20200508Avista to Staff 10-38.pdf

AVISTA CORPORATION RESPONSE TO REQUEST FOR INFORMATION

JURISDICTION: IDAHO DATE PREPARED: 05/04/2020 CASE NO: AVU-E-19-01 WITNESS: Clint Kalich REQUESTER: Staff RESPONDER: James Gall TYPE: Production Request DEPARTMENT: Energy Resources
REQUEST NO.: Staff-010 TELEPHONE: [phone] REQUEST: Please explain how changing from a 20-year IRP time frame to a 25-year time frame affects the
portfolio cost and resource selections. RESPONSE: Going to a 25-year time period does not change the portfolio cost, nor resource selections. Avista
has conducted a 25-year analysis of the portfolio as part of its last several IRPs, however only
presented the 20 year analysis in its IRP. Avista made this change in the past to account for end of period effects in the 20-year time horizon, this change was discussed in past IRP technical advisory committee meetings.
RECEIVED
2020 May 8AM9:32
IDAHO PUBLIC
UTILITIES COMMISSION
AVISTA CORPORATION RESPONSE TO REQUEST FOR INFORMATION

JURISDICTION: IDAHO DATE PREPARED: 05/04/2020 CASE NO: AVU-E-19-01 WITNESS: Clint Kalich REQUESTER: Staff RESPONDER: James Gall TYPE: Production Request DEPARTMENT: Energy Resources
REQUEST NO.: Staff-011 TELEPHONE: [phone] REQUEST: In reference to: “Moving to 25 years led to removing some of the cost estimates for resources
beyond 20 years.” IRP at 2-12. Please provide the cost estimates that were removed and explain the effect of removing these costs. RESPONSE:
The reference to removing the cost estimates for resources beyond 20 years has two components.
The first component, as part of past IRP processes, Avista modeled only 20 years in Aurora, however created a portfolio for 25 years as described in Avista’s response to Production Request 10. Since this data for the last five years of the plan was estimated using a trend analysis; this data was replaced with modeled data in the 2020 IRP. The second component was the removal of
selected resource costs beyond 25 years in the objective function. In this case, the prior IRPs
included resource fixed costs beyond 25 years, these costs were removed for model simplification as they did not have a material effect on results due to present value discounting.
AVISTA CORPORATION RESPONSE TO REQUEST FOR INFORMATION

JURISDICTION: IDAHO DATE PREPARED: 05/04/2020 CASE NO: AVU-E-19-01 WITNESS: Clint Kalich REQUESTER: Staff RESPONDER: James Gall TYPE: Production Request DEPARTMENT: Energy Resources
REQUEST NO.: Staff-012 TELEPHONE: [phone] REQUEST: In reference to: “Avista assigned peak credits to renewable and storage resources depending on
their ability to meet peak loads using its Avista Reliability Assessment Model (ARAM).” IRP at 2-11. Please explain how the ARAM model assigns peak credits and how this method is different than past IRPs.
RESPONSE:
Avista uses an Effective Load Carrying Capability (ELCC) method to determine the amount of dependable capacity a resource provides customers during peak periods. In this case referred to as Peak Credit. To determine a resource’s Peak Credit, first the ARAM model simulates the Avista portfolio to achieve a 5 percent Loss of Load Probability (LOLP) using existing resources plus
new natural gas CTs.
To test a resource’s Peak Credit, a portion of the natural gas CTs are removed and replaced with the resource under consideration until the LOLP reaches 5 percent. For example, if 20 MW of CTs are removed, but takes 900 MW of solar to replace this capacity to achieve a 5 percent LOLP, the
ELCC or Peak Credit is 2.2 percent (20/900). Avista covered ELCC methodology at the 2nd Technical Advisory Committee Meeting. Further, this methodology was requested by many of the Committee members after the 2017 IRP process.
Avista agreed with the change because it was giving no Peak Credit to solar and local wind and 7.5
percent for Montana wind. For storage resources full capacity credit was given (Avista had only modeled small amount with three and six hour duration). Avista’s prior methodology for local wind was to give it zero value to due to its non-dispatchable
characteristic and Avista had only a single project (not a diversified portfolio of wind). Avista now
gives local wind a five percent Peak Credit using the new methodology. For solar, a Peak Credit of zero was given previously because it is dark when Avista typically peaks, however now the resource receives two percent based on the new methodology. Lastly for Montana wind, in the prior IRP, a review of actual production at a single wind facility in Montana was compared to
Avista peak load hours. The 7.5 percent reflects the average capacity factor in these peak hours.
The 2020 IRP gives these resources 36 percent. In the 2017 IRP storage was given 100 percent Peak Credit since a sound methodology wasn’t available at the time to determined its benefit to the system, and available storage was limited in size and scope and had little impact on the previous IRP results. The 2020 IRP methodology calculated a Peak Credit of 15 percent for four-hour
storage and page 9-27 of the IRP includes peak credits for other storage durations, Avista does not limit storage quantities in this IRP.
AVISTA CORPORATION RESPONSE TO REQUEST FOR INFORMATION

JURISDICTION: IDAHO DATE PREPARED: 05/04/2020 CASE NO: AVU-E-19-01 WITNESS: Clint Kalich REQUESTER: Staff RESPONDER: James Gall TYPE: Production Request DEPARTMENT: Energy Resources
REQUEST NO.: Staff-013 TELEPHONE: [phone] REQUEST: Please provide the workpapers used to create Peak Credit table 9.11 in Excel format with formulas
enabled. RESPONSE: Please see Avista's response 013C, which contains TRADE SECRET, PROPRIETARY or CONFIDENTIAL information and exempt from public view and is separately filed under
IDAPA 31.01.01, Rule 067 and 233, and Section 9-340D, Idaho Code. Avista conducted over 140 studies in preparation of Table 9.11; attached in Staff-PR-013C Confidential Attachment A which includes a log of these studies and calculations of peak credits.1
Not all studies were retained due to the quantity and size of the files. Staff-PR-013C Confidential
Attachment B is the Preferred Resource Strategy study as an example of the ARAM model.

1 As noted in the 2020 IRP document, Avista has not conducted ELCC studies for hydro resources at this time.
AVISTA CORPORATION RESPONSE TO REQUEST FOR INFORMATION

JURISDICTION: IDAHO DATE PREPARED: 05/04/2020 CASE NO: AVU-E-19-01 WITNESS: Clint Kalich REQUESTER: Staff RESPONDER: James Gall TYPE: Production Request DEPARTMENT: Energy Resources
REQUEST NO.: Staff-014 TELEPHONE: [phone] REQUEST: Please describe how ARAM validates resource adequacy and resource peak contributions.
RESPONSE: Resource Adequacy: ARAM validates if Avista’s resource portfolio is adequate by conducting 1,000 simulations using
historical temperatures, hydro conditions, wind shapes, and forced outages to determine whether
or not the portfolio can serve 100 percent of load and reserve requirements in approximately 950 of the simulators or greater (i.e. 5 percent LOLP). Resource Peak Contributions:
To determine a resource’s peak credit, first the ARAM model simulates the Avista portfolio to
achieve a 5 percent Loss of Load Probability (LOLP) using existing resources plus new natural gas CTs. To test a resource’s peak credit, a portion of the natural gas CTs are removed and replaced with the
resource under consideration until the LOLP reaches 5 percent. For example, if 20 MW of CTs are removed, but takes 900 MW of solar to replace this capacity to achieve a 5 percent LOLP, the ELCC or Peak Credit is 2.2 percent (20/900).
AVISTA CORPORATION RESPONSE TO REQUEST FOR INFORMATION

JURISDICTION: IDAHO DATE PREPARED: 05/04/2020 CASE NO: AVU-E-19-01 WITNESS: Clint Kalich REQUESTER: Staff RESPONDER: James Gall TYPE: Production Request DEPARTMENT: Energy Resources
REQUEST NO.: Staff-015 TELEPHONE: [phone] REQUEST: Related to reliability analysis, please respond to the following:
a. Please explain why 2030 was chosen as the year to test reliability.
b. Please explain why the Company did not perform its reliability analysis over multiple years. Please explain the difficulties, benefits, and costs to perform of performing this type of analysis over multiple years.
c. Please explain why the Company did not perform its reliability analysis on multiple or all
resource portfolios. Please explain the difficulties, benefits, and costs to perform this analysis on all or multiple resource portfolios.
d. In Audit Request Response No. 16, the Company stated, “Future IRPs may contain broader reliability analysis.” Please provide additional detail on the Company’s plan for a broader
reliability analysis, which includes details on the purpose, objectives, and scope for the plan.
e. In Audit Request Response No. 16, the Company stated, “Avista used the same peak credit and planning margin on all portfolios in the IRP.” Please explain the risk of some portfolio scenarios over or under building resources and exceeding or failing reliability tests due to the use of a single planning margin for all portfolios. Please include an explanation on the effect of
the total cost of the portfolios and all issues with comparing the portfolios to each other.
RESPONSE: a. To adequately test resources for ELCC and whether or not new resources satisfy LOLP
requirements require a future year with significant resource deficits, Avista chose 2030 as it
met these qualifications. Avista would prefer to test all years, but due to the time to conduct each study a single year was used.
b. Difficulties: Unfortunately the current ARAM model takes three to four days to simulate the 1,000 iterations per study, not including set up time. Avista performed over 150 studies for this
IRP equating to nearly 11,000 computer hours. Due to the numerous studies required for ELCC studies it is not practical to conduct additional years of analysis or alternative portfolios.
Benefits: At this time Avista uses the single year results and assumes the results of this study are consistent across all years due to the time constraints discussed. For example, if the planning margin in 2030 is 17 percent for a 5 percent LOLP, then it is used for all years. Due to
this approach, the Company was able to complete a significant number of portfolio scenarios.
To further complicate this issue, the region (Northwest Power Pool) is looking to begin a Regional Adequacy (RA) program. If this effort is successful, resources will be assigned a peak credit and as long as the utility has enough resources to meet its requirement (planning margin), it will be considered resource sufficient. With this program the need to run ARAM
may no longer be necessary with the exception to measure regional market reliance risk.
Costs: The single year approach represents the planning margin in other years so long as the portfolio is similar. In scenarios where there is a removal of natural gas turbines and high
renewable/storage penetration there is a probability the planning margin would be different.
Avista agrees with these concerns and is exploring options (see part d).
c. See answer to part b.
d. To address the issue of a single year for reliability studies, Avista is developing a new version of the model to increase the time period, although the solution speed has not improved. Due to
this concern Avista is exploring several options to address this issue:
1) Developing a new system may significantly improve the speed but will require consulting services because it requires extensive optimization programming. Avista is currently in the process of getting a bid for this work, but has not determined if it will proceed with the contract due to budget constraints. If this project is as fast as
projected, the study time period can be extended to at least 20 years1.
2) Another option the Company is exploring, is to use Aurora to complete the reliability studies. In the past Aurora was unable to adequately model Avista flexibility in the hydro system for this type of study. Since then, enhancements to Aurora may
overcome these challenges. Currently this method is under investigation.
Avista proposes to discuss these issues and make a recommendation at the first IRP TAC meeting.
e. Portfolio 5 is the only portfolio with potential for underbuilding capacity in 2045 since it retires all natural gas plants and relies on storage and renewables by 2045. Scenarios with
extended natural gas or coal resources such as 7, 8, 14, and 15 may over build for the period
these resources are extended, but will need the higher planning margins toward the end of the plan as modeled. Avista did not test a portfolio relying on only natural gas turbines for new capacity to validate portfolio cost with lower planning margins.
To estimate impacts of planning margin changes, the higher or lower capacity additions would
change the results of the portfolio by the capacity payment (see Table 11.6- column “Capacity $/kW-yr”). The change in planning margin from the last IRP is +70.5 MW of capacity in 2030 (1,763 MW winter peak load). Using $116.6/kW-yr this equates to $8.2 million system costs for 2030.
The issues discussed above are why Avista is interested in combining the PRiSM model and
the ARAM model. Without a combined model, Avista will need to test each portfolio in the ARAM model. Without a faster tool, this will limit the number of scenarios that can be evaluated.
1 Depending on model speed improvements, a potential enhancement is to integrate the PRiSM model logic for
resource selection. This would eliminate the need for estimating a planning margin and the model would solve for both reliability and resources at the same time.

AVISTA CORPORATION RESPONSE TO REQUEST FOR INFORMATION

JURISDICTION: IDAHO DATE PREPARED: 05/05/2020 CASE NO: AVU-E-19-01 WITNESS: Clint Kalich REQUESTER: Staff RESPONDER: Grant D. Forsyth, PhD TYPE: Production Request DEPARTMENT: Financial Planning & Analysis
REQUEST NO.: Staff-016 TELEPHONE: [phone] REQUEST: Please explain why the Company believes it is reasonable to replace use per customer elasticity
estimates with “academic assumptions and estimates”. RESPONSE: In the residential UPC regressions used by the Company, price (the average price per Kwh
including fixed charges adjusted for inflation excluding energy prices) is not statistically
significant and the sign on the price coefficient is unstable. Therefore, to build a simulation around changes in price, the Company needs to use academic research to identify a reasonable estimate for both short- and long-run own-price elasticity. Identifying own-price elasticity at the utility-level can be problematic. Academic research from the RAND institute finds that elasticity estimates
become hard to identify as one moves from super-regional or regional electricity usage and price to
utility specific usage and price. The study can be found at https://www.rand.org/pubs/technical_reports/TR292.html . On page 50 of the study, the authors’ note, “Overall, we cannot conclude much from the utility-level analysis, other than the large amount of variation in price elasticities suggests that it may be useful to delve further into
analyzing utility-level electricity demand.”
AVISTA CORPORATION RESPONSE TO REQUEST FOR INFORMATION

JURISDICTION: IDAHO DATE PREPARED: 05/05/2020 CASE NO: AVU-E-19-01 WITNESS: Clint Kalich REQUESTER: Staff RESPONDER: Grant D. Forsyth, PhD TYPE: Production Request DEPARTMENT: Financial Planning & Analysis
REQUEST NO.: Staff-017 TELEPHONE: [phone] REQUEST: The Use Per Customer Regression Equation (Eq. 3.2) contains no autoregressive term; however,
the accompanying text discusses inclusion of the Autoregressive Integrated Moving Average (ARIMA) terms in the model. Please provide the revised version of Equation 3.2 that was actually used to model Usage Per Customer. Please answer the following questions regarding the model that was actually used to model Usage Per Customer:
a. What is the highest order autoregressive term used in this model?
b. Were the autoregressive predictors lagged values of Usage Per Customer, or were they lagged values of the prediction error? RESPONSE:
Equation 3.2 is meant to be a general expression for a UPC regression model that is used for the
medium-term forecast (out to 2025). The forecast feeds into the company’s revenue and earnings models. From this medium-term forecast, the remaining IRP years are “bootstrapped” to 2045 using long-run relationships.
By way of example, consider the residential schedule 1 for ID. The monthly load forecast for schedule 1 is the product of the use-per-customer (UPC) forecast and the customer forecast. At the time of IRP forecast, the schedule 1 UPC regression model used for the medium-term part of the IRP forecast was from the Spring 2019 load forecast used for the Company’s budget model. This
equation was:

Note that the ARIMA modeling is applied to the error term—in this case, a 10 month AR lag with no differencing, seasonal differencing, or moving average error terms. This style of model is often
called a “transfer function.” The dummy variables control for non-degree day seasonality or large outliers (a persistent problem with billed usage). HDD, CCD, and GAS are, respectively, heating degree days, cooling degree days, and an estimate of natural gas penetration. The UPC forecasts from this model are then multiplied by those generated by the following monthly schedule 1 customer forecast for ID:

As with the UPC model, this is a transfer function model with an ARIMA error term with a 6 month AR term and first-differencing, but no seasonal differencing or moving average terms. As
before, the dummy variables control for seasonality and outliers and POP is the population of Kootenai County (the main MSA of our ID service area) that has been interpolated to a monthly value. This same approach was used for the majority of residential schedules.
AVISTA CORPORATION RESPONSE TO REQUEST FOR INFORMATION

JURISDICTION: IDAHO DATE PREPARED: 05/05/2020 CASE NO: AVU-E-19-01 WITNESS: Clint Kalich REQUESTER: Staff RESPONDER: Grant D. Forsyth, PhD TYPE: Production Request DEPARTMENT: Financial Planning & Analysis
REQUEST NO.: Staff-018 TELEPHONE: [phone] REQUEST: On page 3-11, the Company states that it made its forecast for Residential Schedule 1 customer
growth using an ARIMA time series model. The Company also states that, if the growth rates generated from this approach differ from forecasted population growth, the forecasts are adjusted to match forecasted population growth. Please answer the following questions: a. Why is an ARIMA model justified, and what is the meaning of the ARIMA terms in the model?
b. What is the highest order autoregressive term in the model?
c. Given that the Company adjusts the forecast to match the forecast of population growth, why was population growth not used directly? RESPONSE:
a.-c.
See the Company’s response to Staff_PR_017.
For the IRP, load forecasts across all residential schedules are then aggregated and collapsed to an
annual number out to 2025. From 2026 forward, the total residential load is extrapolated by
applying long-run forecasts of own-price elasticity, customer growth, gas penetration, and the penetration of solar and electric cars. Note the residential process described above is the same IRP forecasting process that was used for
the commercial and industrial schedules in WA and ID.
AVISTA CORPORATION RESPONSE TO REQUEST FOR INFORMATION

JURISDICTION: IDAHO DATE PREPARED: 05/05/2020 CASE NO: AVU-E-19-01 WITNESS: Clint Kalich REQUESTER: Staff RESPONDER: Grant D. Forsyth, PhD TYPE: Production Request DEPARTMENT: Financial Planning & Analysis
REQUEST NO.: Staff-019 TELEPHONE: [phone] REQUEST: The Company's Residential Long-Run Forecast Relationship is expressed in Equation 3.3. Please
confirm that ly, cy, and uy are expressed in terms of percentage growth. RESPONSE: Staff’s interpretation is correct, these variables are measured as growth rates.

AVISTA CORPORATION RESPONSE TO REQUEST FOR INFORMATION

JURISDICTION: IDAHO DATE PREPARED: 05/05/2020 CASE NO: AVU-E-19-01 WITNESS: Clint Kalich REQUESTER: Staff RESPONDER: Grant D. Forsyth, PhD TYPE: Production Request DEPARTMENT: Financial Planning & Analysis
REQUEST NO.: Staff-020 TELEPHONE: [phone] REQUEST: In the discussion of Monthly Peak Load Forecast Methodology (IRP at 3-18 through 3-23), the
equation numbers and text references do not appear to match. a. Please confirm that the textual reference to Equation 3.9 is referring to Equation 3.4. b. Please confirm that the textual references to Equations 3.10 and 3.11 are referring to Equations 3.7 and 3.8.
c. Please provide corrections to any other textual references that may be in error.
RESPONSE: Part a: To clarify the typo on page 3-18, there is a statement at the end of the first paragraph at the top of the page that notes, “Equation 3.9 shows the current peak load regression model.” Staff is
correct, it should read, “Equation 3.4 shows the current peak load regression model.” This
correction is also noted in the response to Staff Request 22. Part b: To clarify the typo on page 3.20, there is a statement, “With over 100 years of average maximum and minimum temperature data, Equations 3.10 and 3.11 applied to each month t will
produce over 100 simulated values of peak load that can be averaged to generate a forecasted average peak load for month t in the current year, yc.” It should read, “With over 100 years of average maximum and minimum temperature data, Equations 3.5 and 3.6 applied to each month t will produce over 100 simulated values of peak load that can be averaged to generate a forecasted
average peak load for month t in the current year, yc.”
Part c: No other significant typos could be identified.
AVISTA CORPORATION RESPONSE TO REQUEST FOR INFORMATION

JURISDICTION: IDAHO DATE PREPARED: 05/05/2020 CASE NO: AVU-E-19-01 WITNESS: Clint Kalich REQUESTER: Staff RESPONDER: Grant D. Forsyth, PhD TYPE: Production Request DEPARTMENT: Financial Planning & Analysis
REQUEST NO.: Staff-021 TELEPHONE: [phone] REQUEST: In Peak Load Regression Model Equation 3.4, please explain the following:
a. The notation ed,g,y for t,y=June 2004↑. b. Please explain why June 2004 was used as a start date for the data series. RESPONSE:
Part a: The notation means the error term associated with the peak load on “day d, in month t, in
year y.” The “June 2004 ↑” means the historical data series is restricted to June 2004 forward. Part b: The data series is restricted to the period shown because of a limited time series available for Inland Empire Paper Company (IEP). IEP is a large paper company based near Spokane, WA.
The peak load forecast is performed by first removing the calendar loads reported in Avista’s
Nucleus data base for Clearwater Paper (CWTR) based in Lewiston, ID and IEP. They are initially removed from the regression modeling because their large loads, which can be idiosyncratic, can skew the results of the peak-load regression estimates. In the case of IEP, their calendar loads in Nucleus are only available back to 2004. In other words, there is no easily accessible source of
IEP’s load data prior to 2004.
AVISTA CORPORATION RESPONSE TO REQUEST FOR INFORMATION

JURISDICTION: IDAHO DATE PREPARED: 05/05/2020 CASE NO: AVU-E-19-01 WITNESS: Clint Kalich REQUESTER: Staff RESPONDER: Grant D. Forsyth, PhD TYPE: Production Request DEPARTMENT: Financial Planning & Analysis
REQUEST NO.: Staff-022 TELEPHONE: [phone] REQUEST: On pages 3-18 and 3-19, the Company states that "...a series of peak forecasts from the current year, yc , are generated out N years by using forecasted levels of GDP as shown in Equation 3.3."
Staff notes that Equation 3.3 includes no term representing a forecasted level of GDP. Please explain how GDP forecasts were incorporated into Equation 3.3.
RESPONSE:
To clarify a typo on page 3-18, there is a statement at the end of the first paragraph at the top of the page that notes, “Equation 3.9 shows the current peak load regression model.” It should read, “Equation 3.4 shows the current peak load regression model.” To clarify why there is no explicit mention of GDP in equation 3.3, the company notes on page 3-12 that equation 3.3 is for the
long-run residential forecast, where GDP is not an explicit forecast driver. The impact of GDP
shows up in the industrial load forecast as described on page 3-9. GDP is an explicit driver only for the medium-term part of the IRP forecast (the medium term forecast goes out to 2026; see also the response to Requests 17 and 18). Beyond that period, the simulation model smooths over business cycle events, which is equivalent to assuming GDP growth will be at a long-run average
of just under 2%. In response to staff questions (during the TAC process) about the impact of the business cycle, simulations of periodic recessions and expansions did not fundamentally alter the results of the smoothed long-run forecasts.
AVISTA CORPORATION RESPONSE TO REQUEST FOR INFORMATION

JURISDICTION: IDAHO DATE PREPARED: 05/03/2020 CASE NO: AVU-E-19-01 WITNESS: Clint Kalich REQUESTER: Staff RESPONDER: John Lyons TYPE: Production Request DEPARTMENT: Energy Resources
REQUEST NO.: Staff-023 TELEPHONE: [phone] REQUEST: The text on page 3-21 states that Table 3.5 shows estimated peak load growth rates with and
without the two large industrial customers; however, Table 3.5 includes only forecasts including large industrial customers. Please provide the forecast without the two large industrial customers. RESPONSE:
Avista does not anticipate any peak load growth for its two large industrial customers, therefore
the growth rates are the same with and without the two large industrial customers. It is possible either of the customers may participate in future Demand Response programs that may ultimately lower peak loads.
Avista DSM savings history summary
0.15
0.28
online savings
Cumulative
Gross (aMW
1978 0.15 1,340,000 1,340,000 0.15
1979 0.28 2,450,000 3,790,000 0.43
1980 1.00 8,770,000 12,560,000 1.43
1981 7.45 65,300,000 77,860,000 8.89
1982 4.73 41,440,000 119,300,000 13.62
1983 6.79 59,510,000 178,810,000 20.41
1984 2.84 24,860,000 203,670,000 23.25
1985 2.26 19,810,000 223,480,000 25.51
1986 1.72 15,030,000 238,510,000 27.23
1987 0.80 7,000,000 245,510,000 28.03
1988 0.17 1,460,000 246,970,000 28.19
1989 0.33 2,850,000 249,820,000 28.52
1990 0.48 4,180,000 254,000,000 29.00
1991 1.31 11,480,000 265,480,000 30.31
1992 5.81 50,920,000 316,400,000 36.12
1993 17.54 153,610,000 470,010,000 53.65
1994 10.18 89,160,000 559,170,000 63.83
1995 3.65 31,956,907 591,126,907 67.48
1996 2.89 25,333,122 616,460,029 70.37
1997 2.21 2.21 19,394,002 635,854,031 72.59
1998 1.88 4.09 16,458,851 652,312,882 74.46
1999 3.90 7.99 34,177,145 686,490,027 78.37
2000 4.55 12.54 39,864,163 726,354,190 82.92
2001 16.46 29.01 144,215,194 870,569,384 99.38
2002 3.98 32.99 34,882,381 905,451,765 103.36
2003 4.12 37.11 36,092,892 941,544,657 107.48
2004 3.96 41.08 34,733,154 976,277,811 111.45
2005 6.65 47.72 58,245,373 1,034,523,184 118.10
2006 5.29 53.01 46,316,227 1,080,839,411 123.38
2007 6.13 59.14 53,695,391 1,134,534,802 129.51
2008 8.55 67.69 74,861,160 1,209,395,962 138.06
2009 9.23 76.91 80,830,008 1,290,225,970 147.29
2010 10.14 87.06 88,842,528 1,379,068,497 157.43
2011 14.27 101.33 125,009,672 1,504,078,169 171.70
2012 8.89 110.22 77,875,000 1,581,953,169 180.59
2013 8.69 118.91 76,147,944 1,658,101,113 189.28
2014 7.75 126.66 67,873,466 1,725,974,579 197.03
2015 5.96 132.61 52,170,725 1,778,145,304 202.98
2016 16.43 149.04 143,904,911 1,922,050,215 219.41
2017 15.90 164.94 139,290,868 2,061,341,083 235.31
2018 8.74 173.68 76,531,910 2,137,872,993 244.05
Cumulative
Net (aMW
secondary
Cumulative
Net aMW
(secondary
0.15
0.43
1.43
8.89
13.62
20.41
23.25
25.51
27.23
28.03
28.19
28.52
29.00
30.31
36.12
53.65
63.83
67.48
70.22
72.15
73.03
69.48
69.30
78.97
80.11
81.97
84.22
90.07
95.19
100.99
109.06
116.98
121.31
118.04
116.76
121.80
126.66
130.40
144.95
156.95
161.13
aMW
5 aMW
10 aMW
15 aMW
20 aMW
25 aMW
An
n
u
a
l

S
a
v
i
n
g
s
1,034.00 2018 Load
1,195.13 Load w/o conservation
12.1%Conservation Percentage
aMW
20 aMW
40 aMW
60 aMW
80 aMW
100 aMW
120 aMW
140 aMW
160 aMW
180 aMW
200 aMW
220 aMW
240 aMW
Cu
m
u
l
a
t
i
v
e

S
a
v
i
n
g
s
AVISTA CORPORATION RESPONSE TO REQUEST FOR INFORMATION

JURISDICTION: IDAHO DATE PREPARED: 05/05/2020 CASE NO: AVU-E-19-01 WITNESS: Clint Kalich REQUESTER: Staff RESPONDER: Ryan Finesilver TYPE: Production Request DEPARTMENT: Energy Efficiency
REQUEST NO.: Staff-024 TELEPHONE: [phone] REQUEST: Table 1-2 of the Company's 2016-2017 Idaho Electric Energy Efficiency report states that Gross
Verified Savings were 42,223,004 kWh, or approximately 4.82 aMW. Please provide workpapers, with electronic links intact, showing how the Company obtained the 155 aMW savings value shown on page 5-1 of the Company's 2019 IRP.
RESPONSE:
The 155 aMW as stated on page 5-1 of the Company’s 2019 IRP represents the cumulative historic conservation achieved through the energy efficiency program and is not intended to represent a one or two year achievement level. Figure 5.1 on page 5-2 of the 2019 IRP provides an illustration of the cumulative savings. Note that the 155 aMW represents the “remaining” savings of measures
that are not past the average measure life of 18-years.
Please see Staff-PR-024 Attachment A which shows the historic conservation achieved by Avista’s Energy Efficiency program.
Feeder State Construction Start
Date
Construction End
Date
Baseline Report
Date
Baseline Report
Version
Estimated Annual
Pri. Reconductor
MWh Savings
Estimated Annual
Transformer Loss
MWh Savings
Total Estimated
Annual MWh
Savings3,4,5
9CE 12F4 WA -2009
BEA 12F1 WA 2012 2012
F&C 12F2 WA 2012 2012
BEA 12F5 WA 2013 2013
CDA 121 ID 2012 2013
WIL 12F2 WA 2013 2015
OTH 502 WA 2015 2015
M23 621 ID 2014 2015 3/20/2015 Version 4 412.6 163.2 575.8
RAT 231 ID 2014 2015 3/17/2015 Version 3 0.0 148.7 148.7
WAK 12F2 WA 2015 2016 3/3/2015 Version 7 40.3 135.3 175.6
MIL 12F2 WA 2016 2017 3/10/2015 Version 4 21.0 164.8 185.8
SPI 12F1 WA 2015 2019 4/1/2015 Version 2 31.6 83.2 114.8
RAT 233 ID 2016 2019 3/17/2015 Version 5 90.3 381.4 471.7
SPR 761 WA 2017 2019 9/17/2015 Version 3 49.9 55.7 105.6
ORO 1280 ID 2017 2017 10/19/2015 Version 1 3.5 108.2 111.7
TUR 112 WA 2017 2018 5/6/2016 Version 2 140.1 92.7 232.8
PDL 1201 WA 2017 2017 5/27/2016 Version 2 23.5 165.5 189.0
MIS 431 ID 2018 2023 8/22/2006 Version 1 128.8 128.3 257.1
F&C 12F1 WA 2018 2019 11/16/2016 Version 1 1.8 258.5 260.3
HOL 1205 ID 2018 2018 3/30/2017 Version 1 0 65.5 65.5
BEA 12F2 WA 2019 2020 10/13/2017 Version 1 8.8 260.5 269.3
M15 514 ID 2020 2023 4/30/2018 Version 1 0 245.6 245.6
SIP 12F4 WA 2020 2022 12/14/2018 Version 1 10.5 272.8 283.3
ROS 12F5 WA 2021 2021 5/31/2019 Version 1 6.1 145.9 152.1
3 Additional MWh savings estimated through Distribution Automation improvements are not included in these figures
4 Additional MWh savings estimated through the removal of Open Wire Secondary districts are not included in these figures
5 Additional MWh savings estimated through power factor correction initiatives with capacitors, IVVC, or CVR are not included in these figures
Updated 6/19/19 by Shane Pacini
1 Completed under the DREE Program. Annual MWh Energy Savings may have been estimated and provided by others, however they did not follow the same analysis
process and documentation that was started by Grid Modernization in late 2013, and may not be able to be recreated
2 Completed under the Feeder Upgrade Program. Annual MWh Energy Savings may have been estimated and provided by others, however they did not follow the same
analysis process and documentation that was started by Grid Modernization in late 2013, and may not be able to be recreated
Annual MWh Energy Savings were not estimated or documented at this time1
Annual MWh Energy Savings were not estimated or documented at this time2
Annual MWh Energy Savings were not estimated or documented at this time2
Annual MWh Energy Savings were not estimated or documented at this time2
Annual MWh Energy Savings were not estimated or documented at this time
Annual MWh Energy Savings were not estimated or documented at this time2
Annual MWh Energy Savings were not estimated or documented at this time2
AVISTA CORPORATION RESPONSE TO REQUEST FOR INFORMATION

JURISDICTION: IDAHO DATE PREPARED: 05/05/2020 CASE NO: AVU-E-19-01 WITNESS: Clint Kalich REQUESTER: Staff RESPONDER: Ryan Finesilver TYPE: Production Request DEPARTMENT: Energy Efficiency
REQUEST NO.: Staff-025 TELEPHONE: [phone] REQUEST: Please provide the analysis and workpapers that Avista used to determine estimated savings for
feeder upgrades in 2020 as 269 MWh, and 152 MWh in 2021. IRP at 5-5. RESPONSE:
For the purpose of establishing distribution savings for the Company’s energy efficiency program,
the Energy Efficiency team worked with members of distribution engineering to identify expected grid modernization projects and the estimated energy savings derived from those efforts. Staff_PR_025 Attachment A provides an overview of the historic and expected future projects that
have an end date between 2009 and 2023. It was identified that three projects would be completed
during the 2020-2021 timeframe. Those include one project in Washington for 2020 and one project in Washington for 2021.
AVISTA CORPORATION RESPONSE TO REQUEST FOR INFORMATION

JURISDICTION: IDAHO DATE PREPARED: 05/04/2020 CASE NO: AVU-E-19-01 WITNESS: Clint Kalich REQUESTER: Staff RESPONDER: John Lyons TYPE: Production Request DEPARTMENT: Energy Resources
REQUEST NO.: Staff-026 TELEPHONE: [phone] REQUEST: Please describe how Avista determined that Energy Efficiency will meet 71% of future load
growth. If Demand Response is a factor in the estimate, please break out the contributions from Energy Efficiency and Demand Response potential separately. RESPONSE:
The 2020 IRP identified that 71 percent of new load growth would be met with energy efficiency
on page 11-10. This was determined by adding the amount of forecasted energy efficiency back into the load forecast and comparing the two forecasts with and without energy efficiency; the difference between the two over the 2020 IRP forecast is 71 percent. No Demand Response is included in the 71 percent.
AVISTA CORPORATION RESPONSE TO REQUEST FOR INFORMATION

JURISDICTION: IDAHO DATE PREPARED: 05/05/2020 CASE NO: AVU-E-19-01 WITNESS: Clint Kalich REQUESTER: Staff RESPONDER: John Lyons TYPE: Production Request DEPARTMENT: Energy Resources
REQUEST NO.: Staff-027 TELEPHONE: [phone] REQUEST: As part of an aggressive plan to meet future load growth, please describe the Company plan for Demand Response Rates. IRP at 6-4. RESPONSE:
As discussed on page 6-4 of the IRP, Demand Response (DR) pricing programs included in AEG’s study require AMI as an enabling technology. For the purposes of the IRP study, AEG assumed Avista would be installing AMI metering in Idaho, finishing in 2025.
The Company recognizes that DR pricing programs will play a key role in meeting future load
growth. At this time, the Company does not yet have a plan developed for implementing DR rates. The Company will be actively evaluating DR pricing programs as it completes the rollout of its AMI metering in Washington and later in Idaho.
As the Company develops plans for DR pricing programs, it will share the plans with Commission
Staff when available.
AVISTA CORPORATION RESPONSE TO REQUEST FOR INFORMATION

JURISDICTION: IDAHO DATE PREPARED: 05/04/2020 CASE NO: AVU-E-19-01 WITNESS: Clint Kalich REQUESTER: Staff RESPONDER: James Gall TYPE: Production Request DEPARTMENT: Energy Resources
REQUEST NO.: Staff-028 TELEPHONE: [phone] REQUEST: Regarding Liquid Air Storage, the Company states on pages 9-14 that round-trip efficiencies can
be improved using the waste heat from existing natural gas-fired turbines. Given the clean energy goals enumerated by the Company on page 1-6, what thermal resources will be available to boost energy efficiency? Please provide an estimate of the heat that will be available when Liquid Air Storage systems are producing electricity.
RESPONSE: The Liquid Air Storage resources may operate independently from host heat resources. If Avista chooses to co-locate this technology at an existing thermal resources and use host heat, the most likely site is the Rathdrum CT; but other plant locations are discussed below. Avista’s clean energy
goal is to be 100 percent clean by 2045; while this may seem to preclude use of Avista’s current
natural gas resources or heat from these resources, it is possible these resources could still operate and supply heat with either renewable natural gas or clean sourced hydrogen or other fuel. Rathdrum CT- The most likely facility available for use in conjunction with Liquid Air Energy
Storage (LAES) would be the Rathdrum Combustion Turbine Facility. The reason for this is that these are large frame simple-cycle combustion turbines and they have the most available waste heat. Rathdrum CT has heat sufficient to facilitate a LAES facility in the 50-100 MW range. Energy storage capacity in MWh is decoupled from MW capacity.
Coyote Springs 2- While this facility uses an even larger frame combustion turbine than is used at the Rathdrum CT, it already uses a highly efficient combined cycle configuration. With stack exhaust temperatures in the 200 degree F range, there is much less, if any, waste heat available for LAES.
Kettle Falls- Both the wood-fired unit and the combustion turbine are much smaller than the Rathdrum CT. Furthermore, the main wood unit does not have anywhere near the exhaust heat available compared to the Rathdrum CT. Small LAES installations might be possible.
Northeast Combustion Turbine- These turbines are approaching their end of life. They are not efficient and they are almost exclusively used for reserve requirements. They would most likely not be suitable for use as part of a LAES installation.

AVISTA CORPORATION RESPONSE TO REQUEST FOR INFORMATION

JURISDICTION: IDAHO DATE PREPARED: 05/04/2020 CASE NO: AVU-E-19-01 WITNESS: Clint Kalich REQUESTER: Staff RESPONDER: James Gall TYPE: Production Request DEPARTMENT: Energy Resources
REQUEST NO.: Staff-029 TELEPHONE: [phone] REQUEST: On pages 9-24 and 9-25, the Company describes analyses that it conducted in order to determine
the intermittent generation costs of wind and solar resources. These costs include a $5 per MWH for wind, $1.80 for solar, and a 10% capacity value adder determined by Avista's ELCC studies. Please provide the studies, including a description of study methodology, that were used to determine these values.
RESPONSE: The wind integration costs are based upon a regression of the data from the 2007 Wind Integration Study. The analysis is provided in the IRP in Appendix H. These estimates include an estimated 40 percent reduction in costs from participating in the EIM, based on information provided by the
CAISO. Avista is currently in the process of conducting a new renewable integration cost study to
include both wind and solar. Regarding the 10 percent capacity adder for ELCC studies, this is in reference to when an intermittent generator is generating. The model must carry regulating reserves equal to 10 percent
of its potential generation. The 10 percent is based on the 2007 Wind Integration study and is similar to Avista operational history with its intermittent generation.
AVISTA CORPORATION RESPONSE TO REQUEST FOR INFORMATION

JURISDICTION: IDAHO DATE PREPARED: 05/06/2020 CASE NO: AVU-E-19-01 WITNESS: Clint Kalich REQUESTER: Staff RESPONDER: John Lyons TYPE: Production Request DEPARTMENT: Energy Resources
REQUEST NO.: Staff-030 TELEPHONE: [phone] REQUEST: Please explain why the Company did not test multiple closure dates for Colstrip besides both units
closing in 2025, both units closing in 2035, and unit #3 closing in 2025 while unit #4 closing in 2035. RESPONSE:
As previously described in the response to Staff-AR-05, Avista believes Colstrip could operate
until 2045 with continual capital investment and maintenance before reaching its end of life. Considering the state of the current fuel supply, the projected need for additional environmental controls, such as selective catalytic reduction (SCR) in the future, and the general lack of support of many of the owners for investing additional capital, the Company estimated that the likely
economic operational life of Colstrip is through 2035.
Given the reasons just stated, the results of the scenarios modeled for the 2020 IRP that indicated it was an economic benefit to shut the plant down by the end of 2025, as well as the legal requirement of Avista is to no longer have Colstrip in Washington rates after 2025, the Company believed there
were no additional insights to be gained from running additional analyses. The scenarios that were included in the 2020 IRP were discussed on multiple occasions with the Technical Advisory Committee and the Company did not receive any requests for additional
studies or for alterations to the studies that were completed. There was also the practical matter of
limitations to how many studies can be run in the amount of time available to develop an IRP. The Company can run other study scenarios with different dates regarding Colstrip closure if requested by Commission Staff. Avista suggests additional specific Colstrip scenarios can be studied and discussed in a workshop to inform the 2021 IRP Colstrip scenarios.

AVISTA CORPORATION RESPONSE TO REQUEST FOR INFORMATION

JURISDICTION: IDAHO DATE PREPARED: 05/06/2020 CASE NO: AVU-E-19-01 WITNESS: Clint Kalich REQUESTER: Staff RESPONDER: James Gall TYPE: Production Request DEPARTMENT: Energy Resources
REQUEST NO.: Staff-031 TELEPHONE: [phone] REQUEST: With the current coal supply agreement set to expire Dec. 31, 2025, is Colstrip fuel expense after
2025 modeled in any portfolio? Please explain why or why not. RESPONSE: Avista completed all modeling for the IRP “prior” to final negotiations of the Colstrip fuel supply
agreement. That stated, Avista did include costs similar to those in the final agreement in our
modeling. For scenarios extending Colstrip beyond 2025, Avista included coal prices with annual price increases in line with the final agreement. Avista included these costs in response to Staff-AR-05.
AVISTA CORPORATION RESPONSE TO REQUEST FOR INFORMATION

JURISDICTION: IDAHO DATE PREPARED: 05/06/2020 CASE NO: AVU-E-19-01 WITNESS: Clint Kalich REQUESTER: Staff RESPONDER: James Gall TYPE: Production Request DEPARTMENT: Energy Resources
REQUEST NO.: Staff-032 TELEPHONE: [phone] REQUEST: Please explain how Avista will calculate the additional cost associated with assigning a new
capacity deficiency date, in part driven by the Company’s decision to eliminate Colstrip capacity to meet Washington legislative initiatives and Avista’s clean energy goals. RESPONSE:
Avista anticipates being capacity deficit in 2026 regardless of a decision to shut down or
potentially transfer its ownership of Colstrip. The Lancaster PPA expires in October 2026 and the loss of Colstrip increases the quantity of the shortfall, however, does not significantly accelerate the timing of a shortfall at this time. Please also note that the decision to eliminate Colstrip as shown in the IRP is primarily due to the economics of the plant compared to other resources, and
not clean energy policies.
When Avista exits Colstrip, it plans to acquire replacement capacity/energy for the loss of this resource to maintain system adequacy and reliability. The required replacement capacity/energy may not be the full 222 MW of the project as it will depend on the need at the time the plant is no
longer economic compared to other resource options. The Company will make future resource decisions based on the best economic information available at the time, and will seek guidance from both the Idaho and Washington Commissions as to the best approached to recover the cost of existing and new resources. One option to consider is to seek recovery using the existing P/T ratio
methodology for all future costs, however more dialogue is needed with the states to evaluate
alternative recovery mechanisms.
AVISTA CORPORATION RESPONSE TO REQUEST FOR INFORMATION

JURISDICTION: IDAHO DATE PREPARED: 05/01/2020 CASE NO: AVU-E-19-01 WITNESS: Clint Kalich REQUESTER: Staff RESPONDER: James Gall TYPE: Production Request DEPARTMENT: Energy Resources
REQUEST NO.: Staff-033 TELEPHONE: [phone] REQUEST: Please provide the capacity shortfall and related costs associated with an early Colstrip retirement
in 2025. Please include an analysis of financial impacts to Idaho customers. Audit Request Response Nos. 4 and 5. RESPONSE:
Please see Avista's response 033C, which contains TRADE SECRET, PROPRIETARY or CONFIDENTIAL information and exempt from public view and is separately filed under IDAPA 31.01.01, Rule 067 and 233, and Section 9-340D, Idaho Code. Avista demonstrated in Staff-AR-04 the total PVRR (Present Value of Revenue Requirement)
change to Idaho customers is $6.6 million higher if Colstrip remains online through 2035 as
compared to 2025 (See “Column O”). The change in Avista’s capacity position is a loss of 208 MW.1 Included in Attachment Staff_PR_033C Confidential Attachment A is the response to AR No. 4 with the addition of the
capacity deficit positions shown in Columns R and S2 assuming an 18 percent planning margin in the PRS. Avista proposes to hold a workshop with Staff to discuss these scenarios and answer any additional questions. In regards to Staff-AR-05, Avista identified two reasons for the 2035 likely end date. The first is
by 2035 the mine may no longer be able to supply the plant with coal. The second reason is economic because the plant will likely require an SCR by 2035 to meet Regional Haze Program goals. Colstrip is highly unlikely to remain economical if an SCR is required given the projected market prices of electricity and the capital costs of this equipment. Avista’s share in this capital investment (estimated in the 2017 IRP) is $79 million (2014 dollars).
To further justify the 2035 economic life, the plant will need to replace both the turbine and generators in the next 15 years to be capable of operating to 2045. Operating to a 2035 end date is possible with the existing equipment, but with higher expected forced outages until the shutdown.
In addition to the likely required investment in a SCR, replacing existing equipment to add only
approximately 10 years of life may not be economic.

1 Colstrip is 222 MW, but Avista avoids 14 MW of losses with Colstrip exiting the portfolio. 2 PRiSM models only January and August Peak events, Avista’s 2026 deficit with Colstrip remaining online begins
after the Lancaster PPA expires in Oct 2026.
AVISTA CORPORATION RESPONSE TO REQUEST FOR INFORMATION

JURISDICTION: IDAHO DATE PREPARED: 05/06/2020 CASE NO: AVU-E-19-01 WITNESS: Clint Kalich REQUESTER: Staff RESPONDER: James Gall TYPE: Production Request DEPARTMENT: Energy Resources
REQUEST NO.: Staff-034 TELEPHONE: [phone] REQUEST: Is Avista considering the sale of Colstrip ownership shares in exchange for Power Purchase
Agreement(s) for renewable energy? Please explain Audit Request Response No. 6. RESPONSE: Please see Avista's response 034C, which contains TRADE SECRET, PROPRIETARY or CONFIDENTIAL information and exempt from public view and is separately filed under
IDAPA 31.01.01, Rule 067 and 233, and Section 9-340D, Idaho Code.
AVISTA CORPORATION RESPONSE TO REQUEST FOR INFORMATION

JURISDICTION: IDAHO DATE PREPARED: 05/04/2020 CASE NO: AVU-E-19-01 WITNESS: Clint Kalich REQUESTER: Staff RESPONDER: James Gall TYPE: Production Request DEPARTMENT: Energy Resources
REQUEST NO.: Staff-035 TELEPHONE: [phone] REQUEST: Please provide an economic analysis that demonstrates when it is no longer economically
beneficial to operate Colstrip. Please allow the retirement date to float based on the economics of the resource as opposed to adjusting retirement dates to 2025 or 2035 and provide all workpapers with formulas enabled.
RESPONSE:
Please see Avista's response 035C, which contains TRADE SECRET, PROPRIETARY or CONFIDENTIAL information and exempt from public view and is separately filed under IDAPA 31.01.01, Rule 067 and 233, and Section 9-340D, Idaho Code.
In order to perform the analysis in this request, Avista made the following assumptions:
1) Avista’s PRiSM model and the 2020 IRP includes Social Cost of Carbon (SCC) in the objective function for Washington’s share of the existing and potential resource options. For this analysis, Avista removed the SCC to not influence resource selection or removal of
Colstrip. This removal does not change the Revenue Requirement in the IRP, but rather removes this cost from the decision making logic in the model’s resource selection. Without this change, Colstrip may be uneconomic in all years of the study due to Washington portion of Colstrip receiving a SCC adder. 2) Colstrip costs identified as capital costs are assumed to be recovered over 5 years with the
exception of any costs between 2031 and 2035 being amortized over the remaining life of the asset. 3) The plant will retire on Dec 31, 2035 or float to close on any date prior to 2035 per the request of this study. As discussed in Staff_PR_030, Avista does not expect the plant to continue operating beyond 2035. The IRP did not estimate Colstrip costs beyond 2035.
Avista can model a 2045 shutdown date in a process outside the IRP and estimate the cost and benefits of additional plant investment for another 10 years of operating life. 4) The decision to keep the plant operating or not is for both state’s customers- the model is not set up to only apply resources to specific states; the model assumes all resources in the
portfolio are for all customers regardless of their location. The next IRP will evaluate the
impacts of transferring Washington’s portion of Colstrip to the shareholders in 2026. 5) These scenarios do not take into account the regional effects of shutting the plant down early without replacement capacity. Without Colstrip, the region is not likely to be resource adequate. While Avista may remain resource adequate based on its planning margin
requirement, it would be at greater risk, since there would be less regional market power to
rely on during an adequacy event. Avista proposes to discuss regional issues regarding Colstrip in a future workshop with Staff. 6) This scenario should be used for informational purposes only and is subject to change
based on market conditions and cost forecast changes.

Results
Avista conducted the following scenarios for this study to get a complete understanding of this
request: 1) Optimized Portfolio with No Social Cost of Carbon and Colstrip ends in 2025, Staff_PR_035C Confidential Attachment A. The 25 year PVRR is $11,759 million. This
scenario serves as a baseline for the next two scenarios. 2) Optimized Portfolio with No Social Cost of Carbon and Colstrip retirement by unit may float between 2021 and 2035, Staff_PR_035C Confidential Attachment B. The 25 year PVRR is slightly lower then case #1 at $11,744 million. The model retires one Colstrip unit by 1/1/2021 the other in 1/1/2022, resulting in a savings to customers of $15 million on a
system basis. 3) Avista studied a scenario “Optimized Portfolio with No Social Cost of Carbon and Colstrip retirement by unit may float between 2021 and 2035 with lower O&M expense for Colstrip continued operation beyond 2025” to see if the plant would remain economic with lower
O&M growth. In this case the plant assumes O&M spending is lowered to 2.2 percent
(GDP implicit price deflator) vs 11 percent assumption in the IRP. The IRP assumption reflects increased higher O&M after 2026 as the plant is shifting from capital spending to expensing costs as it ages toward the end of the life. This scenario tests whether or not the closure decision would be different with modest O&M growth after 2025 as compared to
the higher cost assumption used in the IRP scenarios, Staff_PR_035C Confidential
Attachment C. The model keeps both units operating with the 25 year PVRR of $11,716 million with this assumption change. Avista proposes to hold a workshop with Staff to discuss these scenarios and answer any
additional questions.
AVISTA CORPORATION RESPONSE TO REQUEST FOR INFORMATION

JURISDICTION: IDAHO DATE PREPARED: 05/06/2020 CASE NO: AVU-E-19-01 WITNESS: Clint Kalich REQUESTER: Staff RESPONDER: James Gall TYPE: Production Request DEPARTMENT: Energy Resources
REQUEST NO.: Staff-036C TELEPHONE: [phone] REQUEST: In reference to the state-specific study mentioned on page 12-5 in the IRP, please provide the
workpapers (with formulas enabled) used to calculate the increase in Idaho rates. RESPONSE: Please see Avista's response 036C, which contains TRADE SECRET, PROPRIETARY or CONFIDENTIAL information and exempt from public view and is separately filed under
IDAPA 31.01.01, Rule 067 and 233, and Section 9-340D, Idaho Code. The file used to create Figure 12.2 is attached as Staff_PR_036C Confidential Attachment A.
AVISTA CORPORATION RESPONSE TO REQUEST FOR INFORMATION

JURISDICTION: IDAHO DATE PREPARED: 05/06/2020 CASE NO: AVU-E-19-01 WITNESS: Clint Kalich REQUESTER: Staff RESPONDER: James Gall TYPE: Production Request DEPARTMENT: Energy Resources
REQUEST NO.: Staff-037 TELEPHONE: [phone] REQUEST: Please explain why portfolios scenarios without CETA (Scenarios #2 and #7) show new resource
additions prior to the 2026 deficit year identified in the IRP. RESPONSE: Resources selected prior to 2026 are selected due to economics. In the case of Scenario #2 and #7,
acquiring Montana wind with federal tax incentives is economic for customers due to price of the
expected PPA as compared to future market prices and the need for new capacity later in the planning period. For example, Montana wind acquisition (at the busbar) prior to Jan 1, 2022 with full PTC is $21.43 per MWh as compared to $47.49 per MWh in 2026 without any PTC. Avista’s IRP renewable resource costs were informed by the previous RFP where it selected Rattlesnake
Flat wind project. Avista will conduct a competitive RFP before acquiring any new resource,
which will determine the economics of acquiring resources prior to 2026. After conducting an RFP, Avista will have a better understanding of renewable pricing but does not have to select a proposed project if the pricing isn’t favorable.
The benefits of issuing a renewable RFP in the near future include:
• The expiring Production Tax Credit (PTC) is lowering prices as compared to price quotes after 2020. o The PTC is scheduled to be reduced or expire after 2020 (and similarly the
investment Tax Credit (ITC) in 2022). Pricing may increase once tax credits expire.
o Many developers have projects that will qualify for tax opportunity programs and are seeking buyers.
• Other market indicators such as pricing and developer activity.
o Developer activity along with industry market insights provide Avista
opportunities to observe and analyze changes in renewable energy technology and pricing. o Recent indicative and actual pricing for renewables in the west suggest renewable resources are competitive in the wholesale electric market.
• Competing RFPs.
o There are other utilities in the northwest actively pursuing renewable resources. Portland General Electric and Puget Sound Energy engaged in RFP’s in 2018 and PacifiCorp is planning to issue an RFP in 2020 as well. o This demand could create competition for preferred projects.
• Further advancing renewable technology and competition for least cost resources.
o With the advances of machine technologies and the sun-setting of tax credits, pricing for renewables continues to be competitive with market based alternatives. Pricing is forecast to increase when tax credit opportunities expire.

AVISTA CORPORATION RESPONSE TO REQUEST FOR INFORMATION

JURISDICTION: IDAHO DATE PREPARED: 05/06/2020 CASE NO: AVU-E-19-01 WITNESS: Clint Kalich REQUESTER: Staff RESPONDER: James Gall TYPE: Production Request DEPARTMENT: Energy Resources
REQUEST NO.: Staff-038 TELEPHONE: [phone] REQUEST: Please describe how the Company will define, separate, and track Idaho and Washington financial
obligations for Colstrip and CETA costs. Audit Request Response No. 4. RESPONSE: At this time, given that the CETA rulemakings are just beginning, it may be premature to
determine exactly how financial obligations for Colstrip and CETA will be defined. As CETA
requirements become more developed, the Company will work with both Idaho and Washington to determine the best way to track Colstrip and CETA costs. The Company suggests this effort be conducted in a new process outside the IRP.