Loading...
HomeMy WebLinkAbout20210423Clearwater to Avista 1-3.pdfAVISTA CORPORATION RESPONSE TO REQIIEST FOR INFORMATION RECEIVED 202|April 2j, AM 10:21 IDAHOPUBLIC UTILITIES COMMISSION JTJRISDICTION CASE NO: REQIJESTER: TYPE: REQUEST NO.: IDAHO DATE PREPARED: O4/1012021 AW-E-21-01/AW-G-21-01 WTINESS: ElizabethAndrews Clearwater Paper RESPONDER: Paul Kimball Production Request DEPARTMENT: Regulatory AffairsCP-001 TELEPHONE: (s09) 49s-4s84 REQUEST: Please provide, in electonic format with all formulae intact where possible, all workpapers and other documents used in the development of Avista's Application in this matter. RESPONSE: Avista will be providing all case filings and workpapers to Mr. Peterson using the OneDrive application. If you would like to also receive the documents in this form please provide me with an email address to receive the files. ruRISDICTION: CASE NO: REQUESTER: TYPE: REQUEST NO.: AVISTA CORPORATION RESPONSE TO REQUEST FOR TNFORTVTATTON IDAHO DATE PREPARED: 0411012021 AVU-E-21-01 / AW-G-21-01 WITNESS: Elizabeth Andrews Clearwater Paper RESPONDER: Paul Kimball Production Request DEPARTMENT: Regulatory AffairsCP-002 TELEPHONE: (s09) 495-4584 REQUEST: Please provide copies of all communications with the Idatro Public Utilities Commission and its Staffregarding Avista's Application in this matter. RESPONSE: The Company has provided all communications with the Commission to date and will continue to provided them. Question No. 1: In Response to Staff s Production Request No. 56, Avista stated that the Company does not apply integration on a per-MWh basis, but instead charge $1.067/kW-month. The rate used is contained in Sta[PR_056 Attachment A - "Avista 2007 Wind Integration Cost Summary (PR56).xlsx." Please answer the following questions. a. Please explain why the Company does not apply integration on a per-MWh basis, but on a per kW-month basis. The 2007 Wind Integration study was in $/MWh. However, since then we learned that the cost is really driven by the size of the project not the energy, therefore we now apply this on a per kW-month basis as opposed to a per MWh basis. Please refer to an update version PR56 I sent to Ms. Yin on March 29 (LIVE CALCS VERSION StaflPR_056 Attachment A - Avista 2007 Wind Integration Cost Summary (PR56).xlsx) that demonstrates the sources and calculations to arrive at the rates we use. b. Column E of Sta[PR_056 Attachment A - "Avista 2007 Wind Integration Cost Summary @R56).xlsx." lists different integration charges at different wind penetration levels. Please explain how the values in Column E were determined. The values in Column E were taken from the wind study. The study explains how as pe,netration of wind grows, integration costs rise. Each case has a level of wind penetration as described and therefore has increased reserve obligations that translate to higher integration costs. c. Were the values in Column E calculated in the 2007 Wind Integration Study? If so, please reference the page numbers. Yes, but they were published only on a per-MWh basis. To determine the rates, Attachment A of PR 56 uses the Low Market Prices case Integration Costs on a per- MWh basis from Table 27 on Page 50 of the Wind Integration Study and "translates" them to per-kW using the capacity factor for that wind penetration scenario from Table l0 on page 16.. The Low Market Prices case is used because current market prices more closely match the Low Market Prices case d. The note in Cell F7 states that Avista has more than 200 MW of wind on its system starting n2021. Please provide Avista's current wind penehation level in MW. Avista's current wind penetration level is 265.75 MW, comprised of its Palouse and Rattlesnake projects. e. Please explain why Avista does not use $0.994lkW-mo, the integration charge at the 200 MW penetation level, but use $1.067/kw-month, the integration charge at the 400 MW penefration level. Integration costs increase as wind penetration crosses the thresholds studied. Because our portfolio now contains more than 200 MW of wind, we use the higher incremental cost. Wlnd lntegratlon Study Results has morc than 200 MW of wlrd on lts sy$em startlng ln 2021 f. Has this $1.067/kW-month integration charge used in this case? If so, please explain how and where this number has been used. Avista does not model integration costs in its power supply cost proforma. We use the estimate when comparing variable generation resource options as they enter our portfolio to ensure we are comparing on an "apples to apples" basis with other resource options. Integration costs are reflected in the case substantially by the use of a de- optimized S-year historical shaping of our hydro operations. In other words, absent wind and solar variability our hydro facilities likely would have been able to generate on average more power across higher value peak periods than they did with wind in our portfolio. This de-optimization is exactly what the 2007 Wind Integration Study endeavored to calculate. g. Please explain why Avista does not use the integration charges approved in Order No. 30500, which is based on the Settlement Stipulation in the 2007 case (AW-E-07-02). Order 30500 has integration costs based on 2007 market conditions and a period where Avista had no wind on its systern. Therefore the $2.75lMWh rate was applicable. Since that time we have added significant wind resources and market prices have changed. The 2007 Wind Integration Study was intended to provide scenarios keeping its work relevant under various future conditions. Given all of the changes, it could be appropriate to modifu integration charges for QF facilities; however our preference would be to await the outcome of our presently VER integration study results. Prkc Prke Isrntrrld SEe NCF S/Mwh , $/tw+no S/hr-rno 03231m 2m :!l.o9a 33.99a 0.9ca /t00 30.516 1.!1 0.328 2.57t 0.661 3.E81 0.854 6m :10.593 3.98: 0.885 O3r:11 Question No. 2: Line 149 in Mr. Kalich's workpaper "Transmission Expense - Account 565" includes an adjusfrnent of $92,500 for BPA PTP for Colstrip, Coyote Spring 2, Lancaster. Please explain what the adjusfrnent is. This BPA PTP for Colstrip, Coyote Springs 2, Lancaster amount is for the additional 50 MWs for Coyote Springs 2 which were necessary due to plant enhancements that increased capacity. However, this additional amount was denied due to constraints in the Tri-Cities area. This amount will be removed from the pro-forma period when forwarded prices are updated and before the rate case is finalized. Question No. 3: Please walk me through the "Conf Gas Contracts MTM" tab in Exhibit No. 9. Question No. 4: Line77 of Schedule 2 of Exhibit No. 9 contains 'oSurplus AECO to Malin Transportation". a. Please explain why it is determined by the difference between the'oConf Fuel Costs" tab and "Conf Gas Contracts MTM" tab. b. Please explain why the 2019 acfial ($53,356,000) is significantly greater than the pro forma amount ($5,0 I 7,000). Question No.3: Please walk me through the "Conf Gas Contracts MTM" tab in Exhibit No. 9. Will be addressed on April 6 call Question No. 4: Lne77 of Schedule 2 of Exhibit No. 9 contains "Surplus AECO to Malin Transportation". a. Please explain why it is determined by the difference between the "Conf Fuel Costs" tab and "Conf Gas Contracts MTM" tab. Will be addressed on April 6 call. b. Please explain why the 2019 actual ($53,356,000) is significantly greater than the pro forma amount ($5,0 I 7,000). Will be addressed on April 6 call. Question No. 5: Page l8 of Mr. Kalich states that Avista does not model index contracts since they do not impact power supply costs. There is only one such contract in this year's filing, the 2021Morgan Stanley Renewable Energy Credit (REC) sale. However, Production Request Response No. 73 lists six index deals in the following table and states these index deals are entered into Aurora as index so these contracts are automatically updated when new forward prices are calculated and input into Aurora. lndex 277274 Morgan Stanley CapitalGroup lnc. UortfrWestern Energy PacifiCorp euget iound Energy, lnc. Talen Energy Montana, LLC 277544 lndex 277545 286577 Morgan Stanley CapitalGroup lnc. Morgan Stanley PCC-1 REC deal tndex lndex a. Please reconcile the two statements and explain which deal(s) are considered in this rate 277546 z.tls+l Nichols tticttoit rVi.f',of s Nichols Pump index Pump index Pump index Pump index sale sale sale saleln ln dex J"* DescriptionDealTyp KEY Counterparty case. Nichols pumping (one contract with multiple Colstrip owners shown above as 4 individual keys) is included in the pro forma as a contract as opposed to a term deal. See line 62 in the pro forma in 'Schedule 2' and under contracts in 'Schedule lC'. The Morgan Stanley deals are excluded from the pro forma period (see Kalich testimony page l9line 8) and probably shouldn't have been included in response to PR73. b. Which deal(s) have impacts on power supply costs? These have an indirect impact on authorized power supply costs because the Nichols Pumping contracts enable us to purchase less non-firm Northwestem transmission for delivering Colstrip output to load. c. For the deal(s) that do not have impacts on power supply costs, are they included in annual the PCA filing? The Morgan REC deals flow through the PCA filing in actuals. d. By "Morgan Stanley index sale" on Schedule 2 of Mr. Kalich's Exhibit 9, does Avista refer to "Morgan Stanley-Clearwater REC deaf' or "Morgan Stanley PCC-I REC deal" or both? By Morgan Stanley index sale - Avista is referring to both. e. Are all six deals based on Mid-C prices? The Nichols Pumping conffact price is discounted $0.50 to reflect the lower value of power there relative to Mid-C prices. f. What does'Nichols Pump index sale" mean? Each Colstrip owner provides a prorated share of the energy needed for cooling water pumped from an adjacent river at the Colstrip facility. Avista supplies all pumping power for those utilities at the plant in exchange for index-based payments. Question No. 6: Schedule 3 of Mr. Kalich's Exhibit 9 includes a note for "Douglas PUD": Contract has no direct power supply impact. Reflected in beneficial impact on shape of portfolio hydro operations in Aurora model. a. Please explain why it has no direct power supply impact. There is no direct power supply impact but due to the energy exchange at different times, there is an indirect impact to power supply costs. b. Please explain what it means by "reflected in beneficial impact on shape of portfolio hydro operations in Aurora model". The power we get from Douglas is scheduled for delivery mostly during on-peak hours with Avista controlling when we acquire the energy. Energy is returned on a flat product schedule. This exchange benefits customers. Question No. 7: Schedule 3 of Mr. Kalich's Exhibit 9 includes a note for "Pend Oreille PUD": Not included in rate period proforma. However, $5000 is included in the pro forma period on Schedule 2. Please reconcile. The "not included in rate period pro forma" note on Schedule 3 was included in error. While the monthly amounts appear to be zero, there is a total of $5,000 included in the annual pro forma period. Question No. 8: Response to Question No. I states that "Order 30500 has integration costs based on2007 market conditions and a period where Avista had no wind on its system. Therefore the $2.75l\{Wh rate was applicable." Please explain why $2.75 is used in the context of Order No. 30500. (The order uses three percentages with a cap of $6.50/MWh.) This was a mistake. The reference to $2.75lMWh reflected the study base case assuming our then-low level (i.e., zero MW) of wind penetration. And to contrast that base case with the new case reflecting lower market prices and higher system penetration. We have never needed to apply the charge to a QF contract, but would use this math were we needing to. Tier I Tier 2 Tier 3 Amount of Wind Online 0 to 199 MW 200 to 299 MW 300 MW and above Integration Charge (cap) 7o/o ($6.504.1Wh) 8% ($O.SoA{Wh) e% ($0.s04{wh) Question No. 1: Please answer the following questions regarding "Short-Term Market" under Account 555. a. Please explain why there is a significant reduction from its 2019 actual value of 40,814 thousand dollars to its pro forma value of 3,201thousand dollars. The drop fiorn 2019 reflects reduced load (-48 aMW), greater hydro (+64 aMW), and more VER ( 1 50 aMW) offset by fewer contract resources (-55 aMW). So the net is tnore than I 00 aMW of additional energy in the system. This large change brings actual purchases down greatly along with a commensurate reduction in short-term market purchases. Comparison of Project Generation vs. Actual Modeled GWh b. Is the pro forma value of 3,201 thousand dollars calculated based on the purchase from Mid-C Market Resource created under the single-zone method? Yes c. Does the pro forma value of 3,201thousand dollars match the Company's experience in reality? If yes, please provide evidence. If not, please explain how to address the issue. Please see the response to a. above. 2020 m21-2,2 20r9 2019 Case 2o21 Ca*Delta to 2019 Item Actual Proiected Actual aMW Load g.{t2 I,111 9,346 055 fttt Clark Fork 2.565 2,771 2.599 I 188 21 17ipokane9551,073 1,051 1,103 16 Mid-C 1,006 1,231 1,273 233 26 Total Hvdro 4.52t's.075 4,923 089 563 u 't.$2 1.220 1.575 nColstrip1,625 cs2 1,891 1,984 1,767 2,120 229 26 Lancaster 1,798 1,707 1,6E5 1 5E6 (24 Kettle Falls 316 312 265 299 17 (2 265 22',1 221 355 90 10Gas Peakers Total Thermal 5.852 5,849 s,158 5 9 Palouse wind w2 315 370 316 14 2 Rattlesnake Wind 35 37 /169 /t69 * 42 45 qz (5Lind Solar 41 5lTotal VER yt4 350 452 Small Power/C.oqen 560 232 672 26 (/t0 WNP-3 173 (20 (504 e,47 5Other LT Contracts (616)(446) Other ContracG 49 (384)n6 231 (5i ST Purchases 1,42 177 1,W2 114 (152 ST Sales Q.t 42 (1,917)(2.796)Q,314 36 fi.5{t01 fi.7401 n.4141 fi.014)(1t61Market Transactions Station Service (101 m m fil Total Supply 9.171 I,140 9,345 9,057 14111 (47) EE@trI Question No. 2: Please explain whether "Lancaster PPA" includes o'Lancaster Heat Rate Tracker" (an item included in AVU-E-19-04 for costs associated with difference between contractual and actual heat rate efficiency) in both the 2019 actual value of 28,141 thousand dollars and the pro forma value of 28,467 thousand dollars. If not, please explain why the fracker is not included. Yes, the Lancaster PPA does include the former line item "Lancaster Heat Rate Tracker". See also sheet 'Schedule 3' for line item 6 - Lancaster PPA includes contract costs of capital, O&M and heat rate tracker. Question No.3: Please define and explain what "BPA Point-to-Point for Colstrip, Coyote Springs 2 &Lancaster" is under Account 565. Transmission expense based on contracted capacity at the tariffed rate plus a 3olo escalator effective October 1,2021. This line item includes the 50 MW contract for Coyote Springs 2 as well. We will remove 50 MW CS2 proforma expense for final rates as BPA did not award us the transmission. See also 'Schedule 3' for line item 46. Question No. 4: Please explain what "WNP-3" under Account 555, "WA WNP3 - reconciling items (not in ERM/PCA)" under Account 555, and "WNP-3" under Account 565 are. In addition, please explain why the pro forma value for each one of thern is 0. WNP-3 is a contract that expired June 30,2019. It was included in the pro forma as 0 since it was expired and wouldn't occur in the pro forma period, however, the line item was included because it was in 2019 actuals. See also 'Schedule 3' for line item 44. Question No. 6: Mr. Kalich's confidential workpaper "Fuel Costing worksheet notes" states that "we have capacity rights of 69,388 Dth/day from Kingsgate to points south-26,388 of which we would like to sell at Malin-leaving 43,000 to utilize at the plants." Please explain how 26,388 Dth/day is determined. This is Avista's contracted GTN capacity rights - 60,592 dekatherms per day for AECO- Kingsgate, 43,000 dekatherms per day for Kingsgate to Stanfield, and 26,388 dekatherm per day for Kingsgate to Malin. Question No. 7: Mr. Kalich's confidential workpaper "Fuel Costing worksheet notes" states that o'we buy 8,796 at Kingsgate each day giving us a total of 69,388 Dth/day to nominate." Is the8,796 Dth/day based on a contract? If not, please explain how the amount is determined. For the benefit of our customers, to fully utilize our transportation from AECO to Malin, this intermediary pipeline path is contracted for to complete the path for our rights. Question No. 8: Please answer the following questions regarding the tab "Conf Aurora Fuel Output" of Mr. Kalich's Exhibit 9. a. When the AURORA model dispatches natural gas plants based on natural gas prices, does the model consider fuel loss on pipelines and transportation costs of each plant? Yes it does. Both items are reflected in the cost of fuel delivered to the plants. b. If not, does the model have the capabilities to do that or is it the modeler's decision to not consider fuel loss on pipelines and transportation costs of each plant in dispatch? N/A c. If it is the latter, please explain why the modeler makes that decision. Costs for losses and transport to plants are based on line items identified in the transportation tariff (e.g., mileage-based fuel losses) and/or known taxes (i.e., Washington -3.873o/o tax on end-use natural gas consumption). d. When the AURORA model dispatches natural gas plants, does the model consider the consfiaint ofpipeline capacities below, which is mentioned on the "Conf Fuel Costs" tab? Aurora dispatches the plant based on the Malin gas price. However, the gas is re-priced based on lower-cost AECO gas in "Conf Fuel Costs" tab of file "Kalich Exhibit 9, Schedule l C.xlsc" GTN capacity(dth/dav) AE-KG KG-ST KG-MA 60,592 43,000 26.388 e. Please explain why Rathdrum_l is dispatched less frequently than Rathdrum_2 during the pro forma period. Consistent with how our Rathdrum units are dispatched and as stated in testimony on page l7 line 4, only one unit is dispatched at a time in order to cover unanticipated outages. Therefore, in this rate case, we constrained the Rathdrum units to only allow one to run at a time. This is different from how we modeled it in the last rate and explained in testimony. f. Please explain why Northeast_A and Northeast_B are not dispatched during the pro forma period. Northeast, even if cost-effective to run relative to market prices, is limited to 100 hours per year due to regulation by the Spokane Air Pollution Control Board. ln testimony on page l7 line 4, the Company explains how it holds the units back for emergency or near- emergency operations. This is different from the last rate case as we allowed the Northeast units to dispatch up the 100 hour limit. g. What is Lancaster_DB? What is the difference between Lancaster_DB and Lancaster and why is the Lancaster_DB dispatched less frequeirtly than Lancaster? The duct burner has a substantially higher heat rate than the main unit; so, given the higher heat rate, it dispatches less. h. Please explain why the Start Fuel Usage by Day is zero for the following plants: Northeast_B, Northeast_A, Lancaster_DB, Kettle_Falls_CT_CCCTMode, Coyote_Springs_2_DB, Boulder_Park_l, Boulder_P ark 2, Boulder_Park_3, Boulder_Park_4, Boulder_Park_5, and Boulder_Park_6. There was no start fuel associated with these units. For example, the Northeast units didn't dispatch as the 100 hours were saved for reserve as spoken to in testimony. Reciprocating engine and duct fire units are quick start and don't require warrn up time; therefore, no start up fuel is used. i. ForNortheast B, Kettle_Falls_CT_CCCTMode, and Boulder_Park_5, please explain why every day in the pro forma period has zero Start Fuel Usage by Day, but the totals in Line 387 are 90.30641 thousand dekatherm/day, 55.18849 thousand dekatherm/day, and 47.52767 thousand dekatherm/day. How are the grand totals calculated? The grand totals were hard-coded values from a previous version. This row should have been removed. However, it should be noted that the grand totals are not used for any of the calculations, so the inclusion did not change the outcome of the results. j. It appears for some plants, Start Fuel is higgered once if the plant generates continuously, such as Lancaster and Coyote_SpringsJ, while for other plants, Start Fuel is higgered multiple times even if the plant generates continuously, such as Rathdrum_2. Please explain how each plant's Start Fuel is calculated. The work paper shows the summarized daily startup costs. For example, Coyote Springs 2 may start and continuously run but only show the one starfup. For Rathdrum, the plant may start I or 2 times each day and shut off, therefore the data only shows what happened each day. The hourly Rathdrum data shows how the plant dispatches across the day. The start fuel is calculated by for each start a fuel quantity is consumed and then priced at the local price of natural gas. We can provide the hourly data if you'd like to explore this further. Question No.9: Please explain why the calculations on the "Conf Fuel Costso'tab of Mr. Kalich's Exhibit 9 only considers the GTN pipeline, not other pipelines such as the Northwest pipeline. Also, why do the calculations only consider Kingsgate, Stanfield, and Malin, and not other ffading hubs such as Sumas. The Thermal side of the business has rights to pipeline capacity solely on the TransCanada/GTN pipeline, therefore Thermal only purchases at market points along the Trans Canada/GTN pipeline: AECO, Kingsgate, Stanfield and Malin. Question No. 10: The note for Cell Ela @ipe) states that "Kl - use King-Stan capacity, King segment. S - use King-Stan capaaty, Stan segment. K2 - use King-Malin capacity, King segmento'. Please confirm that King segment and Stan segment mean the gas is purchased at Kingsgate and Stanfield, respectively and that the pipeline capacity at each segme,nt is the same, which is 43,000 dekatherm/day. That is correct. The King-Stanfield segment can move up to 43,000 (all units are Dth/day) either through forward or backward haul. Look at January 23,2022 as an example. Boulder Park requires 2,301 and Lancaster requires 46,043 for a total of 48,343 Dth/day. We forward haul 43,000 from Kingsgate to the plants (cell EUl T) and purchase the remaining 5,343 from Stanfield and backhaul it to Lancaster and Boulder Park (cells EUlS &EUzl) because the 43,000 forward haul has been exhausted. CS2 requires 51,779 which comes from a combination of 43,000 (cell EU30) purchased and forward hauled from Stanfield (the 43,000 segment from Stanfield to CS2 was not utilized by the other plants-known as pipeline segmentation), plus 8,779 of the Kingsgate to Malin 26,388 capacity that we wanted to sell at Malin, but in this case need for CS2 (cell EU3l ). Question No. 11: Please answer the following questions related to the Gas Nomination Qty section on the "Conf Fuel Costs" tab of Mr. Kalich's Exhibit 9. a. Please confirm that the Stanfield Hub can deliver gas to the north and to the south at the same capacity level at the same time. Yes, true. The pipeline has rules for back hauling, none of which are violated in this modeling. b. Please explain why there is no surplus amounts sold at Stanfield in Row 33. c. Surplus gas in the Kingsgate-Stanfield segment is sold (on Row 33) when the demand at the gas plants is reduced, as is typical in Q2 (cell JB33 for example). There are 56 days in the test period where surplus gas is sold at Stanfield. When Row 33 is zero, there is no surplus gas. Question No. 12: Please explain why a "buy Stan - no burn" scenario is not considered. A "buy Stan-no burn" scenario is buying gas at Stanfield and selling it at Malin. We have rights to 26,388 Dth/day between AECO and Malin that is 100% utilized each day, so there is no additional capacity to move gas from Stanfield to Malin. We would always buy AECO rather than Stanfield in using the 26,388. Question No. 13: Please explain why "buy King -burno'on September l, 2021is -15 thousand dekatherm. There do not appear to be any negative values in Row 38 in the exhibit as filed nor should they ever be. Question No. 14: Please explain what the numbers in the green box on the "Conf Gas Contracts MTM' represent and how these numbers are calculated. MTM, or mark-to-market, is the position of the financial trades made for the Thermal side of the business, with positive values representing expense (out of the money) and negative values representing revenue (in the money). The detail behind the numbers shaded in green are in Column O. For example, the deal in Row 5 shows a basis sale of Malin at a price of -$0.5375. The MTM basis price, based on the prices used in the Case, is -$0.0463. The difference in the basis is -$0.4912, multiplied by 75,000 Dth, and that deal is $36,843 of expense. Question No. 15: Under the single-zone method, does the total load in the zone (which includes the Mid-C Market Load) have to be met by the total resources in the zone (which includes the Mid-C Market Resource) so that demand is equal to the supply? No. As explained in testimony on page 9 starting at line 13, the total load just needs to be big enough to absorb all potential surplus sales from Avista resources when they are lower cost to operate than the market price of power. ruRISDICTION CASE NO: REQUESTER: TYPE: REQUEST NO.: AVISTA CORPORATION RESPONSE TO REQUEST FOR TNFORMATTON IDAHO DATE PREPARED: 0411412021 AW-E-21-01 / AW-G-21-01 WITNESS: Elizabeth Andrews Clearwater Paper RESPONDER: Paul Kimball Production Request DEPARTMENT: Regulatory AffairsCP-003 TELEPHONE: (s09) 49s-4s84 REQUEST: Please provide copies of all responses to production requests (both formal and informal) provided to any other party to this proceeding. RESPONSE: Please see Avista's response 003C, which contains TRADE SECRET, PROPRIETARY or CONFIDENTIAL information and exempt from public view and is separately filed under IDAPA 31.01.01, Rule 067 and233, and Section 9-340D, Idaho Code. The Company has provided all production requests from all parties to date and will continue to provide therr. See CP_PR_003 Attachments A-G and CP_PR_003C Confidential Attachments A & B for the informal responses provided to Staff.