PROCEEDING NO.

17A-0797E

RE: IN THE MATTER OF THE § BEFORE THE PUBLIC

APPLICATION OF PUBLIC SERVICE §
COMPANY OF COLORADO TO § UTILITIES COMMISSION OF THE
MODIFY THE DEPRECIATION §
SCHEDULES FOR THE EARLY § STATE OF COLORADO
RETIREMENT OF COMANCHE 1 AND §
COMANCHE 2 GENERATING UNITS, §
ESTABLISH A REGULATORY ASSET §
TO COLLECT INCREMENTAL §
DEPRECIATION, REDUCE THE §
RENEWABLE ENERGY STANDARD §
ADJUSTMENT COLLECTION TO ONE §
PERCENT, AND IMPLEMENT A §
GENERAL RATE SCHEDULE §
ADJUSTMENT, CONTINGENT ON §
THE APPROVAL OF THE COLORADO §
ENERGY PLAN PORTFOLIO IN §
PROCEEDING NO. 16A-0396E. §

CROSS-ANSWER TESTIMONY OF CHARLES S. GRIFFEY

ON BEHALF OF

The Coalition of Ratepayers

June 15, 2018

 PROCEEDING NO. 17A-0797E

RE: IN THE MATTER OF THE § BEFORE THE PUBLIC

APPLICATION OF PUBLIC SERVICE §
COMPANY OF COLORADO TO § UTILITIES COMMISSION OF THE
MODIFY THE DEPRECIATION §
SCHEDULES FOR THE EARLY § STATE OF COLORADO
RETIREMENT OF COMANCHE 1 AND §
COMANCHE 2 GENERATING UNITS, §
ESTABLISH A REGULATORY ASSET §
TO COLLECT INCREMENTAL §
DEPRECIATION, REDUCE THE §
RENEWABLE ENERGY STANDARD §
ADJUSTMENT COLLECTION TO ONE §
PERCENT, AND IMPLEMENT A §
GENERAL RATE SCHEDULE §
ADJUSTMENT, CONTINGENT ON §
THE APPROVAL OF THE COLORADO §
ENERGY PLAN PORTFOLIO IN §
PROCEEDING NO. 16A-0396E. §

EXECUTIVE SUMMARY

Mr. Griffey’s Cross Answer testimony rebuts Sierra Club witness Tyler Comings’ claim

that “Comanche 1 and 2 are clearly uneconomic when compared to other options available to the

Company.” Mr. Griffey testifies that the bid results made available in Public Service Company of

Colorado’s (PSCo) 120 Day Report and updated Renewable Energy Standard Adjustment (RESA)

projections reinforce the facts supporting a determination that continued operation of Comanche

1 and 2 is the most economic result for ratepayers. He demonstrates that PSCo’s claim that the

Preferred Colorado Energy Plan portfolio (Preferred CEP) will save $215 million on a net present

value (NPV) basis relative to the Preferred Electric Resource Plan (Preferred ERP) does not

accurately represent the costs of the CEP to ratepayers.

Cross-Answer Testimony of Charles S. Griffey

Page 2
 Because the unspent RESA balance is ratepayer money that provides value to customers,

he testifies that its use to pay off the accelerated depreciation costs associated with early retirement

of Comanche 1 and 2 should be reflected in the NPV costs of the CEP. In addition, the Commission

should examine when any purported CEP cost savings would occur. Data provided by PSCo

reveals the Preferred CEP portfolio would not break even on a NPV basis until the year 2046—

almost 30 years from now—and the calculated NPV savings are only $42 million. Using PSCo’s

RESA balance presented in the 120 Day Report, Mr. Griffey demonstrates that ratepayers are better

off by $468 million in 2033 and $365 million in 2035 on a nominal basis and $255 million and

$223 million on a present value basis, respectively, with continued operation of Comanche 1 and

2. Further, the ability of the CEP to break even on an NPV basis at all is the result of PSCo’s cost

assumptions for the generic plant that would replace Comanche 1 and 2 in the Preferred ERP

portfolio and the replacement power from purchase power contracts from this ERP that expire in

the backend of the planning period. He also offers testimony that, on a portfolio basis, PSCo’s

costs of transmission for the Preferred ERP portfolio relative to the Preferred CEP portfolio are

questionable.

Based upon his 30 years of experience as a utility planner and decision-maker, he observes

that utility decisions to build plants that are premised on NPV savings that are not projected to

materialize for several decades frequently do not materialize at all. The cost of such projects are

incurred upfront and paid by ratepayers, but no one is accountable for ensuring the provision of the

speculative future benefits. This puts excessive risk on the ratepayers. Just ask the ratepayers who

have had to support uneconomic nuclear and integrated gasification combined cycle (IGCC)

investments that were premised on savings 15 to 25 years into the future. He recommends that the

Commission give little weight to the speculative and small NPV savings estimates that arise from

Cross-Answer Testimony of Charles S. Griffey

Page 3
forecasted costs that will not be incurred for 20 to 30 years, if ever. Finally, he agrees with Mr.

Comings that there are many coal units across the United States that are uneconomic to continue

operating, but there is no data to indicate that Comanche 1 and 2 are among those units. Accordingly,

the Commission should reject the Preferred CEP and the related rate proposals in this case and defer

further review of early retirement of the units until a subsequent electric resource planning proceeding.

Cross-Answer Testimony of Charles S. Griffey

Page 4
 PROCEEDING NO. 17A-0797E

RE: IN THE MATTER OF THE § BEFORE THE PUBLIC

APPLICATION OF PUBLIC SERVICE §
COMPANY OF COLORADO TO MODIFY § UTILITIES COMMISSION OF THE
THE DEPRECIATION SCHEDULES FOR §
THE EARLY RETIREMENT OF § STATE OF COLORADO
COMANCHE 1 AND COMANCHE 2 §
GENERATING UNITS, ESTABLISH A §
REGULATORY ASSET TO COLLECT §
INCREMENTAL DEPRECIATION, §
REDUCE THE RENEWABLE ENERGY §
STANDARD ADJUSTMENT §
COLLECTION TO ONE PERCENT, AND §
IMPLEMENT A GENERAL RATE §
SCHEDULE ADJUSTMENT, §
CONTINGENT ON THE APPROVAL OF §
THE COLORADO ENERGY PLAN §
PORTFOLIO IN PROCEEDING NO. 16A- §
0396E. §

TABLE OF CONTENTS

EXECUTIVE SUMMARY

I. INTRODUCTION AND SUMMARY ................................................................................8

II. CONTINUED OPERATION OF COMANCHE 1 AND 2 IS ECONOMIC. ...................15

III. CONCLUSION ..................................................................................................................29

AFFIDAVIT

Cross-Answer Testimony of Charles S. Griffey

Page 5
 LIST OF EXHIBITS

CSG-CA-1 Impact of Accounting for Accelerated Depreciation Using PSCo Data

CSG-CA-2 Impact of Accounting for Differences in RESA Payments/Accrual Using PSCo

Data

LIST OF FIGURES

CSG-CA-1 Economics of Operating Comanche 1 and 2 Using PSCo Data from 120 Day
Report - Accelerated Depreciation Method

CSG-CA-2 Economics of Operating Comanche 1 and 2 Using PSCo Data from 120 Day
Report - RESA Accrual Method

CSG-CA-3 Comparison of Transmission Cost Between Preferred ERP and Preferred CEP
Portfolios

Cross-Answer Testimony of Charles S. Griffey

Page 6
 GLOSSARY OF ACRONYMS AND DEFINED TERMS

Acronym/Defined Term Meaning

CCGT Combined Cycle Gas Turbine

CT Combustion Turbine
GRSA General Rate Schedule Adjustment
IGCC Integrated Gasification Combined Cycle
ITC Investment Tax Credit
NPV Net Present Value
PPA Power Purchase Agreement
Preferred CEP Preferred Colorado Energy Plan
Preferred ERP Preferred Energy Resource Plan
PSCo Public Service Company of Colorado
PTC Production Tax Credit
PVRR Present Value Revenue Requirements
RESA Renewable Energy Standard Adjustment

Cross-Answer Testimony of Charles S. Griffey

Page 7
 PROCEEDING NO. 17A-0797E

RE: IN THE MATTER OF THE § BEFORE THE PUBLIC

APPLICATION OF PUBLIC SERVICE §
COMPANY OF COLORADO TO MODIFY § UTILITIES COMMISSION OF THE
THE DEPRECIATION SCHEDULES FOR §
THE EARLY RETIREMENT OF § STATE OF COLORADO
COMANCHE 1 AND COMANCHE 2 §
GENERATING UNITS, ESTABLISH A §
REGULATORY ASSET TO COLLECT §
INCREMENTAL DEPRECIATION, §
REDUCE THE RENEWABLE ENERGY §
STANDARD ADJUSTMENT §
COLLECTION TO ONE PERCENT, AND §
IMPLEMENT A GENERAL RATE §
SCHEDULE ADJUSTMENT, §
CONTINGENT ON THE APPROVAL OF §
THE COLORADO ENERGY PLAN §
PORTFOLIO IN PROCEEDING NO. 16A- §
0396E. §

2 CROSS-ANSWER TESTIMONY OF CHARLES S. GRIFFEY

3 I. INTRODUCTION AND SUMMARY

4 Q. PLEASE STATE YOUR NAME, OCCUPATION, AND BUSINESS ADDRESS.

5 A. My name is Charles S. Griffey, and I am a consultant providing services to the energy

6 sector. My address is 2918 Todville Road, Seabrook, Texas 77586.

7 Q. ARE YOU THE SAME CHARLES S. GRIFFEY WHO PREVIOUSLY TESTIFIED

8 ON BEHALF OF THE COALITION OF RATEPAYERS IN THIS PROCEEDING

9 AND PROCEEDING NO. 16A-0396F?

10 A. Yes.

11 Q. WHAT IS THE PURPOSE OF YOUR TESTIMONY?

Cross-Answer Testimony of Charles S. Griffey
Page 8
 1 I respond to Sierra Club witness Tyler Comings’ claim that “Comanche 1 and 2 are clearly

2 uneconomic when compared to other options available to the Company. Expediting their

3 retirement frees up the Company to pursue these less expensive options.”1

4 Q. IS THIS CLAIM IN MR. COMINGS’ TESTIMONY ACCURATE?

5 A. No.

6 Q. HAVE YOU PREVIOUSLY ADDRESSED THE ECONOMICS OF OPERATING

7 COMANCHE 1 AND 2?

8 A. Yes. I presented analysis in Docket No. 16A-0396E on this topic. My analysis showed

9 that, using PSCo’s own modeling assumptions, the CEP portfolio will be far more costly

10 than operating Comanche 1 and 2 through the end of their useful lives.2 Any claimed NPV

11 savings are an artifact of PSCo’s analytic methods and the company’s faulty decision to

12 ignore the impact of its assumption that the RESA will be 2% if the plants continue to

13 operate but only 1% in the CEP portfolio.3 I also demonstrated that continued operation of

14 Comanche 1 and 2 was economic in my Answer Testimony in this case. My Answer

15 Testimony used the cost data then available from PSCo and showed that continued

16 operation of Comanche 1 and 2 saved ratepayers approximately $180 million in present

17 value through the end of Comanche 1 and 2’s current lives of 2033 and 2035, respectively.

18 My Answer Testimony further reinforced that any purported net present value savings from

19 early retirement was simply the result of PSCo’s speculative modeling assumptions for

20 what might replace Comanche 1 and 2 in the year 2034. The approximately $180 million

1
Answer Testimony of Mr. Comings at 7.
2
Answer Testimony of Charles Griffey and Surrebuttal Testimony of Charles Griffey in Docket No. 16A-0396E.
3
Id.
Cross-Answer Testimony of Charles S. Griffey
Page 9
 1 figure was based on PSCo’s estimates of the outcome of its bid evaluation process. I also

2 projected that the actual outcome would not be meaningfully different based on the average

3 bid prices PSCo reported from the refreshed bids because the actual bids selected would

4 need to have been at an unreasonably significant discount to the average bids.

5 Q. ARE THE ACTUAL RESULTS OF THE SOLICITATION NOW AVAILABLE?

6 A. Yes. The results were made available on June 6, 2018 in PSCo’s 120 Day Report.

7 Q. DO THE ACTUAL BID RESULTS CHANGE YOUR OPINION REGARDING

8 WHETHER CONTINUED OPERATIONS OF COMANCHE 1 AND 2 IS

9 ECONOMIC?

10 A. No. The actual results reinforce my determination that continued operation of Comanche

11 1 and 2 is the most economic result for ratepayers. Using PSCo’s Preferred ERP portfolio

12 and Preferred CEP portfolio, the economic results from the solicitation do not change the

13 conclusion that the most economic result at this time is continued operation of Comanche

14 1 and 2. In fact, the economic results through the year 2035 are little changed from what I

15 showed in my cross-answer testimony.

16 Q. PLEASE EXPLAIN.

17 A. The Preferred ERP portfolio contemplates continued operation of Comanche 1 and 2

18 through 2033 and 2035, respectively, and purchases of 322 MW of solar resources, 789

19 MW of wind, 50 MW of batteries, and a 301 MW asset purchase of combustion turbines.

20 The Preferred CEP portfolio contemplates retirement of Comanche 1 and 2 on an

21 accelerated basis and acquisition of 707 MW of solar, 1131 MW of wind, 275 MW of

22 batteries, and 383 MW of existing combustion turbines. Data provided in PSCo’s 120 Day

Cross-Answer Testimony of Charles S. Griffey

Page 10
 1 Report shows that, when including accelerated depreciation as a cost in the Preferred CEP

2 portfolio, the savings to ratepayers from continued operating of Comanche 1 and 2 through

3 2033 is $340 million on a nominal basis and $192 million on a present value basis. The

4 nominal amount is $232 million and the present value amount is $160 million through

5 2035.4 The Preferred CEP portfolio does not break even on a present value basis until 2042,

6 well after the status quo retirement of Comanche 1 and 2.

7 Q. DO THESE COST CALCULATIONS REFLECT ANY ADJUSTMENTS TO THE

8 DATA PSCo PRESENTED IN THE 120 DAY REPORT?

9 A. No. These calculations are based on PSCo’s own figures for annual revenue requirements

10 and the impact of accelerated depreciation,5 and do not include any other adjustments that

11 I believe should be made.

12 Q. ARE THERE ADJUSTMENTS THAT YOU BELIEVE SHOULD BE MADE?

13 A. Yes. PSCo’s RESA balance calculations show that the RESA balance in the Preferred ERP

14 case is significantly higher than what it otherwise would be if the funds are used to pay off

15 the costs of the CEP. PSCo is now showing that there is an impact to RESA balances in

16 the CEP in addition to the 1% General Rate Schedule Adjustment (GRSA). As I have

17 previously testified, PSCo’s NPV cost calculations do not account for the value of the

18 RESA to ratepayers. The RESA balance is ratepayer money and its use to pay off the

4
These calculations are based on the annual costs presented by PSCo in Appendix N to the 120 Day Report and the
impact of including accelerated depreciation shown by PSCo in Appendix L to the 120 Day Report. PSCo now utilizes
2016 as the present value base for its calculations, while the present value calculation in my cross answer testimony
used 2017 as a base year. Therefore, the $180 million NPV reported in my cross-answer testimony needs to be
discounted by 6.8% for an additional year for a 2016 NPV. Thus my cross-answer testimony figure is $169 million
NPV (2016) compared to $160 million based on the data in the 120 Day Report.
5
I cannot agree with PSCO’s calculation of the impact of accelerated depreciation at this time because they have not
shown its derivation. However, their present value figure appears to be low relative to the $211 million nominal
amount of accelerated depreciation that should be taken.
Cross-Answer Testimony of Charles S. Griffey
Page 11
 1 accelerated depreciation costs associated with the CEP should be reflected in the costs of

2 the CEP. The fact that there are additional RESA impacts should also be considered when

3 comparing the portfolios.

4 Q. HOW WOULD THE DIFFERENCE IN THE RESA VALUES FOR THE ERP AND

5 CEP PORTFOLIOS IMPACT PSCo’S ESTIMATES OF CEP SAVINGS?

6 A. It would decrease them significantly. PSCo now shows that the accumulated RESA

7 balances differ by more than just the 1% GRSA between the CEP and the case with

8 continued operation of Comanche 1 and 2. Indeed, PSCo now shows that by 2026 the

9 unspent RESA balance will be $487 million (an increase from the $440 million shown

10 earlier in this case) with continued operation of Comanche 1 and 2 and $209 million in the

11 CEP case (a decrease from the $267 million previously shown), a difference of $279

12 million.6 This difference is much greater than simply the difference in the GRSA. PSCo

13 does not account for this difference when comparing the present value savings of the CEP

14 compared to continued operation of Comanche 1 and 2, yet it is economically inappropriate

15 to ignore the value to ratepayers of the hundreds of millions in cash sitting in the RESA

16 balance. It is ratepayer money and can be used to provide value to the ratepayers.

17 Under PSCo’s own approach and calculations using the GRSA to offset accelerated

18 depreciation while accounting for the difference in RESA balances, ratepayers would be

19 worse off by $468 million on a nominal basis and $255 million on a present value basis in

20 2033. When accounting for the value in the RESA over collections in the case of continued

21 operation of Comanche 1 and 2, the CEP does not break even on a present value basis until

6
Current projections are from the Corrected Appendix D to the 120 Day Report filed on June 7, 2018, while the
previous projections were from Mr. Trowbridge’s testimony in this case at 25-29.
Cross-Answer Testimony of Charles S. Griffey
Page 12
 1 2046. This is not only well after Comanche 1 and 2 would have otherwise retired, it is even

2 after the contracts from the current solicitation would have expired.

3 Q. HOW DO YOU ADDRESS PSCo’S CLAIM THAT THE NPV OF THE

4 PREFERRED CEP IS $215 MILLION LOWER THAN THE PREFERRED ERP

5 PORTFOLIO?

6 A. PSCo’s calculation continues to ignore the accelerated depreciation/RESA differential.

7 When correctly accounting for payments that customers make into the RESA/GRSA, the

8 net present value between the Preferred ERP and Preferred CEP using PSCo’s data would

9 only be $42 million. Further, that $42 million benefit has nothing to do with the operation

10 of Comanche 1 and 2. As shown by the breakeven years I discussed above, the Commission

11 should not give the NPV benefit much weight. The present value benefits in this case are

12 due to speculative modeling assumptions comparing current and future contracts to what

13 PSCo assumes today will replace Comanche 1 and 2 in the year 2034. That is a poor basis

14 on which to force ratepayers to support a $2.5 billion investment today.

15 Q. HAVE RATEPAYERS BEEN HARMED BY SIMILAR POLICY DECISIONS IN

16 THE PAST?

17 A. Yes. For instance, ratepayers elsewhere that have had to support uneconomic nuclear and

18 IGCC investments can certainly attest to the fact that promised savings 15 to 25 years into

19 the future frequently do not materialize.7 High capital cost resources with low energy cost

20 are particularly susceptible to justifications based on distant purported savings. The

21 Commission should weigh the relative certainty of the fact that customers will be worse

7
For example, the now-canceled Kemper County IGCC in Mississippi was determined to be the best resource in utility
company evaluations from 2008-2012 based on expectations of savings that did not occur for over a decade.
Cross-Answer Testimony of Charles S. Griffey
Page 13
 1 off for the next twenty to twenty-five years to the speculative nature of the promised later

2 benefits. Under the CEP, customers pay now and for the next two decades, and no one is

3 accountable for providing the promised benefits later.

4 Q. DOES ANY OTHER INFORMATION IN THE 120 DAY REPORT SUPPORT

5 CONTINUED OPERATION OF COMANCHE 1 AND 2?

6 A. Yes. PSCo states that it expects that solar generation will continue to be low cost and may

7 continue to fall in cost. 8 This is another reason ratepayers are better off if solar purchases

8 are delayed. Since PSCo’s own data shows that it is currently more economic to operate

9 Comanche 1 and 2, the economically reasonable choice is to continue to operate Comanche

10 1 and 2 and to delay additional solar purchases beyond what is shown in the ERP results

11 until some future period when they become economic compared to the operation of

12 Comanche 1 and 2. To do otherwise is economically irrational. Put another way, the

13 Commission need not decide today to operate Comanche 1 and 2 through their expected

14 lives compared to early retirement. That can be looked at in each resource planning cycle.

15 Q. MR. COMINGS CLAIMS THERE ARE POSSIBLE FUTURE ENVIRONMENTAL

16 COSTS THAT ARE NOT IN THE MODELING RUNS.9 DOES THIS NEED TO BE

17 ACCOUNTED FOR NOW?

8
See 120 Day Report at 12 (“…the acquisition of solar and solar with battery storage projects, which we anticipate
will continue to be available at even lower prices in the next ERP cycle with the full 30% Investment Tax Credit
(ITC). While this last solar with battery storage project is cost-effective, it results in the acquisition of capacity before
it is needed by the system. Even assuming the price of solar and storage remain flat through the next ERP, delaying
this last contract until the next ERP cycle would save customers approximately $20 million. As noted above, we expect
these prices to continue their decline so this savings estimate is conservative.”).
9
Answer Testimony of Mr. Comings at 10.
Cross-Answer Testimony of Charles S. Griffey
Page 14
 1 A. No. The only choice that need be made today is whether to continue operation of

2 Comanche 1 and 2 through to the next ERP cycle or to commit to billions of dollars of

3 Power Purchase Agreements (PPAs) to replace them. There are no known or reasonably

4 forecasted environmental compliance costs that are unaccounted for with respect to that

5 decision. The economic outcome is to continue to operate the Comanche units and revisit

6 that decision in each ERP cycle if additional compliance costs arise. The Commission

7 should avoid the false choice of having to decide whether those units should operate for

8 the next 15 to 17 years compared to purchasing billions in high capital cost alternatives

9 today.

10 II. CONTINUED OPERATION OF COMANCHE 1 AND 2 IS ECONOMIC.

11 Q. WHAT DOES MR. COMINGS RELY UPON TO CLAIM THAT OPERATION OF

12 COMANCHE 1 AND 2 IS UNECONOMIC?

13 A. He cites the illustrative calculations made by PSCo using the annuity method,10 the belief

14 that the actual solicitation results would be substantially lower than the illustrative bids,11

15 and a belief that PSCo understated the costs of continuing to operate Comanche 1 and 2

16 by not including operation and maintenance costs.12

17 Q. HAS THE COMMISSION ALREADY RULED UPON THE RELIANCE THAT

18 SHOULD BE PLACED ON CALCULATIONS USING THE ANNUITY METHOD?

19 A. Yes. The Commission ruled that the results using the annuity method should be given little

20 weight with regard to evaluating the CEP. The Commission stated “[w]e also agree with

21 the Ratepayers Coalition that it is appropriate to give little weight to the annuity method

10
Direct Testimony of Mr. Comings at 8, n. 3.
11
Id. at 9–10.
12
Id. at 11.
Cross-Answer Testimony of Charles S. Griffey
Page 15
 1 applied to the CEP portfolio for purposes of determining the cost-effectiveness of the early

2 retirements of Comanche 1 and 2. Although the annuity method provides relevant and

3 essential information when comparing bids against utility-owned resources across various

4 portfolios, the annuity method could provide skewed information when comparing new

5 proposed resources to the existing Comanche 1 and 2 units.”13

6 Q. DID PSCo ACTUALLY FORGET TO INCLUDE OPERATION AND

7 MAINTENANCE COST IN EVALUATING THE ECONOMICS OF CONTINUED

8 OPERATION OF COMANCHE 1 AND 2?

9 A. No. Mr. Comings cites to a calculation that PSCo provided in this case, but the intent of

10 that calculation was to show the purported capital cost savings from retiring Comanche 1

11 and 2 early,14 not to show whether it was overall in the best interest of ratepayers to retire

12 Comanche 1 and 2. In addition to not including the cost of operations and maintenance

13 expense for Comanche 1 and 2, the calculation does not include any fuel savings associated

14 with the operation of Comanche 1 and 2; nor does it include any of the capital or operating

15 costs of replacement power. Mr. Comings seems to recognize this when he notes that the

16 calculation in Attachment MAM-1 is not comparable to a calculation made in an ERP

17 analysis.15 Given that, it is unclear why he conflates the calculation in Attachment MAM-

18 1, which only purports to look at capital cost savings 16 with the type of analysis necessary

19 to determine whether it is economic to continue operating Comanche 1 and 2.

13
Decision No. C18-0191 at paragraph 99.
14
Mr. Comings cites to Attachment MAM-1 in this case. Answer Testimony of Mr. Comings at 10.
15
Answer Testimony of Mr. Comings at 11.
16
Answer Testimony of Charles S. Griffey at 19–27. Note that I disagree with the manner in which PSCo made this
specific analysis.
Cross-Answer Testimony of Charles S. Griffey
Page 16
1 Q. DOES THE USE OF THE ACTUAL SOLICITATION RESULTS DEMONSTRATE

2 THAT COMANCHE 1 AND 2 SHOULD BE RETIRED EARLY, AS MR.

3 COMINGS CLAIMS?

4 A. No. Instead, the actual results demonstrate that, at this time, it is economic to continue

5 operation of Comanche 1 and 2.

6 Q. PLEASE DEMONSTRATE WHY CONTINUED OPERATION OF COMANCHE 1

7 AND 2 IS ECONOMIC.

8 A. Please refer to Figure CSG-CA-1 (also included in more detail as Exhibit CSG-CA-1)

9 below:

Cross-Answer Testimony of Charles S. Griffey

Page 17
 Figure CSG-CA-1
Economics of Operating Comanche 1 and 2
Using PSCo Data from 120 Day Report
Accelerated Depreciation Method

C1 C2 C3 C4 C5 C6
Revenue Revenue Nominal
Requirement Requirement Impact of Cumulative Cumulative
Portfolio #3 Portfolio #6 Accelerated Nominal Present Value
Preferred ERP Preferred CEP Depreciation Difference Difference
Year $ Millions $ Millions $ Millions $ Millions $ Millions
2016 1,473 1,474 - (1) (1)
2017 1,882 1,882 - (1) (1)
2018 1,917 1,917 (16) (17) (15)
2019 1,921 1,923 (30) (48) (40)
2020 1,934 1,938 (28) (80) (65)
2021 1,989 2,007 (26) (123) (96)
2022 1,943 1,959 (28) (167) (126)
2023 1,996 2,006 (7) (184) (137)
2024 2,034 2,042 (7) (200) (146)
2025 2,077 2,072 (8) (202) (147)
2026 2,118 2,124 1 (207) (150)
2027 2,175 2,177 (1) (211) (151)
2028 2,193 2,179 (1) (198) (146)
2029 2,339 2,317 - (176) (136)
2030 2,400 2,373 - (149) (126)
2031 2,448 2,505 - (206) (147)
2032 2,499 2,563 - (270) (169)
2033 2,530 2,600 - (340) (192)
2034 2,695 2,648 - (293) (178)
2035 2,781 2,720 - (232) (160)
2036 2,907 2,806 - (131) (133)
2037 2,935 2,834 - (30) (108)
2038 3,003 2,898 - 75 (83)
2039 3,102 3,001 - 176 (60)
2040 3,177 3,070 - 283 (38)
2041 3,272 3,153 - 402 (15)
2042 3,521 3,409 - 514 5

1 To make these calculations, I took the annual revenue requirements from Appendix N to

2 the 120 Day Report and the annual figures that PSCo reports for the impact of considering

3 the accelerated depreciation of Comanche 1 and 2 in the economic analysis from Appendix
Cross-Answer Testimony of Charles S. Griffey
Page 18
 1 L to the 120 Day Report.17 I add the difference in the annual revenue requirements to the

2 annual accelerated depreciation impact and then calculate a running total of those amounts

3 in Column 5. In Column 6, I show the running total of the present value of the revenue

4 requirement difference between the two portfolios. Comanche 1 retires in the year 2033

5 and Comanche 2 retires in the year 2035 in the ERP preferred portfolio. In the year 2033,

6 on a nominal basis, ratepayers are cumulatively $340 million better off and, in 2035, they

7 are cumulatively $232 million better off from continued operation of Comanche 1 and 2.

8 The present value amounts for those figures are $192 million and $160 million for the years

9 2033 and 2035, respectively.

10 Q. DOES PSCo’S ANALYSIS IN THE 120 DAY REPORT PROVIDE ANOTHER

11 WAY TO PUT THE PREFERRED ERP PORTFOLIO AND THE PREFERRED

12 CEP PORTFOLIO ON AN APPLES-TO-APPLES BASIS?

13 A. Yes. The total present value revenue requirements (PVRR) for each of these portfolios

14 shown in Appendix B to the 120 Day Report are incomplete because they do not include

15 the impacts of accelerated depreciation or the proposed GRSA or the RESA charges and

16 balances. I have already discussed why the differences in RESA balances should be

17 accounted for — i.e., because they represent an asset to customers that can fund rate

18 reductions or the purchase of a resource of equal or greater value — so I will not discuss

19 the point in further detail.18 In Figure CSG-CA-1 above, I attempt to put the PVRRs of the

20 Preferred CEP and ERP portfolios on an apples-to-apples basis by using PSCo calculations

21 that assume there is no GRSA, the RESA charge is the same between the two portfolios,

17
As noted earlier, I do not agree that PSCo’s calculations on the impact of accelerated depreciation are correct, but
for purposes of this example I will take them as a given.
18
Answer Testimony of Charles S. Griffey at 16-19, testimony of Charles S. Griffey in Docket 16A-0396E.
Cross-Answer Testimony of Charles S. Griffey
Page 19
1 and accelerated depreciation is taken as it occurs. However, this calculation does not

2 account for different RESA balances based on the different impact that the Preferred ERP

3 and Preferred CEP portfolios have on the RESA balance. PSCo has now provided data

4 that can be used to compare the portfolios on an apples-to-apples basis, i.e., assuming that

5 the GRSA is granted to retire the accelerated depreciation as PSCO proposes and tracking

6 the relative RESA balances between the two portfolios. By including the value in the

7 higher RESA balance in the Preferred ERP portfolio, a comparison of the total economic

8 impact to ratepayers can be made. This is shown in Figure CSG-CA-2 below, with full

9 calculations shown in Exhibit CSG-CA-2:

Figure CSG-CA-2
Economics of Operating Comanche 1 and 2
Using PSCo Data from 120 Day Report
RESA Accrual Method

C1 C2 C3 C4 C5 C6
Cumulative
Nominal
Revenue Revenue Difference
Requirement Requirement Nominal After Cumulative
Portfolio #3 Portfolio #6 Difference in Accounting Present Value
Preferred ERP Preferred CEP RESA Accrual For RESA Difference
Year $ Millions $ Millions $ Millions $ Millions $ Millions
2016 1,473 1,474 - (1) (1)
2017 1,882 1,882 - (1) (1)
2018 1,917 1,917 - (1) (1)
2019 1,921 1,923 (1) (4) (4)
2020 1,934 1,938 (4) (12) (10)
2021 1,989 2,007 (50) (80) (59)
2022 1,943 1,959 (53) (150) (106)
2023 1,996 2,006 (38) (198) (136)
2024 2,034 2,042 (41) (247) (165)
2025 2,077 2,072 (44) (285) (186)
2026 2,118 2,124 (47) (338) (214)
2027 2,175 2,177 - (340) (215)
2028 2,193 2,179 - (326) (208)
2029 2,339 2,317 - (304) (199)
Cross-Answer Testimony of Charles S. Griffey
Page 20
 2030 2,400 2,373 - (277) (188)
2031 2,448 2,505 - (334) (210)
2032 2,499 2,563 - (398) (232)
2033 2,530 2,600 - (468) (255)
2034 2,695 2,648 - (421) (240)
2035 2,781 2,720 - (360) (223)
2036 2,907 2,806 - (259) (196)
2037 2,935 2,834 - (158) (170)
2038 3,003 2,898 - (53) (145)
2039 3,102 3,001 - 48 (123)
2040 3,177 3,070 - 155 (101)
2041 3,272 3,153 - 274 (78)
2042 3,521 3,409 - 386 (58)
2043 3,734 3,621 - 499 (38)
2044 3,831 3,719 - 611 (20)
2045 3,919 3,799 - 731 (3)
2046 4,179 4,106 - 804 8
1

2 Q. WHAT DO PSCo’S NUMBERS SHOW?

3 A. PSCo’s numbers show that under the Preferred ERP portfolio where the RESA stays at 2%,

4 ratepayers will pay an additional $40 million to $50 million annually from 2021-2026

5 relative to the Preferred CEP portfolio. That amount is accrued into a higher RESA

6 balance. PSCo never gives ratepayers credit for that accrued value in its PVRR

7 calculations. When those ratepayer funds are included as a value, customers are better off

8 by $468 million in 2033 and $360 million in 2035 on a nominal basis and $255 million and

9 $223 million on a present value basis, respectively. When properly accounting for the

10 difference in RESA alone, the Preferred CEP portfolio does not break even until the year

11 2046.

12 Q. WHY DO YOU FOCUS ON THE YEARS 2033 AND 2035?

13 A. I focus on those years because that is when Comanche 1 and Comanche 2, respectively,

14 are currently slated to retire. If the alternate portfolio cannot show savings prior to those

Cross-Answer Testimony of Charles S. Griffey

Page 21
 1 retirement years, it makes the case that the two units are, in fact, economic to continue

2 operating. It further highlights that the any modeled present value savings arise not from

3 a comparison to the costs of operating the existing Comanche units to the CEP, but from a

4 comparison of the (1) the cost of the CEP to the assumed cost of the plant assumed to

5 replace Comanche 1 and 2 in 2034 , and (2) the assumed cost of the resources that will

6 replace the CEP purchase power agreements when they expire in twenty years to the

7 assumed cost of the plants assumed to replace Comanche 1 and 2. One needs to examine

8 both of those assumptions to determine the weight to give to speculative savings estimates

9 that do not occur for 20 to 30 years.

10 Based on my understanding of PSCo’s method, I recommend that the Commission give

11 little weight to purported savings that occur after the retirement of Comanche 1 and 2. In

12 my over 30 years of experience in the utility industry I have found that plants and contracts

13 that are purported to provide cumulative NPV savings on a planning basis but which rely

14 on savings that do not appear for over twenty years are bad deals. Traditional utility

15 modeling approaches that look at NPV only and disregard savings on an annual basis

16 understate uncertainty and ignore the option value of waiting for improvements in

17 technology. Even PSCo states it expects the cost of solar resources to decline. Why then

18 would a rational planner retire units that are more economic to operate today than the

19 alternatives when it can delay retirement and receive the benefit of lower cost technology

20 in the future? PSCo gives no answer to that question, but creates an either/or decision

21 when such a decision does not need to be made. There is simply no need to make the

22 decision today, and potentially evaporating PTC and ITC credits do not justify an irrational

23 decision. The correct economic choice is to defer the retirement.

Cross-Answer Testimony of Charles S. Griffey

Page 22
 1 Q. WHAT FACTOR IS A MAJOR DRIVER IN PSCo’S PURPORTED SAVINGS OF

2 THE CEP RELATIVE TO THE ERP?

3 A. As discussed in Docket 16A-0396E, PSCo hard-coded a decision to replace Comanche 1

4 and 2 with a CCGT based on current technology in the year 2034. Not surprisingly, the

5 year 2034 is when the CEP starts to consistently provide some benefit (but not yet net

6 benefit) relative to the ERP portfolio.

7 Q. ARE YOU ATTEMPTING TO RELITIGATE THE COMMISSION DECISION IN

8 DOCKET 16A-0396E?

9 A. No. I am pointing out the implications of that decision so the Commission can

10 appropriately weigh the evidence.

11 Q. ONE OF YOUR ISSUES IN DOCKET 16A-0396E WAS THAT THE COMBINED

12 CYCLE GAS TURBINE (CCGT) IN THE YEAR 2034 WAS

13 DISPROPORTIONATELY BURDENED WITH OVER $100 MILLION IN

14 TRANSMISSION COSTS,19 AND THAT THE ASSUMED COST OF THE CCGT

15 APPEARED HIGH. HAS PSCo RECENTLY PROVIDED INFORMATION THAT

16 CASTS DOUBT ON THE ASSUMPTIONS IT USED IN THE MODELING FOR

17 THE CCGT?

18 A. Yes. PSCo provided the estimates of interconnection cost they used to evaluate bids, as

19 well as the levelized cost of the bids in Appendix H to the 120 Day Report. The actual

20 amounts PSCo used to evaluate the bids are highly confidential and are not included in my

21 testimony in this docket. However, the Commission should compare the transmission

19
Answer Testimony of Charles S. Griffey in Docket 16A-0396E at 36.
Cross-Answer Testimony of Charles S. Griffey
Page 23
 1 amounts assigned to the gas turbine and CCGT bids received to the approximate $100

2 million transmission cost assigned by PSCo to the generic CCGT that replaces Comanche

3 1 and 2 in 2034. The differences in transmission cost have a direct effect of biasing the

4 cost of the ERP portfolios higher, which makes continued operation of Comanche 1 and 2

5 appear less favorable on an NPV basis, even though the transmission cost of the CCGT in

6 2034 has nothing to do with the operation of Comanche 1 and 2.

7 Q. PLEASE COMPARE THE LEVELIZED COST OF THE BIDS TO LEVELIZED

8 COSTS ASSUMED FOR THE GENERIC COMBUSTION TURBINE (CT) AND

9 THE CCGT ASSUMED TO REPLACE COMANCHE 1 AND 2.

10 A. Again, the amounts are highly confidential, but the Commission should review the cost

11 PSCo assumed that a new CT would cost to their self-build bids for new CTs and look at

12 the discount to the assumed cost. Similarly, compare the third party bids to the cost

13 assumed by PSCo. Some of the third party facilities bid appear to be at greenfield sites. A

14 comparison of the levelized cost for CCGTs to the cost of the generic CCGT assumed to

15 replace Comanche 1 and 2 in 2034 is also a comparison that should be made to determine

16 the likelihood that savings will actually accrue to ratepayers after Comanche 1 and 2 are

17 retired. After reviewing the bid cost and associated transmission costs relative to the costs

18 included for the hard-coded CCGT in 2034, my conclusion is that little weight should be

19 given to the savings estimates that flow directly from the assumptions associated with the

20 hard-coded 2034 CCGT.

21 Q. WHAT TRANSMISSION COSTS WERE ASSIGNED TO THE PREFERRED ERP

22 AND THE PREFERRED CEP PORTFOLIOS?

23 A. Please refer to Figure CSG-CA-3 below:

Cross-Answer Testimony of Charles S. Griffey
Page 24
 1 Figure CSG-CA-3
2 Comparison of Transmission Cost
3 Between Preferred ERP and Preferred CEP Portfolios
4 $Millions

PREFERRED ERP PREFERRED CEP

PORTFOLIO #3 PORTFOLIO #6

ELECTRIC
INTERCONNECTION FOR 42 73
SELECTED RESOURCES

ELECTRIC DELIVERY
149 113

TRANSMISSION COSTS FOR

GENERIC CCGT 100 0
REPLACING COMANCHE 1
AND 2

TOTAL
(IN MILLIONS) 291 186

6 The Preferred ERP portfolio has 1,111 MW of new renewable resources, a 301 MW

7 purchase of existing gas CTs, the operation of Comanche 1 and 2 through 2033/2035 and

8 then its replacement in 2034 by a greenfield CCGT with $100 million in transmission cost.

9 The Preferred CEP portfolio has 1,838 MW of new renewables and the purchase of 383

10 MW of existing CTs. Just looking at the new renewables, one is struck by the outcome that

11 transmission interconnection and upgrade costs for 1,111 MW cost $191 million in the

12 ERP portfolio while it only costs $186 million for 1,838 MW in the CEP portfolio. The

Cross-Answer Testimony of Charles S. Griffey

Page 25
 1 Commission should recall that no other units in the analysis have any transmission

2 interconnection costs assigned to them. This disparity should raise questions.

3 Q. HOW DOES PSCo ARRIVE AT THAT OUTCOME?

4 A. In relative terms, the interconnection costs for only the renewables between the two

5 portfolios are not too far apart (although in absolute terms, it is still striking that PSCo

6 assumes it can interconnect 1,838 MW of renewables for $73 million, but that a 700 MW

7 combined cycle would cost $100 million to interconnect, as megawatt-mile costs do not

8 depend on how power is made). Aside from the CCGT in 2034, the major difference lies

9 in the network upgrade costs between the two portfolios.

10 Q. HOW DID PSCo CALCULATE THE NETWORK UPGRADE COSTS?

11 A. It is unclear, as the values presented in the Bid Portfolio Analysis Summary Results do not

12 match the procedure laid out in Appendix I to the 120 Day Report considering transmission

13 planning. The Preferred CEP portfolio shows 1325 MW of generation utilizing the 345kW

14 backbone system and 800 MW on the Rush Creek generation tie. Using the information

15 in Appendix I, that should lead to $110-$125 million in transmission upgrade costs.20 But

16 there should also be an additional $10 million to mitigate “flicker” in the Pueblo area due

17 to the retirement of Comanche 1 and 2.21 Thus, according to Appendix I, the transmission

18 upgrade cost for the CEP should be $120 - $135 million. The Preferred ERP portfolio

19 shows 749 MW of new generation and 660 MW of Comanche 1 and 2 on the 345 kW

20 backbone and 499 MW on Rush Creek. Using Appendix I, that should give $132 million

21 in transmission upgrade costs, so even using PSCo’s approach, the Preferred ERP

20
$100 million for upgrades to the system backbone plus $10 million for upgrades to Rush Creek.
21
Appendix I at p. 1.
Cross-Answer Testimony of Charles S. Griffey
Page 26
 1 portfolio’s transmission cost appears overstated and the Preferred CEP portfolio’s

2 transmission cost appears understated.

3 Q. DO YOU THINK THE COSTS ASSOCIATED WITH SYSTEM UPGRADES ARE

4 MODELED CORRECTLY?

5 A. No. First of all, the Preferred ERP portfolio has more generation than is needed, with a

6 17.7% reserve margin in 2023. Further, the network transmission upgrade cost was not

7 included in the optimization of determining how much generation to add. PSCo simply

8 modeled a range of network upgrade costs for 200-800 MW of new generation if

9 Comanche 1 and 2 continue to operate.22 PSCo’s cost estimates show that simply dropping

10 only 50 MW from the Preferred ERP portfolio would eliminate $32 million in network

11 upgrade costs under PSCo’s approach. Further, PSCo ignores that there is a ten year

12 difference in the Comanche 1 and 2 retirement dates in the Preferred CEP and Preferred

13 ERP portfolios. The timing difference should be reflected in the network upgrade cost

14 comparison, but it is not. Rather, PSCo assumes that the full cost of network upgrades are

15 needed for the Preferred ERP portfolio, but not for the Preferred CEP portfolio. This does

16 not make sense.

17 Q. PLEASE SUMMARIZE YOUR FINDINGS FROM THE ACTUAL BID DATA.

18 A. The assumption that the CCGT hard-coded to replace Comanche 1 and 2 will incur $100

19 million in transmission interconnection cost is totally unsupported by and in conflict with

20 the transmission interconnection cost used by PSCo to evaluate the bids and to create the

21 Preferred CEP. Based on what PSCo used to evaluate the bids, it is reasonable to believe

22
120 Day Report at Appendix I.
Cross-Answer Testimony of Charles S. Griffey
Page 27
 1 that the capital and transmission cost assumed for the CCGT replacing Comanche 1 and 2

2 are too high, and that savings calculations based on assumptions are likely too high by over

3 $100 million. Further, on a portfolio basis, PSCo’s approach greatly overestimates the

4 cost of transmission in the ERP portfolio compared to the CEP portfolio, $291 million

5 compared to $186 million. It does this by, in addition to overestimating the cost to

6 interconnect a new CCGT in 2034 relative to what it assigns to the bids in this case,

7 overstating network transmission upgrades in the ERP cases, and by apparently

8 understating the cost of network upgrades in the CEP cases.

9 Q. DOES THE ACTUAL BID DATA SUPPORT MR. COMINGS’S CLAIM THAT IT

10 IS UNECONOMIC TO CONTINUE OPERATING COMANCHE 1 AND 2?

11 A. No. Using the actual bids and portfolios preferred by PSCo as a result of the solicitation,

12 PSCo’s own calculations show that continued operation of Comanche 1 and 2 through their

13 planned retirement dates will save customers hundreds of millions of dollars on both a

14 nominal and present value basis. When properly accounting for RESA impacts/accelerated

15 depreciation, even under PSCo’s modeling the Preferred CEP portfolio does not break even

16 until after twenty-five to thirty years into the future, and the calculated NPV of savings is

17 only $42 million. These small present value benefits shown by PSCo’s modeling are very

18 uncertain and appear to be driven largely by assumptions made as to the capital and

19 transmission costs of the generic CCGT assumed to replace Comanche 1 and 2, other

20 transmission cost, and assumptions regarding degradation or the lack thereof in wind, solar,

21 and battery output. Any present value savings shown in the modeling are very suspect as

22 a result and likely will not occur. I agree with Mr. Comings that there are many coal units

Cross-Answer Testimony of Charles S. Griffey

Page 28
 1 across the United States that are uneconomic to continue operating, but there is no data to

2 indicate that Comanche 1 and 2 are among those units.

3 III. CONCLUSION

4 Q. DO YOU HAVE ANY CONCLUDING REMARKS?

5 A. When the recovery of the accelerated depreciation is appropriately included in the

6 calculation, the early retirement of Comanche Units 1 and 2 will cost ratepayers $340

7 million on a nominal basis and $190 million on a present value basis through 2033 using

8 PSCo’s own figures. When incorporating other RESA impacts, those figures become $468

9 million and $255 million, respectively. Any benefits after that time are very speculative

10 and are driven solely by modeling assumptions about resources that are unlikely to ever

11 actually be built. In any case, even using what I consider to be assumptions that bias against

12 the Preferred ERP portfolio, the Preferred CEP portfolio does not break even on a present

13 value basis until the year 2046.

14 Furthermore, if, as PSCo suggests, solar prices are expected to continue to decrease, then

15 there is no reason to make a decision to retire Comanche 1 and 2 at this time. They are

16 clearly more economic to operate than the alternatives in the near-term, and the economic

17 decision would be to defer retirement, operate the two units, and wait for solar costs to

18 continue their decline. In such a situation, the RESA could be lowered, thus giving

19 ratepayers a rate decrease. Ratepayers would also get the benefit of the lower costs

20 associated with operating Comanche 1 and 2 and the production tax credit (PTC) benefits

21 and low operating cost associated with the nearly 800 MW of new wind in the ERP

22 portfolios. They would not be saddled with the higher reserve margins and certain high

23 capital costs associated with the Preferred CEP portfolio.

Cross-Answer Testimony of Charles S. Griffey
Page 29
1 Q. HAS PSCo’S 120 DAY REPORT ALTERED YOUR POSITION REGARDING

2 WHETHER THE CEP AND THE ASSOCIATED COST RECOVERY

3 MECHANISMS PRESENTED IN THIS PROCEEDING ARE IN THE PUBLIC

4 INTEREST?

5 A. No, the Preferred CEP places excessive risk on ratepayers and largely insulates PSCo’s

6 shareholders from risk. Given the real costs to ratepayers and the lack of tangible benefits,

7 it is not in the public interest.

8 Q. DOES THAT CONCLUDE YOUR TESTIMONY?

9 A. Yes.

Cross-Answer Testimony of Charles S. Griffey

Page 30

20884493v.8