Loading...
HomeMy WebLinkAbout20180430Link Supplemental Rebuttal.pdfRECEIVED 20\SAPR30 AM 9:38 BEFORE THE IDAHO PUBLIC UTILITIES COMMISSION IN THE MATTER OF THE APPLICATION )CASE NO.PAC-E-17-07OFROCKYMOUNTAINPOWERFORA)CERTIFICATE OF PUBLIC )SUPPLEMENTALREBUTTALCONVENIENCEANDNECESSITYAND)TESTIMONY OF RICK T.LINKBINDINGRATEMAKINGTREATMENT)FOR NEW WIND AND TRANSMISSION )FACILITIES ) O ROCKY MOUNTAIN POWER CASE NO.PAC-E-17-07 APRIL 2018 O l Q.Are you the same Rick T.Link who previously provided testimony in this case on 2 behalf of Rocky Mountain Power,a division of PacifiCorp? 3 A.Yes. 4 PURPOSE AND SUMMARY OF SUPPLEMENTAL REBUTTAL TESTIMONY 5 Q.What is the purpose of your supplemental rebuttal testimony in this proceeding? 6 A.My testimony supports the company's application for certificates of public convenience 7 and necessity ("CPCNs")and binding ratemaking treatment for the Aeolus-to- 8 Bridger/Anticline line and network upgrades ("Transmission Projects")and the Ekola 9 Flats,TB Flats I and II,Cedar Springs,and Uintaprojects.These are the four new wind 10 resources ("Wind Projects")included on the final shortlist of the 2017R Request for 11 Proposals ("2017R RFP"),(collectively,the "Combined Projects").Specifically,my 12 testimony responds to the April 11,2018,testimony filed by Anthony Yankel on behalf 13 of the Idaho Irrigation Pumper Association ("IIPA"),Nicholas L.Phillips,on behalf of 14 Monsanto,Bradley G.Mullins on behalf of PacifiCorp Idaho Industrial Customers 15 ("PIIC"),and Michael Louis and Michael Eldred on behalfof the Idaho Public Utilities 16 Commission Staff ("Staff"). 17 Q.Please summarize your testimony. 18 A.I respond to claims that PacifiCorp does not have a resource need.I address criticisms 19 of PacifiCorp's 2017R RFP bid evaluation and selection process and criticisms of the 20 company's economic analysis,which shows that the Combined Projects will generate 21 significant customer benefits.In response to claims that the Combined Projects may 22 not be the least-cost,least-risk resource option,I also summarize the economic analysis 23 used to finalize PacifiCorp's 2017S Request for Proposals ("2017S RFP")bid-selection O Link,Supp-Reb -1 Rocky Mountain Power l process.My supplemental rebuttal testimony demonstrates that: 2 Even after accounting for the updated load forecast that is summarized in my 3 supplemental direct testimony,PacifiCorp has a 595-MW capacity deficit in 4 2021 that grows to 3,395 MW in 2036 and that the Combined Projects are least- 5 cost,least-risk resources that will partially meet this need. 6 As supported by independent evaluators that were appointed,retained,and 7 managed by two different state regulatory commissions,the 2017R RFP was 8 fair,transparent,and unbiased. 9 These independent evaluators found that the bids selected to the 2017R RFP 10 final shortlist represent the top offers that are viable under current transmission 11 planning assumptions,and one of the experts concluded that the final shortlist 12 should result in significant savings for customers.O 13 The company has performed over 1,300 20-year simulations of PacifiCorp's 14 system to thoroughly evaluate how the net benefits of the Combined Projects 15 are affected by a broad range of variables and uncertainties. 16 While solar resources may provide customer benefits,contrary to claims from 17 certain parties,solar resource bids submitted into the 2017S RFP are not a 18 superior resource alternative to the Combined Projects. 19 Solar resources are best viewed as an incremental opportunity,not as an 20 alternative to the Combined Projects. 21 During the evaluation of bids in the 2017S RFP,PacifiCorp analyzed valuation 22 risks that are unique to the procurement of solar resources and determined that 23 solar resource costs are likely to continue to fall. O Link,Supp-Reb -2 Rocky Mountain Power 1 Given these solar resource-valuation risks,expected cost declines,and 2 availability of the 30-percent investment tax credit ("ITC")for solar projects 3 coming online as late as 2021,PacifiCorp does not need to act now and has 4 decided not to select any of the solar power-purchase agreement ("PPA")bids 5 to the 2017S RFP final shortlist. 6 PacifiCorp will continue to assess potential economic benefits from solar- 7 resource opportunities through bi-lateral opportunities and in the 2019 8 Integrated Resource Plan ("IRP"),including a thorough review of valuation 9 risks with full stakeholder engagement,to determine whether a new competitive 10 solicitation process for projects capable of achieving commercial operation by 11 the end of 2021 will providecustomer benefits. 12 In contrast,the phase-out of production tax credit ("PTC")benefits that areO13availableforqualifyingwindprojectsoccurssoonerthantherampdownofITC 14 benefits that are available for solar resources,which requires that PacifiCorp 15 act now to deliver the new wind and needed transmission investments that will 16 produce both near-term and long-term benefits for customers. O Link,Supp-Reb -3 Rocky Mountain Power l RESOURCE NEED 2 Q.Messrs.Louis,Phillips,Mullins,and Yankel continue to question the need for any 3 new resources.(Louis Supp.Direct,page 3,lines 20-25;Mullins Supp.Direct,page 4 35,lines 3-12;Phillips Supp.Direct,page 2,line 23 and page 28,line 5:Yankel 5 Supp.Direct,page 1,lines 20-22,page 4,lines 12-20,and page 5,lines 1-12.)Please 6 summarize how the 2017 IRP identified a resource need that can be met by the 7 Combined Projects. 8 A.In my rebuttal testimony,I explained in detail that the company has an immediate 9 resource need and that the Combined Projects would displace higher cost,higher risk 10 front-office transactions ("FOTs")in the near term and defer the need for other,higher- 11 cost resources in the 2028 time frame.Therefore the Combined Projects meet a near- 12 term and long-term resource need as identified in the 2017 IRP.(Link Rebuttal,page 13 7,line 9 to page 17,line 8.) 14 Q.Mr.Mullins claims that the company's position is imprudent because it 15 "disregards market access"when determining resource sufficiency.(Mullins 16 Supp.Direct,page 35,lines 14-19.)Is this true? 17 A.No.Mr.Mullins implies that the company just ignores FOTs in its IRP modeling,which 18 is the exact same modeling used in this case.In fact,as I described in my rebuttal 19 testimony,FOTs must compete against all other resource options,including the 20 Combined Projects.The fact that the SO model selected the Combined Projects over 21 FOTs demonstrates that the Combined Projects are a superior resource choice to meet 22 the capacity shortfall identified in the 2017 IRP,and moreover to meet all system 23 requirements.The implication of Mr.Mullins's position is that the company should rely O Link,Supp-Reb -4 Rocky Mountain Power l on FOTs to meet its resource needs regardless of cost and risk,which is the truly 2 imprudent course of action. 3 Q.Mr.Mullins and Mr.Yankel claim that the capacity need identified in the 2017 4 IRP no longer exists when the company's resource need assessment is updated to 5 account for the most recent,lower load forecast.(MullinsSupp.Direct,page 36, 6 line 3 to page 37,line 14,and Yankel Supp.Direct,page 4,line 12 to page 5,line 7 12.)Is this true? 8 A.No.In 2021,the first full year that the Wind Projects are in service,the 2017 IRP shows 9 a capacity deficit of 1,023 MW.The updated load forecast summarized in my 10 supplemental direct testimony shows a 428-MW reduction to the coincident peak load 11 forecast in 2021 relative to the load forecast used in the 2017 IRP (Link Supp.Direct, 12 page 18,line 10 to page 19,line 8).Consequently,accounting for the updated load 13 forecast,PacifiCorp's capacity deficit in 2021 is now 595 MW (1,023 MW capacity 14 deficit less 428 MW reduction in coincident peak load).Accounting for the updated 15 load forecast,PacifiCorp's capacity need grows to 3,395 MW by 2036.The capacity 16 contribution of the Wind Projects is 207 MW (1,311 MW nameplate capacity times 17 15.8 percent capacity contribution),which is well below the 595 MW of capacity need 18 in 2021 and the 3,395 MW of capacity need in 2036 even after accounting for the 19 updated load forecast. O Link,Supp-Reb -5 Rocky Mountain Power l Q.Mr.Louis justifies his proposed conditions based upon a distinction between a 2 resource need and capacity need,pointing out that the company's 2017 IRP does 3 not show a capacity need until 2028.(Louis Supp.Direct,page 6 line 16 to page 7 4 line 4.)How do you respond? 5 A.I strongly disagree with the fundamental premise of Mr.Louis's view of resource need. 6 As discussed above,even after accounting for an updated load forecast,PacifiCorp has 7 a 595-MW capacity deficit in 2021 that grows to 3,395 MW by 2036.This means that 8 if the company did not procure any new resources,it would not have sufficient capacity 9 to reliably meet customer demand throughout the entire forecast horizon.The IRP 10 models are used to identify the least-cost,least-risk mix of resources,among all 11 resource alternatives (i.e.,FOTs,demand-side management,energy storage,and 12 generating assets),that can be used to fill this capacity deficit.In every scenario that 13 the company has analyzed,the IRP models choose the proposed Wind Projects,among 14 all other resource alternatives.This means that the proposed Wind Projects are lower 15 cost than all other resource alternatives. 16 Mr.Louis's attempt to distinguish between resource need and the need for 17 capacity to meet load is misguided and not supported.The fact is that there is no 18 difference-a resource need is defined as the need for capacity to meet load.This 19 capacity can come in many different forms (i.e.,FOTs,demand-side management, 20 energy storage,and generating assets),and the company's economic analysis clearly 21 demonstrates that the Wind Projects are the least-cost,least-risk resource alternative. O Link,Supp-Reb -6 Rocky Mountain Power l Q.Why is it importantto recognize the parties'flawed and contradictory positions 2 on resource need? 3 A.Messrs.Louis,Mullins,and Phillips have recommended several unprecedented 4 conditions that the Commission should apply if it grants CPCNs for the Combined 5 Projects,including disallowance of rate-basetreatment for any turbine not in service in 6 time to receive 100-percent PTCs,a capital-cost cap that results in an automatic 21- 7 percent disallowance,a lifetime cap on O&M and capital expenditures,imputation of 8 the full estimated PTC benefits over the next 10 years,and total disallowance if the 9 Combined Projects are not completed (Phillips Corrected Supp.Response,page 59, 10 line 7 to page 60,line 21.)They justify these conditions because they claim that the 11 "Combined Projects are an opportunity investment for RMP"and therefore "it is 12 appropriate to apply the traditional regulatory compact in reverse"to effectively 13 guarantee customer benefits and eliminate customer risk.(Phillips Direct,page 7,line 14 21 to page 8,line 5;see also Phillips Direct,page 34,line 21 to page 35,line 2 15 (Combined Projects are "discretionary,and not designed to fulfill any resource 16 requirement or other needs[.]");Phillips Direct,page 3,lines 17-19 ("Wind Projects 17 are not being pursued by RMP as a matter of need;rather,they are a discretionary 18 project predominantly intended to harvest tax credits and increase RMP's rate base 19 which might provide savings to ratepayers.")).However,because the Combined 20 Projects clearly meet a resource need in the traditional sense there is no basis for these 21 conditions. O Link,Supp-Reb -7 Rocky Mountain Power 1 Q.Have these parties demonstrated that the Combined Projects pose greater risk to 2 customers than increased reliance on FOTs in the near-term and the acquisition 3 of a resource in 2028? 4 A.No.While they recommend that customers be relieved of virtuallyall risk related to the 5 Combined Projects,they have not demonstrated that customers will be exposed to 6 higher or unreasonable risk because of the Combined Projects relative to the next best 7 resource options.Just as there is no reason customers should be relieved of all risk 8 related to FOTs,there is no reason customers should be relieved of all risk associated 9 with the Combined Projects.Because the Combined Projects meet an identified 10 resource need,there is no basis to apply conditions that represent a dramatic and 11 unprecedented departure from well-established and long-standing regulatory 12 principles. 13 Q.Mr.Phillips notes the Oregon independent evaluator's recommendation for 14 ratemaking treatment for the Combined Projects to support his proposed 15 conditions.(Phillips Supp.Direct,page 57,lines 1-35.)How do you respond? 16 A.Mr.Phillips's proposed conditions go far beyond the recommendation of the Oregon 17 independent evaluator.For example,the Oregon independent evaluator recommends a 18 hard cap on the capital and O&M costs for the Combined Projects.Mr.Phillips 19 recommends a hard cap and a 21-percent disallowance.Moreover,the Oregon 20 independent evaluator's recommendation was intended to provide a comparable risk 21 profile for utility-owned and PPA resources.Mr.Phillips's conditions are designed to 22 remove customer risk regardless of the commercial structure,as evidenced by the fact 23 his conditions were proposed before he knew whether the 2017R RFP would result in O Link,Supp-Reb -8 Rocky Mountain Power l PPAs or utility-owned resources.Ultimately,the company believes that the 2 Commission's existing ratemaking tools provide robust customer protections that do 3 not require the imposition of unprecedented conditions on the Combined Projects. 4 Q.Mr.Yankel claims that "PacifiCorp's need for more internal generation has 5 moved back farther than the 10 years originally mentioned in the filing."(Supp. 6 Direct,Page 1,lines 21-22.)How do you respond? 7 A.As described at length in my rebuttal testimony,the fact that the IRP includes FOTs 8 means that there is a resource need that is not met by existing resources.If PacifiCorp 9 can meet that need with resources that are lower cost and lower risk than FOTs,it is 10 reasonable to do so.The 2017 IRP demonstrates that there is a near-term resource need 11 that can be met with FOTs or with new wind investments enabled by the Aeolus-to- 12 Bridger/Anticline transmission line.The company's economic analysis in this 13 proceeding,developed using the same IRP models,continue to validate results in the 14 2017 IRP.Just like the 2017 IRP,the economic analysis in this proceeding demonstrates 15 that a resource portfolio that includes the proposed new wind and transmission 16 investments is the least-cost,least-risk portfolio.This conclusion has been confirmed 17 and strengthened over the course of this case. 18 PacifiCorp's 2017 IRP analysis compared new wind and transmission 19 investments to all other available resource options,including FOTs,thermal resources, 20 other renewable resources,and additional demand-side resources.The robust analysis 21 in the IRP,which was confirmed in this case,demonstrates that wind resources are 22 least-cost,least-risk even after accounting for their intermittency and resulting 23 capacity-contributionvalue. O Link,Supp-Reb -9 Rocky Mountain Power 1 Q.Mr.Phillips claims that shareholders not customers are the ones who will benefit 2 from the Combined Projects.(Phillips Supp.Direct,page 26 line 1 to page 28 line 3 2.)How do you respond? 4 A.PacifiCorp's resource decisions consider the customer costs associatedwith a particular 5 resource decision and do not,and should not,consider whether one provider or another 6 benefits from that decision.To be very clear,PacifiCorp simply selected the lowest- 7 cost,lowest-risk resources regardless of shareholder impact. 8 2017R RFP MODELING AND RESULTS 9 Q.Monsanto,PIIC and Staff claim that the 2017R RFP was unfair and biased.(See, 10 e.g.,Phillips Supp.Direct,page 13,line 21 to page 14,line 5;Mullins Supp.Direct, 11 page 17,lines 15-20;Eldred Supp.Direct,page 3 lines 8-18.)What is your general 12 response to this contention? 13 A.I disagree.More importantly,these witnesses'assertions are directly contrary to the 14 conclusions of the independent evaluators who monitored the 2017R RFP.Both 15 independent evaluators provided their own independent analysis and carefully 16 scrutinized the process and results.And both independent evaluators concluded that the 17 2017R RFP was transparent,fair,and unbiased. 18 Q.Please provide more detail on the role of the independent evaluators. 19 A.The 2017R RFP was overseen by two independent evaluatorsone appointed by the 20 Public Utility Commission of Oregon ("Oregon Commission")and retained by 21 PacifiCorp,and one appointed and retained by the Public Service Commission of Utah 22 ("Utah Commission").In accordance with the statutes,rules,and policies in Oregon 23 and Utah,the independent evaluator is an independent expert appointed and managed O Link,Supp-Reb -10 Rocky Mountain Power 1 by the commission (not PacifiCorp)to ensure that the RFP process was conducted in a 2 fair and unbiased manner and the final shortlist projects are reasonable and consistent 3 with the modeling results used to evaluate bids. 4 In the 20l7R RFP,both independent evaluators were involved from the 5 beginning-providing feedback and recommendations regarding the design and 6 content of the 2017R RFP and actively participatingin every stage of the RFP.For its 7 part,PacifiCorp ensured that the independent evaluators had complete and unrestricted 8 access to all information related to the 2017R RFP and kept both independent 9 evaluators informed of developments as they occurred. 10 Q.Did the independent evaluators provide an assessment of PacifiCorp's benchmark 11 resources bid into the 2017R RFP (i.e.,TB Flats I and II,Ekola Flats,and 12 McFadden Ridge II)? 13 A.Yes.Because the 2017R RFP included benchmark resources,both independent 14 evaluators provided detailed assessments of the benchmark bids to ensure that they 15 were reasonable and would not bias the solicitation in favor of utility-owned resources. 16 The benchmark review process occurred before any other bids were received to provide 17 additional assurance that the benchmarks were not provided an unfair advantage. 18 Oregon's final independent evaluator report,issued February 16,2018,is provided as 19 Highly Confidentialand Confidential Exhibit No.67 ("Oregon IE Report"),and Utah's 20 independent evaluator report is Highly Confidential and Confidential Exhibit No.68 21 ("Utah IE Report"). O Link,Supp-Reb -11 Rocky Mountain Power l Q.Did the independent evaluators review confirm the reasonableness of the 2 benchmark bids? 3 A.Yes.As described in my second supplemental direct testimony,the Utah independent 4 evaluator concluded that (1)PacifiCorp provided detailed information related to the 5 benchmarks that exceeded industry standards,(2)cost estimates were reasonable,and 6 (3)the review,assessment,and scoring of the benchmark resources was conducted in 7 a fair and equitable manner with no outward perception of bias.(Link Second Supp., 8 page 29,line 20 to page 30 line 12.) 9 The Oregon independent evaluator also conducted a thorough assessment of the 10 benchmarks,noting that when "assessing a utility's own bids in response to the RFP, 11 our greatest concern is that the utility will incorporate cost estimates that have been 12 aggressively estimated and do not characterize the costs of the project accurately." 13 (Oregon IE Report at 10.)To make its assessment,the Oregon independent evaluator 14 "looked at a detailed breakdown of each of the benchmarks costs to determine if any 15 items have been improperly omitted from the cost calculation,and at overall capital 16 cost levels by comparing them to publicly-available data on recent wind generation 17 capital costs."(Oregon IE Report at 10.)This "comparison provided a measure of the 18 overall reasonablenessof the Benchmark capital costs and capacity factors."(Oregon 19 IE Report at 10.)The Oregon independent evaluator ultimately found that the 20 benchmarks were acceptable based on three items: 21 First,the benchmarks were not deliberately underpriced through omission of 22 any capital cost components. Link,Supp-Reb -12 Rocky Mountain Power l Second,the benchmark capital and operating costs appeared reasonable when 2 compared with public data on U.S.wind projects. 3 Third,the capacity factors of the benchmarks were reasonable when compared 4 with public data and were supported by credible third-party analysis. 5 (Oregon IE Report at 10-11.) 6 Q.Did the independent evaluators provide any overall conclusions related to the 7 2017R RFP? 8 A.Yes.The Oregon independent evaluator recommended that the Oregon Commission 9 approve PacifiCorp's final shortlist based on the following conclusions: 10 The selected bids represent the top offers that are viable under current ll transmission planning assumptions and provide the greatest benefits to 12 ratepayers. 13 The selected bids represent the best viable options from a competitive 14 perspective,based on the 59 bid options presented. 15 The independent evaluator's analysis confirmed that the selected bids were 16 reasonably priced and,while not the lowest-cost offers,were the lowest-cost 17 offers that were viable under current transmission planning assumptions.The 18 independent evaluator's analysis included its own cost models for each bid 19 option and a review of PacifiCorp's models. 20 The independent evaluator took special care to confirm the selection of 21 PacifiCorp's benchmark resources.The independent evaluator confirmed the 22 accuracy of the benchmark costs and scoring.The independent evaluator noted 23 that the benchmark bids were disciplined by the fact that a third-party bidder O Link,Supp-Reb -13 Rocky Mountain Power l submitted a competing offer for a build-transfer agreement ("BTA")for 2 benchmark projects. 3 The independent evaluator confirmed that the 2017R RFP aligns with the 4 2017 IRP. 5 (Oregon IE Report at 2-3.) 6 The Utah independent evaluator also supported the final shortlist projects based 7 on the followingconclusions: 8 The 2017R RFP was fair,reasonable,and generally in the public interest.(Utah 9 IE Report.) 10 The bid evaluation and selection processes were designed to lead to the 11 acquisition of wind-generated electricity at the lowest reasonable cost based on 12 the detailed state-of-the-art portfolio evaluation methodology used,the steps 13 taken to achieve comparability between utility cost-of-service resources and 14 third-party firm priced bids,the flexibility afforded bidders via a range of 15 eligible resource alternatives,and the attempt to allow for equal terms for PPA 16 and BTA resources.(Utah IE Report at 71.) 17 PacifiCorp's modeling demonstrates that the Combined Projects "should result l 8 in significant savings for customers."(Utah IE Report at 83.)Further,because 19 PTCs will flow through to customers in the first ten years,the "near-term 20 benefits to customers should be significant."(Utah IE Report at 83.) O Link,Supp-Reb -14 Rocky Mountain Power l Q.Please respond to Mr.Phillips's and Mr.Eldred's claims that PacifiCorp's changes 2 to its economic modeling for purposes of developing the final shortlist for the 3 2017R RFP unfairly biased the results.(Phillips Rebuttal,page 6,lines 3-11; 4 Eldred Supp.Direct,page 6,lines 11-15.) 5 A.As explained in my supplemental direct testimony,when comparing bids in the 2017R 6 RFP portfolio development phase,for self-build and BTA bids,PTC benefits were 7 applied on a nominal basis rather than a levelized basis to better reflect how the PTC 8 benefits flow through customer rates.(Link Supp.Direct,page 25,line 11 to page 26, 9 line l 1.)This refinement better aligns project costs and benefits and impacts only the 10 System Optimizer ("SO")model and Planning and Risk model ("PaR")results through l 1 2036.This modeling refinement had no impact on the nominal revenue requirement 12 calculations that were also reported in my supplemental direct and second supplemental 13 direct testimony.This change did not bias the results of the 2017R RFP as Mr.Phillips 14 and Mr.Eldred claim,as described in more detail below. 15 Q.Did you continue to use levelized capital costs during the portfolio development 16 phase of the 2017R RFP bid evaluation and selection process? 17 A.Yes. 18 Q.Is the treatment of PTCs and capital costs consistent with how PacifiCorp has 19 analyzed specific resource decisions using its IRP models in the past? 20 A.Yes.When the company has historically conducted economic analysis of specific 21 resource decisions,it treats costs that are not spread over the life of the asset on a 22 nominal basis.Typically,this means that capital costs are levelized,while other costs, 23 such as operations and maintenance ("O&M")costs,are nominal.The company used O Link,Supp-Reb -15 Rocky Mountain Power 1 this approach without controversy when it requested CPCNs to install emission control 2 equipment at its Jim Bridger Unit 3 and Unit 4 coal units and when it conducted coal- 3 plant analysis in its IRPs.The refined modeling used here simply conforms the 4 treatment of PTCs to the treatment of other costs and benefits that are not spread out 5 over the life of the asset. 6 Q.Does PacifiCorp intend to model PTCs in this manner in its IRPs? 7 A.Yes.Because modeling PTCs on a nominal basis better reflects how they are treated in 8 rates,PacifiCorp intends to use this approach in future IRPs. 9 Q.Did the independent evaluators overseeing the 2017R RFP object to PacifiCorp's 10 refined modeling? 11 A.No.Both independent evaluators overseeing the 2017R RFP were aware of 12 PacifiCorp's decision to model PTC benefits on a nominal rather than levelized basis, 13 and neither concluded that the refinement biased the bid-evaluationresults.In fact,the 14 sensitivity analysis requested by the independent evaluators that I described in my 15 supplemental direct testimony,(Link Supp.Direct,page l 1,line 1 to page 12,line 2), 16 was designed to specifically test whether the refined modeling of PTC benefits 17 unreasonably biased the resource selection.The Oregon independent evaluator's report 18 supports the conclusions I reported regarding this sensitivity.According to the Oregon 19 independent evaluator,levelizing the PTC benefits caused the SO model to select PPAs 20 instead of self-build and BTA bids.(Oregon IE Report at 30.)But "looking at the actual 21 flow of cost recoveries,treating both PTCs and costs as incurred"out through 2050, 22 demonstrated that each portfolio produced virtuallyidentical net benefits.(Oregon IE 23 Report at 32.)The Oregon independent evaluator also noted that the PPA portfolio was O Link,Supp-Reb -16 Rocky Mountain Power l more expensive in the early years.(Oregon IE Report at 32.)Thus,PacifiCorp's refmed 2 PTC modeling did not unreasonably bias the selection of resources.The Oregon 3 independent evaluator also specifically noted that the PTC-modeling refinement "had 4 no impact on winning projects selected in this RFP"because several of the PPAs that 5 were selected in the sensitivity requested by the independent evaluators were ultimately 6 non-viableprojects.(Oregon IE Report at 5.) 7 Q.Did the Utah independent evaluator discuss this treatment of PTCs in the portfolio 8 development phase of the 2017R RFP? 9 A.Yes.The Utah independent evaluator noted a concern that the PTC modeling could 10 produce a bias in favor of utility-owned resources "if only a portion of the capital costs 11 associated with the benchmarks and BTAs are recovered during the 20-year evaluation 12 period,since these projects have a 30-year life and capital cost recovery period."(Utah 13 IE Report at 62.)In response,the Utah independent evaluator described the additional 14 analysis provided by the company,along with several meetings with the independent 15 evaluators to discuss this issue.The Utah independent evaluator observed in his report 16 that PacifiCorp "refuted the basis for evaluating PTCs on a levelized cost basis since 17 [PacifiCorp]would flow through all the customer costs in the near-term."(Utah IE 18 Report at 62.)Further,according to the Utah independent evaluator,PacifiCorp "also 19 provided a 30-year analysis of the costs and benefits of the initial portfolio [i.e.,the 20 portfolio with utility-owned resources]and the updated portfolio [i.e.,the portfolio with 21 PPAs]...to demonstrate that the original portfolio would still provide greater benefits 22 over a 30-year timeframe."(Utah IE Report at 62.) Link,Supp-Reb -17 Rocky Mountain Power 1 When PacifiCorp presented its final shortlist to the independent evaluators,the 2 Utah independent evaluator provided additional discussion of this issue: 3 PacifiCorp also addressed two of the IEs concerns raised in 4 discussions on shortlist evaluation and selection.The first issue 5 dealt with the application of the PTCs in the evaluation 6 methodology.As noted,PacifiCorp's analysis assumes that the 7 PTC inputs to the SO model would be based on nominal dollar 8 values since the actual benefits would be flowed through to 9 customers.The Oregon IE requested a sensitivitywhere the PTC 10 benefits produced by BTA and benchmark options would be 11 levelized over the full 30-year life of the project.A second issue 12 raised by the IEs was whether the term of the analysis through 13 2036 (approximately 16 years)and the real levelized cost 14 treatment for capital revenue requirements adequately reflects 15 all the capital costs associated with utility ownership options 16 over a thirty-year project life.In response,PacifiCorp completed 17 an analysis of the expected benefits and costs through 2050 O 18 comparing the results of PacifiCorp's selected portfolio and the 19 IE sensitivitycase.In its presentation,PacifiCorp concluded that 20 the PVRR(d)benefits through 2036 from the final shortlist 21 portfolio total $343 million and the benefits from the IE 22 Sensitivity with the PPA included in the bid portfolio total $277 23 million.Through 2050,the benefits from the final shortlist bid 24 portfolio of $223 million are closely aligned with the IE 25 Sensitivity bid portfolio that provides an estimated $224 million 26 in benefits through 2050.The revised shortlist portfolio provides 27 greater near-term benefits. 28 (Utah IE Report at 65.) 29 Q.Did the Utah independent evaluator provide any conclusions related to whether 30 the self-build or BTA bids received a preference as a result of PacifiCorp's 31 modeling? 32 A.Yes.The Utah independent evaluator concluded that the results of the sensitivity 33 (discussed above)"indicated that there did not appear to be an inherent advantage Link,Supp-Reb -18 Rocky Mountain Power 1 associated with a utility-ownership bid due to the shorter evaluation period for purposes 2 of evaluating and selecting a portfolio of resources."(Utah IE Report at 75.)The 3 independent evaluator explained that the "net benefits approach used may eliminate the 4 costs for a longer-term resource but also eliminates the revenue side of the equation, 5 which would likely be escalating over time."(Utah IE Report at 75.)Thus,the 6 company's modeling "allows for a consistent and fair evaluation of bids of different 7 technologies and terms and is a reasonabletool for initial evaluation of bids."(Utah IE 8 Report at 75.) 9 Q.Mr.Phillips and Mr.Eldred claim that the use of nominal pricing for the PTCs 10 and levelized pricing for the capital costs create an improper mismatch that biased l 1 the resources selected in the 2017R RFP.(Phillips Supp.Direct,page 6,lines 12- 12 23;Eldred Supp.Direct,page 7 line 10-15.)Do you agree? 13 A.No.Moreover,neither of the independent evaluators that monitored the 20l7R RFP 14 agree either,as discussed above.Mr.Phillips and Mr.Eldred claim that the use of 15 nominal PTC pricing together with levelized capital costs improperly reduced the net 16 present value ("NPV")of utility-owned resources making it more likely that the SO 17 model would select self-build or BTA bids.(Phillips Supp.Direct,page 6,lines 20-21; 18 Eldred Supp.Direct,page 3,lines 8-18.). 19 The IRP models select least-cost portfolios based on present-value system costs. 20 And it would not be appropriate to include nominal revenue requirement from capital 21 investments for assets having a depreciable life that extends beyond the 20-year IRP 22 study period in any present-value calculation.It would only be appropriate to include 23 capital revenue requirement on a nominal basis in present-value calculations when O Link,Supp-Reb -19 Rocky Mountain Power l those calculations cover the full life of the proposed new wind facilities.In contrast,it 2 is appropriate to consider nominal PTC benefits in the IRP models because all of these 3 benefits will be realized within the 20-year time frame of those studies.This is 4 consistent with how PTCs will flow through to customer rates,and consequently, 5 PacifiCorp's IRP models appropriately weight the front-end loaded PTC benefits 6 without disproportionatelyweighting capital costs in its present-value calculations. 7 It is also disingenuous for Mr.Phillips to imply that PacifiCorp's modeling 8 change was improperly motivated when he argues that the nominal revenue 9 requirement results,which have always applied PTCs on a nominal basis,"most closely 10 depict how project costs and benefits will pressure rates."(Phillips Supp.Direct,page 11 7,lines 5-7.) 12 Q.Did Mr.Phillips refute the sensitivity analysis requested by the independent 13 evaluators that was presented in your supplemental direct testimony and 14 discussed at length in the independent evaluators'reports? 15 A.No. 16 Q.Mr.Phillips also claims that the company improperly used "real"levelization 17 instead of uniform levelization.(Phillips Corrected Supp.Direct,page 11,lines 1- 18 6.)Is this true? 19 A.No.I explained in my direct testimony that it is important to levelize capital revenue 20 requirement in the SO model and PaR to avoid potential distortions in the economic 21 analysis of capital-intensive assets that have different lives and in-service dates.(Link 22 Direct,page 26,lines 13 to page 27,line 22.)As noted by Mr.Phillips,the company 23 uses an inflation-adjusted real-levelized method rather than using a uniform- O Link,Supp-Reb -20 Rocky Mountain Power l levelization method.The inflation-adjustedreal-levelized method more closely aligns 2 with the fact that benefits for capital investments generally increase over time. 3 Consequently,and similar to the problems associated with using a nominal revenue 4 requirement approach in the SO model and PaR,the application of a uniform- 5 levelization method would also create potential distortions in resource selections for 6 capital-intensive assets that have different lives and in-service dates. 7 Q.Mr.Phillips suggests that the Wind Projects are higher risk than PPAs because 8 customers are insulated from risks when the company executes PPAs,whereas 9 customers bear risks for utility-ownedresources (e.g.,the risk of construction cost 10 over-runs).(Phillips Supp.Direct page 14,lines 17-24,page 15,lines 1-18.)Mr. 11 Mullins makes a similar point.(MullinsSupp.Direct,page 14,lines 7-22.)How do 12 you respond? 13 A.I disagree.Mr.Phillips ignores the fact that customers also receive upside benefits for 14 utility-owned resources that they do not receive under a PPA.For example,I described 15 in my previous testimony the potential upside benefits associated with renewable 16 energy credits ("RECs"),reduced O&M costs,and increased energy production.(Link 17 Second Supp.Direct,page 15,line 4 to page 16,line 13;Exhibit No.38;see also Teply 18 Rebuttal,page 16,lines 2-15.)In each of these cases,customers will receive the 19 increased benefits because of the nature of cost-of-service ratemaking.Under a PPA 20 structure,on the other hand,project owners receive all the upside benefits.PPAs can 21 provide some amount of certainty,but that certainty can both benefit and harm 22 customers. O Link,Supp-Reb -21 Rocky Mountain Power l Moreover,a utility self-build or BTA project provides substantial long-term 2 benefits that customers never receive under a PPA.Once a PPA term expires,customers 3 walk away with nothing.If the utility owns the resource,however,customers will 4 continue to receive the benefits of that resource for as long as it operates,and even after 5 the resource is no longer operational,customers retain the value associated with the 6 land and facilities that have lives that extend beyond the life of the generating resource. 7 To use Mr.Phillips's example of a home mortgage,under a utility-owned bid, 8 customers pay the mortgage and,after 30 years,they own the home.Under a PPA, 9 customers pay the mortgage and,after 30 years,customers have nothing. 10 Q.Mr.Phillips also complains that he had insufficient time to review the Combined 11 Projects.(Phillips Supp.Direct,page 5,lines 11-22,page 6,lines 1-2.)Do you 12 agree? 13 A.No.Parties have had nine months to review the proposed resource decision in this case. 14 Over that time,the Aeolus-to-Bridger/Anticline line has not changed in any material 15 way.While it is true that the results of the 2017R RFP were disclosed fairly recently, 16 PacifiCorp's modeling has remained virtually unchanged,and three of the four 17 resources included in the company's initial filing were also included in the fmal 18 shortlist. 19 Q.Mr.Mullins disputes the fact that the least-cost bids were selected due to 20 interconnection queue positions.(MullinsDirect Supp.,page 2,lines 11-14,and 21 page 9,lines 7-9.)How do you respond? 22 A.As explained by Mr.Rick A.Vail,bids with generator interconnection queue positions 23 lower than Q0712 (i.e.,projects with higher interconnection queue numbers)required O Link,Supp-Reb -22 Rocky Mountain Power l the construction of Gateway South for interconnection and therefore could not be 2 completed by 2020.Mr.Mullins's claim that the company could have allowed projects 3 with higher interconnection queue numbers (i.e.,projects with lower interconnection 4 queue positions)to bypass projects with lower interconnection queue numbers (i.e., 5 projects with higher interconnection queue positions)is contrary to PacifiCorp's open 6 access transmission tariff ("OATT"). 7 Q.Did the independent evaluators address this issue? 8 A.Yes.Both independent evaluators agreed with PacifiCorp's assessment that projects 9 with interconnection queue positions lower than Q0712 were non-viable.The Oregon 10 independent evaluator explained that PacifiCorp's "transmission arm,which assesses 11 interconnection costs,must,by law,assume that each queue project is interconnected 12 in order received so each project assumes that all projects ahead of it in the queue are 13 interconnected."(Oregon IE Report at 32.)Thus,"[a]s more projects in the Wyoming 14 area are interconnected it puts more strain on the transmission system until eventually 15 major upgrades such as the Gateway West and South projects are needed."(Oregon IE 16 Report at 32.)In this case,the major upgrades were required for all projects with queue 17 positions lower than Q0712.The Oregon independent evaluator concluded that it 18 "understand[s]and appreciate[s]PacifiCorp's position and do[es]not disagree with 19 their transmission department's findings (beyond noting the obvious fact that many 20 projects will likely drop out of the queue and that actual interconnection costs will 21 differ from projected)."(Oregon IE Report at 35.)According to the independent 22 evaluator,"[t]o go forward with projects that cannot meet the proposed online date O Link,Supp-Reb -23 Rocky Mountain Power l without major accelerated transmission investment would not seem to be the wisest 2 course of action."(Oregon IE Report at 35.) 3 Q.Is the fact the independent evaluators disagree with Mr.Mullins's claim 4 particularlynotable? 5 A.Yes.While Mr.Mullins appears to rely on the independent evaluators,neither support 6 his conclusion. 7 Q.Mr.Mullins claims that the company never disclosed the possibility that a bidder's 8 interconnection queue position could impact the viabilityof its project.(Mullins 9 Supp.Direct,page 11,lines 11-25.)Is this true? 10 A.No.The fact that there was limited interconnection capability was known at the 11 beginning of the 2017R RFP process,which is why PacifiCorp's initial minimum bid 12 eligibility screen included a requirement for an interconnection system impact study. 13 Commenters and bidders requested that this requirement be removed from the 14 minimum bid eligibility screen to allow broader participation.At the recommendation 15 of the independent evaluators,this restriction was changed to generators who had begun 16 the interconnection study processi.This change increased the number of projects that 17 could bid into the 2017R RFP,which resulted in robust participation,including 18 numerous bids that were not dependent on the construction of the Aeolus-to- 19 Bridger/Anticline line.Although transmission constraints ultimately rendered some 20 bids non-viable,neither of the independent evaluators indicated that the 2017R RFP 21 process was biased or unreasonable because of this fact. 'See Application of Rocky Mountain Power for Approval ofSolicitation Processfor Wind Resources,Utah PSC Docket No.17-035-23,Hearing Transcript,page 56,lines 4-10 (Sept.19,2017). Link,Supp-Reb -24 Rocky Mountain Power l Q.Mr.Mullins also claims that the company's "treatment of transmission costs"was 2 inconsistent with its communications with bidders in the period leading up to the 3 2017R RFP.(MullinsSupp.Direct,page 12,lines 1-13.)Is this true? 4 A.No.Mr.Mullins confuses transmission costs with interconnection costs.Mr.Mullins is 5 correct that the company informed bidders that costs associated with the Aeolus-to- 6 Bridger/Anticline transmission line,which relieves congestion and enables 7 interconnection would not be assigned to individual projects.PacifiCorp did not inform 8 bidders that interconnection costs required to receive interconnection service,which 9 are specific to any individual wind facility,would not be accounted for in the 10 company's bid selection and evaluation process.In fact,one of the minimum bid- 11 eligibility requirements explicitly identified in the 2017R RFP clearly states that bids 12 could be disqualified if bidders failed to provide interconnection costs.In specifying 13 this minimum bid-eligibility requirements,the 2017R RFP document further states that 14 cost estimates are required even if a study from the transmission provider was not 15 completed or available at the time bids were due.Clearly,PacifiCorp would not have 16 established this minimum bid-eligibility requirement,which if not met could disqualify 17 a bid,if it did not intend to use this information to evaluate bids submitted into the 18 2017R RFP. 19 Q.Mr.Mullins claims that the company should have "either equalize[d]or 20 mitigate[d]the bidding advantage otherwise available to a bidder with a higher 21 queue position."(MullinsSupp.Direct,page 13,lines 9-14.)Can PacifiCorp do 22 what Mr.Mullins recommends? 23 A.No.As described above,such action would be inconsistent with the company's OATT. O Link,Supp-Reb -25 Rocky Mountain Power 1 Mr.Mullins's apparent misunderstanding of this fact is not a reflection on the accuracy 2 of the 2017R RFP process nor an indicationthat the process was unfair. 3 Q.Mr.Mullins claims that because "PacifiCorp applied incremental transmission 4 costs to the bids whose queue position exceeded the incremental transmission 5 capacity,the higher queue position resources had no way of being selected by the 6 model."(MullinsSupp.Direct,page 13,lines 15-18.)Is this true? 7 A.No.In fact,my supplemental direct testimony describes the bid evaluation and 8 selection process that was completed before considering the results of the 9 interconnection restudy process.The original final shortlist of bids summarized in that 10 testimony included the same projects selected to the updated final shortlist summarized 11 on my second supplemental direct testimony except that the original final shortlist 12 included the McFadden Ridge II benchmark bid.In direct contradiction to the claimsO13madebyMr.Mullins,the original bid evaluation and selection process performed by 14 PacifiCorp and monitored by two independent evaluators demonstrates that the 15 interconnection restudy process did not prevent,in any way,the selection of projects 16 becauseof their interconnection queue number. 17 Q.Based on this understanding,Mr.Mullins then argues that there is no way to know 18 if the best resources were actually selected to the final shortlist.(MullinsSupp. 19 Direct,page 13,lines 15-23.)Is this true? 20 A.No.As discussed above,Mr.Mullins's assertion is contrary to basic facts and,therefore, 21 fundamentally flawed.Before considering results of the interconnection restudy 22 process,the only interconnection-related constraint was the assumption that total 23 interconnection capability with the addition of the Aeolus-to-Bridger/Anticline Link,Supp-Reb -26 Rocky Mountain Power l transmission line would be 1,270 MW.The interconnection restudies performed after 2 the original final shortlist was determined resulted in the followingconclusions: 3 (1)That the TB Flats I and II and Cedar Springs projects could interconnect 4 with the addition of the Aeolus-to-Bridger/Anticline transmission line and no 5 other elements of the company's long-term plan; 6 (2)That McFadden Ridge II could not interconnect without additional elements 7 of the company's long-term transmission plan,namely Gateway West and 8 Gateway South;and 9 (3)That additional interconnection capability would be created with the 10 addition of the Aeolus-to-Bridger/Anticline transmission line,which allowed 11 McFadden Ridge II to be replaced with Ekola Flats. 12 Rather than limiting the outcome of the 2017R RFP,the interconnection restudy 13 process provided new information that allowed the inclusion of a more economic 14 project because of increased interconnection capability.The only thing that was 15 preventing the models from choosing Ekola Flats over McFadden Ridge II in 16 development of the original final shortlist was the original 1,270-MW limit on 17 interconnection capability. 18 Mr.Mullins also ignores the fact that the interconnection considerations 19 resulted in PacifiCorp proposing to replace only one shortlist bid,with all other shortlist 20 bids remaining unchanged.More specifically,the interconnection restudy process 21 provided new,more updated information that caused PacifiCorp to exclude the 22 McFadden Ridge II benchmark bid.While the new and more updated information from 23 the interconnection restudy process demonstrates that projects with an interconnection 24 queue number greater than Q0712 would not be viable at this time,this information 25 had no impact on selection of the best resources other than allowing the more economic 26 Ekola Flats benchmark bid to replace the McFadden Ridge II benchmark bid. Link,Supp-Reb -27 Rocky Mountain Power 1 This single shortlist change resulting from interconnection restudies can hardly 2 be described as interfering with the value of the company's entire competitive 3 solicitation process.Allowing participation without regard to interconnection queue 4 position or study status resulted in a robust competitive solicitation,including 5 numerous bids that were not enabled by construction of the Aeolus-to- 6 Bridger/Anticline transmission line.Interconnection considerations,based on the most 7 current and up-to-date information,causing the replacement of a single project did not 8 unravel those benefits.What Mr.Mullins really appearsto be arguing is that the original 9 (pre-interconnection considerations)shortlist should have included lower-queued 10 projects for other,non-interconnection-related reasons,not that interconnection queue 11 considerations caused those projects to be eliminated from the shortlist in the first 12 place.These arguments should be disregarded because they are inconsistent with the 13 results of the economic evaluation of the bids. 14 Q.Mr.Mullins claims that PPA bids were better alternatives and that these 15 alternatives were eliminated based only on their interconnection queue position. 16 (MullinsSupp.Direct,page 13,line 18 to page 14,line 6.)Is this true? 17 A.No.As described above,the preliminary shortlist of bids that was selected before the 18 interconnection restudy processwas finalized included virtuallythe same resources that 19 are included in the updated final shortlist.Moreover,as discussed in my supplemental 20 direct testimony,at the request of the independent evaluators,PacifiCorp conducted a 21 sensitivity to specifically test whether the highest performing PPAs bid into the RFP 22 could displace the bids selected to the preliminary shortlist.This sensitivity study, O Link,Supp-Reb -28 Rocky Mountain Power l developed before the interconnection restudy process was completed,show that the 2 PPAs were not superior resource selections. 3 ECONOMIC ANALYSIS 4 Q.Mr.Phillips argues that the Commission should give greater weight to the nominal 5 revenue requirement analysis,which was performed through2050.(Phillips Supp. 6 Direct,page 7,lines 1-13.)Mr.Mullins agrees.(MullinsSupp.Direct,page 18, 7 lines 11-20.)Do you agree? 8 A.No.Both types of analysis-thesystem modeling results through 2036 and the nominal 9 revenue requirement results through 2050-are useful in assessing the economics of 10 the Combined Projects.The system modeling results provide a view of economic 11 analysis that is consistent with the planning period and approach used to identify a 12 least-cost,least-risk preferred portfolio in the IRP.This type of analysis was used to 13 identify new wind and transmission projects as an element of PacifiCorp's least-cost, 14 least-risk plan in the 2017 IRP and has been used to evaluate past resource acquisitions 15 and plant investments.For instance,the same IRP models used to evaluate the 16 Combined Projects in this proceeding,configured to simulate PacifiCorp's system over 17 a 20-year time frame with the application of levelized capital costs,were used to 18 support the company's acquisition of the Chehalis combined-cycle plant,support 19 selection of the Lake Side 2 combined-cycle plant through an RFP process,and to 20 support the company's Wyoming CPCN application for the installation of selective 21 catalytic reduction equipment at Jim Bridger Unit 3 and Unit 4. 22 The nominal revenue requirement analysis provides a sense of how the 23 Combined Projects might impact customer rates,relative to alternative resource O Link,Supp-Reb -29 Rocky Mountain Power l procurement scenarios,over time.While an extension of system benefits associated 2 with the Combined Projects through 2050 enables a present-value revenue-requirement 3 differential ("PVRR(d)")to be calculated,as with any long-term study,longer-term 4 results are increasingly more difficult to project.Moreover,I noted in my second 5 supplemental direct testimony that the long-term extrapolationof system benefits used 6 in the nominal revenue requirement analysis is conservative because the extrapolation 7 approach yields projected benefits that do not reach the levels observed in the model in 8 2036 until 2047. 9 Q.Mr.Mullins claims that there is no reason to analyze levelized costs at all in this 10 case because the "[u]se of levelized costs might be appropriate when considering 11 the costs of multipledifferent resources in a capital expansion model and where 12 the study period does not align with the useful life of the resource."But,according 13 to Mr.Mullins,the company is not doing that in this case.(MullinsSupp.Direct, 14 page 18,lines 14-20.)Is this true? 15 A.No.Mr.Mullins appears to fundamentallymisunderstand PacifiCorp's modeling used 16 to both select the bids to the final shortlist and to evaluate the customer benefits of the 17 Combined Projects.Mr.Mullins implies that the company has somehow hardwired the 18 Wind Projects into its models and then measured the benefits using the SO model and 19 PaR without considering the costs of multiple resource alternatives.To the contrary,in 20 every price-policy scenario,the company used the SO model (i.e.,Mr.Mullins's 21 referenced "capital expansion model")to determine which,if any,of the bids submitted 22 into the 2017R RFP were selected as an element of the least-cost mix of resources over 23 a 20-year study period.Wind bids were forced to compete with all other resource O Link,Supp-Reb -30 Rocky Mountain Power l options (including solar resources in the solar sensitivity discussed below)and were 2 selected only if they were least-cost,which is precisely how the company conducts its 3 IRP modeling.In every scenario studied,the SO model selected the Wind Projects. 4 Q.Mr.Phillips claims that the economics of the Combined Projects are no better than 5 when originally proposed.(Phillips Supp.Direct,page 45,lines 1-5.)Do you 6 agree? 7 A.No.Mr.Phillips concedes that PacifiCorp's updated nominal revenue requirement 8 analysis shows that the benefits under the medium natural gas,medium CO2 SCenaTIO 9 increased from $137 million to $167 million-an increase of over 20 percent.(Phillips 10 Supp.Direct,page 46,line 3-7.)Mr.Phillips's claim that the company's testimony was 11 "erroneous and misleading"on this point is unsupported. 12 Q.Mr.Phillips claims that the updated nominal revenue requirement analysis shows 13 that the NPV savings over the first20 years is lower than in the company's original 14 analysis.(Phillips Supp.Direct,page 46,line 8 to page 47,line 16.)How do you 15 respond? 16 A.It is not surprising that the updated nominal revenue requirement analysis,reflecting 17 winning bids from the 2017R RFP and changes in federal tax law,produces a different 18 net-benefitprofile than what was shown in my original analysis,which reflected proxy 19 wind resources and higher federal tax rates for corporations.Importantly,and as stated 20 in my second supplemental direct testimony,with reduced costs from the winning bids 21 from the 2017R RFP,the Combined Projects generate substantial near-term benefits 22 despite a reduction in PTC benefits associated with changes in federal tax law,and O Link,Supp-Reb -31 Rocky Mountain Power l generate net benefits in 23 years out of the 30 years that the proposed owned-wind 2 resources are assumed to operate.(Link Second Supp.,page 20,lines 2-9.) 3 Q.Mr.Phillips also claims that the updated nominal revenue requirement analysis 4 shows that the majorityof the customer benefits occur later and therefore the 5 Combined Projects are now riskier as compared to the original filing.(Phillips, 6 Supp.Direct,page 48,lines 1-10.)Is this a fair metric for measuring risk? 7 A.No.As noted above,Mr.Phillips is simply stating that updated nominal revenue- 8 requirement analysis produces a different net-benefit profile than what was shown in 9 my original analysis,which primarily reflects changes in Wind Project costs and 10 associated network upgrades,federal income tax rates applicable to corporations,and 11 updated system assumptions (i.e.,more current price-policy scenario assumptions and 12 an updated load forecast).This does not mean that project risks have increased.In fact, 13 project risks have been materially reduced since the company's original filing.For 14 instance,when the company made its initial filing,it was uncertain whether federal tax- 15 reform legislation would be introduced and how that legislation might impact PTC 16 benefits,which are critical to the economic benefits of the Combined Projects. 17 Similarly,at that time,the company had not yet issued the 2017R RFP and had not 18 received firm pricing for wind resource bids solicited through a competitive bidding 19 process.At this time,these uncertainties have been eliminated and replaced with known 20 tax law changes and firm,competitive wind resource pricing,and the updated economic 21 analysis of the Combined Projects continues to demonstrate that these investments will 22 generate substantial customer benefits. O Link,Supp-Reb -32 Rocky Mountain Power l Q.Mr.Phillips and Mr.Eldred claim that the onlyway the company can claim a $167 2 million customer net benefit using its nominal revenue-requirement analysis is to 3 include a terminal value benefit in 2050 that was not included in the original 4 analysis (Phillips Supp.Direct,page 49,line 1-7:Eldred Supp.Direct,page 9 lines 5 1-8.)How do you respond? 6 A.It is reasonable to include a terminal value benefit for projects where the company 7 retains control of the site at the end of the asset life and,contrary to Mr.Phillips's claim, 8 the company's analysis does not rely heavily on 2050 results to demonstrate a positive 9 net benefit.Even if the terminal value were completely eliminated,which would not be 10 appropriate,the Combined Projects would still produce $124 million in net customer 11 benefits before accounting for the conservative extrapolationmethodology used by the 12 company,conservative CO2 emissions cost savings,potential upside in O&M cost 13 savings,and upside from renewable energy credit ("REC")potential revenue. 14 Q.Why did the company include a terminal value benefit for utility-owned 15 resources? 16 A.The terminal value benefit recognizes the fact that at end of a utility-owned resource's 17 life,there is residual value that accrues to customers.For a PPA,the terminal value 18 accrues to the project owner,not customers.That terminal value includes the facilities 19 supporting the resources,like transmission facilities,that have longer useful lives and, 20 in the case of generation tied to natural resources such as wind resources,there is 21 inherent value in the site itself-particularly resources located in high-capacity-factor 22 geographic areas like eastern Wyoming.These high-value,renewable-resource O Link,Supp-Reb -33 Rocky Mountain Power l locations are often scarce or unique in their suitability for generation permitting and 2 construction,as well as proximity to transmission. 3 Q.Did the independent evaluators comment on the inclusion of the terminal value 4 benefit in the 2017R RFP modeling? 5 A.Yes.The Utah independent evaluator observed that the terminal value is typicallyequal 6 to the net salvage value of the resource,but for wind resources there are additional 7 "assets associated with the wind site,such as land,site characteristics and generation 8 interconnection and transmission facilities"that may provide additional value.(Utah 9 IE Report at 33.)The independent evaluator explained that the terminal value benefits 10 reflected the depreciated value of assets that have not fully depreciated at the end of the 11 assumed 30-year life for the wind facilities,such as transmission assets,and the 12 appreciated value of other elements of the project that remain at the end of the 30-year 13 life,such as development rights. 14 The Oregon independent evaluator also noted that the terminal value was 15 included to account for the fact that the company would own the site at the end of the 16 project's useful life.(Oregon IE Report at 15.) 17 Q.Did the independent evaluators comment on the size of the terminal value benefit? 18 A.Yes.The Utah independent evaluator noted that the terminal value was "relatively low." 19 (Utah IE Report at 42.)Likewise,the Oregon independent evaluator found that the 20 "terminal value adders were fairly small."(Oregon IE Report at 17.) O Link,Supp-Reb -34 Rocky Mountain Power l Q.Mr.Phillips questions the terminal value calculations included in the company's 2 analysis,claiming this benefit is speculative.(Phillips Supp.Direct,page 52,line 3 8-11.)How do you respond? 4 A.I disagree.Notably,as described above,both of the independent evaluators confirmed 5 and validated the company's bid-selection and evaluation process,and proposed no 6 adjustment. 7 Q.Does Mr.Mullins challenge the company's terminal value used in the economic 8 modeling? 9 A.Yes.While Mr.Mullins does not challenge the magnitude of terminal values associated 10 with the new wind projects,and does "not necessarily disagree"that utility-owned 11 resources provide a terminal value that PPAs do not,he argues that,with regard to the 12 transmission project,the company needed to also consider the ongoing capitalO13maintenanceandinvestmentrequiredtoachievetheterminalvalueassumedinthe 14 economic analysis.(MullinsSupp.Direct,page 19,lines 9-14.) 15 PacifiCorp's analysis recognizes that the useful life of the transmission project 16 extends more than 30 years beyond the useful life of the new wind projects.Mr.Mullins 17 is correct that costs of the transmission project are not included beyond 2036 in the 18 system modeling,nor are they included beyond 2050 in the nominal revenue 19 requirement analyses.However,the company also did not include any incremental 20 benefits of the proposed transmission project beyond 2036 in the levelized view,or 21 beyond 2050 in the nominal view. O Link,Supp-Reb -35 Rocky Mountain Power l Q.Mr.Phillips argues that the Combined Projects are higher risk now,compared to 2 the original filing,because of the changes in the federal corporate tax rate,lower 3 load forecasts,and low natural-gas prices.(Phillips Supp.Direct,page 51,lines 3- 4 14.)How do you respond? 5 A.I disagree.It is true that each of the factors identified by Mr.Phillips decreased 6 customer benefits.But the decrease associated with these factors was more than offset 7 by other factors,such as lower installed capacity costs associated with the Wind 8 Projects.In total,when all of the changes are considered,the company's analysis shows 9 that risks have decreased and customer benefits have increased since the initial filing. 10 Q.Mr.Phillips claims that the company has not assessed the risk associated with ll wind variability.(Phillips Corrected Supp.Response,page 53,line 1-7.)Is this 12 true?O 13 A.No.PacifiCorp performed robust risk analysis of wind variability,including the 14 retention of a third-party expert to verify the wind-production estimates for every bid 15 selected to the initial shortlist in the 2017R RFP.Mr.Chad A.Teply also provided 16 testimony explaining that the company's existing wind projects in the Medicine Bow 17 area of Wyoming have out-performed pre-construction estimates.(Teply Rebuttal,page 18 16,line 9 to page 17,line 6.) O Link,Supp-Reb -36 Rocky Mountain Power l Q.Mr.Phillips claims that there is a risk that future qualifying facility ("QF") 2 development may cause curtailment of the Wind Projects,thereby reducing their 3 production.(Phillips Corrected Supp.Response,page 54,line 7-21.)Is this a 4 reasonable concern? 5 A.No.Mr.Phillips describes curtailment risk associated with a 320-MW QF project in 6 eastern Wyoming that has an executed interconnection agreement.This interconnection 7 agreement requires additional transmission upgrades,which includes all of Energy 8 Gateway West and Energy Gateway South,scheduled to occur in 2024.Mr.Phillips 9 then correctly explains that the company did not reserve interconnection capacity for 10 this QF project when performing its economic analysis of the Combined Projects. 11 PacifiCorp did not reserve any of the incremental interconnection capability 12 associated with the Aeolus-to-Bridger/Anticline transmission line for this particular 13 320-MW QF project because the project can only interconnect if the transmission 14 upgrades identified in this QF project's executed interconnection agreement are built, 15 including all of Energy Gateway West and Energy Gateway South.The upgrades are 16 required for this 320-MW QF project to proceed would increase interconnection 17 capacity in the region and would increase the transfer capability out of eastern 18 Wyoming.Consequently,if this QF project moves forward,it would mean that all of 19 Energy Gateway West and Energy Gateway South have been built,which would 20 mitigate,not increase,any potentialcurtailment of the proposed Wind Projects. O Link,Supp-Reb -37 Rocky MountainPower l Q.Mr.Phillips is concerned that the company has not thoroughly evaluated the 2 Combined Projects,(Phillips Supp.Direct,page 5,lines 21-22),and faults the 3 company for not conducting any capital cost over-run or load forecast sensitivities 4 in its updated analysis.(Phillips Supp.Direct,page 56,lines 3-8.)How do you 5 respond? 6 A.The company's economic analysis in this docket has been thorough and extensive.The 7 updated economic analysis summarized in my second supplemental direct testimony 8 alone includes 26 SO model simulations and 26 PaR simulations.Each PaR simulation 9 considers 50 different iterations of system performance with variations in stochastic 10 variables,which includes variations in load.Accounting for the stochastic system 11 simulations performed using PaR,the economic analysis summarized in my second 12 supplemental direct testimony represents over 1,300 simulations of PacifiCorp's 13 system over a 20-year forecast time frame.Through these studies,the company has 14 assessed how the net benefits of the Combined Projects are affected by the proposed 15 wind repowering project,solar resource opportunities,selection of alternative wind- 16 turbine equipment,alternative natural-gas price assumptions,alternative CO2 price 17 assumptions,and application of alternative assumptions for O&M cost and REC 18 revenues. 19 It is also important to recognize that the winning bids selected to the 2017R 20 RFP final shortlist are based on firm-pricing proposals through a competitive 21 solicitation process with oversight from two independent evaluators.The company also 22 provided evidence that its prior two large-scale transmission projects were 19 percent 23 and six percent under budget.(Vail Rebuttal,page 15,Table 1.) O Link,Supp-Reb -38 Rocky Mountain Power 1 Q.Are all of the risks identified by Mr.Phillips asymmetrical,i.e.,can the risks only 2 run against customer interests? 3 A.No.Variability of the factors described by Mr.Phillips can favor customers too.Project 4 performance can be better than expected,as Mr.Teply indicates has occurred.Capital 5 costs can be lower than expected,as Mr.Vail indicates has occurred.And ongoing 6 O&M costs can be less than expected,as I previously discussed. 7 Q.Mr.Eldred calculates the percentage increase in capital costs that would eliminate 8 net benefits for each price-policy scenario.(Edred Supp.Direct,page 16 to page 9 18,line 12).Are Mr.Edred's calculations correct? 10 A.No.Mr.Eldred's calculations have two errors.First,he includes run-rate O&M costs 11 for the Aeolus-to-Bridger/Anticline transmission line when calculating the percentage 12 increase in capital costs that would result in a break-even PVRR(d).Second,he does 13 not account for the transmission revenue credits from the Aeolus-to-Bridger/Anticline 14 transmission line,which would increase if capital costs increase.These two errors 15 understate the increase in capital costs that would result in a break-even PVRR(d)by 16 approximately$9 million in each price-policy scenario.Consequently,in the medium 17 natural gas,medium CO2 price-policy scenario capital costs would need to increase by 18 $205 million (approximately9.1 percent)to eliminate net benefits,not the $196 million 19 (approximately 8.7 percent)figure calculated by Mr.Eldred. 20 Q.Is there also a risk that natural-gas prices will be higher than expected? 21 A.Yes.In my direct testimony,I noted that the low natural-gas price forecast assumed 22 stagnant liquefied natural gas ("LNG")exports.(Link Direct,page 32,line 13 to page 23 33,line 2.)According to the U.S.Energy InformationAdministration's Annual Energy O Link,Supp-Reb -39 Rocky Mountain Power l Outlook2018 ("AEO 2018"),published on February 6,2018,the United States is now 2 a net exporter of natural gas and its reference case shows increased LNG exports in the 3 coming years as additional terminals come into service.The increased exports will 4 likely put pressure on future natural gas prices,meaning that over the next 32 years 5 (i.e.,until 2050),it is unlikelythat natural gas prices will remain as low as the low case 6 used here. 7 Q.Mr.Yankel claims that the low natural-gas price scenarios are the most likely to 8 occur.(Yankel Supp.Direct,page 6,line 18 to page 9,line 14.)Do you agree? 9 A.No.PacifiCorp's medium natural-gas price scenarios are the most likelyto occur.These 10 forecasts are based on observed market forward prices and base-case projections from 11 third-party experts.Moreover,for the reasons discussed above,pressure on future 12 natural gas prices may actually be higher than what is assumed in the low natural-gasO13pricescenario. 14 Q.Is there a price risk associated with long-term PPA contracts? 15 A.Yes.Recently in the context of avoided-cost pricing,the Commission reduced QF 16 contract terms to two years.The Commission ruled:"Based upon our record,we find 17 that 20-year contracts exacerbate overestimations to a point that avoided cost rates over 18 the long-term period are unreasonable and inconsistent with the public interest.We find 19 shorter contracts reasonable and consistent with federal and state law for multiple 20 reasons.First,shorter contracts have the potential to benefit both the QF and the 21 ratepayer.By adjusting avoided cost rates more frequently,avoided costs become a 22 truer reflection of the actual costs avoided by the utility and allow QFs and ratepayers 23 to benefit from normal fluctuations in the market."In the Matter ofIdaho Power Co.'s O Link,Supp-Reb -40 Rocky Mountain Power l Petition to Modify Terms and Conditions ofPURE4 Purchase Agreements,et al.,Case 2 Nos.IPC-E-15-01,AVU-E-15-01,PAC-E-15-03,Order No.33357 at 23 (Aug.20, 3 2015). 4 The same is true here-there is no bias in the medium natural-gas price forecast 5 and therefore,actual future natural-gas prices are as likely to be higher as they are 6 lower.However,all of the parties'testimony has been very asymmetrical only focusing 7 on the risk associated with the company's proposal and ignoring the fact that seven of 8 the nine scenarios demonstrate that customers will benefit from the Combined Projects. 9 Q.Do Mr.Phillips and Mr.Yankel continue to rely on the low natural-gas price 10 scenario? 11 A.Yes.Mr.Phillips reiterated that the low natural-gas price forecast is the "status quo" 12 and appears to continue to rely heavily on the low natural-gas price scenarios for his 13 analysis.(Phillips Supp.Direct,page 20,lines 17-18.)Mr.Yankel stated that "the two 14 scenarios where customers are worse off,have the most likelihood of occurring." 15 (Yankel Supp.Direct page 2,lines 3-4.)This is despite the lack of bias in the company's 16 price forecasts,as noted above,and assumes that current "price-floor"conditions will 17 persist for the next 32 years. 18 Q.Mr.Mullins claims that PacifiCorp's economic analysis has not taken into 19 consideration declining market prices.(MullinsSupp.Direct,page 22,lines 2-8.)Is 20 this true? 21 A.No.Mr.Mullins correctly notes that PacifiCorp's December 2017 official forward price 22 curve ("OFPC")reflects 72 months of market forwards followed by 12 months of a 23 forwards-fundamental blend that transitions to a pure fundamentals-based forecast in O Link,Supp-Reb -41 Rocky Mountain Power l month 85.Consequently,the first seven years of the December 2017 OFPC reflects or 2 is influenced by observed market forwards as of December 29,2017.This was the most 3 current OFPC available at the time the company was finalizing its 2017R RFP bid 4 evaluationand selection process and is representative of current market conditions. 5 Q.Mr.Mullins goes on to explain that the company relies on a third-partyforecast 6 from November 21,2017,and is concerned that the December OFPC does not 7 consider the effects of tax reform.(MullinsSupp.Direct,page 23,lines 8-17.)How 8 do you respond? 9 A.As noted above,the OFPC reflects or is influenced by observed market prices through 10 the first seven years (through 2024).The December 2017 OFPC that the company used 11 in its medium price-policy scenarios reflects market forwards as of December 29,2017, 12 which is after President Trump signed the tax reform bill.This means that through the 13 first seven years of the December 2017 OFPC,observed prices account for tax reform. 14 Moreover,I have reviewed observed forward prices,which are updated each trading 15 day,throughout December 2017,and there is no indication that would suggest there 16 was any material change in forward prices that coincide with the timing of when tax 17 reform legislation was passed by Congress and subsequently signed by President 18 Trump.Consequently,I would not expect a material change in forecasted prices beyond 19 the first seven years of the December 2017 OFPC when prices are based on a third- 20 party forecast. O Link,Supp-Reb -42 Rocky Mountain Power 1 Q.Mr.Mullins claims that forward market prices for calendar year 2022 in the 2 December 2017 OFPC declined by approximately 35 percent relative to prices in 3 the June 2017 OFPC,which were used in PacifiCorp's original economic analysis 4 (MullinsSupp.Direct,page 24,lines 1-3.)Is this accurate? 5 A.No.The average Henry Hub natural gas price for calendar-year 2022 from the June 6 2017 OFPC is $2.92/MMBtu.The average Henry Hub natural gas price for calendar- 7 year 2022 from the December 2017 OFPC,which reflects observed market prices as of 8 December 29,2017 (not January 2,2018 as claimed by Mr.Mullins),is $2.89/MMBtu. 9 By my calculations,this reflects a one-percent reduction in prices.This is a far cry from 10 the 35-percent reduction calculated by Mr.Mullins. 11 Q.Mr.Mullins asserts that year-on-year changes in prices from the December OFPC 12 over the 2023-2026 time frame is possibly due to use of a stale forecast.(Mullins 13 Supp.Direct,page 24,lines 8-10.)How do you respond? 14 A.The reason for an increase in prices over this time frame is not caused by the use of a 15 stale forecast.As described above,PacifiCorp used the most current OFPC available at 16 the time the company was finalizing its 2017R RFP bid evaluation and selection 17 process,and it is representative of current market conditions.The increase in prices 18 over the 2023-2026 time frame is consistent with an expectation of increased LNG 19 exports.As I noted earlier in my testimony,according to the AEO 2018,published on 20 February 6,2018,the United States is now a net exporter ofnatural gas and its reference 21 case shows increased LNG exports in the coming years as additional terminals come 22 into service.The increased exports is expected to put pressure on future natural gas 23 prices. O Link,Supp-Reb -43 Rocky Mountain Power 1 Q.Mr.Mullins estimates that had PacifiCorp used more current price forecasts, 2 present-value net benefits would be reduced by approximately $359 million. 3 (MullinsSupp.Direct,page 26,lines 2-5.)Is this analysis reasonable? 4 A.No.Mr.Mullins did not supply work papers with his testimony,so I was not able to 5 validate the accuracy of his calculations.Nonetheless,Mr.Mullins's description of his 6 calculations highlights methodological deficiencies.It is not clear from Mr.Mullins's 7 testimony whether he calculated his $359 million adjustment through 2036 or through 8 2050.Based on his statement that this adjustment would eliminate net benefits in the 9 low natural gas,zero CO2 price-policy scenario,I assume his calculation is based on 10 benefits calculated through 2036. 11 From what I can tell,Mr.Mullins calculated an implied market heat rate off of 12 the December 2017 OFPC by dividing Palo Verde electricity prices by Henry Hub 13 natural gas prices.He produced an alternative electricity price forecast by multiplying 14 this implied market heat rate derived from the December 2017 OFPC by a lower natural 15 gas price forecast.He then states that he multiplied the difference in Palo Verde prices 16 (the difference between prices in the December 2017 OFPC and his estimated lower 17 price forecast)by the volume of wind energy associatedwith the Combined Projects to 18 arrive at his estimate of reduced benefits. 19 Assuming I understand Mr.Mullins's methodology correctly,his oversimplified 20 approach grossly overstates the impact of a reduced natural gas price forecast on the 21 net benefits from the Combined Projects.This methodology assumes that the energy 22 benefits from the Combined Projects are valued at the Palo Verde market curve.This is 23 simply not the case.In the company's analysis,the energy benefits from the Combined O Link,Supp-Reb -44 Rocky Mountain Power l Projects are heavily driven by avoided system fuel costs,particularly coal costs, 2 through the 2027 time frame.Over this period,market prices have less of an impact on 3 system energy benefits than over later time frames (i.e.,beyond 2027).This is a result 4 of system constraints that limit access to markets,particularly prior to those years 5 where coal unit retirements are assumed. 6 For instance,in the company's economic analysis that uses medium natural gas, 7 medium CO2 price-policy assumptions,system energy benefits over the 2021-2027 8 time frame (before the Dave Johnston coal plant is assumed to retire)average 9 $23.40/MWh.In the low natural gas,zero CO2 price-policy scenario,system energy 10 benefits average $20.13/MWh (approximately 14 percent lower than in the medium 11 natural gas,medium CO2 price-policy scenario).Henry Hub natural gas prices over this 12 time frame average $3.54/MMBtu in the medium natural gas,medium CO2 price-policyO13scenarioand$2.69/MMBtu in the low natural gas,zero CO2 price-policy scenario 14 (approximately 24 percent lower than in the medium natural gas,medium CO2 pfÎCC- 15 policy scenario).These results demonstrate that a change in natural gas price 16 assumptions does not proportionately impact system energy benefits from the 17 Combined Projects. l 8 Clearly,it is inappropriate to assume that a dollar-for-dollar change in price 19 curve assumptions equatesto a dollar-for-dollar change in system energy benefits.This 20 approach,which was used by Mr.Mullins,will grossly overstate the impact of a change 21 in natural gas price assumptions on net benefits from the Combined Projects,and 22 consequently,Mr.Mullins's estimate and associated conclusions are not valid. O Link,Supp-Reb -45 Rocky Mountain Power l Q.Did Mr.Mullins present all of the natural gas price forecasts he received from the 2 company throughdiscovery in Confidential Figure 3 of his supplemental rebuttal 3 testimony? 4 A.No.PacifiCorp also provided an update to the November 2017 natural gas price 5 forecast that was used in the company's December 2017 OFPC.This updated forecast 6 was issued on February 18,2018 and is actually slightly higher than the November 7 2017 forecast used in the company's economic analysis.However,Mr.Mullins omitted 8 this forecast in Confidential Figure 3 of his supplemental rebuttal testimony. 9 Q.Mr.Mullins restates his opinion that market prices have consistently been lower 10 than utilities'long-term forecasts.(Mullins Supp.Direct,page 26,lines 15-17.) 11 How do you respond? 12 A.Assuming Mr.Mullins's reference to the term "market prices"is synonymous with 13 "actual prices,"I agree that market forwards exceededactual prices over the 2010-2015 14 time frame,when structural expansion in natural gas supply outpaced market 15 expectations.However,beyond 2015,the ratio of forward market prices to actuals has 16 come down significantly and now hovers near one.Simply stated,market expectations 17 are catching up with supply capabilities.This is reasonable given that technological 18 progress and efficiencies in natural gas production continue to increase,but at a slower 19 rate.As such,historical variances between forward market prices and spot prices are 20 not good indicators of future price developments. O Link,Supp-Reb -46 Rocky Mountain Power l SOLAR RESOURCE SENSITIVITY 2 Q.Please summarize the solar resource sensitivity provided in your previous 3 testimony. 4 A.My second supplemental direct testimony provided robust modeling results through 5 2036 using the SO model and PaR based on preliminary bid analysis from the 2017S 6 RFP.(Link Second Supp.Direct,page 21,line 4 to page 25,line 3.)Those modeling 7 results supported two important conclusions. 8 First,solar PPAs providedfewer benefits than the Combined Projects under the 9 medium natural gas,medium CO2 price-policy scenario,and slightly fewer benefits 10 under the low natural gas,zero CO2 price-policy scenario using PaR,and slightly more 11 benefits under the low natural gas,zero CO2 price-policy scenario using the SO model. 12 In other words,under the medium natural gas,medium CO2 price-policy scenario,the 13 Combined Projects are superior,and under the low natural gas,zero CO2 price-policy 14 scenario the Combined Projects are roughly equal to the solar PPAs. 15 Second,when analyzed together,the Combined Projects and solar PPAs 16 produced greater customer benefits under both the medium natural gas,medium CO2 17 price-policy scenario and low natural gas,zero CO2 price-policy scenario relative to 18 scenarios where either the Combined Projects or solar PPAs are procured on their own. 19 Significantly,none of wind or solar bids were hard coded into the model,and 20 when solar bids were selected in the models,they did not displace the wind bids.These 21 conclusions indicated that it is not a question of whether the company should pursue 22 the Combined Project or the solar PPAs,but rather a question of whether the company 23 should pursue the Combined Projects and the solar PPAs. O Link,Supp-Reb -47 Rocky Mountain Power l Q.Did the company provide the solar sensitivity to the independent evaluators who 2 monitored the 2017R RFP? 3 A.Yes.The Oregon independent evaluator noted in his report:"In all cases the 4 combination of solar and shortlisted [wind]resources provided more net benefits." 5 (Oregon IE Report at 36.)Although the Utah independent evaluator did not specifically 6 comment on the solar sensitivity,he did not challenge it in his final report.(see Utah 7 IE Report at 61.) 8 Q.Messrs.Eldred,Mullins,and Phillips argue that solar PPAs represent a superior 9 resource option for customers and therefore the Combined Projects are contrary 10 to the public interest.(Eldred Supp.Direct,page 11,line 6 to page 12 line 5; 11 Mullins Supp.Direct,page 16,line 8-16;Phillips Supp.Direct,page 30,lines 2-6.) 12 Do you agree? 13 A.No.PacifiCorp has now completed its bid evaluation and selection process for the 14 2017S RFP,and the complete analysis and results confirm the company's earlier 15 assessment that solar-PPA bids do not displace the economic benefits of the Combined 16 Project.While the base economic analyses of solar bids show that there are potential 17 customer benefits associated with a 1,320 MW portfolio of solar PPAs from the 2017S 18 RFP,subsequent sensitivity analyses show a risk,unique to solar resource 19 opportunities,that the projected benefits for the solar PPAs in the base economic 20 analysis are overstated,as I will discuss below. 21 In addition,driven by uncertainties regarding tariff and tax reforms,current 22 solar resource pricing likely reflects a risk premium,and solar project costs are 23 expected to decline.Because the 30-percent ITC is available for solar resources that O Link,Supp-Reb -48 Rocky Mountain Power l come online by 2021,PacifiCorp expects that solar pricing received in late 2019 for 2 projects that could come online in 2021 will be lower than pricing received in the 2017S 3 RFP and would avoid the current risk premium associated with the tariff and tax reform 4 uncertainties.Thus,PacifiCorp does not need to act now and has decided not to select 5 any of the 2017S RFP bids to the final shortlist. 6 PacifiCorp will continue to assess potential economic benefits from solar 7 resource opportunities in the 2019 IRP and through bi-lateral discussions with 8 developers,including a thorough evaluation of hourly price-profile and capacity- 9 contribution risks (discussed below)with full stakeholder engagement and a more 10 orderly assessment of the potential customer benefits of solar generation.Should 11 subsequent analysis in the 2019 IRP demonstrate that solar resource opportunities 12 provide economic benefits for customers,or if there is an opportunity to mitigate 13 evaluationrisks,there will be sufficient time to initiate a new competitive solicitation 14 process or to pursue bi-lateral contracts for projects capable of achieving commercial 15 operation by the end of 2021 that can qualify for the 30-percent ITC.This potential 16 solicitation could consider storage bids as a means to mitigate valuation risks and allow 17 sufficient time for participants to be further along in the transmission interconnection 18 process. 19 Q.Did PacifiCorp inform the independent evaluator overseeing the 2017S RFP of its 20 final shortlist results? 21 A.Yes.PacifiCorp summarized its 2017S RFP final shortlist bid evaluation and selection 22 analysis with London Economics International,LLC,the independent evaluator 23 retained by the company to monitor the 2017S RFP,on March 12,2018.This summary O Link,Supp-Reb -49 Rocky Mountain Power l is included in the final report of the independent evaluator for the 2017S RFP,which is 2 provided as Confidential Exhibit No.69 ("Solar IE Report"). 3 Q.Did the independent evaluator for the 2017S RFP agree with the company's 4 conclusions? 5 A.Yes.The independent evaluator concluded that the company's decision to not accept 6 any solar bids was not unreasonable and that PacifiCorp's concerns over conditions in 7 the solar market that reflected uncertainties over tax reform and tariffs were reasonable. 8 In addition,the independent evaluator concluded that the 2017S RFP was conducted in 9 a manner that was consistent with general procurement best practices,unbiased,that 10 the selection of the shortlisted resources was fair,and that the company's modeling 11 reflected industry best practices.Solar IE Report at 4-5. 12 Q.What additional sensitivity analyses did PacifiCorp perform in the 2017S RFP toO13betterassessthepotentialcustomerbenefitsandvaluationrisksassociatedwith 14 the solar resource bids? 15 A.PacifiCorp performed two additional sensitivities.First,the company refined how it 16 converts its forward market prices into hourly prices to more accurately reflect hourly 17 market-price variation in those hours when solar resources are producing energy. 18 Second,the company performed a capacity-contribution sensitivity to assess how 19 changes in the assumed ability of solar resource to meet peak load during periods when 20 there is an increased probability of loss-of-load events affect the overall customer 21 benefits. O Link,Supp-Reb -50 Rocky Mountain Power l Q.Please describe the hourly price-profile sensitivity developed to analyzebids in the 2 2017S RFP. 3 A.PacifiCorp uses hourly price scalars,which are applied to monthly on-peak and off- 4 peak prices in the forward price curve,to derive hourlymarket price profiles that vary 5 by month and day type (i.e.,weekdays,Saturdays,and Sundays/holidays).PacifiCorp 6 currently uses five years of hourly Powerdex price data to develop price scalars.The 7 company's review of the Powerdex data shows that the five-year price history is not 8 supported by a significantvolume of reported transactions (many hours have no market 9 pricing inputs)and that the resulting hourly price shapes do not align with prices 10 observed in operations that are being increasingly influenced by growth in solar l 1 resources across the region.Thus,for the hourly price-profile sensitivity,PacifiCorp 12 developed an alternative set of price scalars that are derived from one year of day-ahead 13 hourlyprices available from the California Independent System Operator ("CAISO"). 14 The figure below illustrates the differences between the Powerdex-derived 15 scalars and the CAISO-derivedscalars. O Link,Supp-Reb -51 Rocky Mountain Power 1 Figure 1-SR:HourlyPrice-Scenario Sensitivity Current Method (Powerdex)Sensitivity (CAISO Day Ahead) 10 g i 10 0 ------'*----0%0 -----------0% i 2 3 4 5 6 7 5 9101:12131415161718192021222324 1 2 3 4 5 6 7 8 91011121314151617ts192021222324 2 -2o21 pac,proni.2oas arre proni.---Pordono 2 cr -2o2:erw.noni.-2ose we.noni.---perdono 2 ce 3 The figure at top left shows representative average hourly price profiles as 4 derived from historical Powerdex data and used in the bid evaluationprocess of the 5 2017S RFP.The figure at top right shows representative average hourlyprice profiles 6 derived from historical CAISO data and used in this sensitivity.In both figures,the 7 hourlyprice profile is based on the average hourly prices from representative months 8 (January,April,July,and October)and shown alongside the average hourly energy 9 profile of bids included in a solar-PPA bid portfolio.The price profile used in the 10 sensitivity shows that when accounting for the growth of solar resources across the 11 region,prices are lower during those hours when the resources in the solar-PPA bid 12 portfolio are expected to generate electricity. 13 Q.Does the company intend to use the CAISO-derived scalars in future resource 14 analyses? 15 A.Yes.The company intends to use the refined scalars in the 2017 IRP Update,future 16 IRPs,and future regulatory filings. O Link,Supp-Reb -52 Rocky Mountain Power 1 Q.How do the refined hourly price scalars impact the benefits of the solar-PPA 2 resources? 3 A.The use of the CAISO-derived hourly price scalars decreasedthe benefits of the solar 4 PPAs.This outcome was observed regardless of whether these price scalars were 5 applied to studies evaluating solar-PPA bids with or without the Combined Projects. 6 When analyzed in isolation from the Combined Projects,20-year PaR studies (through 7 2036)show that applicationof the CAISO-derivedhourly price scalars decreasedsolar- 8 PPA benefits from $174 million to $108 million (a reduction of $66 million)based on 9 stochastic-mean PaR results and from $183 million to $114 million (a reduction of 10 $69 million)based on risk-adjusted PaR results in the medium natural gas,medium 11 CO2 price-policy scenario. 12 When analyzed under the low natural gas,zero CO2 price-policy scenario,the 13 CAISO-derived hourly price scalars decreased the benefit of the solar PPAs from 14 showing a $45 million net benefit to showing a $10 million net cost (a $55 million 15 reduction in benefits)based on stochastic-mean PaR results and from showing a 16 $48 million net benefit to showing a $10 million net cost (a $58 million reduction in 17 benefits)based on risk-adjusted PaR results. 18 The price-policy scenario assumptions used to analyze solar-PPA bids in the 19 2017S RFP are identical to those used to analyze the Combined Projects in my second 20 supplemental direct testimony,with the exception that the medium CO2 price 21 assumptions were correctly applied as a nominal cost instead of real costs in 2012 22 dollars. O Link,Supp-Reb -53 Rocky Mountain Power 1 Q.Are there any other issues to consider related to the price-profileused to evaluate 2 the solar-PPA bids? 3 A.Yes.The expected increase in solar generation,coupled with correlation among 4 expected solar resource generation profiles across the west,has had a significant impact 5 on hourly prices and will continued to do so as solar development increases.S&P 6 Global Market Intelligencetracks power-plant capacity,and reports that solar capacity 7 in the Western Electricity Coordinating Council ("WECC")region,which represents 8 capacity that is online or announced to go onlinehaving obtained regulatory approvals, 9 will grow from 16.8 gigawatts ("GW")in 2017 to 29.8 GW by 2023 (growth of 10 approximately 77 percent over six years).Similarly,the AEO 2018 Reference Case 11 trends closely with the S&P Global Market Intelligence data,and shows continued 12 growth of solar capacity in the WECC,which reaches 46.8 GW by 2050.By the end of 13 a 25-year solar PPA (2045),the AEO 2018 Reference Case predicts that solar capacity 14 in the WECC region will grow to 41.3 GW,which is 2.5 times the amount of solar 15 capacity reported for 2017. 16 The rapid increase in solar capacity across the region over the past five years 17 has significantly impacted hourly market prices,and continued growth in new solar 18 capacity could further affect the market value of solar energy beyond what has been 19 analyzed in the price-profile sensitivitydescribed above.Moreover,proxy solar profiles 20 from the National Renewable Energy Laboratory ("NREL")show a high degree of 21 correlation among potential solar sites across the WECC region,indicating that the 22 potential impacts on hourly price profiles are likely regardless of where new solar is O Link,Supp-Reb -54 Rocky Mountain Power l added.The figure below illustrates the expected growth in solar generation and the 2 correlated generation profiles throughout the region. 3 Figure 2-SR:Growth in Solar Generation and Correlation of Generation Profiles Solar Capacity the WECC Region Representative Solar Profiles (NREL Data) 20 t j 30% 15 ,#O 20% 10 5 10% 0 0% an on a w h o m to on rw an 00 w w N O 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 248888888888888888rwryNontormNNNNNNNN«¾--CentralCA(Sacrarnento)--SouthernCA(PalmSprings) 5&PGlobalMarketintelligence ---AEO 2018Reference Case --Arizona (Scottsdale)SouthCentral OR 4 ¯'°' 5 Q.Did the independent evaluator for the 2017S RFP comment on the hourly price 6 sensitivity? 7 A.Yes.The independent evaluator concluded that the "alternative price profile was a 8 reasonable way to examine potential downside risks to customers of committing to 9 solar resources."Solar IE Report at 25. 10 Q.Mr.Yankel claims that the company's analysis should account for the impact 11 increased wind generation will have on market prices similar to the adjustment 12 applied to solar.(Yankel Supp.Direct,page 11 lines 3-19.)Do you agree? 13 A.No.The company's price curves,which includes monthly on-peak and off-peakpower 14 prices,account for the growth of renewable resources throughout the WECC over time. 15 For instance,across the WECC region,the fundamentals-based component of 16 PacifiCorp's December 2017 OFPC includes 9.4 GW of new solar resources and 5.4 17 GW of new wind resources by 2025.New solar capacity increasesto 18.0 GW and new O Link,Supp-Reb -55 Rocky Mountain Power l wind capacity grows to 12.3 GW by 2035.This OFPC does in fact account for the 2 impact of increased solar and wind resources over time. 3 The price-profile sensitivity does not assess how new solar resources affect 4 overall monthly prices.As discussedabove,the impact of new solar and wind resources 5 is already captured in the company's OFPC.Rather,the price-profile sensitivity is 6 intended to evaluate how changes to the hourly price profile within a given 24-hour 7 period is impacted if more accurate market data from the CAISO is used to more closely 8 align these hourly price profiles with those observed in operations.To be clear,the price 9 profiles developed for this sensitivity are based on market data and are not based on 10 assumed levels of solar penetration.It is not surprising that the hourlyshape of these ll market data show the effects increased solar generation,which only produce energy 12 during day-light hours.In contrast,wind resources produce energy across all hours of 13 the day.Therefore the energy output from wind resources is less likely to materially 14 impact the hourly shape of market prices. 15 Q.Please describe the capacity-contributionsensitivity used in the 2017S RFP bid 16 evaluation and selection process. 17 A.The capacity-contribution sensitivity is designed to assess the risks associated with 18 overstating the capacity contribution of solar resources when evaluatingthe potential 19 customer benefits of solar-PPA bids.The capacity contribution of solar resources, 20 represented as a percentage of resource capacity,is a measure of the ability for these 21 resources to reliably meet demand.The company's base economic analysis used to 22 evaluate bids submitted into the 2017S RFP and used to support the solar sensitivity 23 studies in my supplemental direct and second supplemental direct testimony applied Link,Supp-Reb -56 Rocky Mountain Power l the capacity-contributionvalues for solar resources developed for the 2017 IRP (59.7 2 percent for the solar PPAs located in Utah),and therefore,the base economic analysis 3 assumes that the 1,320 MW of solar-PPA capacity included in the 2017S RFP bid 4 portfolio can displace the need for approximately 788 MW of system capacity 5 (59.7 percent multiplied by the 1,320 MW of solar-PPA capacity). 6 As more highlycorrelated solar generation is added to the system,the energy 7 output from these resources is more likely to shift the timing of potential loss-of-load 8 events to evening hours when solar irradiance is low and generation levels are greatly 9 reduced or zero.Consequently,solar capacity-contributionvalues are highlysensitive 10 to increasing solar penetration levels.The figure below illustrates study results 11 concluding that additional solar generation reduces the capacity contribution of solar 12 resources. 13 Figure 3-SR:Capacity Contribution Compared to Penetration 80 -NV Power:Perez et al (2008) O Case Study:Mills and Wiser (2012) 70 -CA Case Study:Jones (2012) 60 APS -Tracking:R.W.Beck (2009) --APS -Fixed:R.W.Beck (2009) 50 ---Westconnect:GE Energy (2010) 40 -Toronto:Pelland and Abboud (2008) -PGE:Perez et al (2008) 30 20 10 0 0 5 10 15 20 25 30 PV Penetration (%annual energy) 14 15 Source:Mills,Andrew,and Ryan Wiser.2012."An Evaluation of Solar Valuation Methods Used in Utility Planning and 16 Procurement Processes."LBNL-5933E,Berkeley,CA:Emest Orlando Lawrence Berkeley National Laboratory. Link,Supp-Reb -57 Rocky Mountain Power l For PacifiCorp,the addition of 1,320 MW of solar capacity would more than double 2 the amount of solar resources on its system.The capacity-contribution sensitivity 3 evaluates the economic impact of halving the capacity-contribution value from 59.7 4 percent to 29.9 percent when applying medium natural gas,medium CO2 and low 5 natural gas,zero CO2 price-policy assumptions.Considering that the company will 6 begin using the hourlyprice profiles derived from day-ahead CAISO data in the 2017 7 IRP Update,future IRPs,and future regulatory filings,the capacity-contribution 8 sensitivity also includes the CAISO-derived hourly price profile. 9 Q.What were the results of this capacity-contributionsensitivity used to evaluate 10 bids in the 2017S RFP? 11 A.With the capacity-contribution assumption reduced from 59.7 percent down to 12 29.9 percent,the amount of system capacity that the 1,320 MW of solar resource 13 capacity can displace is reduced from 788 MW to 394 MW.This reduces the resource- 14 deferral value of the solar-PPA resources,which in turn reduces the net benefits of the 15 solar-PPA bids. 16 The combined effect of the hourly price-profile and capacity-contribution 17 assumptions,when solar-PPA bids are analyzed in isolation of the Combined Projects 18 over a 20-year time frame in PaR,is to decrease the solar-PPA benefits from 19 $174 million to $69 million (a reduction of $105 million in benefits)based on 20 stochastic-mean PaR results,and from $183 million to $73 million (a reduction of 21 $110 million in benefits)based on risk-adjusted PaR results in the medium natural gas, 22 medium CO2 price-policy scenario. O Link,Supp-Reb -58 Rocky Mountain Power l When analyzed under the low natural gas,zero CO2 price-policy scenario,the 2 combined effect of the hourly price-profile and capacity-contributionassumptions is to 3 decrease the benefit of the solar PPAs from showing a $45 million net benefit to 4 showing a $56 million net cost (a $101 million reduction in benefits)based on 5 stochastic-mean PaR results,and from showing a $48 million net benefit to showing a 6 $58 million net cost (a $106 million reduction in benefits)based on risk-adjusted PaR 7 results. 8 Again,the price-policy scenario assumptions used to analyze solar-PPA bids in 9 the 2017S RFP are identical to those used to analyze the Combined Projects in my 10 second supplemental direct testimony,with the exception that the medium CO2 price 11 assumptions were correctly applied as a nominal cost instead of real costs in 2012 12 dollars.O 13 Q.When assessing the impact of the hourly price-profile sensitivity for the 2017S 14 RFP,did the company consider how the CAISO-derived hourly price scalars 15 might affect the economic analysis of the Combined Projects? 16 A.Yes.The table below summarizes how the CAISO-derived hourly price-scalar 17 assumptions impact the Combined Projects and,separately,how these assumptions 18 impact the 1,320 MW bid portfolio that includes solar PPAs without the Combined 19 Projects when applying medium natural gas,medium CO2 price-policy assumptions. O Link,Supp-Reb -59 Rocky Mountain Power l Table 1-SR:Solar-OnlyCompared to Combined Projects 2 Hourly-PriceSensitivity System Modeling Results 3 (Medium Gas,Medium CO2) Stochastic-Risk-Adjusted Mean PaR PaR PVRR(d) PVRR(d)(Benefit)/Cost (Benefit)/Cost $million Combined Projects Benchmark Analysis (Second Supplemental Direct3 $(357)$(386) Hourlv Price-Profile Sensitivity &Nominal CO2 $(328)$(343) Depresse in Net Renefits $29 $43 2017S Solar-PPA Bid Portfolio Benchmark Analysis (Current HourlyScalars)$(237)$(248) Hourlv Price-Profile Sensitivity $(160)$(168) Decrease in Net Benefits $77 $80 4 This analysis shows that the new hourly prices-profile decreases the customerO5benefitsoftheCombinedProjectsonastand-alone basis and decreasesthe customer 6 benefits of the solar PPAs on a stand-alone basis.But,importantly,the reduction in net 7 benefits associatedwith the hourly-priceprofile sensitivityis between 1.9 and 2.7 times 8 greater for the solar PPAs than it is for the Combined Projects when applying medium 9 gas,medium CO2 price-policy assumptions.The disproportionate impact is consistent 10 with the fact that solar generation profiles are more highlycorrelated with the impact 11 solar resources are having on hourly price profiles relative to wind.While both types 12 of technologies are faced with the same reduction in the market value of energy during 13 the middle of the day,the wind generation produces energy during the early morning 14 and late evening hours,when the market value of energy is higher. O Link,Supp-Reb -60 Rocky Mountain Power 1 Q.Did you conduct this same analysis for the low gas,zero CO2 price-policy 2 scenario? 3 A.Yes.The table below summarizes how the CAISO-derived hourly price-scalar 4 assumptions impact the Combined Projects and the 1,320 MW solar-PPA bid portfolio 5 when applying low gas,zero CO2 price-policy assumptions. 6 Table 2-SR:Solar-OnlyCompared to Combined Projects 7 Hourly-Price Sensitivity System Modeling Results 8 (Low Gas,Zero CO2) Stochastic-Risk-Adjusted Mean PaR PaR PVRR(d) PVRR(d)(Benefit)/Cost (Benefit)/Cost $million Combined Projects Benchmark Analysis (Second Supplemental Direct)($150)($156) O Hourlv Price-Profile Sensitivity ($125)($130) Decrease in Net Renefits $25 $26 2017S Solar-PPA Bid Portfolio Benchmark Analysis (Current HourlyScalars)($125)($131) Hourlv Price-Profile Sensitivity ($69)($72) Decrease in Net Benefits $56 $59 9 Similar to the medium gas,medium CO2 price-policy scenario,the results show 10 that the net benefits associated with both the Combined Projects and the solar PPAs 11 decreased,but,again,the reduction in net benefits associated with the hourly-price 12 profile sensitivity is approximately 2.2 to 2.3 times greater for the solar PPAs than it is 13 for the Combined Projects when applying low gas,zero CO2 price-policy assumptions. 14 Q.What conclusions can you draw from these results? 15 A.The solar PPAs are more sensitive to the refined hourly price-profile and therefore 16 present a greater risk that the customer benefits of the solar PPAs are overstated relative Link,Supp-Reb -61 Rocky Mountain Power l to the Combined Projects. 2 Q.Did the company apply the capacity-contributionsensitivity to the Combined 3 Projects? 4 A.No.Unlike solar resources,wind resources are expected to generate in all hours of the 5 day,and thus the energy output from wind resources are not likely to shift the timing 6 of potentialloss-of-load events to hours when the wind is not generating.Consequently, 7 the capacity-contribution value for wind resources (15.8 percent for east wind as 8 reported in the 2017 IRP)is less likely to be materially impacted with increasing 9 penetration of either new wind or solar resources. 10 Q.How do the economics of the Combined Projects with CAISO-derived hourly 11 price scalars compare to the economics of the solar-PPA bid portfolio that reflects 12 the combined effects of the alternative hourly-price and capacity-contribution 13 assumptions? 14 A.The table below summarizes how these assumptions impact the Combined Projects and 15 the 1,320 MW solar-PPA bid portfolio when applying medium natural gas,medium 16 CO2 price-policy assumptions. Link,Supp-Reb -62 Rocky Mountain Power l Table 3-SR:Solar-OnlyCompared to Combined Projects 2 Capacity-ContributionSensitivity System Modeling Results 3 (Medium Gas,Medium CO2) Stochastic-Risk-Adjusted Mean PaR PaR PVRR(d) PVRR(d)(Benefit)/Cost (Benefit)/Cost $million Combined Projects Benchmark Analvsis (Second Supplemental Direct)($357)($386) Hourlv Price-Profile Sensitivity &Nominal CO2 ($328)($343) Decrease in Net Renefits $29 $43 2017S Solar-PPA Bid Portfolio Benchmark Analysis (Current HourlyScalars/Cap Cont ì ($237)($248) Hourlv Price-Profile/Cap Cont.Sensitivity ($93)($97) Decrease in Net Renefits $144 $151 4 As set forth above,the combined effect of the hourlyprice-profile and capacity-O 5 contribution assumptions is to reduce the net benefits of the solar-PPA bids by between 6 $144 million and $151 million in the medium gas,medium CO2 price-policy scenario, 7 which is approximately 3.5 to 5.0 times greater than the impact of the hourly price- 8 profile on the Combined Projects. 9 Q.What do these sensitivities show when applying low gas,zero CO2 price-policy 10 assumptions? 11 A.The table below summarizes how hourly price-scalar and capacity-contribution 12 sensitivity assumptions affect the Combined Projects and the 1,320 MW solar-PPA bid 13 portfolio when applying low natural gas,zero CO2 price-policy assumptions. O Link,Supp-Reb -63 Rocky Mountain Power l Table 4-SR:Solar-OnlyCompared to Combined Projects 2 Capacity-ContributionSensitivity System Modeling Results 3 (Low Gas,Zero CO2) Stochastic-Risk-Adjusted Mean PaR PaR PVRR(d) PVRR(d)(Benefit)/Cost (Benefit)/Cost $million Combined Projects Benchmark Analvsis (Second Supplemental Direct)($150)($156) Hourlv Price-Profile Sensitivity ($125)($130) Decrease in Net Renefits $25 $26 2017S Solar-PPA Bid Portfolio Benchmark Analysis (Current Hourly Scalars/Cap Contì ($125)($131) Hourlv Price-Profile/Cap Cont.Sensitivity ($8)($8) Decrease in Net Renefits $117 $123 4 The combined effect of the hourly price-profile and capacity-contributionO5assumptionsistoreducethenetbenefitsofthesolar-PPA bids by between $117 million 6 and $123 million in the low natural gas,zero CO2 price-policy scenario,which is 7 approximately 4.7 times greater than the impact of the hourly price-profile on the 8 Combined Projects. 9 Q.What conclusions can you draw from these sensitivities? 10 A.The sensitivities set forth above demonstrate that there is risk that the customer benefits 11 from the solar PPAs are overstated because the assumed capacity-contributionvalue 12 and associated resource-deferral benefits are likely to be lower than what is assumed in 13 the base analysis.Importantly,this same risk does not apply to the Combined Projects. 14 In fact,the Combined Projects will bring additionaltransmission capacity and a diverse 15 resource that is uncorrelated to solar production (i.e.,wind production occurs in all O Link,Supp-Reb -64 Rocky Mountain Power l hours,not just daylight hours).Moreover,solar-resource opportunities do not displace 2 the benefits of the Combined Projects,and similarly,the Combined Projects do not 3 displace the potential benefits of solar-resource opportunities.Solar resources are best 4 viewed as an incremental opportunity to the Combined Projects,not as an alternative. 5 Q.Did the company perform an annual revenue requirement analysis to assess how 6 these risks affect the Combined Projects and the 1,320 MW solar-PPA bid 7 portfolio? 8 A.Yes.Figure 4-SR provides these annual revenue requirement results when applying 9 medium natural gas,medium CO2 price-policy assumptions.The figure also shows the 10 cumulative PVRR,where the PVRR for each year representsthe present value of annual 11 revenue requirement from that year and all prior years. O 12 Figure 4-SR:Annual Revenue Requirement Results increase/(Decrease)in Nom.Rev.Req.Increase/(Decrease)Cumulative PVRR 14 As Figure 4-SR illustrates,the PVRR(d)benefits of the Combined Projects, 15 reflecting an hourly price profile derived from the CAISO day-ahead data,when 16 calculated from nominal revenue requirement results is $127 million.The PVRR(d) 17 benefits of the solar PPAs,reflecting an hourlyprice profile derived from the CAISO 18 day-ahead data and reflecting a 29.9 percent capacity-contribution value,is Link,Supp-Reb -65 Rocky Mountain Power 1 $149 million.The Combined Projects have a higher net cost relative to the solar PPAs 2 for two years;however,with PTCs,the net costs drop below the solar-PPA bids 3 beginningyear three and the Combined Projects begin producing net benefits by 2025. 4 The solar PPAs do not begin producing net benefits until 2029.Beyond the first few 5 years,the cumulative PVRR of the Combined Projects is favorable relativeto the solar- 6 PPA bids through 2035.Over the long term,more speculative benefits that reflect no 7 further deterioration to hourly price profiles or capacity-contribution value drive the 8 cumulative PVRR benefits of the solar-PPA bids below wind.In 2050,the terminal 9 value assumed for owned assets (applicable to 1,011 MW of the new wind)improves 10 the cumulative PVRR for the Combined Projects. 11 Q.In addition to the risk associated with hourly prices and capacity contribution,are 12 there any other risks associated with obtaining solar PPAs now as a result of the 13 2017S RFP? 14 A.Yes.As shown in Figure 5-SR,solar resource costs have been steadily declining and 15 the trend is expected to continue. O Link,Supp-Reb -66 Rocky Mountain Power l Figure 5-SR:Solar Resource Costs 2017 USD per Watt DC Utility-ScalePV Utility-ScalePV.Fixed Tilt (100 MW)One-Axis Tracker (100 MW) caSoft Costs -Others (PII.Land Acquisition.Sales Tex.Overhead,and Net Profit)e Soft Costs -Install LaboraHardwareBOS-Structural and Electncel Componentsainverter 2 °Modu' 3 4 5 Source:Fu,Ran,David Feldman,Robert Margolis Mike Woodhouse,and Kristen Ardani."U.S.Solar Photovoltaic 6 System Cost Benchmark:Q1 2017."NationalRenewable Energy Laboratory.September 2017. 7 As illustrated above,solar resource costs have fallen over time with a 8 77-percent reduction in utility-scale solar photovoltaic system costs for fixed-tilt 9 systems over the 2010-2017 time frame and an 80-percent reduction for single-axis 10 tracker systems.Stemming from increases in module costs due to a global shortage of 11 Tier 1 module supply,tax-reform uncertainty,and tariff uncertainty,solar costs 12 increased for the first time in the third quarter of2017 since the Solar Energy Industry 13 Association and GTM Research began publishing market cost reports in 2010; 14 however,cost reductions are expected to continue over the long term.By the second 15 half of 2019,tariff and tax risks,including implications on tax-equity markets,are 16 expected to have been mitigated and module costs are expected to fall to as low as O Link,Supp-Reb -67 Rocky Mountain Power l 30 cents-per-watt on a direct-current basis by 20192.Additional reductions to the cost 2 of inverters,tracking structures,and other balance-of-system components are expected 3 to further reduce total-system costs in 2019 and 2020. 4 Q.How do these changes in solar resource costs impact the company's assessment of 5 the 2017S RFP resources? 6 A.When considering the relatively long lead time between contract execution of 2017S 7 RFP solar resource bids with commercial operation dates in late 2020,and the fact that 8 the 30-percent ITC is available for solar projects coming online as late as 2021,current 9 pricing for solar resources likely reflects a risk premium,by both bidders and their tax- 10 equity investors,related to tariff and tax-reform uncertainties.Solar pricing received in 11 late 2019 for projects that could come online in 2021 and qualify for the 30-percent 12 ITC should reflect expected cost reductions and avoid the current risk premium 13 associated with tariff and tax-reform uncertainties. 14 Q.Mr.Phillips claims that the company was misleading because it did not present 15 the nominal revenue requirement results through 2050 for the solar sensitivity 16 presented in the second supplemental direct testimony.(Phillips Supp.Direct, 17 page 19,lines 4-22.)How do you respond? 18 A.I disagree that the company's testimony and analysis was misleading.As I described in 19 my second supplemental testimony,the company's system-modeling analysis 20 demonstrated that the combined benefits of the solar resources and the Combined 2 "Why Solar Is on a Path to Dominance,"Greentech Media,Yuri Horwitz,February 15,2018 (available at https://www.areentechmedia.com/articles/read/solar-is-going-to-win-bielv). Link,Supp-Reb -68 Rocky Mountain Power l Projects were higher than the individual benefits of each resource option alone.Mr. 2 Phillips does not dispute that conclusion. 3 As I discussed earlier,the system-modeling results provide a view of the 4 economic analysis that is consistent with the planning period and approach used to 5 identify a least-cost,least-risk preferred portfolio in the IRP.While the nominal 6 revenue-requirement analysis provides a sense of how the Combined Projects and solar 7 resources might impact customer rates over time,longer-term results in this analysis 8 are increasingly difficult to project.The company focused on the system-modeling 9 results when performing its solar resource sensitivities because these studies are more 10 suitable for comparing different resource portfolios,consistent with how resource ll portfolios are evaluated in the IRP. 12 Q.Messrs.Mullins,Phillips and Eldred claim that the nominal revenue-requirement 13 results show that solar PPAs are a superior resource option when compared to the 14 Combined Projects.(Mullins Supp.Direct,page 16,lines 8-16;Phillips Supp. 15 Direct,page 20,lines 3-13;Eldred Supp.Direct,page 11,lines 6-15.)How do you 16 respond? 17 A.First,as noted above,Mr.Phillips does not dispute that the customer benefits of the 18 Combined Projects and the solar resources together are higher than each resource 19 option alone when analyzed over a 20-year time frame,consistent with evaluation of 20 resource portfolios in the IRP.That is the key finding reported in my solar sensitivity 21 analysis. O Link,Supp-Reb -69 Rocky Mountain Power l Second,as described above,there is a risk that benefits of the solar PPAs 2 reported in my second supplemental testimony are overstated,as demonstrated by the 3 additional sensitivities discussed above,and that these risks could increase over time. 4 Q.These witnesses also claim that the solar option is also less risky than the 5 Combined Projects because the solar resources are PPAs.(Phillips Supp.Direct, 6 page 21,lines 1-13 and page 24,line 9 to page 25,line 2;Eldred Supp.Direct,page 7 10,lines 16-22.)Is this true? 8 A.No.These parties'focus on only the commercial structure is overly simplistic.As 9 described above,solar resources generally present additionalrisks that do not apply to 10 wind resources.Specifically,solar resources tend to generate most duringthe day,when 11 demand and prices are relatively low.Because the generation profile of solar resources 12 is consistent across the west,the increasing penetration of solar resources throughoutO13theregionwilllikelyfurtherdepresspricesduringtheperiodwhensolargenerates. 14 Thus,there is a risk with solar that the value of the generation provided will be less 15 than current forecasts and could be less than projected in the hourly price-profile 16 sensitivities. 17 Moreover,the capacity contribution of solar resources is likely decreasing as 18 solar penetration increases.As discussed above,this is a risk that is unique to solar 19 resources and means that the customer benefits for solar resources are likely overstated. 20 Q.Are there any other risks associated with pursuing solar resources now? 21 A.Yes.These parties also claim that the solar PPAs are less risky because they do not 22 require the Aeolus-to-Bridger/Anticline transmission line.But,as described by Mr. 23 Vail,that transmission line will provide substantial customer benefits independent of O Link,Supp-Reb -70 Rocky Mountain Power l the fact that it will enable interconnection of the Wind Projects.And,as described by 2 Mr.Vail,the company currently anticipates construction of the line by 2024 even 3 without the Combined Projects.Thus,far from reducing customer risk,if the company 4 selected the solar PPAs instead of the Combined Projects,it would create a very real 5 risk that customers would ultimately bear the cost of the Aeolus-to-Bridger/Anticline 6 line without the subsidy provided by the PTC-eligible Wind Projects.And if the costs 7 to construct the Aeolus-to-Bridger/Anticline line are considered in the solar-PPA 8 analysis,with the addition of PTC-eligible wind resources,the benefits of those solar- 9 PPA resources would decrease dramatically and would be substantially less than the 10 benefits of the Combined Projects. ll Q.Mr.Phillips also argues that the solar PPAs are superior because they provide no 12 equity returns to PacifiCorp.(Phillips Supp.Direct,page 26,line 1 to page 27,line 13 8.)Should the amount of equity returns have any bearing on the resource decision 14 at issue here? 15 A.No.PacifiCorp's resource planning considers the costs associated with a particular 16 resource decision and does not,and should not,consider whether a component of a 17 resource's cost is an equity return to PacifiCorp's shareholders or an equity return to a 18 shareholder of an independent power producer.There is no logical reason that 19 PacifiCorpwould select a more expensive or higher-risk resource simply because it did 20 not include an equity return to the company. Link,Supp-Reb -71 Rocky Mountain Power l Q.Mr.Phillips claims that the company would not have issued the 2017S RFP if the 2 Utah Commission had not suggested doing so and that this demonstrates serious 3 flaws in the 2017 IRP.(Phillips Supp.Direct,page 18,lines 1--9.)How do you 4 respond? 5 A.As discussed above,the 2017S RFP provided a great deal of useful market information 6 that will inform future IRPs.The recommendation from the Utah Commission to 7 consider solar-resource opportunities does not in any way suggest that the 2017 IRP 8 was flawed.Moreover,the 2017S RFP will not ultimately result in the acquisition of 9 solar resources because benefits of waiting are greater than the risks of moving forward 10 at this time. 11 ENERGY IMBALANCE MARKET MODELING 12 Q.Mr.Mullins again argues that the Company has not accounted for energy 13 imbalance market ("EIM")uninstructed imbalance charges.(Mullins Supp. 14 Direct,page 32,lines 18-20.)Can you please explain uninstructed imbalance 15 charges? 16 A.Yes.First,I will provide more context for the explanation and how EIM settlements are 17 calculated for PacifiCorp's resources.In the EIM,the company provides a base 18 schedule for all of its participating and non-participatingresources,including variable 19 energy resources such as wind facilities.The base schedules are hourlyand are used by 20 the CAISO for purposes of a balancing test to ensure that the company has scheduled 21 its resources within one percent of its expected demand in the upcoming hour.The next 22 step in the scheduling process is the fifteen-minute schedule,which is generated 23 approximately 30 minutes before the operating interval for each resource in O Link,Supp-Reb -72 Rocky Mountain Power l PacifiCorp's system.This fifteen-minute schedule is considered an advisory schedule 2 because it is not used for dispatch purposes.Lastly,there is a five-minute schedule, 3 which is a dispatch instruction to each of PacifiCorp's resources,including expected 4 wind output for the five-minute interval.Each of these three schedules--hourly,fifteen- 5 minute and five-minute--is used to calculate the instructed imbalance market 6 settlements for a resource. 7 For the uninstructed imbalance settlement,the CAISO utilizes the variance in 8 the actual submitted meter data for a resource,the five-minute dispatch instruction and 9 the five-minute locational marginal price at the resource node.The difference between 10 the five-minute dispatch instruction and the actual meter data is multiplied by the 11 locationalmarginal price and divided by 12 (division by 12 is required because the time 12 frame is a five-minute interval,and there are 12 five-minute intervals in an hour).This 13 calculation results in a charge to a resource if it produced less energy relative to the 14 schedule.Conversely,this calculation results in a payment to a resource if it produced 15 more energy relative to its schedule. 16 Q.In the company's rebuttal testimony filed in December 2017,Mr.Vail testified that 17 the company expects that the uninstructed imbalance charges should be neutral 18 over the life of the resource.(Vail Rebuttal,page 23,line 15 to page 24,line 14.) 19 Mr.Mullins argues that Mr.Vail was wrong (MullinsSupp.Direct,page 33,lines 20 3-10.)How do you respond? 21 A.As explained by Mr.Vail,the uninstructed imbalance charges are a reflection of 22 forecast error (actual meter data minus a five-minute forecast).Assuming that the 23 forecast,which is produced less than 30 minutes before the interval,has an equal O Link,Supp-Reb -73 Rocky Mountain Power l chance of being higher or lower over the life of a resource,the net charges should be 2 close to zero. 3 Mr.Mullins provides evidence related to two resources over a short period of 4 time to argue that there is an inherent bias in the forecasting.But the alleged bias is 5 simply the result of Mr.Mullins's reliance on a limited data set and is not reflective of 6 long-term expectations,which are that the net outcome will be closer to zero. 7 Q.Are there any other flaws in Mr.Mullins's analysis? 8 A.Yes.The existence of uninstructed imbalance charges assigned to certain resources 9 does not mean that there is an actual cost (or revenue)that is passed through to 10 customers.Uninstructed imbalance reflects the movement of resources and load that 11 are outside of the CAISO's dispatch and,therefore,PacifiCorp is required to manage 12 that variation using its regulating resources as the balancing area authority.PacifiCorp 13 must manage its area-control error as close to zero as possible to maintain its balancing 14 and frequency requirements in accordance with the National Electric Reliability 15 Council's standards.Thus,if a wind resource was five MW above its CAISO dispatch 16 (five-minute forecast),then another resource,likely a regulating resource,on the 17 PacifiCorp system would need to decrease by five megawatts to maintain system 18 balance. 19 Q.When the regulating resource moves in the opposite direction of the wind resource, 20 is that considered uninstructed imbalance? 21 A.Yes.The movement would be uninstructed imbalance because it was not part of the 22 CAISO's dispatch solution.When PacifiCorp regulates with its resources for changes 23 in wind,solar,and load outside of the CAISO's dispatch,that is considered regulation O Link,Supp-Reb -74 Rocky Mountain Power l and is maintained by keeping several of PacifiCorp thermal units in "regulating mode" 2 to make sure that PacifiCorp's system-balancing requirements are met. 3 Q.Does that mean there is a reciprocal cost or revenue for PacifiCorp's regulating 4 resources? 5 A.Yes.While Mr.Mullins includes a table that shows a cost for the wind facilities' 6 uninstructed imbalance,what he does not show is the corresponding revenue that was 7 received by one of PacifiCorp's regulating resources. 8 Q.Is there a cost for regulating for variable energy resources? 9 A.Yes.There is a cost for regulating for variable-energy resources,which is why 10 PacifiCorp includes an integration cost in its economic analysis,consistent with the 11 treatment of including an integration cost in the IRP. 12 Q.If the Commission used Mr.Mullins's assessment of the uninstructed imbalance 13 costs for the new wind facilities,would that be double counting the costs of 14 integration? 15 A.Yes.As noted above,integration costs are already included in the company's economic 16 analysis. 17 Q.One of Mr.Mullins's conditions is that in all future ratemaking proceedings,the 18 company's dispatch model used to set net power costs should include 300 MW of 19 transmission "link"between Jim Bridger and Walla Walla,consistent with the 20 EIM benefit assumption used in economic analysis of the Combined Projects. 21 (MullinsSupp.Direct,page 4,lines 28-32.)What is your response? 22 A.This is not the correct forum to establish modeling assumptions for all future 23 ratemaking proceedings that rely on the company's dispatch models to set net power O Link,Supp-Reb -75 Rocky Mountain Power l costs.PacifiCorp's transmission-topology assumptions used to capture EIM benefits in 2 this proceeding are valid,as described in my direct testimony,and are based on the best 3 available information at this time.It may be appropriate to incorporate this type of 4 assumption in future ratemaking proceedings that rely on the company's dispatch 5 models to set net power costs,but that analysis and determination should be made at 6 that point in time and should not be pre-determined in this proceeding. 7 CONCLUSION 8 Q.Please summarize the conclusions of your supplemental rebuttal testimony. 9 A.As confirmed by two different independent evaluators,the 2017R RFP was fair, 10 transparent,and unbiased.The independent evaluators found that the bids selected to 11 the 2017R RFP final shortlist represent the top offers that are viable under current 12 transmission planning assumptions,and the Utah independent evaluator found that the 13 final shortlist of bids should result in significant savings for customers.While solar- 14 resource bids submitted into the 2017R RFP may provide customer benefits,contrary 15 to claims from certain parties,solar-resource bids are not a superior resource alternative 16 to the Combined Projects.When considering solar resource valuation risks,expected 17 cost declines,and availability of the 30-percent ITC for solar projects coming online as 18 late as 2021,PacifiCorp does not need to act now and has decided not to select any of 19 the solar-PPA bids to the 2017S RFP final shortlist.PacifiCorp will continue to reassess 20 potential economic benefits from solar-resource opportunities through bi-lateral 21 opportunitiesand in the 2019 IRP,considering a thorough assessment of valuation risks 22 with full stakeholder engagement,to determine whether a new competitive solicitation 23 process for projects capable of achieving commercial operation by the end of2021 will O Link,Supp-Reb -76 Rocky Mountain Power l provide customer benefits. 2 In contrast,the phase out of PTC benefits that are available for qualifying wind 3 projects occurs sooner than the ramp down of ITC benefits that are available for solar 4 resources,which requires that PacifiCorp must act now to deliver the new wind and 5 needed transmission investments that will produce both near-term and long-term 6 benefits for customers.This conclusion is supported by thorough and extensive 7 economic analyses that is based on over 1,300 20-year simulations of PacifiCorp's 8 system,which have been used to evaluate how the net benefits of the Combined 9 Projects are affected by a variety of variables and uncertainties. 10 Q.Does this conclude your supplemental rebuttal testimony? ll A.Yes. O O Link,Supp-Reb -77 Rocky Mountain Power