20031021Said Direct.pdf

BEFORE THE IDAHO PUBLIC UTILITIES COMMISSION
IN THE MATTER OF THE APPLICATION
OF IDAHO POWER COMPANY FOR
AUTHORITY TO INCREASE ITS RATES
13
AND CHARGES FOR ELECTRIC SERVICE
TO ELECTRIC CUSTOMERS IN THE STATE)
OF IDAHO. )
)
IDAHO POWER COMPANY
DIRECT TESTIMONY
OF
GREGORY W. SAID
CASE NO. IPC-E-03-
1
2
Q.
A.
Please state your name and business address.
My name is Gregory W. Said and my business
3 address is 1221 West Idaho Street, Boise, Idaho.
4 Q. By whom are you employed and in what
5 capacity?
6 A. I am employed by Idaho Power Company as the
7 Manager of Revenue Requirement in the Pricing and
8 Regulatory Services Department.
9
10
Q.
A.
Please describe your educational background.
In May of 1975, I received a Bachelor of
11 Science Degree with honors from Boise State University. In
12 1999, I attended the Public Utility Executives Course at
13 the University of Idaho.
14 Q. Please describe your work experience with
15 Idaho Power Company.
16 A. I became employed by Idaho Power Company in
17 1980 as an analyst in the Resource Planning Department. In
18 1985, the Company applied for a general revenue requirement
19 increase. I was the Company witness addressing power
20 supply expenses.
21 In August of 1989, after nine years in the
22 Resource Planning Department, I was offered and I accepted
23 a position in the Company's Rate Department. With the
SAID, DI 1
Idaho Power Company
1 Company's application for a temporary rate increase in
2 1992, my responsibilities as a witness were expanded.
3 While I continued to be the Company witness concerning
4 power supply expenses, I also sponsored the Company's rate
5 computations and proposed tariff schedules in that case.
6 Because of my combined Resource Planning and
7 Rate Department experience, I was asked to design a Power
8 Cost Adjustment (PCA) which would impact customers' rates
9 based upon changes in the Company's net power supply
10 expenses. I presented my recommendations to the Idaho
11 Public Utilities Commission in 1992 at which time the
12 Commission established the PCA as an annual adjustment to
13 the Company's rates. I have sponsored the Company's annual
14 PCA adjustment in each of the years 1996 through 2003.
15 In 1996, I was promoted to Director of
16 Revenue Requirement. At year-end 2002, I was promoted to
17 the senior management level of the Company.
18 Q. What topics will you discuss in your
19 testimony in this proceeding?
20 A. I will discuss changes in loads and
21 resources since the Company's last general rate case and
22 the impact of those changes on the Company's power supply
23 expenses. I will sponsor the exhibits that provide the
SAID, DI 2
Idaho Power Company
1 basis for determining the Company's normalized net power
2 supply expenses for ratemaking purposes. I will also
3 discuss how the new normalized power supply expenses impact
4 future PCA computations until the Company's next general
5 rate case.
6 Q. Please describe the change in the Company's
7 system loads since the last general rate case, IPC-E-94-5.
8 A. The Company's 1993 annual normalized system
9 load used in the IPC-E-94-5 case was 14.5 million megawatt-
10 hours (MWh) . The Company's 2003 annual normalized system
11 load used in this case is 14.1 million MWh. The annual
12 system load served today is approximately the same as it
13 was ten years ago.
14 Q. Over the last ten years, what changes in
15 loads combined to result in a 2003 annual system load that
16 is so similar to the 1993 annual system load?
17 A. While there has been load growth within most
18 customer classes, the Company has also experienced load
19 decline in a couple of distinct areas. Ten years ago, FMC
20 was Idaho Power's single largest customer with a load of
21 1.7 million MWh per year. FMC, which later became known as
22 Astaris, discontinued operation leaving only a small
23 residual industrial load being served as a Schedule 19
SAID, DI 3
Idaho Power Company
1 customer. Idaho Power also had some FERC jurisdictional
2 contract loads amounting to approximately 1.4 million MWh
3 that were intended to be served by surplus resources that
4 existed at that time, but were scheduled for discontinuance
5 as the Company's state jurisdictional loads grew to match
6 generation capability. As planned, those FERC
7 jurisdictional contracts have reached their conclusion.
8 The 3.1 million megawatt-hour reduction in annual system
9 loads have been replaced by 2.7 million MWh of load growth
10 within other customer classes.
11 Q. Has the monthly shape of the annual load
12 changed in the last ten years?
13 A. Yes. The FMC contract as well as the
14 concluded FERC contracts that existed ten years ago
15 provided the Company with relatively consistent monthly
16 loads that were somewhat flat throughout the year. The FMC
17 load had an interruptible component. Load growth within
18 the various customer classes has tended to be much more
19 seasonal and dependent upon weather. As a result of the
20 loss of relatively flat loads and the addition of non-
21 interruptible seasonal loads, the Company's Integrated
22 Resource Plan now shows the need for summer peaking
23 resources (June, July, and August) and winter peaking
SAID, DI 4
Idaho Power Company
1 resources (November and December).
2 Q. Please define the term "power supply
3 expenses" as the Company and the Commission have used the
4 term historically.
5 A. The Company and the Commission have used the
6 term "power supply expenses" to refer to the sum of fuel
7 expenses (FERC accounts 501 and 547) and purchased power
8 expenses (FERC account 555) excluding PURPA qualifying
9 facilities (QF) expenses minus surplus sales revenues (FERC
10 account 447). For ratemaking purposes, QF expenses have
11 been quantified separately from other power supply expenses
12 and are treated as fixed inputs to power supply modeling
13 rather than variable outputs.
14 Q. How would you expect power supply expenses
15 to be affected by the changes in loads, as you have
16 described, that resulted in approximately the same annual
17 load, but with seasonal shifts in loads and higher peak
18 hour requirements?
19 A. I would expect power supply expenses to rise
20 as a result of the seasonal and peak hour load shifts that
21 the Company has experienced over the last ten years.
22 Additional loads during the peak hours of the summer season
23 will need to be served by higher cost resources.
SAID, DI 5
Idaho Power Company
1 Q. How have market prices of energy changed in
2 the last ten years?
3 A. Market prices for energy are generally
4 higher than market prices ten years ago. In the IPC-E-94-5
5 case it was assumed that the highest monthly market price
6 that the Company might encounter would be $27 per MWh,
7 which is equivalent to 27 mills per kilowatt-hour (kWh) or
8 2.7 cents per kWh. Ignoring the run-up in market prices
9 that occurred in the 2000-2001 time period, the Company has
10 routinely seen market prices in the $40 to $50 per MWh
11 price range during the last two drought years. It has been
12 quite some time since the Company and the region
13 experienced high water conditions, but if high water was to
14 occur, I would expect that market prices would be
15 significantly lower than the $40 to $50 per MWh range, but
16 not as low as the $7 to $17 per MWh range expected to
17 accompany high water conditions ten years ago.
18 Q. What affect on power supply expenses would
19 you envision as a result of the upward movement in the
20 market price for energy?
21 A. As I have mentioned, I believe that a
22 relationship between hydro conditions and the market price
23 of energy still exists. When the Company and the region
SAID, DI 6
Idaho Power Company
1 have abundant water, higher cost generating plants are not
2 required to satisfy Company or regional loads. The
3 marginal resource at such times is likely a low cost coal
4 unit or even on occasion hydro generation. As a result,
5 the market price for energy will fall to the incremental
6 cost of the marginal resource. Conversely, when the region
7 is in a drought condition, as is the current situation,
8 higher cost coal units and gas-fired units will be the
9 marginal resources influencing market prices.
10 As a result of the supply and demand
11 relationship, the Company will continue to encounter higher
12 market prices when both the Company and the region are
13 resource deficient and conversely will encounter lower
14 market prices when both the Company and the region have
15 abundant resources. Power supply expenses are reduced by
16 higher valued market sales, but are increased by higher
17 valued market purchases. I would expect overall upward
18 pressure on power supply expenses as a result of an upward
19 trend in market prices especially when considering the
20 seasonal and peak period load shifts that I discussed
21 earlier.
22 Q. How have the fuel costs of the Company's
23 coal-fired resources changed over the last ten years?
SAID, DI 7
Idaho Power Company
1 A. My response to this question includes known
2 and measurable changes to fuel costs, which I will discuss
3 later in my testimony. Including known and measurable
4 adjustments, the fuel cost for the Bridger units has
5 increased at an annual average rate of 1.0 percent per year
6 over the last ten years from $11.51 per MWh to $12.75 per
7 MWh. The fuel cost for the Boardman plant has increased at
8 an annual average rate of 0.5 percent per year over the
9 last ten years from $12.59 per MWh to $13.25 per MWh. Due
10 to the renegotiation and replacement of coal contracts for
11 the Valmy plant, the fuel cost for the Valmy units has
12 decreased by 31 percent from $21.19 per MWh in 1993 to
13 $14.7 per MWh in the test year 2003.
14 Q. Due to the changes in the fuel costs of the
15 Company's coal-fired resources, what effect would you
16 expect to see with regard to power supply expenses?
17 A. With only modest increases in the fuel costs
18 for Bridger and Boardman and significant decreases in the
19 fuel cost for Valmy, I would expect some downward movement
20 in the Company's power supply expenses. Lower per unit
21 fuel costs at Valmy will reduce the fuel expense at Valmy
22 when it is dispatched to serve system loads, but also will
23 provide for more frequent opportunities to sell Valmy
SAID, DI 8
Idaho Power Company
1 surpluses into the market. Both of these impacts serve to
2 reduce net power supply expenses.
3 Q. Are there any resource additions that have
4 occurred in the last ten years that would reduce power
5 supply expenses?
6 A. Yes. The addition of any resource has the
7 effect of reducing power supply expenses. This results
8 because of economic dispatch principals. If additional
9 resources can be dispatched at costs lower than
10 alternatives, then dispatch of the new resources occurs
11 thus reducing power supply expenses. If the additional
12 resource cannot be dispatched at costs lower than
13 alternatives, no additional power supply expense occurs.
14 In the last ten years, the Company has added the Danskin
15 gas-fired plant, located at the Evander Andrews complex
16 near Mountain Horne, Idaho and has also received energy from
17 additional PURPA QF projects. In 2004, the Company will
18 acquire additional generation from the PPL Montana Power
19 Purchase Agreement (PPA) and from a new QF project called
20 the Tiber Montana LLC (Tiber) project. The costs of QF
21 projects have not historically been included in "power
22 supply expenses" and thus power supply expenses are reduced
23 by new QF projects as they reduce the need for resources
SAID, DI 9
Idaho Power Company
1 that are reflected in power supply expenses.
2 Q. Have you supervised the preparation of power
3 supply modeling to reflect the changes in test year
4 characteristics that you have described in your testimony?
5 A. Yes. Under my supervision and at my
6 request, two power supply simulations representative of the
7 test year 2003 under a variety of water conditions were
8 prepared. The first simulation is for the test year 2003
9 prior to known and measurable power supply adjustments.
10 This simulation reflects the load changes, market price
11 changes, fuel cost changes and resource changes that have
12 occurred in the last ten years since the last test year
13 1993. The second simulation modifies the first simulation
14 of the test year to reflect known and measurable power
15 supply adjustments that I will describe later in my
16 testimony. As has been the case in the past, the power
17 supply modeling results reflect the average power supply
18 expenses associated with multiple hydro conditions that are
19 representative of the possible circumstances the Company
20 might encounter. This year the analyses include water
21 conditions corresponding to years 1928 through 2003. The
22 average of the expenses related to each of the 76 water
23 conditions represents the normalization of power supply
SAID, DI 10
Idaho Power Company
1 expenses.
2 Q. Have you supervised the development of an
3 exhibit showing the results of the power supply expense
4 normalization for test year 2003 prior to any known and
5 measurable power supply adjustments?
6 A. Yes. Exhibit 32 shows the results of the
7 power supply expense normalization prior to known and
8 measurable power supply adjustments. Page 1 of Exhibit 32
9 shows the summary results containing the 76-year average
10 power supply generation sources and expenses. Pages 2
11 through 77 contain results for each of the 76 individual
12 water conditions 1928 through 2003.
13 Q. Please summarize the sources and disposition
14 of energy as shown on page 1 of Exhibit 32.
15 A. From the summary information contained on
16 page 1 of Exhibit 32 it can be seen that for the test year
17 2003, hydro generation supplies 8.8 million MWh while
18 thermal generation supplies 6.7 million MWh (Bridger 5.0,
19 Boardman 0.4, Valmy 1.3) from Company-owned generation
20 resources. Danskin, as a peaking plant, operates
21 intermittently, but offers significant contribution at
22 important times when resources and purchases are inadequate
23 to serve peak loads. Purchases of power come from three
SAID, DI 11
Idaho Power Company
1 sources: market purchases, contract purchases other than
2 QF and QF purchases. QF purchases are assumed at fixed
3 normalized levels amounting to 783,635 MWh. Because the
4 fixed QF purchases are fixed inputs to power supply
5 modeling, they are not shown on the variable output
6 summary, however, when combined with the market and other
7 contract purchases, total purchases amount to 1.1 million
8 MWh. As a result, hydro generation contributes
9 approximately 53 percent (8.8 I 16.6) of the generation
10 mix, thermal generation contributes approximately 40
11 percent (6.7 I 16.6) and purchases contribute approximately
12 7 percent (1.1 I 16.6). Of the over 16.6 million MWh
13 consumed, 14.1 million MWh are utilized for system loads
14 while over 2.5 million MWh are sold as surplus.
15 Q. Please describe the expense and revenue
16 information associated with the normalized operation that
17 you have described as shown in Exhibit 32.
18 A. Exhibit 32 contains variable expense and
19 revenue information limited to FERC accounts 501, Fuel
20 (coal); 547, Fuel (gas); 555, Purchased Power; and 447,
21 Sales for Resale. Hydro generation has no assumed fuel
22 expense. Coal expenses of $89.9 million are comprised of
23 Bridger at $63.7 million, Valmy at $20.8 million and
SAID, DI 12
Idaho Power Company
1 Boardman at $5.4 million. Gas expenses amount to $3.2
2 million. Purchased power expenses not including QF amount
3 to $10.6 million while surplus sales amount to $54.1
4 million. Altogether, net power supply expenses amount to
5 $49.6 million (89.9 + 3.2 + 10.6 - 54.1).
6 Q. How do these power supply expenses compare
7 to the 1993 normalized amounts approved by the Commission
8 at the conclusion of the IPC-E-94-5 case.
9 A. Fuel expenses (entirely coal related) for
10 the 1993 normalized test year were $61.5 million.
11 Purchased power not including QF was $11.0 million and
12 surplus sales were at a $24.5 million level. The Company
13 had no gas fuel expenses in 1993. Net power supply
14 expenses were $48 million (61.5 + 11 - 24.5). While
15 normalized surplus sales revenues have increased by $29.6
16 million (54.1 - 24.5), fuel costs have also increased by
17 $31.6 million (89.9 + 3.2 - 61.5). While market prices
18 have increased, reliance on purchases has decreased,
19 resulting in little change to non-QF purchased power
20 expenses. The net change in normalized power supply
21 expenses before known and measurable adjustments is only a
22 $1.9 million increase from 10 years ago.
23 Q. Please describe the types of "known and
SAID, DI 13
Idaho Power Company
1 measurable" power supply adjustments that you recommend in
2 this proceeding.
3 A. I propose two types of known and measurable
4 adjustments to normalized power supply expense
5 computations; (1) changes in purchased power contracts and
6 (2) changes in fuel costs. These adjustments have not only
7 a direct impact on specific expenses, but also have
8 indirect impacts on the Company's market purchase expenses
9 and market sales revenues.
10 Q. Please describe your proposed changes to
11 purchased power contracts that will have a known and
12 measurable impact on the power supply expenses of the
13 Company.
14 A. I propose the inclusion of two power
15 purchase contracts that will become effective in 2004 as
16 new rates are implemented. The first contract, as I
17 mentioned earlier in my testimony, is a PURPA QF contract
18 with Tiber Montana LLC for the acquisition of 29,144 MWh at
19 a cost of $1.2 million. First deliveries of power from
20 Tiber are scheduled for May 2004. The second contract,
21 also mentioned earlier in my testimony, is a PPA with PPL
22 Montana for the purchase of 99,360 MWh at a cost of $4.4
23 million. The first delivery of power from PPL Montana is
SAID, DI 14
Idaho Power Company
1 scheduled for June 2004. This Commission has approved both
2 of these contracts.
3 Q. Please describe your proposed changes to
4 fuel costs that will have a known and measurable impact on
5 power supply expenses.
6 A. I have been informed by employees in the
7 Company's Power Supply Department that certain minor known
8 and measurable changes in coal prices will occur in 2004 as
9 a result of contract provisions, train lease agreements and
10 depreciation. A change of greater significance results
11 from the expiration of a long-term coal contract at Valmy.
12 For two plants, Boardman and Valmy the known and measurable
13 adjustments result in lower per unit fuel costs. Boardman
14 fuel costs drop from $13.66 per MWh to $13.25 per MWh.
15 Valmy fuel will drop from $16.2 per MWh to $14.7 per MWh.
16 At Bridger, the fuel cost rises slightly from $12.65 per
17 MWh to $12.75 per kWh.
18 Q. Have you supervised the development of an
19 exhibit showing the results of the power supply expense
20 normalization when the known and measurable power supply
21 adjustments are included?
22 A. Yes. Exhibit 33 shows the results of the
23 power supply expense normalization once the known and
SAID, DI 15
Idaho Power Company
1 measurable power supply adjustments have been included.
2 Page 1 of Exhibit 33 shows the summary output containing
3 the 76-year average power supply generation sources and
4 expenses. The following pages 2 through 77 show the
5 individual water conditions 1928 through 2003 output as
6 those water conditions would impact the test year 2003.
7 Q. Have you supervised the development of an
8 exhibit to quantify the extent to which the normalized
9 power supply expenses change as a result of including the
10 known and measurable adjustments you have proposed?
11 A. Yes. Exhibit 34 details the changes in both
12 normalized power supply expenses that exclude QF expenses
13 and also the change in QF expenses that result from known
14 and measurable adjustments. Net power supply expenses
15 decrease by $1.9 million as a result of changes to fuel
16 costs and additional power purchase contracts. QF expenses
17 increase by $1.2 million as a result of inclusion of the
18 Tiber contract.
19 Q. How do base level PCA expenses differ from
20 test year power supply expenses?
21 A. Base level PCA expenses differ from test
22 year power supply expenses in two ways. First, base level
23 PCA expenses include QF expenses. Second, base level PCA
SAID, DI 16
Idaho Power Company
1 expenses are determined for an April through March time
2 frame rather than a calendar year. April represents the
3 beginning of the runoff period that provides the basis for
4 the PCA projection.
5 Q. What are the 2003 test year normalized QF
6 expenses including the Tiber project?
7 A. Including the Tiber project, 2003 test year
8 normalized QF expenses amount to $46.4 million.
9 Q. How do 2003 test year normalized QF expenses
10 compare to 1993 test year QF expenses?
11 A. The 2003 test year normalized QF expenses of
12 $46.4 million are $12.1 million greater than the $34.1
13 million 1993 test year normalized QF expenses. However,
14 the $46.4 million value is $1.2 million less than the value
15 used in the current PCA projection formula.
16 Q. What is the base level of PCA expenses for
17 test year 2003?
18 A. As I stated earlier in my testimony, the
19 base level of PCA expenses is the sum of the normalized
20 power supply expenses and normalized QF expenses. In this
21 case, normalized power supply expenses amount to $47.7
22 million and normalized QF expenses amount to $46.4 million.
23 The sum, $94.1 million, represents the new base PCA expense
SAID, DI 17
Idaho Power Company
1 level.
2 Q. Have you directed the preparation of an
3 exhibit that shows the derivation of the appropriate new
4 PCA regression formula to be used for projecting the next
5 year's PCA expenses?
6 A. Yes, I directed the preparation of Exhibit
7 35 to show the derivation of the new PCA regression
8 formula.
9
10
Q.
A.
Please describe Exhibit 35.
Exhibit 35 consists of six columns at the
11 top of the page. Column one shows the number of the
12 observation from 1 to 75. Column 2 contains the PCA year
13 corresponding to each observation; observation 1 is 1928,
14 observation 2 is 1929, and so on through observation 75,
15 which is 2002. Because the PCA year is for months April
16 through March of the following year, there are only 75
17 observations instead of the 76 conditions represented in
18 Exhibit 33. Column 3 contains the April through July
19 runoff for each of the observation years 1928 through 2002.
20 Column 4 contains the natural logarithm of the runoff value
21 contained in Column 3. Column 5 contains the observed
22 April through March annual power supply expense based upon
23 data from Exhibit 33, but reflecting PCA totals rather than
SAID, DI 18
Idaho Power Company
1 calendar year totals. Finally, Column 6 contains the
2 regression predicted value of April through March annual
3 power supply expenses.
4 To the right of the columns are summary output of
5 certain regression statistics (such as r-square) and
6 formula coefficients.
7 Q. Please describe the new PCA regression
8 formula based upon Exhibit 35.
9 A. The basic PCA formula takes the following
10 form: Annual PCA expense= Cl - C2 * ln (Brownlee runoff)
11 + C3. The values of Cl, C2 and C3 are constant with the
12 only variable being Brownlee runoff. The equation without
13 C3 is used to predict net power supply expenses and is the
14 direct result of the regression analysis contained in
15 Exhibit 35. The constant Cl represents the prediction of
16 annual net power supply expense that would occur if there
17 was zero April through July Brownlee runoff. The value of
18 Cl is $1,140,615,325. In reality, the lowest April through
19 July Brownlee runoff contained in the observations is 1.97
20 million acre-feet which occurred in the 1992 observation.
21 Because the regression provides a linear fit of a
22 non-linear transformation, the value of C2 is somewhat
23 difficult to explain. Observed Brownlee runoff data in
SAID, DI 19
Idaho Power Company
1 acre-feet is first transformed by the natural logarithm
2 function. For each unit increase in the natural logarithm
3 of the Brownlee runoff data the projection of annual power
4 supply expenses will be reduced by C2, which is
5 $70,685,112. The average natural logarithm of Brownlee
6 runoff values, based upon the observations contained in
7 Exhibit 35, is 15.46. This value corresponds to a runoff
8 of approximately 5.2 million acre-feet (e A 15.46 =
9 5,178,365 million acre-feet). With a runoff of 5.2 million
10 acre-feet and a natural logarithm of 15.46, the projected
11 net power supply expenses would be $47,823,493
12 ($1, 140, 615, 325 - $70, 685, 112 * 15. 46). An increase of 1
13 to the natural logarithm would result if the runoff was
14 approximately 14.1 million acre-feet (ln(14,076,256) equals
15 16.46 which equals 15.46 + 1). With a runoff of 14,076,266
16 million acre-feet, the net power supply expenses would be
17 $70,685,112 less than $47,823,493 making the projection of
18 power supply expenses a negative $22,861,619
19 ($1,140, 615,325 - $70, 685,112 * 16.46).
20 The natural logarithms of observed Brownlee runoff
21 ranged from 14.49 (1992 runoff) to 16.35 (1984 runoff).
22 The difference, 1.86 (16.35 - 14.49), multiplied by
23 $70,685,112 equals approximately $131.5 million, which
SAID, DI 20
Idaho Power Company
1 represents the change in projected power supply expenses
2 from the highest water case (1984) to the lowest water case
3 (1992).
4 The value of C3 is $46,413,000, the normalized
5 expense for QF. Because the normalized expense for QF is
6 quantified separately from net power supply expenses it is
7 added to net power supply expenses to determined the PCA
8 expenses.
9 Q. What is the new PCA regression equation with
10 values inserted for the constants?
11 A. The new PCA regression equation is:
12 Annual PCA expense= $1,140,615,325
13 - $70,685,112 * ln (Brownlee runoff)
14 + $46,413,000.
15 Q. In the past, has the PCA regression equation
16 also contained a constant related to FMC, later Astaris,
17 second block revenues?
18 A. Yes, FMC second block revenues were
19 previously treated as separately identified revenue that,
20 like surplus sales, reduced net PCA expenses. The FMC
21 constant is no longer appropriate due to the cancellation
22 of the FMC contract.
23 Q. How does the range in projected power supply
SAID, DI 21
Idaho Power Company
1 expenses from high condition to low condition resulting
2 from this regression equation compare to the range of
3 projected power supply expenses in the previous regression
4 equation?
5 A. The predictions of power supply expenses
6 based upon the regression observations contained in the
7 previous regression analysis ranged from minus $9.9 million
8 (1984) to $112.4 million (1992), a range of $122.3 million.
9 Q. Do you recommend any additional PCA
10 computational changes with the establishment of the new PCA
11 regression formula?
12 A. Yes. There are three PCA computational
13 factors that need to be updated as a result of the current
14 review of power supply expenses. First, for PCA projection
15 calculations, a new normalized base PCA rate can be
16 determined. Second, a new Idaho jurisdictional percentage
17 can be determined. Third a new expense adjustment rate to
18 be applied to load growth or decline can be determined.
19 Q. Have you supervised the development of an
20 exhibit to determine the PCA computational factors you have
21 just mentioned?
22 A. Yes, Exhibit 36 is a one-page exhibit
23 detailing the appropriate computation of the PCA factors I
SAID, DI 22
Idaho Power Company
1 have outlined.
2 Q. What is the first computation shown on
3 Exhibit 36?
4 A. The first computation recaps the normalized
5 PCA computation that I have discussed thoroughly in my
6 testimony. The new normalized PCA expenses for 2003 test
7 year amount to $94.1 million compared to the previous $73.1
8 million value for the 1993 test year.
9 Q. Please discuss the normalized Base PCA rate
10 computation contained in Exhibit 36.
11 A. First, I would point out that in my opinion,
12 the normalized Base PCA rate has been improperly determined
13 in the past. While expenses are incurred based upon loads,
14 they are recovered based upon sales. Historically, the
15 normalized Base PCA rate of 0.5238 was determined by
16 dividing the $73.1 million of normalized PCA expenses by
17 the normalized system firm load value. My recommendation
18 for the current computation of the normalized Base PCA rate
19 is that the $94.1 million normalized PCA expenses be
20 divided by the normalized system sales value of 12,863,484
21 MWh. The resulting PCA base rate is 0.7315 cents per kWh.
22 Q. Was a similar load/sales error previously
23 corrected by the Commission?
SAID, DI 23
Idaho Power Company
1 A. Yes, PCA true-up rate computations were
2 originally based upon Idaho jurisdictional firm loads
3 rather than Idaho jurisdictional firm sales levels. In
4 1996, the Commission corrected that error in Order No.
5 26455.
6 Q. Please discuss the Idaho jurisdictional
7 percentage computation contained in Exhibit 36.
8 A. The Idaho jurisdictional percentage is
9 derived by dividing the Idaho jurisdictional firm load by
10 the system firm load number. As I mentioned earlier in my
11 testimony, the Company's FERC jurisdictional contract loads
12 have been reduced by 1.4 million MWh while at the same time
13 Idaho jurisdictional loads have grown. As a result, Idaho
14 jurisdictional loads now represent 94.1 percent of the
15 Company's total load.
16 Q. Please discuss the Expense Adjustment rate
17 to be applied to load changes for PCA true-up computations.
18 A. When the PCA was established, the Commission
19 recognized that load growth would provide additional
20 revenue that would in part offset the corresponding
21 additional power supply expenses incurred to serve the
22 additional load. The revenues generated would be the
23 result of rates designed to recover the full embedded costs
SAID, DI 24
Idaho Power Company
1 of serving existing customers including generation costs,
2 distribution costs, transmission costs and other costs of
3 the Company. However, the true cost of serving additional
4 customers is comprised of a blend of new marginal costs
5 incurred to serve new customers and reduced embedded costs
6 when existing facilities allow for additional customers at
7 zero or low cost. The Commission determined that rates
8 paid by new customers would cover all additional costs
9 including $16.84 per MWh of PCA expenses that might occur
10 to serve additional load. The $16.84 per MWh credit was
11 computed by averaging the Boardman and Valmy fuel costs.
12 Using the same computational method the new expense
13 adjustment rate for load changes is $13.98 per MWh.
14 Q. Based upon your understanding of Mr. Keen's
15 testimony in this proceeding, do you believe the $13.98 per
16 MWh rate should be used as the new credit for load growth?
17 A. No. Mr. Keen pointed out that whether
18 looking at generation, distribution, or transmission, the
19 Company has little ability to serve additional customers
20 without investment in new facilities. In my opinion,
21 revenues derived from additional customers served at
22 embedded rates will not be sufficient to recover both the
23 incremental costs of required new facilities and an amount
SAID, DI 25
Idaho Power Company
1 greater than the embedded cost of PCA expenses (the PCA
2 base rate). I believe it would be more appropriate to have
3 a load growth credit based upon the normalized PCA base
4 rate of $7.30 per MWh (7.3 mills per kWh). That is the
5 portion of customers' rates that it is contemplated will
6 cover base PCA expenses. The remainder of customers' rates
7 cover the other than PCA expenses that Mr. Keen has
8 suggested will grow at a significant pace in the coming
9 years.
10 Q. Do you have a non-computational
11 recommendation with regard to the PCA?
12 A. Yes. Mr. Gale, Ms. Brilz and I have
13 discussed Ms. Brilz' recommendations in this proceeding to
14 create seasonal pricing that if accepted would create a
15 seasonal rate change on June 1 of each year. If the PCA
16 rate change date were to continue to occur on May 16 of
17 each year, customers would see two rate changes within 16
18 days. If Ms. Brilz' seasonal pricing recommendations are
19 approved, then in order to eliminate back-to-back rate
20 changes, I recommend that the PCA recovery period be moved
21 from a May 16 through May 15 period to a June 1 through May
22 31 time period. No other changes to PCA time frames would
23 be required. PCA projection and true-up computations would
SAID, DI 26
Idaho Power Company
1 still be based upon an April 1 through March 31 time frame
2 and the Company would still file its PCA request by April
3 15 each year.
4
5
Q.
A.
Does that conclude your testimony?
Yes.
SAID, DI 27
Idaho Power Company