RECAP model contains special inputs to model
market imports.The Company,however,did not use those designated inputs for market
imports in its RECAP analysis.The Company set the market import capability in RECAP to
zero,and instead modeled market purchases as if they were a variable energy resource,based
on a historical profile of market purchases.Based on the way that the RECAP model generates
stochastics,treating historical market purchases the same way as an intermittent resource does
not accurately reflect the actual capability the Company derives from the market,and
accordingly,is a flawed methodology.
In contrast to the methodologyused by the Company,my opinion is that it is more
appropriate to consider regional resource adequacy when evaluatingthe amount of available
market capacity.The Northwest Power and Conservation Council (the "Council")recently
published the Seventh Power Plan.In that document the Council found that "[i]n more than 90
percent of future conditions,cost-effective efficiency met all electricity load growth through
2030 and in more than half of the futures all load growth for the next 20 years."6/While the
Council's recommendation was based on an ambitious forecast for demand response additions,
the Council's report suggests that there will be sufficient regional capacity,an indication that
surplus capacity in the region will likely be available through bilateral markets.
Due to these deficiencies,I recommend that the Commission not accept the RECAP
model for evaluatingthe resource adequacy requirements of the Company.On its face,the
RECAP model produces results that are unreasonable relative to the Company's past planning
practices and fails to adequately consider unique aspects of the power supply system in the
Northwest.
2.A Traditional Approach Based on a Planning Reserve Margin Should be Used to
Consider Resource Adequacy Requirements
Rather,I recommend that the Commission evaluate the resource adequacy requirements
of the Company using a traditional approach,relying upon a deterministic PRM.This sort of
Northwest Power and Conservation Council,Seventh Power Plan at 1-1 (emphasis in original).
O
LC 66 -2016 IntegratedResource Plan of Portland General Electric Company
Initial Comments of Bradley G.Mullins RMP Cross Exhibit
On Behalf of the Industrial Customers of Northwest Utilities
Page 5 of 16
approach is a simpler way to gain an understanding of the Company's resource needs and to
evaluate resource alternatives to satisfy those needs.
Appendix P of the 2016 IRP contains a high-level load and resource balance.However,
I have identified several problems with the load and resource balance presented in Appendix P,
many of which will be discussed below.
As a result of these problems,I developed an independent load and resource balance,in
an attempt to better understand the resource needs of the Company in the coming years.My
analysis is more detailed than that presented by the Company in Appendix P and provides a
more realistic assessment of the Company's capacity position followingthe retirement of
Boardman.My analysis has been provided in Confidential Attachment A and is summarized in
Table 1,below.
TABLE 1
Preliminary Load and Resource Balance (MW,Winter Peak)
No Resource Additions
2018 2019 2020 2021 2022 2023 2024 2025 2030 2035 2040
Resoumes
Coal 809 809 809 296 296 296 296 296 296 -
NaturalGas 1,851 1,851 1,851 1,851 1,851 1,851 1,851 1,851 1,851 1,851 1,851
O Hydroelectne 804 804 804 804 804 804 804 584 574 574 574
Renewable Resources 84 84 84 84 84 84 84 84 81 81 72
QualifyingFacilities 82 82 82 82 82 82 82 82 82 81 81
Seasonal Contracts 100 100
Total Existing Rsms 3,730 3,730 3,630 3,117 3,117 3,117 3,117 2,897 2,885 2,587 2,578
Net Izad +PRM 3,633 3,645 3,611 3,660 3,671 3,686 3,705 3,729 3,904 4,108 4,343
Net Position(Befoæ FOT)97 85 18 (543)(554)(569)(588)(831)(1,019)(1,520)(1,765)
Available Front Office 300 300 300 300 300 300 300 300 300 300 300Transactions*
NetPosition(AfterfDT)397385318(243)))(ÄÃ)1)),5Ã)(,SÃ)
*Prelirrunary assumption to be updated in Final Comments
As can be seen from Table 1,the Company is faced with little,to no,near-term peak
load growth,after accounting for incremental demand-side management and distributed
generation resources.The above analysis also shows that,after the retirement of Boardman
beginningin 2021,the Company is faced with a capacity need of approximately243 MW.
This need is substantially less than the 850 MW capacity resource addition proposed by the
Company.Some of the differences between my load and resource analysis and the Company's
are detailed as follows.
Planning Reserve Margin
As noted,the Company's analysis assumed an approximate 19.4%PRM.While I
believe that a 12%PRM used in the 2013 IRP is too high--particularlygiven the fact that the
O
LC 66 -2016 IntegratedResource Plan of PortlandGeneral Electric Company
Initial Comments of Bradley G.Mullins RMP Cross Exhibit
On Behalf of the Industrial Customers of Northwest Utilities
Page 6 of 16
Council has determined that regional adequacy margins are closer to zero---for purposes of this
proceeding,I used a 12%PRM to evaluate the Company's expected resource needs in
Confidential Attachment A.That PRM is reflected in the Net Load values detailed above.
Demand side resources are also reflected as an offset to Net Load in the above table,in contrast
to the tables detailed in Appendix P,which separately detail demand-side resources in a manner
similar to supply-side resources.
Front Office Transactions
In Appendix P,the Company assumed that only 98 MW of market transactions would
be availablein 2021 to meet peak loads.This assumption is largely unsupported in the 2016
IRP,and based on my experience,is arbitrarily low.
For purposes of the initial analysis summarized in Table 1,I assumed that the Company
had the ability to import approximately300 MW of winter peaking capacity from the Mid-
Columbia and California-OregonBorder ("COB")markets.This is a preliminary assumption
and may still be unrealistically low,given the Company's transmission access to regional
markets.Accordingly,the Company should provide further information regarding available
market capacity in its Reply Comments.
In its Reply Comments,I particularly request that the Company consider that it has a
substantial amount of import capability from the COB market.In response to ICNU Data
Request 17,for example,the Company indicated that it has rights to approximately727 MW of
transmission from COB.Customers are paying for transmission from COB,and thus,should be
reaping the reliabilitybenefits associated with access to that market.Because California is
summer peaking,the Company should able to realize a significant amount of capacity from this
transmission link.Just as PacifiCorp includes import capability from COB in its IRP,the
Company should also consider COB imports in its IRP.
Load Forecast
Both the load and resource balance and RECAP analysis prepared by the Company
appear to have been based on an outdated load forecast.In the Company's final MONET
update in Docket No.UE 308,for example,2017 peak loads were approximately97.5 MW
lower than reflected in Appendix P to the 2016 IRP.Similarly,loads in the RECAP model
were overstated by an even greater amount,approximately 187.7 MW,compared to the load
forecast provided in MONET in Docket No.UE 308.The difference between the peak load
forecast in RECAP and the peak load forecast in the Company's load and resource balance may
offer one explanationfor the excessively high planning reserve margin assumed in the
Company's analysis.
Based on the final MONET update,my analysis makes a 97.5 MW downward
adjustment to peak loads in Attachment A.
O
LC 66 -2016 IntegratedResource Plan of Portland General Electric Company
Initial Comments of Bradley G.Mullins RMP Cross Exhibit
On Behalf of the Industrial Customers of Northwest Utilities
Page 7 of 16
These variances in the Company's load forecast are particularly concerning given a
recent report by the BerkeleyNational Laboratorythat demonstrated that many utilities,
including the Company,have systematically overstated load growth rates in historical IRPs.A
copy of that report,titled "Load Forecasting in Electric UtilityIntegrated Resource Planning"
can be accessed as of January 2017 through the followinghyperlink:
https://emp.lbl.cov/sites/all/files/lbnl-1006395 0.pdf.
Load forecasts are one of the more important aspects to consider when evaluating
resource adequacy,and thus,should be updated based on the best information available to the
Company.It certainly would not be prudent for the Company to follow through with its plan to
acquire a major resource,based on an outdated load forecast.This is one of the reasons I
believe it is appropriate for the Company to continue to monitor its load forecast and resource
needs for a few more years,prior to taking any concrete resource actions for 2021.
Hydro Capability
In forming its peak load and resource balance,the Company has understated the
capacity available from its hydro electric facilities.It appears that the Company has used
average annual energy to assess the capacity contribution of run-of-river hydro systems.
Because the Company is winter peaking,it would be more accurate for the Company to assume
average energy during the winter timeframe.My load and resource balance calculates capacity
contribution based on average energy in the months of January -February.
I also assumed that the Portland Hydro Project would be renewed.The Portland Hydro
Project,which consists of approximately 33 MW of run-of-river hydro capacity on the Bull Run
River,was assumed to be extended and was reflected in rates in Docket No.UE 308,the 2017
Annual Power Cost Update Tariff filing.Thus,it is appropriatelyincluded in the load and
resource balance in the IRP.
Renewable Resources
From the Company's load and resource balance,it is not clear what capacity
contribution values were assumed for wind and solar resources.ICNU was hopeful that-
following Docket UM 1719,an Investigationto Explore Issus Related to a Renewable
Generator's Contribution to Capacity-the IRP would report capacity contribution values based
upon the Effective Load Carrying Capability methodology.Notwithstanding,ICNU was
unable to identify in the IRP where the Company calculated the contribution to peak of these
resources.Absent the analysis,and based on my experience,my analysis assumes a 10%
capacity contribution for wind and a 20%capacity contribution for solar.I request that the
Company provide further information regarding the capacity contribution of wind and solar
resources in its Reply Comments.
O
LC 66 -2016 IntegratedResource Plan of Portland General Electric Company
Initial Comments of Bradley G.Mullins RMP Cross Exhibit
On Behalf of the Industrial Customers of Northwest Utilities
Page 8 of 16
Distributed Solar
While the Company's load and resource balance included approximately 118 MW of
capacity for non-renewable distributed generation,including the Dispatchable Stand-by
Generation program,it appearsto have excluded any contribution to peak associated with the
approximate 65 MW of distributed solar resources included in the IRP.27 My analysis includes
a provision for capacity provided by distributed solar resources as an offset to load in the load
and resource balance detailed in Confidential Attachment A.
Gas Plants Capacity
The Company appearsto have slightlyunderstated capacity available from gas
resources.Based on capability in the month of January,my analysis based on the Company's
most recently filed MONET model shows a slightlygreater amount of capacity available from
gas plants,than the 1,810 MW assumed in the Company's load and resource balance.While I
have not necessarily reconciled the difference,it may be that the Company has measured the
capability of gas plants in a different month of the year.In my opinion,January is the most
appropriate month to measure nameplate capacity of gas plants because the Company load
typicallypeaks when weather is coldest.In fact,it is possible that the use of average January
temperatures may actually understate the capacity from gas plants because the average is not
representative of the coldest hours in the month,when the Company is peaking.
In summary,based on the above analysis,existing resources will provide the Company
with reasonable resource adequacy at least until the retirement of Boardman at the end of 2020.
Consistent with the results of the Council's Seventh Plan,I believe that the Company should
take no immediate action with respect to potential resource needs in 2021.Rather,I
recommend that,prior to pursuing a physical resource,the Company should 1)continue to
monitor its load forecast;2)review availability of market imports;and,3)continue to review
demand-side options as alternatives to physical resource acquisition.
III.PORTFOLIO ANALYSIS
In addition to the problems associated with the use of the RECAP model,I believe that
the portfolio analysis in the 2016 IRP is also based upon a methodologythat is fundamentally
flawed.Other utilities use models,such as PacifiCorp's System Optimizer model,to develop
least-cost portfolios given a set of inputs.These type of models are designed to optimize the
type and timing of resource additions in each portfolio for purposes of satisfying peak load
requirements.
The Company's analysis,however,did not attempt to optimize the type and timing of
resource additions.Rather,the Company performed a scenario analysis in the AURORAxmp
model based on a series of predetermined resource portfolios,in an attempt to determine which
2016 IRP,Volume II,Appendix D,at 384
O
LC 66 -2016 IntegratedResource Plan of Portland General Electric Company
Initial Comments of Bradley G.Mullins RMP Cross Exhibit
On Behalf of the Industrial Customers of Northwest Utilities
Page 9 of 16
preselected portfolio was the lowest cost.This type of analysis,however,does not necessarily
reflect an optimal resource portfolio,unless one believes that the Company preselected the
optimal resource portfolio.As I demonstrate below,the portfolios preselected by the Company
are not optimal,particularly considering the resource need discussed in the prior section.
Similarly,the Company's attempt to consider the risk of portfolios is also analytically
flawed.A risk analysis can only be said to be reasonable if it reflects a reasonable distribution
of likely outcomes.While the Company considers some outcomes-such as a high gas price
scenario-it did not consider portfolio performance under other potential outcomes.
Faced with these deficiencies in portfolio modeling,it is difficult to analyze the
reasonableness of any aspect of the Company's proposed action plan.Notwithstanding,I have
attempted to work within the Company's scenario analysis framework,using the
AURORAxmp model,in order to try to determine the portfolio actions that the Company might
take in order to best satisfy its potential capacity needs in coming years.
1.The Company's Portfolio Analysis Did Not Evaluate Resource Type or Timing
One purpose of the IRP is to determine the lowest cost resources to acquire,as well as
the proper timing of such acquisitions.This often results in complex trade-offs between various
baseload,peaker,storage,and market resources.Yet the Company's approach of evaluating
portfolios is based on a limited number of pre-defined portfolios that fails to provide an
adequate answer to this question of resource type and timing.
In Attachment B,I have constructed a simple Excel model that builds resource
portfolios based on the load and resource table provided in Attachment A.The analysis
assumes that all capacity shortfalls,after accounting for market purchases and other new
resources,must be met by a combined cycle combustion turbine ("CCCT"),as a last resort.
Based on this analysis,I was able to test whether it is less costly to acquire other types of
resources or to rely more heavily on market purchases.I also prepared some sensitivities
surrounding RPS compliance.These portfolios were input into a reference case scenario in
AURORAxmp,and the resulting portfolio costs were considered in a net present value revenue
requirement analysis.The results of that analysis are detailed in Table 2,below.
O
LC 66 -2016 IntegratedResource Plan of Portland General Electric Company
RMP Cross ExhibitInitialCommentsofBradleyG.Mullins ----
On Behalf of the Industrial Customers of Northwest Utilities
Page 10 of 16
TABLE 2
AURORAxmpPortfolio Sensitivities (2018 --2037)
Present Value Revenue Requirement ($000)Deltas from Portfolio 1
Reference Gas,No Carbon Constraint
10-year PVRR Delta 20-year PVRR Delta
Portfolio 1 Base Case (2021 Combined Cycle 10,836,781 -20,639,777 -
Combustion Turbine)
Portfolio 2 Low Market Capacity 10,865,208 28,427 20,670,108 30,331
Portfolio 3 High Market Capacity 10,576,839 (259,942)20,135,192 (504,584)
Portfolio 4 Wells Not Extended 10,896,214 59,432 20,875,291 235,515
Portfolio 5 Simple Cycle Combustion Turbine 2021 10,812,003 (24,778)20,555,660 (84,117)
Portfolio 6 150 MW Direct Access 10,662,544 (174,238)20,206,268 (433,509)
Portfolio 7 RPS Early Action (515 MW Wind 2018)11,147,303 310,522 21,111,530 471,753
Portfolio 8 No Unbundled Renewable Energy 10,910,010 73,229 20,922,906 283,129
Portfolio 9
onRnae
able Portfolio Standard 10,779,267 (57,514)20,124,937 (514,839)
Portfolio O Company Preferred:Efficient Capacity 11,418,357 581,576 21,689,254 1,049,478
2021
Several interesting things can be discerned from the AURORAxmp runs detailed above.
While I plan to present a fuller portfolio analysis in Final Comments,below are some of my
preliminary conclusions.
The base portfolio (Portfolio 1)in my analysis is approximately $1.0 billion less
expensive,on a 20-year present value revenue requirement basis,than the
Company's preferred portfolio (Portfolio 0).Thus,the Company's proposal for
850 MW of traditional capacity and 515 MW of renewable capacity could result
in substantial and unnecessary costs to ratepayers.
Portfolio 3-a portfolio which assumes 700 MW of market capacity is available
to the Company-is the lowest cost portfolio.This indicates that it is crucial for
the Company and the Commission to consider the amount of capacity available
from the Mid-Columbia and COB markets,prior to pursuing physical resource
acquisition.
O
LC 66 -2016 IntegratedResource Plan of Portland General Electric Company
Initial Comments of Bradley G.Mullins RMP Cross Exhibit
On Behalf of the Industrial Customers of Northwest Utilities
Page 11 of 16
Portfolio 6 indicates that there are significant capacity benefits to remaining
customers if 150 MW of load were to migrate to direct access and permanently
opt-out of cost of service rates.Such a portfolio could defer the need to acquire
a physical resource until 2025,saving ratepayers $433.5 million on a 20-year net
present value revenue requirement basis.
Portfolio 5 indicates that,if physical capacity is truly required in 2021,it would
be more cost effective to ratepayers to acquire a smaller,Simple Cycle
Combustion Turbine ("SCCT"),rather than a larger,CCCT.Ratepayers save
$84.1 million in a portfolio that includes a 2021 SCCT,delaying the need to
build a larger,CCCT until at least 2025.
Portfolio 7 cost ratepayers $471.8 million more on a 20-year net present value
revenue requirement basis than the base portfolio,indicating it is not a least cost
strategy to pursue early action of an RPS resource at this time.
These conclusions contradict many of the conclusions reached by the Company in its
IRP.These conclusions also indicate that it could cost ratepayers substantially,and
unnecessarily,if the Company is to pursue the supply-side actions to acquire 850 MW of
traditional capacity and 515 MW of renewable capacity.For that reason,I do not agree that the
Company is justified in pursuing those supply-side actions.While the Company may be
O justified in acquiring a smaller,SCCT resource in 2021,the economics of that strategy are
contingent on the level of market capacity actually available,as well as potential opportunities
for large customers to opt out of cost of service rates.These options must be understood before
a resource decision can be made.
2.Large Customer Opt-OutShould be Considered as a Resource Option
The Company assumes that no customers will elect to opt out of cost of service rates in
the study period.Yet,if a large customer were to opt out of cost of service rates,it may allow
the Company to avoid acquiring expensive capacity and renewable resources.Based on the
portfolio analysis conducted above,ratepayers would save approximately$433.5 million,on a
20-year net present value revenue requirement basis,if 150 MW of additional load were to elect
to opt out of cost of service rates.Given this potential savings,the Company should consider
options that provide proper economic signals to,and eliminate barriers for,large customers who
may be interested in opting out of cost of service rates.The current transition adjustment
methodology,which focuses solely on short-term marginal costs,does not consider the long-
term capacity benefit that the Company,and its remaining customers,receive when a large
customer chooses to opt out of cost of service rates.There are also caps that have been put in
place that restrict the amount of load that can migrate to direct access.Prior to building a new
resource,the Company should consider raising the current direct access cap and adopting
transition adjustments that consider the long-term capacity benefits associated with incremental
direct access customers,while still ensuring that remaining customers are unharmed.
O
LC 66 -2016 IntegratedResource Plan of Portland General Electric Company
Initial Comments of Bradley G.Mullins RMP Cross Exhibit
On Behalf of the Industrial Customers of Northwest Utilities
Page 12 of 16
3.A 34-year Planning Period is Too Long
When performing the above scenarios,I limited the planning period to 20 years.A 34-
year planning period is too long and puts too much weight on speculative assumptions about
distant future conditions to be used for planning.The Commission's IRP guidelines do not
require an analysis beyond 20 years.
Moreover,the proposed 34-year planning period is considerably longer than the 20-year
planning period used in the Company's 2013 IRP.While such a long study period may present
some useful information,modeling portfolio performance that far into the future is problematic
and the costs of distant resources should not form the basis for near-term resource acquisitions.
Resource decisions often involve a trade-off between paying more rates today,in order to
achieve dispatch savings expected over a long period of time.
Forecasting conditions far into the future is inherently speculative.Ten years ago,when
the Company issued the 2006 IRP for example,there was no contemplation of the rapid
expansion of the Energy Imbalance Market ("EIM")throughoutthe West.If one goes back
even further,34 years ago to 1982,the utilities were not even contemplating the liquid bilateral
markets that are ubiquitous today.The concept of something such as the EIM would have been
outside of the range of any future possibility in 1982.A lot can change over a ten-year period,
let alone the 34-year planningperiod proposed by the Company,and for that reason it makes
sense to place less weight on benefits and costs expected far into the future.For purposes of
making resource decisions today,a twenty-yearplanning period is sufficient to make informed W
resource decisions that form the basis for the Company's action plan.Additionally,I also gave
greater weight to the first ten years of the analysis in order to reflect the greater certainty of cost
predictions over this period.Thus,in the above table I also calculated levelized portfolio costs
over a 10-year period,as a metric used to benchmark against 20-year results.
4.The Company's Risk Analysis is Analytically Improper
While the Company attempts to evaluate the risk profile of various portfolios,it does
not properly evaluate risk.Stochastic modeling is typicallyused to evaluate risk associated
with a resource portfolio.The aim of this type of modeling is to change model inputs based
upon a distribution of expected outcomes.The Company,however,did not use a distribution of
expected outcomes,and rather,relies on a skewed set of scenarios that do not properly reflect a
balanced set of possibilities.
This can be noted by the failure of the Company to model a low natural gas price
scenario.In addition to modeling median and "high"gas price scenarios,the Company needs
to model a "low"natural gas price scenario.As noted in Figure 3.8,the Company models
natural gas prices based on a reference case and a high natural gas price scenario,both of which
assume prices will increase considerably in the study period.The Company,however,does not
model any scenario to evaluate the impact of the very real possibility that natural gas prices will
continue to remain low into the future.
O
LC 66 -2016 IntegratedResource Plan of PortlandGeneral Electric Company
Initial Comments of Bradley G.Mullins RMP Cross Exhibit
On Behalf of the Industrial Customers of Northwest Utilities
Page 13 of 16
The omission of a low gas price scenario is,in my opinion,a critical analytical flaw in
the Company's IRP risk analysis.The exclusion of a low gas price scenario is problematic
primarily because it skews the Company's risk modeling,in favor of a future state that is based
on high gas prices.If one only includes two future natural gas states,a reference gas price state
and a high gas price state,the median outcome of the risk modeling will be representative of
prices that are between the reference state and the high gas state.That is,if onlya high and
reference gas prices are used,the reference case is no longer the median outcome.Thus,if a
high price scenario is to be used,at a minimum,a low price scenario also needs to be modeled
in order to maintain the reference case as a median outcome.
5.The Company Appears to Double Count Carbon Costs in AURORA
The Company develops a carbon price for its reference case scenario based on a report
prepared by Synapse Energy Economics.The Synapse report estimates carbon costs based on
the price of allowances under a mass-based state implementation plan with the new source
complement.The Company notes,however,that this is not its preferred implementationplan
because a source-level rate based implementation plan would result in no incremental cost to
customers.
In addition to modeling carbon costs directly in AURORAxmp,however,it appearsthat
the Company has also established a region-wide cap on carbon emissions as a constraint in the
O model.While not necessarily opposed to understanding the impact of potential carbon costs on
portfolios,modeling a carbon price,as well as a carbon cap,did not seem to be a consistent way
to model potential carbon costs.I request that the Company provide further information about
how it has modeled these costs in its Reply Comments.
In addition,the Company currently expects to comply with the Clean Power Plan at no
incremental cost to customers.Accordingly,the reference case should assume zero carbon
costs to the Company,at least through the Clean Power Plan compliance period.In modeling
prices in the AURORAxmp model,it may be appropriate to model carbon costs in the reference
case forward price curve in regions where the Clean Power Plan is expected to impose
additional costs.However,assigning an incremental carbon cost to the Company's portfolio in
the reference case contradicts my understanding of how compliance with the Clean Power Plan
might impact customers of the Company.
For these reasons,the above scenarios were analyzed in a case with zero carbon costs,
though I am interested in evaluatingthe sensitivity of those portfolios to carbon costs in its
Final Comments.
IV.RENEWABLE PORTFOLIO STANDARDS COMPLIANCE
Based upon the portfolio analysis detailed in Table 2 above,I disagree the Company is
justified in pursuing early action of a renewable resource.My analysis demonstrates that early
action will ultimately cost ratepayers $471.8 million on a net present value revenue requirement
O
LC 66 -2016 IntegratedResource Plan of PortlandGeneral Electric Company
Initial Comments of Bradley G.Mullins RMP Cross Exhibit
On Behalf of the Industrial Customers of Northwest Utilities
Page 14 of 16
basis over a 20-year period.Over a 10-year period,early action will cost ratepayer
approximately$310.5 million on a present value revenue requirement basis.Thus,a just-in-
time RPS resource acquisition strategy is a less costly strategy-and as discussed below,a less
risky strategy-for complying with RPS requirements.
1.The Company Should Adopt a Just-in-Time Acquisition Strategy for RPS Resources,
Delaying RPS Resource Action Until 2030
ICNU continues to support a "just-in-time"acquisition strategy for RPS resources,
including utilization of unbundled Renewable Energy Certificates ("RECs")up to the 20%
statutory maximum level.Consequently,I have several concerns with the Company's proposal
for early action with respect to renewable resource additions in the IRP.Not only does the
Company's early-action proposal disregard a number of costs and risks associated with building
a renewable resource prior to the time that such a resource is needed,an early-action strategy
should not be considered in isolation from other rate impacts,including the fact that ratepayers
will likely be faced with upward rate pressures as a result of replacing lost capacity associated
with the retirement of Boardman.
While there may be instances where it is appropriate to pay more in rates today in order
to achieve long-term rate savings,these sorts of projects with a long-term pay-back are not
preferred in periods when ratepayers are already subject to substantial upward rate pressures.
Additionally,the long-term rate savings should be far more certain to occur than suggested in
the Company's analysis.
Moreover,early-action strategies also rely on long-term resource and planning
assumptions,which carry significant forecasting risk for customers.For example,if solar costs
continue on their current trajectory,the Company may be placed in a situation where it is
considerably less expensive to acquire renewable energy ten years from now,than it is today.
Similarly,if PTCs are ultimately extended,an early-action strategy could also cost ratepayers
greatly.The rapid pace of technological change in the energy industry today creates a
significant risk that acquiring new generation,renewable or otherwise,before it is needed will
impose substantial stranded costs on customers.
2.The Company Should Assume that the Use of Unbundled RECs Will Delay the Need
for New RPS Additions until 2030
As noted in Docket No.UM 1773,if the Company continues to rely on unbundled
RECs,it will not need to acquire a new RPS resource until 2030.In the IRP,the Company
attempts to model a scenario that relies on unbundled RECs to meet 20%of its RPS
requirements.That portfolio,titled "Efficient Capacity 2021 20%UnbundledRECs,"however,
only assumes that an RPS resource can be delayed until 2025.This is because the Company
assumed it could only rely on unbundled RECs over the period 2016 through 2021.
Given the low price of unbundled RECs and the Company's history of relying on
unbundled RECs to meet 20%of its RPS obligation to date,it would be more appropriate for
O
LC 66 -2016 IntegratedResource Plan of Portland General Electric Company
Initial Comments of Bradley G.Mullins RMP Cross Exhibit
On Behalf of the Industrial Customers of Northwest Utilities
Page 15 of 16
the Company to assume that it can rely on unbundled RECs over the entire IRP study period,
which should delay the physical compliance need until 2030,while maintaining a sufficient
REC bank balance.This analysis has been provided in Attachment C.For purposes of the
portfolio analysis detailed above,I assumedthat the Company could acquire RECs at a nominal
levelizedprice of $10/MWh.Based on my analysis,the nominal levelized REC price would
have to exceed $32.75/MWh (the "tipping point")before it became more economic to acquire a
physical resource.
3.Early RPS Build Scenarios Should Include the Cost of Incremental Production Tax
Credit Carry-Forwards
As also noted in Docket No.UM 1773,the Company currently lacks sufficient taxable
income necessary to utilize all of the production tax credits generated from the Biglow Canyon
and Tucannon River Wind facilities.While unused productiontax credits can be "carried-
forward"to be used on a future tax return,the growing balances have presented ratemaking
concerns.Specifically,the Company has historically argued that it should be allowed to earn a
return on the carry-forward balances at its full cost of capital.
The growth in the Company's production tax credit carry-forward balance was expected
to slow,and potentially reverse,when the production tax credits generated from the Biglow
Canyon facility begin to expire over the period 2018 through 2020.If the Company acquires a
O 515 MW wind resource in 2018,however,the growth in the carry-forward balance will not
slow,but rather,will begin to accelerate at a problematic rate.As demonstrated in Attachment
E,for example,I forecast that the production tax credit carry-forward balance is expected to
grow to in excess of $400 million if the Company acquires a 515 MW wind facility in 2018.I
also demonstrate that the return on this balance could cost ratepayers approximately $233.0
million on a present value revenue requirement basis.This additional cost has been reflected in
the cost associated with an early-action strategy in the portfolio analyses detailed in Table 2,
above.
The Company's portfolio analysis,however,has made no effort to quantify the impacts
on rates associated with these tax attributes in the IRP.In the scenarios that rely on an early-
action strategy,the Company includes no additional cost associated with its inability to utilize
production tax credits on its tax return,which consequently overstates the benefits of early
action to acquire the full value of the production tax credit.Giventhe level of risk associated
with an early-action strategy,all costs and risks must be accounted for in the Company's
analysis,including the cost of production tax credit carry-forwards.
V.CONCLUSION
I appreciate the opportunity to provide these comments on behalf of ICNU.I also
appreciate the large amount of work and analysis conducted by the Company in preparing the
IRP.Notwithstanding,there are still fundamental questions that need to be considered before
taking the supply side resource actions proposed by the Company.
O
LC 66 -2016 IntegratedResource Plan of PortlandGeneral Electric Company
Initial Comments of Bradley G.Mullins RMP Cross Exhibit
On Behalf of the Industrial Customers of Northwest Utilities
Page 16 of 16
With respect to the renewable resource addition,I disagree with the economic analysis
proposed by the Company to justify the 515 MW of near-term renewable resources,as my
analysis demonstrates that it will not be beneficial to ratepayers to pursue an early-action
strategy at this time.
In addition,the Company's proposal for approximately 850 MW of supply-side capacity
resources should also not be acknowledged,as the Company has not justified such a substantial
need at this time.Given the smaller magnitude of the capacity need demonstrated by my
analysis-andthe potential that the Company may be able to avoid the need altogether through
existing market capability or other demand-side alternatives-myrecommendation is that the
2021 need be further studied and analyzed by the Company prior to issuing an RFP.
I look forward to working with parties to further addressthe Company's future resource
strategy.
Sincerely,
/s/Bradlev Mullins
Bradley Mullins
Consultant,Energy &Utilities
333 SW Taylor Street,Suite 400
Portland,Oregon 97204
O
LC 66 -2016 IntegratedResource Plan of PortlandGeneral Electric Company