20100323Kalich Di.pdf

r~::: 1
DAVID J. MEYER
VICE PRESIDENT AND CHIEF COUNSEL OF
REGULATORY & GOVERNMENTAL AFFAIRS
AVISTA CORPORATION
P.O. BOX 3727
1411 EAST MISSION AVENUE
SPOKANE, WASHINGTON 99220-3727
TELEPHONE: [phone]
FACSIMILE: [phone]
!H~'1 -,f ...v /I: 04
BEFORE THE IDAHO PUBLIC UTILITIES COMMISSION
) CASE NO. AVU-E-10-01
)
)
)
)
)
)
)
DIRECT TESTIMONY
OF
CLINT G. KALICH
IN THE MATTER OF THE APPLICATION
OF AVISTA CORPORATION FOR THE
AUTHORITY TO INCREASE ITS RATES
AND CHARGES FOR ELECTRIC AND
NATURAL GAS SERVICE TO ELECTRIC
AND NATURAL GAS CUSTOMERS IN THE
STATE OF IDAHO
FOR AVISTA CORPORATION
(ELECTRIC ONLY)
1
2
I. INTODUCTION
Q.Please state your nam, the nam of your
3 employer, and your business address.
4
5
A.My name is Clint Kalich. I am employed by Avista
Corporation at 1411 East Mission Avenue,Spokane,
6 Washington.
7
8
Q.In what capacity are you emloyed?
A.I am the Manager of Resource Planning & Power
9 Supply Analyses, in the Energy Resources Department of
10 Avista Utilities.
11 Q.Please state your educational background and
12 professional exprience.
13 A.I graduated from Central Washington University in
14 1991 with a Bachelor of Science Degree in Business
15 Economics. Shortly after graduation, I accepted an analyst
16 posi tion with Economic and Engineering Services, Inc. (now
17 EES Consulting, Inc.), a Northwest management-consulting
18 firm located in Bellevue, Washington.While employed by
19 EES, I worked primarily for municipalities, public utility
20 districts, and cooperatives in the area of electric utility
21 management.My specific areas of focus were economic
22 analyses of new resource development, rate case proceedings
23 involving the Bonneville Power Administration, integrated
24 (least-cost) resource planning, and demand-side management
25 program development.
Kalich, Di i
Avista Corporation
1 In late 1995,I left Economic and Engineering
2 Services, Inc. to join Tacoma Power in Tacoma, Washington.
3 i provided key analytical and policy support in the areas
4 of resource development, procurement, and optimization,
5 hydroelectric operations and re-licensing, unbundled power
6
7
supply rate-making,contract negotiations, and system
operations.I helped develop, and ultimately managed,
8 Tacoma Power's industrial market access program serving
9 one-quarter of the company's retail load.
10 In mid-2000 I joined the Company and accepted my
11 current position assisting in resource analysis, dispatch
12 modeling,resource procurement,integrated resource
13 planning, and rate case proceedings. Much of my career has
14 invol ved resource dispatch modeling of the nature described
15 in this testimony.
16 Q.What is the scope of your testimony in this
17 proceeding?
18 A.My testimony will describe. the Company's use of
19 the AURORAXMP dispatch model, or "Dispatch Model."I will
20 explain the key assumptions driving the Dispatch Model's
21 market forecast of electricity prices.The discussion
22 includes the variables of natural gas, Western Interconnect
23 loads and resources, and hydroelectric conditions.I will
24 describe how the model dispatches our resources and
Kalich, Di 2
Avista Corporation
1 contracts in a manner that maximizes benefits to customers
2 and tracks their values for use in pro forma calculations.
3 I will then present the modeling results provided to
4 Company witness Mr. Johnson for his power supply pro forma
5 adjustment calculations.Addi tionally, in support of
6 Company witness Ms. Knox, I detail the Company's demand
7 classification calculations.
8 Q.Are you sponsoring any exhibi ts in this
9 proceeding?
10 A.Yes. I am sponsoring ExhibitS, Schedules 1 and
11 2, as well Confidential Schedule 3. Schedule 1 provides a
12 forecast of Company load and resource positions from 2011
13 through 2020.Schedule 2 is the spreadsheet used to
14 calculate the demand classification. Confidential Schedule
15 3 provides summary output from the Dispatch Model.All
16 information contained in the exhibit was prepared by me or
17 prepared under my direction.
18 II. THE DISPATCH MODEL
19 Q.What model is the Comany using to dispatch its
20 portfolio of resources and obligations?
21 A.The Company uses EPIS, Inc.'s Dispatch Model for
22 determining power supply costs.The model optimizes
23 dispatch of Company-owned resources and contracts in each
24
25
hour of the pro forma year.The pro forma period is
October 1, 2010 through September 30, 2011.It reflects
Kalich, Di 3
Avista Corporation
1 true system operations by evaluating future resource
2 decisions on an hourly basis.
3 Q.What AURORA version and database is the Company
4 using for this case?
5 A.The Company is using AURORAxMP version 9.6.1033,
6 and its associated database (North_American_DB_2009-02).
7
8
9
Q.Please briefly describe the Dispatch Model.
The Dispatch Model was developed by EPIS, Inc. ofA.
Sandpoint,Idaho.It is a fundamentals-based tool
10 containing demand and resource data for the entire Western
11 Interconnect.It employs multi-area,transmission-
12 constrained dispatch logic to simulate real market
13 condi tions.Its true economic dispatch captures the
14 dynamics and economics of electricity markets--both short-
15 term (hourly, daily, monthly) and long-term. On an hourly
16 basis, the Dispatch Model develops an available resource
17 stack, sorting resources from lowest to highest cost.It
18 then compares this resource stack with load obligations in
19 the same hour to arrive at the least-cost market-clearing
20 price for the hour.Once resources are dispatched and
21 market prices are determined, the Dispatch Model singles
22 out Avista resources and loads and values them against the
23 marketplace.
Kalich, Di 4
Avista Corporation
1 Q.What experience does the Comany have using
2 AUR0RA ?
3
4
A.The Company purchased a license to use the
Dispatch Model in April 2002.AURORA~ has been used for
5 numerous studies, including all of our integrated resource
6 plans and rate filings after 2001. The tool is also used
7 for various resource evaluations, market forecasting, and
8 requests-for-proposal evaluations.
9 Q. Who else uses AUR0RA?
10 A. AURORAXMP is used all across North America and in
11 Europe. In the Northwest specifically, AURORAXMP is used by
12 the Bonneville Power Administration, the Northwest Power
13 and Conservation Council, Puget Sound Energy, Idaho Power,
14 Portland General Electric, Seattle City Light, Grant County
15 PUD, Snohomish County PUD, and Tacoma Power.
16 Q.What benefits does the Dispatch Model offer for
17 this type of analysis?
18 A.The Dispatch Model generates hourly electricity
19 prices across the Western Interconnect, accounting for its
20 specific mix of resources and loads.The Dispatch Model
21 reflects the impact of regions outside the Northwest on
22
23
Northwest market prices,limited by known transfer
(transmission) capabilities.Ultimately, the Dispatch
24 Model allows the Company to generate price forecasts in-
25 house instead of relying on exogenous forecasts.
Kalich, Di 5
Avista Corporation
1 The Company owns a number of resources, including
2 hydroelectric plants and natural gas-fired peaking units,
3 which serve customer loads during more valuable on-peak
4 hours.By optimizing resource operation on an hourly
5 basis, the Dispatch Model is able to appropriately value
6 the capabilities of these assets. For example, actual 2008
7 and 2009 on-peak prices were 23 percent higher than off-
8 peak prices.2007 on-peak prices were 25 percent higher.
9 Forward on-peak prices for 2011 were 27 percent higher than
10 off-peak prices at the time this case was prepared.For
11 comparison, Dispatch Model on-peak prices for the pro forma
12 period average 29 percent higher than off-peak prices. In
13 summary, the Dispatch Model appropriately values the energy
14 from Avista's resources during on-peak periods in a manner
15 similar to that recently experienced in the Northwest
16 region.
17 Q.On a broader scale, what calculations are being
18 performd by the Dispatch Model?
19 A.The Dispatch Model's goal is to minimize overall
20 system operating costs across the Western Interconnect,
21 including Avista's portfolio of loads and resources.The
22 dispatch model generates a wholesale electric market price
23 forecast by evaluating all Western Interconnect resources
24 simultaneously in a least-cost equation to meet regional
25 loads. As the Dispatch Model progresses from hour to hour,
Kalich, Di 6
Avista Corporation
1 it "operates" those least-cost resources necessary to meet
2 load.With respect to the Company's portfolio, the
3 Dispatch Model tracks the hourly output and fuel costs
4 associated with portfolio generation.It also calculates
5 hourly energy quantities and values for the Company's
6 contractual rights and obligations.In every hour the
7 Company's loads and obligations are compared to available
8 resources to determine a net position.This position is
9 balanced using the simulated wholesale electricity market.
10 The cost of energy purchased from or sold into the market
11 is determined based on the electric market-clearing price
12 for the specified hour and the amount of energy necessary
13 to balance loads and resources.
14 Q.How does the Dispatch Model determne electric
15 market prices?
16 A.The Dispatch Model calculates electricity prices
17 for the entire Western Interconnect, separated into various
18 geographical areas such as the Northwest and Northern and
19 Southern California. The load in each area is compared to
20 available resources, including resources available from
21 other areas that are linked by transmission corridors, to
22 determine the electricity price in each hour. Ultimately,
23 the market price for an hour is set based on the last
24 resource in the stack to be dispatched.This resource is
25 referred to as the "marginal resource."Given the
Kalich, Di 7
Avista Corporation
1 prominence of natural gas-fired resources on the margin,
2 this fuel is a key variable in the determination of
3 wholesale electricity prices.
4 Q.How does the Dispatch Model operate regional
5 hydroelectric projects?
6 A.The model begins by "peak shaving" loads using
7 system hydro resources.When peak shaving, the Dispatch
8 Model determines which hours contain the highest loads and
9 allocates to them as much hydroelectric energy as possible.
10 Remaining loads are then met wi th other available
11 resources.
12 Q.Has the Company made any modifications to the
13 dataase for this case?
14 A.Yes. Avista's portfolio of resources is modified
15 to reflect actual operating characteristics, natural gas
16 prices are modified to match projected forward prices over
17 the pro-forma period, regional resources are modified where
18 better information is known, and Northwest hydro data is
19 replaced with Northwest Power Pool data.
20
21
22
III . HYDRO MODELING ASSUMTIONS
Q. How has the Company modeled hydroelectric
23 generation for this case?
24 A. As in the past, Avista uses historical stream flow
25 data from the Northwest Power Pool (NWPP) to determine
Kalich, Di 8
Avista Corporation
1 hydroelectric generation for its Clark Fork and Spokane
2 Ri ver systems.Certain adjustments to the NWPP data are
3 necessary to yield a proper estimate of generation from the
4 model.These adjustments include changes to address the
5 NWPP's tendency to overstate generation in high-flow
periods,to account for recent upgrades at our6
7
8
hydroelectric projects,to maintain year-to-year
consistency in project operations,to account for
9 encroachment on our Mid-Columbia project shares, and to
10 allow for 2000 irrigation depletion levels.
11 Q. Why does the NWP overstate generation on the
12 Company's hydroelectric facilities?
13 A. The NWPP's regional hydroelectric model is in many
14 ways simplified and therefore does not account for various
15 project operating characteristics.The NWPP model is not
16 granular enough to account for intra-month flow changes.
17 This impact is most significant during the spring months.
18 For example, the Noxon Rapids project has a maximum turbine
19 flow capability of approximately 50,000 cubic feet per
20 second (cfs).The NWPP model will use all water up to
21 50,000 cfs in a given month to generate power. However, a
22 50,000 cfs month is not comprised of 28, 29, 30 or 31 days
23 of 50, OOO-cfs flows.Instead it is made up of flows that
24 range below and above 50,000 cfs. For example, where flows
25 are 20,000 cfs for the first half of the month and 80,000
Kalich, Di 9
Avista Corporation
1 cfs the second half, the average flow for the period is
2
3
50,000 cfs.The NWPP would assume all of this water went
through the generation turbines and made power.In fact,
4 the project would in the first half of the month generate
5 with 20,000 cfs and in the second half of the month it
6 would generate with 50,000 cfs. The additional 30,000 cfs
7 in the second half of the month (80,000 - 50,000 = 30,000),
8 or nearly 30 percent of the monthly total, would be spilled
9 in the actual operation of the project.
10 Q. Does Noxon Rapids have storage capability to
11 account for such variations in flows?
12 A. Noxon does have some storage, but not near enough
13 to convert all of the intra-month variability of flows into
14 electric energy.A study completed by BorisMetrics
15 explained that on average our hydroelectric dams on the
16 Spokane and Clark Fork Rivers generate 3.7 aMW less than
17 the NWPP estimates.This study was reviewed and accepted
18 in previous cases before this Commission.
19 Q. Is the Company now experiencing an even greater
20 difference between actual hydroelectric generation and
21 generation from the NWP model, than that quantified by
22 BorisMetrics?
23 A. Yes.Relative to the NWPP data used in previous
24 cases, hydro generation on the Clark Fork projects has been
25 overstated by a significant amount on average.Over the
Kalich, Di 10
Avista Corporation
1 past 20 years actual hydroelectric generation has been
2 319.72 aMW, 3.2 percent (10 aMW) below the NWPP model
3 resul ts for the 50-year period used in rate modeling. Over
4 the past 10 years generation has been 299.08 aMW, or 10.3
5 percent (31 aMW) below the NWPP modeled results.Lower
6 results in the past 10 years have been driven primarily by
7 lower-than-average stream flows; however, not all of the
8 reduction is driven by lower stream flows. A portion of
9 the overstatement is caused by the design limitations of
10 the model itself.
11 Q. Please provide additional detail as to why the 10-
12 and 20-year averages were below the 50-year NWP study
13 period average?
14 A. There are a number of reasons. Flows in the 1990s
15 were high relative to history, whereas flows in the most
16 recent 10 years have been low relative to average. Also,
17 half of the 20-year average is affected by the use of
18 operating assumptions from our old Clark Fork operating
19 license.New licensing requirements implemented in 2001
20 have negatively affected power production on the Clark Fork
21 projects. Poor hydroelectric conditions also have played a
22 role in a number of recent years.Additionally, the
23 Company continues to shift reserve obligations to the Clark
24 Fork as we lose Mid-Columbia generation capacity, and as we
25 respond to a marketplace greatly affected by new variable
Kalich, Di 11
Avista Corporation
1 generation resources (i. e. , wind).Upgrades at Cabinet
2 Gorge and Noxon Rapids have helped to offset these losses,
3 but the statistics explain that generation levels continue
4 to fall over time.
5 Q. How is hydro generation calculated in this
6 proceeding?
7
8
A. For our Mid-Columbia shares, and for the Spokane
River,there is no change from previous filings.
9 Generation data are taken from the NWPP Headwater Benefits
10 Study, adjusted downward by the results of the BorisMetrics
11 study for the Spokane River and Encroachment for the Mid-
12 Columbia projects.For the Clark Fork River projects we
13 continue to use NWPP data for the historical record (1929-
14 1978). However, instead of using energy levels calculated
15 by their model, and adjusted by the BorisMetrics study for
16 overstated generation, the NWPP flow data is used as an
17 input in a new model: the Clark Fork Optimization Package.
18 Q. Please describe. the Clark Fork Optimzation
19 Package.
20 A. The Clark Fork Optimization Package is a mixed-
21 integer linear programming-based system emulating the
22 operation of the Company's Clark Fork projects.It was
23 developed in support of the Company's system operations,
24 financial forecasting,and hydro upgrade efforts.
25 Operating on an hourly time-step, it accurately represents
Kalich, Di 12
Avista Corporation
1 individual turbine and reservoir operations.License
2 constraints (e.g., minimum flows, elevation limits) are
3 honored in all periods.The Clark Fork Optimization
4 Package is comprised of four components which are described
5 below.
6 Q. In wha t programng language was the model
7 developed?
8 A. The Clark Fork Optimization Package is a suite of
9 database (Microsoft Access) and spreadsheet (Microsoft
10 Excel)programs.The Excel programs benefi t from
11 WhatsBEST!, an Excel Add-In for Linear, Nonlinear, and
12 Integer Modeling and Optimization.WhatsBEST!was
13 developed by Lindo Systems of Chicago, Illinois in 1979.
14 Q. What is the first component of the Clark Fork
15 Optimization Package?
16 A. The first component is the Clark Fork Water Budget
17 Model. It looks over the long-term record and optimizes
18 water flow through the projects to maximize generation
19
20
values.This step is necessary to recognize the storage
capabilities inherent in a hydro project.The long-term
21 optimization is simplified to provide present-day computers
22 wi th the ability to efficiently solve the equations. Each
23 project is represented by one power curve instead of
24 mul tiple curves representing individual turbines.Model
Kalich, Di 13
Avista Corporation
1 granularity is daily instead of hourly. Project elevation
2 and flow constraints are retained.
3 Outputs of the Clark Fork Water Budget Model are
4 weekly beginning and ending project elevations for the
5 Noxon Rapids and Cabinet Gorge projects. These elevations
6 are exported to the second module of the Clark Fork
7 Optimization Package--the Clark Fork Optimization Model
8 Input Database. It is discussed below.
9 Q. What is the source for hydroelectric flows in the
10 Clark Fork Water Budget Model?
11 A. The source is the 2007-08 NWPP Headwater Benefits
12 Study. To shape the monthly NWPP data Avista used a daily
13 study obtained from the Bonneville Power Administration
14 (BPA) .The BPA data were from the U. S. Army Corp of
15 Engineers study re-creating daily historical flows on the
16 Clark Fork River back to 1929 based on today's river
17 system.
18 Because of the need for daily inflow values that
19 the NWPP does not provide, and the fact that the BPA data
20 is daily, Avista elected to shape the NWPP monthly data
21 using the daily shapes of the BPA study in each month.
22 Q. What data does the Clark Fork Optimzation Model
23 Input Database contain?
24 A. The Clark Fork Optimization Model Input Database
25 contains the daily inflows and side flows into the
Kalich, Di 14
Avista Corporation
1 Company's Clark Fork River projects described above.It
2 also contains representative hourly market prices enabling
3 the model to maximize generation levels in the higher-
4 valued on-peak periods.
5 Q. What is the third elemnt of the Clark Fork
6 Optimization Package?
7
8
A. The third element is the Clark Fork Optimization
Model itself.This hourly model uses a mixed-integer
9 optimization routine to maximize the value of the Clark
10 Fork proj ects over time.Each project is represented in
11 detail, including individual turbine efficiency curves,
12 physical and license-constrained reservoir elevations,
13 tailrace elevations,and minimum and maximum flow
14 constraints.
15 The Clark Fork Optimization Model shapes
16 generation into the most economically beneficial time
17 periods using the projects' storage reservoirs.It also
18 maximizes the value of generation by flowing water through
19 the turbines at their most economically efficient points on
20 the power curves.
21 Q. What is the fourth elemnt of the Clark Fork
22 Optimzation Package?
23 A. The fourth element is the Clark Fork Optimization
24 Model Output Database. This database contains results from
25 the Clark Fork Optimization Model, including hourly turbine
Kalich, Di 15
Avista Corporation
1 discharge and spill flows, hourly generation levels, hourly
2 generation values, and hourly reservoir elevations.
3 Q. How did the Company ensure the Clark Fork
4 Optimization Package accurately reflects the operations and
5 value of the Clark Fork projects?
6 A. Once the Clark Fork Optimization Package models
7 were completed, it was benchmarked against the Company's
8 2000-2009 actual results at the Clark Fork projects to
9 ensure its accuracy.
10
11
Q. How did the results comare?
A. The Clark Fork Optimization Package initially
12 over-estimated generation relative to the 2000-2009 periods
13 by approximately 6 percent.This result was expected, as
14 Avista does not operate its projects in isolation. Instead
15 the Company uses the Clark Fork projects to meet its load
16
17
and reserve needs.There are also times where units are
down for maintenance or forced outage.To reconcile the
18 Clark Fork Optimization Package with actual operating
19 history, the power curves for each project were therefore
20 reduced by the 6 percent difference.After the
21 benchmarking process, the model generated just over 100
22 percent of actual generation levels during the 2000-2009
23 period.
24 Q. How is the generation then used for ratemking
25 purposes?
Kalich, Di 16
Avista Corporation
1 A. The generation levels for each project (Mid-
2 Columbia, Spokane River, and Clark Fork) are input into the
3 dispatch model (AURORAxmp) where Avista's portfolio value
4 is quantified for ratemaking purposes.
5 Q. Are the models included in the Company's filing?
6 A. Yes.All four components of the Clark Fork
7 Optimization Package are included in my workpapers,
8 including all input and output data.
9 Q. Does the Clark Fork Optimzation Package account
10 for recent upgrades at the Noxon Rapids project?
11 A. Yes.Once the original model was benchmarked
12 against recent generation years that did not benefit from
13 upgrades at Noxon, the three newly upgraded units (1, 2,
14 and 3) were input into the model to reflect the higher
15 anticipated generation levels.As Unit 2 will not enter
16 service until April 1, 2011, all proforma periods prior to
17 April 2011 include upgrades only to Units 1 and 3.
18 Q. How much additional genera tion did the new units
19 provide based on your modeling?
20 A. The Company evaluated generation levels with the
21 old Noxon units 1 through 3, and the newly upgraded units
22 over the 50-year period for this case.Generation levels
23 from the upgrades increased by a total of 35,778 MWh (4.08
24 aMW) a year, or 1.3 percent.
Kalich, Di 17
Avista Corporation
1 Q. How much additional generation does the new Unit 2
2 provide?
3 A. On an annual basis the new Unit 2 included in this
4 case generates 10,326 MWh per year on average over the 50-
S year period, or 1.18 aMW.
6 Q. Why did the Company not use simlar models in this
7 case for the Spokane River and Mid-Columia projects?
8
9
A. The Clark Fork Optimization Package is the product
of several years of work by Avista.The Company has not
10 yet attempted to build a model for the Mid-Columbia due to
11 those projects' significant reliance on upstream (e. g.,
12 Grand Coulee Dam) projects that greatly affect their
13
14
output.A model for the Spokane River projects is under
development but is not yet ready for use.The Company
15 hopes to have a working version for the Spokane River
16 system prior to its next rate proceeding.We will
17 subsequently examine a model for the Mid-Columbia projects.
18 Q. Please explain why the Company developed the Clark
19 Fork Optimzation Package.
20 A. The Clark Fork Optimization Package is the
21 culmination of nearly ten years of work by the Company to
22 bring in-house a tool to enable true optimization of our
23 hydro facilities.In 2002 the Company acquired the Vista
24 suite from Synexus Global. This tool was used to evaluate
25 system operations and support upgrades at our Noxon Rapids
Kalich, Di 18
Avista Corporation
1
2
and Cabinet Gorge projects.It also was used to evaluate
various Spokane River proj ect upgrades.Because of some
3 problems inherent to the Vista model, and very slow
4 solution times, it. was retired in the middle of the last
5 decade.We then evaluated other options in the
6 marketplace, and the Company acquired Riverware from the
7 University of Colorado at Boulder. After working with this
8 tool over a number of years it became apparent that it
9 cannot meet our need for efficient unit-level dispatch
10 modeling.
11 Due to the apparent lack of a strong package for
12 hydro modeling in the marketplace, the Company began
13 developing the Clark Fork Optimization Package in the
14 middle of 2009.
15 Q. How is the Company using the new Clark Fork
16 Optimzation Package in its business operations, and how
17 does it intend to use the tool into the future?
18 A. The Clark Fork Optimization Package is an
19 essential tool to assist the Company with optimizing hydro
20 system operations, both in short- and long-term planning.
21 Its results are also used for Company budgets, hydro
22 project market valuation studies, and upgrade studies.
23 Gi ven its solution efficiency, it is possible to run large
24 hydro-flow records through it, as is necessary for rate
25 filings such this.
Kalich, Di 19
Avista Corporation
1 The Company anticipates using its new model to analyze
2 opportuni ties to increase the value of the Clark Fork
3 projects and lower overall system costs to customers. With
4 this model there is now a potential to analyze a
5 coordination agreement between Clark Fork River project
6 operators that would be similar to the Pacific Northwest
7 Coordination Agreement. Initiation of discussions on this
8 a potential agreement between the various parties with
9 projects on the river has been hampered to a large extent
10 by the lack of a good means to model the values of
11 coordination.
12 Q.How does the AURORA Dispatch Model operate
13 Company-controlled hydroelectric generation resources?
14 A.The Dispatch Model treats all hydroelectric
15 generation plants wi thin a load area as a single large
16 plant. The Company's hydroelectric plants are on average,
17 however, more flexible than the average plant used in each
18 load area. To account for this additional flexibility, the
19 Company algebraically extracts its plants from the region
20 and develops individual hydro operations logic for them.
21 Company-controlled hydroelectric resources are separated
22 into three river systems:the Spokane River, the Clark
23 Fork River, and individually separate the Mid-Columbia
24 projects.This separation ensures that the flexibility
Kalich, Di 20
Avista Corporation
1 inherent in these resources is credited to customers in the
2 pro forma exercise.
3 Q.Please compare the operating statistics from the
4 Dispatch Model to recent historical hydroelectric plant
5 operations.
6 A.Over the pro forma period the Dispatch Model
7 generates 69 percent of Clark Fork hydro generation during
8 on-peak hours (based on the average of the 50 year hydro
9 record) . Since on-peak hours represent only 57 percent of
10 the year, this demonstrates a substantial shift of hydro
11 resources to the more valuable on-peak hours.This is
12 identical to the 5-year average of on-peak hydroelectric
13 generation at the Clark Fork through 2009.Similar
14 performance is achieved for the Spokane and Mid-Columbia
15 projects.
16
17
18
iv. OTHE KEY MODELING ASSUMTIONS
Q.Please describe your update to pro form period
19 natural gas prices.
20 A.Natural gas prices for this filing are based on a
21 3-month average from October 1, 2009 to December 31, 2009
22 of the rate period forward prices.Natural gas prices
23 used in the Dispatch Model are presented below in Table No
24 i.
25
Kalich, Di 21
Avista Corporation
1 Table No. i - Pro Form Natural Gas Prices
2
3
AECO
CHICAGO
CIG
EL PASO
MALIN
NECT
NWPCRM
5.957 PG&E CITY
6.50 RATHDRUM
5.882 SJUAN BASIN
6.056 SOCAL
6.345 STANFIELD
6.566 SUMAS
5.90 Hen Hub
6.709
6.265
5.975
6.277
6.265
6.372
6.424
Q.What is the Company's assumtion for rate period
4 loads?
5 A.Rate period loads (October 2010 through September
6 2011) used in this case are taken from the Company's load
7 forecast completed in July 2009. As this load is generated
8 using "normal weather," it eliminates the need for a
9 weather-normalization adjustment. Removing the 2009 actual
10 (test year) generation from the Clearwater (previously
11 known as Potlatch) cogeneration facility, from the October
12 2010 to September 2011 proforma period loads, results in
13 system loads of 1,070.4 aMW as filed in this proceeding.
14 The Company's latest energy loads and resources
15 tabulation (L&R) is attached in Exhibit No.5, Schedule 1.
16 Q.Please discuss the availability assumtions for
17 your therml and gas generating facilities.
18 A.For baseload generating facilities such as Coyote
19 Springs 2, Kettle Falls Generating Station, and Colstrip,
20 we use a 5-year average through 2009 to estimate long-run
21 operating performance. The following table summarizes the
Kalich, Di 22
Avista Corporation
1 average forced outage rates for each of the Company's
2 thermal and gas generation facilities.
3 Table No.2 - Equivalent Forced Outage Rates (EFOR) Of
4 Avista Therml and Gas Plants
5
6
Plant EFOR Plant EFOR
Colstrip 9.36%Rathdrum 5.00
Coyote Springs 5.07%Northeast 5.00
Lancaster 3.00 Kettle Falls 1.58%
Boulder Park 15.00 Kettle Falls CT 5.00
Q.Colstrip had an extended outage in 2009. Would
7 it be reasonable to exclude this single year from the
8 average?
9 A.No. In the past, various parties have advocated
10 elimination of years where the Colstrip plant had a high
11 forced outage rate, assuming that such years were abnormal
12
13
and should not be expected to re-occur.This is in fact
not the case.The 5-year average of 9.36 percent falls
14 well below the 11.6 percent lifetime plant average. In the
15 25-year history of Colstrip operations there have been
16 seven years (one event every 3.7 years) where forced outage
17 rates exceed 10 percent. It is therefore not uncommon for
18 some years to have outages like the one experienced in
19 2009. See Chart No. 1 for a history of forced outages at
20 Colstrip.
21
Kalich, Di 23
Avista Corporation
1 Chart No. 1 - Colstri
Colstrip Unit 3 & 4
Forced Outage History (1984200)
50
45 ~~~-~
~!
~;¡:O:~i--i\
~
¡
~~:
;:,t, W"
:~~~;¡"
f¡-~-..
i,,"AnnuaIEFOR ..5yrRollingAvg -1984-2009Avg
40
35
~,30
2
3
4
..on ID ~00 CI 0 ..N ..I on =..00 CI 0 ..N ..~on ID ..I0000000000CICICICICICICICI0000ClCl0
CI CI CI CI CI CI Ol Ol CI CI Ol Ol Ol Ol CI Ol ~~0 Cl ~~~~................................N N
Q.Please provide a sumry of the monthly and
average Northwest forward natural gas and electricity
o
I
i
l~
t'~
ai
o 25LIW
!-$-
,tliit
~
20
15
10
5
5 prices that directly affect proform costs.
6 A.Table No. 3 presents monthly modeled natural gas
7 and electricity prices for this case.
8 Table No. 3 - Dispatch Model Prices Sumry
Oct-10
Nov-10
Dec-10
Jan-11
Felr11
Mar-11
41.15
33.30
31.31
48.91
57.38
55.77
49.19
9
Kalich, Di 24
Avista Corporation
1 Q.Are Mid-Columia electric prices from the
2 Dispatch Model the sam as the Forward Market?
3 A.No,Mid-Columbia electric prices from the
4 Dispatch Model differ from the forward market for a variety
5 of reasons. This being said, they generally are very close
6 as in this filing. Forward market prices are not only an
7 expectation of future prices,but they contain an
8 adjustment for risk or unknown future conditions, based on
9 the premise you can "lock in" prices.The Dispatch Model
10 is a spot market model that forecasts prices for a specific
11 time in the future given load, hydro, and fuel price
12 condi tions .Average annual Mid-Columbia prices in the
13 forward market are $54. 19/MWh on-peak and $42. 56/MWh off-
14 peak (based on average forwards between 10/1/2009 and
15 12/31/2009) .The average Mid-Columbia price from the
16 Dispatch Model is $53. 66/MWh on-peak and $41. 65/MWh off-
17 peak.
18
19
20
V. DEM CLASSIFICATION
Q.Witness Knox explains that the Company is
21 changing its methodology for allocating production costs
22 between capacity and energy based on your work.Please
23 explain your concerns with the present methodology and what
24 you propose as a better way to allocate production costs.
Kalich, Di 25
Avista Corporation
1 A.The historical method to allocate production
2 costs goes through the various FERC accounts and attempts
3 to determine which costs are for demand and which are for
4 energy.As an example, all thermal fuel in FERC account
5 501 is allocated to energy production, and all "Other"
6 production costs are allocated to demand.Unfortunately,
7 the problem is not this simple. Some of the "Other" costs
8 are almost certainly related to the production of energy
9 and, possibly more surprising to some, various fuel costs
10 can be related to providing capacity (demand).
11 Q.How can some of the costs in your examle be
12 considered energy?
13 A.To produce energy it is necessary to maintain a
14 generation plant in a ready state to do so.The "Other"
15 category is an excellent example of a somewhat arbitrary
16 allocation to demand that is done for lack of any better
17 approach.The "Other" category for both production plant
18 (300 series) and O&M (500 series) includes our gas-fired
19 plants and the Lancaster agreement.The "Other" category
20 is allocated 100 percent demand.Because of this the
21 Company has historically removed our Coyote Springs 2 gas-
22 fired CCCT plant from the "Other" category and instead
23 allocated its costs based on the overall Thermal Peak
24 Credi t figure.But other plants are not broken out this
Kalich, Di' 26
Avista Corporation
1 way.Boulder Park, Rathdrum and Northeast are all
2 allocated 100 percent to demand by being in the "Other"
3 category, yet clearly a portion of their plant and O&M
4 costs are attributable to energy production. It is likely
5 that a portion of "Other" expenses are indeed to the
6 benefit of energy production, yet the old allocation method
7 assumed all such costs are attributable to demand.
8
9
Q.How can a fuel cost be classified as demnd?
A.Demand, or capacity, is really the production of
10 energy at the time of system peak. Fuel is consumed during
11 periods of peak operation. It would be unreasonable to not
12 consider this fact. And simply because the majority of a
13 fuel expense is incurred outside of peak operating periods
14 does not mean that no fuel should be allocated to demand.
15 Q.Do you have any other concerns about the present
16 demand allocation methodology?
17 A.Yes. Presently all of our generation assets are
18 melded together to create an allocation. Further, a simple
19 accounting methodology is employed to estimate what it
20 might cost to construct our older facilities today. But it
21 is not realistic to assume that historical investments
22 represent our present costs of capacity (demand).Such
23 allocations should be based on the decisions we are making
24 today, and on the costs we incur today when customers
25 consume electrical energy during times of system peak.
Kalich, Di 27
Avista Corporation
1 Instead of trying to create an incremental demand cost
2 through a complicated and potentially inaccurate escalation
3 of historical expenses, we should instead use present
4 information for plants we are building to meet new customer
5 demands.
6 Q.Please explain the Company's recommnded method
7 for classifying electricity production costs between energy
8 and demnd.
9 A.The Company believes we should link the
10 classification methodology to the Integrated Resource Plan
11
12
(IRP) .The IRP process is an exercise to meet customer
load growth in a least-cost fashion.Central to the
13 equation is the level of our customers' coincident peak
14 demand.We construct a least-cost mix of resources
15 providing both the energy and capacity.
16 Q.What resource does the Company propose be used
17 for splitting demnd and energy costs from overall
18 production expenses?
19 A.We believe that we should use the incremental
20 capaci ty resource from our latest IRP--a gas-fired CCCT.
21 The Company, using its IRP models, calculated the costs of
22 capacity and energy from this resource, and used that
23 figure to allocate overall production costs.
24 Q.How did the Company determne a split between
25 energy and capacity for the incremntal resource?
Kalich, Di 28
Avista Corporation
1 A.For the IRP the Company models the Western
2 Interconnect wholesale power marketplace using AURORAxmp.
3
4
AURORAxmp dispatches available resources against
electricity loads on an hourly basis.The IRP uses
5 AURORAxmp to look at costs out 20 years and "mark-to-
6 market" (MTM) each potential resource option reasonably
7 available to the Company in the future.The dispatched
8 value of the CCCT (i. e., market sales price less fuel and
9 variable maintenance and operation costs) is tracked .hourly
10 over the 20-year IRP timeframe. Additionally, for the IRP
11 the Company models the 20-year future over 250 Monte Carlo
12 iterations to reflect volatility created by various factors
13 including natural gas prices, load variability and forced
14 outage rates.in other words, for each of the 20 years
15 evaluated for the IRP there are 250 MTM values for the
16 CCCT. The annual average MTM figures represent the energy
17 value generated by the plant.Remaining costs not
18 recovered in the wholesale marketplace are defined as
19 capacity.The ratio of those costs remaining after
20 dispatch into the wholesale marketplace (MTM values)
21 relative to the entire cost of the CCCT plant equals the
22 share of production costs then attributed to demand in the
23 cost of service models.
24 Q.What were the results of your analysis?
Kalich, Di 29
Avista Corporation
1
2
A.The analysis allocates 38.1 percent of production
costs to demand.Company witness Knox discusses how this
3 demand allocator compares with that derived from the prior
4 peak-credi t methodology.
5 Q.Where are the calculations referenced above
6 contained?
7 A.The calculations are contained in my work papers
8 in an Excel file called "Demand Classification Final." A
9 summary of the results is presented in Exhibit No.5,
10 Schedule 2.
11 Q. How should the demnd allocation be applied to
12 production costs?
13 A.Because the analysis does not differentiate
14 between fixed and variable costs, but instead evaluates all
15 such costs, it should be applied across the board to all
16 production costs.
17
18
19
VI. RESULTS
Q.Please sumrize the results from the Dispatch
20 Model that are used for ratemking.
21 A.The Dispatch Model tracks the Company's portfolio.
22 during each hour of the pro forma study.. Fuel costs and
23 generation for each resource are summarized by month.
24 Total market sales and purchases, and their revenues and
25 costs, are also determined and summarized by month. These
Kalich, Di 30
Avista Corporation
1 values are contained in Exhibit No.5, Confidential Schedule
2 3 and were provided to Mr. Johnson for use in his
3 calculations.Mr. Johnson adds resource and contract
4 revenues and expenses not accounted for in the Dispatch
5 Model (e.g., fixed costs) to determine net power supply
6 expense.
7 Q.Does this conclude your pre-filed direct
8 testimny?
9 A.Yes, it does.
Kalich, Di 31
Avista Corporation
DAVID J. MEYER
VICE PRESIDENT AND CHIEF COUNSEL OF
REGULATORY & GOVERNMENTAL AFFAIRS
AVISTA CORPORATION
P. O. BOX 3727
1411 EAST MISSION AVENUE
SPOKANE, WASHINGTON 99220-3727
TELEPHONE: [phone]
FACSIMILE: [phone]
DAVID .MEYER~AVISTACORP. COM
BEFORE THE IDAHO PUBLIC UTILITIES COMMISSION
IN THE MATTER OF THE APPLICATION
OF AVISTA CORPORATION FOR THE
AUTHORITY TO INCREASE ITS RATES
AND CHARGES FOR ELECTRIC AND
NATURAL GAS SERVICE TO ELECTRIC
AND NATURAL GAS CUSTOMERS IN THE
STATE OF IDAHO
CASE NO. AVU-E-10-01
EXHIBIT NO. 5
CLINT G. KALICH
FOR AVISTA CORPORATION
(ELECTRIC ONLY)
Lo
a
d

a
n
d

R
e
s
o
u
r
c
e

B
a
l
a
n
c
e

(
a
M
W
20
1
1
;

2
0
1
2
:

2
0
1
3
1
"

2
0
1
4
1

"
2
0
1
s
T
H
2
0
1
6
:

2
0
1
7
:

2
0
1
8
:

2
0
1
9
;

2
0
2
0
'
..

,
.
,
.
.
.
.
,
-
.
'
.
_
.
.
.
,

.
.
.
.
.
.
.
_
.
.
.
.
l
_
.

.
.
.
.
.
.
.
.
.
.
.
"
.
.
_
-
'
.
.
.
.
.
1

.

-

.
,
.
En
~
r
g
y
f
'
(
)
~
l
t
i
o
n
..
.

.
.
,
R
E
J
l
'
l
I
R
E
f
i
E
N
T
S
1¡
N
a
t
i
'
S

L
o
~
.
.2

j
Ç
Q
n
t
r
c
~
t

(
)
~
I
i
g
~
t
i
o
i
i
s
.

.
..
~

¡
I
o
t
t
i
l
.
.
g
~
g
,
l
i
i
~
r
r
~
l
'
t
~
.
..
t
.
.
.
:-
R
E
S
O
U
R
C
E
S
~I
~
~
t
~
g
f
~
9
.
~
t
~
.
.
.
.
.
_
.
_

.
.
.
.
.
.
.
.
51
l
:
d
r
o
..

'
6
;
T
h
e
.
m
a
ì
R
e
s
õ
u
r
c
e
s
"
"
jl
T
o
t
a
l
'
R
e
s
o
u
¡
.
e
s
'
.
'
.
'

.
" .
,
.
¡
~

.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
,
.
.
.
.
"

.
.
.
~
~

'
"

.
_
"
.
-
,
.
.
.
.
.
.
.
.
.

.
.
~
"
.
_
.
.

'
,
_
.

_
.
.
I t..
.
.
.
.
.
.
.

.
.
.
.
.
j
.
.
.
.
.
-
-

.

.
w
.
I
H
~
9
1
.
i
r
.
~
1
~
r
i
i
~
1
.
~
i
§
.
L
H
2
Q
1
?
1
.
.
.

;
i
C
n
l
!

.
"
H
~
~
_
'
-1
,
1
3
0
'

-
1
,
1
5
2
'

-
1
,
1
7
4
:
.
.
1
,
1
9
7
1
-
1
,
2
1
6
)

-
1
,
2
3
6
1

.
.
1
,
2
6
0
)

-
1
,
2
7
8
;

-
1
,
2
9
6
.

-
1
,
3
1
5
;
:1
~
~
L

.
.
1
~
~
L
~
1
~
á
j
_
~
~
.
~
~
T
d

;
§
L
,
~
I
~
'
~
j

.
.
.

.
.
J
1
)
.
.

'
.
;
U

.
.

;
U
,
-1
,
2
6
9
,

-
1
,
2
9
1
:

-
1
,
3
1
3
;

-
1
,
3
3
6
!

-
1
,
2
8
0
1

-
1
,
2
9
9
:

-
1
,
2
7
2
:

-
1
,
2
9
0
1

-
1
,
3
0
7

-
1
,
3
2
6
:
..

.
.
.
:

.
.
.
.
,

.
.
.
:

.

.
.
.
.
-
.
;
~
.
.
.
~
-
,
_
.
;
~

.
'
.
~
.
:
.
'
-
r
.
.
-

¡
.

.
.
.
.
.
,
.
:
.
r

.

H
.
.
.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.

.
.
.

.
.
.
.
.
.
.
.
.
.
,
d.
.
.
,
.
.
.
4
a
r
'

4
t
)
a
¡

4
e
e
T
.
~
e
6
L
.
.
.
~
~
l
r
_
-
~
e
1
J
.
.

~
~
~
r

.
3
3
8
(
.
.
.

3
1
7
.

?
~
2
j
50
5
'

4
9
3
'

4
9
5
1

4
9
5
:

4
9
5
1

4
9
5
1

4
9
5
:

4
9
1
,

4
8
1
!

4
8
1
:
..
o
J
.
.

_
.
.
.
.
.
.
.
.
,
.
~

.
.
.
.
.
,
.
.

.
.
,
.
.
.
ý
.

.
.
.

.
.
,

.
.
¡
.

'
.

'
.

"
'
~
"
-
'

.
;
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
~
_
.
.
.
.
.
.
-
:
.
.
.

.
.
.
.
.
.
.
.
.
"

_
.
.
_
-
.
.
~
.
.
.
i
-
.
-
'
-
-
.

.
-
.
,
"
.
-
"
.
,

.
,
.
"
.
.
,
t
.

.

.
.
.
.
.
.

.
~

,
.
"

,
'
,
,
,
'
,

,
~
.
.
.
.
"

.
.
.
.
.
.
.
.

.
.
.
.

_
.
~
.
,
.
.

"
'
-
'
,
'
"

.
.
.

.
.
.
.
.
.
.
.
.

.
~
.
.
.
.

'
O
A
.

_
.
.
~
.
5~
:

5
?
a
¡
?
?
?
¡
_
_
_
5
3
?
j
.
~
~
L
.
~
~
L
.

5
~
.
~
,
:

S
4
L

5
3
1
,

5
3
2
,
1,
5
2
0
1

1
,
4
8
1
;

1
,
4
8
3
1

1
,
4
9
3
!

1
,
4
2
9
1

1
,
4
0
6
:

1
,
3
6
6
'

1
,
3
7
5
\

1
,
3
3

1
,
3
0
5
.,r
.
.
'
,
'

.
.
.
.
.
.
.
,
.
.
.
.

.
.
.
;
.
.
.
.
~
.
.
.
,
.
~
.
-

.
.
.
.
.

'
.
.
.
.
.
.
.
r
'
.
.
.
.
_
.
.
.
.
"
.
_
~
.
.
.
.
.
.
"
1
.
.
.
.
.
"
.
.
.
.
.
.
_
.
.
.
"
~
_
.
.
.
.
.
"
'
.
.
"
f
"
.
,
,
.
~
-
~
.
.
.
.
.
.
.
.
.
.

.
~
.
"

.
.

.
.
"
-
_
.
,
.
.
.
.
.
.
.
.
.
.
~
.
.
.

.
"
.
.

.
.
.
.
.
.
.
.
.
.
.

.
,
.

,
.

_
.
.
.

.
,
.
"
'
.
.
.
.
.
.
;
'
.

.

.
.
.
.
.
.
.

.
,

"
.
.
.
-
~,
.

l
.
.
"
.
¡
C
O
N
T
I
N
G
E
N
C
Y

P
L
A
N
N
I
N
G
¡:
_
.
~
l
Ç
E
n
j
i
r
i
i
i
~
n
g
i
!
9
!
~
l
.
.
,
.
-

.
.
.
.
d
.
.
d
.
.
d
.
.

.
.
:1
Q
L
I
?
~
k
i
n
~
t
R
~
g
~
r
.
~
~

.
_
_
_
_
.
.
-2
2
1
,

-
2
1
8
,
.~
.
.
~
_
.
.

"
.

.
.
.
.
.

.
.
"
.
_
,
,
"
"
.
.
"
J3
0
L
.

1
3
1
jg
E
~
1
:
~
t
-
~
I
.
¡
.

¡

t

ì
SQ

7
7

6
8

2
A
-2
2
2
;

-
2
2
3
1
.

~
2
0
6
i
'

-
1
8
9
1
14
Õ
¡

1
4
0
1
.
.

1
4
Õ
r
'
.
1
4
0
;
..
,

.
'
.

t
.
"

"
.
,
.

,
.
.
.
.
.

.

~

.
.

.
.
.
.
.
.
.
.

.
.
"
-
-

-
"
.

t
.
.
.
.
.
.
.
.
~
.
.
.
.

,
.
.
.
.
.
_
'
.
.
.
.
_
.
.
1
.7
0
Ex
h
b
i
t

N
O
.
5
Ca
s
e

N
o
.

A

V
U
-
E
-

1
0
-
0

1
C.

K
a
l
i
c
h
,

A

v
i
s
t
a
Sc
h
e
d
u
l
e

1
,

p
.

1

o
f

1
Co
m
b
i
n
e
d
-
C
c
l
e

C
o
m
b
u
s
t
i
o
n

T
u
r
b
i
n
e

(
f
r
o
m

2
0
0
9

I
R
P
)
¡P
r
o
j
e
c
t

S
i
z
e
,
.
¡
ç
~
p
!
t
a
l

C
d
s
t
¡
¡T
r
a
n
s
m
i
s
s
i
o
n

C
o
s
t
¡.
.

.
.
.
'T
o
t
a
l

c
a
p
i
t
a
l

C
o
s
t
rF
i
~
~
d
O
&
M
r

f

.
¡V
~
'
r
i
~
b
l
~

O
S
:
M

f
..
.
.
1
.
.

.
.
.
.
.

.
¡I
n
f
l
a
t
i
o
n
"..
'
f
.
.
.
-
.
o
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Co
I
.
~
i
:
n

N
o
L

1
,

2
.
JN
o
t
e
r.
,
.
:
.

'
~
.
~
.

.
,
~
,

+
.
-
!
j

,

)
.

A
n
n
u
a
l

.
.¡
.

.
.

.
!
!
.

.
¡
G
e
n
e
æ
t
l
s
i
o
..
t
.

;
.
.
,
.

.
.
.
.
.

i
.
.

.
(
I
\
\
N
h
t
iL
e
v
e
l
i
z
e
d

C
o
s
t

I
. ,
,
-
_
.
_
_
.
~
.

-
-
,
.
.
.

,
.
,
.
.

.
-
.
.
.
.

,

.
.
,

~
.
.
-

'
r
_;
.
.

.
,
-
_
.
,
.
~
.
"
,

.
J
.
.
.

-

L
,
.
.
,
.
.
~
¡
:

1

2
0
1
0
;

¡

1
,
4
2
1
,
9
9
6
t.

.
.
,
.
.
.
.
.
.
.

.
+
,
.

.
.
.
.
.
.
.
.
.
.
.
"
:

2
.
2
0
1
1
1

'
1
,
4
3
2
,
1
6
3
¡
.
.

;
.
3
'
i
õ
i
ï
f
"
i
i
s
6
4
;
ï
i
ö

.
.
¡'.
~
.
.
.
.
.
.
~
_
~
.
.
,

"
_
'
¡
;
'
_
'
_
"
'
w
.
_
.
.
.

'
f
~
'

a
"
_

.
.
.
.
.
,
'
,
,
_
.

'
~
'
-
"
"
'
"

.
.
,

-
.
'
.

,
"

.
.
.
.
.
.
.

.
_
.
.
.
.
~
'
:
1

¡

4
:

2
0
1
3
1

:

1
,
6
2
4
.
3
9
9

.
't
.
.
I
.
s
J
.
ï
o
~
4
r

t
i
~
6
5
.
8
.
0
9
~
.
.
.

.
J
:

6

2
0
1
5
\

ì

1
.
6
6
5
.
3
7
2
~..
¡
.
.
.
.
~
.
,
.
.
.
.
¡
u
.
.
,
.
_
_
.
)
.
.
.
.
,
,
_
_
.
.
,
.
.
.
t
.
.
,
"
.
.
.
~
.

.
.
.
.

.
.
.
.

.
_
,
.
.
"
.
.
.
.
.
,
i

i
:

2
0
1
6
,

¡

1
,
6
6
.
0
8
8
."
.
.
.
.

.
+
.
.
_
.
.
,
¡
.
.
.
.
.
,

.
f
'
.
.
.
.
.
.
.

-
t
.
.
.
.
,
;
.
.
'

,
.
,
.
,
.
.
-
.
.
"
-
~
.
"
.
.
,
,
,
:
.
,
'
8
:

2
0
1
7
1

¡

1
,
6
4
6
.
6
2
6
...
.
,
"
"
.
.
.
.
.
.
_
.
.
.
.
.
~
-
-
,
.
.
.
.
.
_
,
-
.
-
.

.
_
"

,
-
,
'
f
"

_
.
.
t
.
.
.
.
.
.
.
,

.
_
.
.
.
.
.
"
.

-
.
.
.
.
'
_
.
.
.
.
'
j
:

9
;

2
0
1
8
!

\

1
,
6
4
1
,
1
6
5

.
:.
.
:
:

.
.

.
.
.
.
:
.
.
.
:
E
:
~
~
ê
~
~
~
~
¡
:
J
.
.
~
:
~
t
:
~
~
~
~
:
,
,
,
1
2
.
2
0
2
1
,

¡

1
,
6
4
2
.
9
8
7

.
L.
.
,
.
.
:
:
.
1
:
J
~
~
~
~
ï
~
F
I
t
~
~
~
2
.
:
.
i
ì

:
i
s
¡
2
0
2
4
i

i

1
.
6
7
5
.
1
1
8

:
r.
:
=
.
.
:
'
~
:
.
=
f
J
l
~
l
i
~
~
r
+
~
;
:
;
~
;
~
~
~
:
,
.
.
.

~
L_
.
_
.
.
_
_
_
L
.
¡
.
~
H
O
.
~
?
l
.
.
I
.
~
'
6
5
t
)
.
~
~
.
,

.
J
;
1
9
!

2
0
2
8
1

'
1
.
6
7
5
,
1
2
3
:
..
.
t
.
.
+
.
.
.
.
.
.

.
.
i
.
.
.
.
.
_
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
1:
2
.
0
L
2
0
2
9

.
.
1
1
.
6
2
9
,
3
?
1
¡
25
0
M
W
l,
5
6
4
$
/
k
w
"

(
2
0
1
0
)
~$
/
k
W

(
2
0
1
0
1
1,
6
1
8
$
/
k
"
!

(
2
C
1
:
i
'
!
)
11
.
2
2
$
!
k
W
-
y
r
(
2
.
0
1
0
t
3.
3
6
.
$
/
t
l
W
h
2.
0
%
a
n
n
u
a
l
i
y
. .
~
.
.
,.

c
:
C
C
T
;
.
.
.
.
.
.
)
Fi
x
e
d

¡

D
e
m
a
n
d

¡
Cd
S
b

L
.

s
h
a
r
ï
l
(.
~
.
.
r
n
.
l
l
.
.
.
.
.
.
.
.
.
_
¡

(
%
1

1
(5
7
.
9
3
1
.
r
"''
'
!
',.
1
1
To
t
a
l
fr
o
m

2
0
0

c
a
p
i
t
a
l
.
IR
P

C
O
L

4
I
fr
o
m

2
0
0
9
'

F
O
&
M

.

;

C
o
l

3

.
IR
P

.
2
5
0

M
W
;

V
O
&
M

+

I
n
f
.

!

C
O
L
.

5
-
8

.
..
t
-
-

.
.
_
.
.
.
_
-
-
.
.

.
.
.
.
.
t
.
.
-
-
-
-
,
.
.
.
~
"
"
.
_
-
.
.
,
.
.
~
.
.
.
.
.
~
.

.
.
,
.
.
.
.
.
.
.
.
.
,
.
.

~
.
.
.
.
"
.
.
.
.
.
.
.
.
.
'
(C
O
L

10
+
11
1
/
2
5
0
/
10
0
0
..
.
.
1
..
j
16

'
..
.
.

.
.
1
!
¡
¡
i
.
,C
o
I
U
/
1
.
C
o
l
.

5

+

7
'

C
O
L

1
5

1
'.

.

.
.
.

.
.
.
.
.
.

.
_
.
.
.

¡
:
1-'L
14
15
.
4.
6
7
Ca
p
i
t
a
l

.
'

r
'
"

.
.
.
.
.

.
Re
c
o
v
e
r
y
.

C
a
p
i
t
a
l

:
F
i
;
~
T

V
a
r
i
a
b
i
ë
'
i
'
.
'
.
"
~
.
F
a
s
;
t
o
r
d

R
~
~
d
v
e
i
:
J

.
F
u
e
l
.
.
.
.
.
.
O
&
M
I
.
ô
&
M
.
I
T
o
~
¡
L
.
T
ø
a
l
.
.
.
.
(~
0
1
I
I
.

L

(
$
m
i
1
t

.
.
.
.

J
~
r
n
i
I
L
¡
.
.

.
J
$
r
r
i
l
L
.
.
i
.
.
.
!
$
.
'
!
i
.
I
L
1
J
$
!
k
w
-
y
r
t

.
.
.
.
($
5
4
.
7
2
)
,

(
$
2
U
5
)
'
J
$
3
.
.
2
.
1
U
.
.

.
.
.

(
S
6
.
,
1
.
1
H
$
?
S
:
!
~
!
:

(
$
3
4
0
.
7
7
:
.
,
~

1
lS
,
~
%
,
J
6
2
.
2
9
i
I
(
ï
5
.
4
5
i
C
(
~
.
a
i
)
t
:
d
Î
~
:
?
7
i
r
i
a
s
.
.
~
l
l
t
(
3
~
1
.
,
2
4
1
.
l
S
O
:
~
6
.
.
.

3
7
:
i

1
4
7
.
6
0
)
1

(
1
9
0
,
3
~
l
i
.
!
1.?
5
"

,
.
.
.
.

!
7
0
:
C
?
~
)
1
.
.
.

(
1
5
,
6
.
5
1
,
.

J
~
.
8
6
I
!
.
.
.
.
.
.
.
.
(
'
!
~
l
L
.
J
.
~
:
:
i
!
h

(
3
.
?
6
.
4
A
)
f
.
1
5
5
.
.
2
.
6
.
:

3
!
l
:
?
!
'

.
.

(
5
5
,
3
0
)
j
(
2
2
U
8
1
L
.
.1
6
.
:
8
.
,
,
:

.
J
E
i
?

.
7
6
E
.
.
.

(
1
5
.
!
l
S
)
L
J
.
2
~
i
.
)
L

.
_
.
.
J
s
,
~
?
I
L
(
~
1
.
9
~
l
+
.
.

l
3
E
i
!
,
m
;
.

.
¡
.

.
.
l
?
'
s
Æ
.
~
.

.
.
4
3
.
,
~
.
?
L
.
j

.
(
.
i
:
C
l
2
.
L

J
l
?
2
:
1
(
)
)
t
.

.
.
16
.
0
%
.

(
6
4
.
9
1
.
)
.

(
1
6
.
2
.
~
!
l
J
2
.
.
9
8
)

I

.
_
J
S
:
~
(
)
I
L
.
.

(
8
~
.
,
~
i
.
)
i

.
1
3
5
.
?
:
6
.
7
1
:

.

.
.

.
1
9
.
~
:
(
)
3
.

;

4
9
,
S
!
.

,

'
.

(
4
(
.
4
1
)
i
J
1
6
.
1
:
6
4
)
;
.
.
.
tS
:
~
"
l

1
6
2
.
.
i
.
.
.
t
.

J
l
?
:
3
Q
l
¡

J
3
.
,
Q
'
a
.

.
J
6
.
l
l
3
.
1
!
.

J
l
l
'
?
.
l
l
¡
J
3
.
5
~
:
Q
l
1
l
.
,
_

.
.
.
.

.
2
~
3
.
:
(
)
3
.
L
~
S
.
,
?
6
.
.

.
.
,

.

1
3
.
2
.
:
?
5
)
,

m
~
:
~
s
.
i
;
.
;
14
.
7
%
:

.

.
.

1
5
9
~
~
)
i
.

(
i
9
:
3
6
.
'
-
¡
.
.
.
l
3
.
,
l
~
l
i
.
.
,
l
6
:
~
s
.
U
,
J
!
:
(
m
.
L

J
3
,
S
,
2
.
i
8
.
l
i

2
~
2
:
!
?
.
.

6
0
:
6
.
?
:

.
.

1
2
.
3
8
.
1
¡

.

(
!
l
l
,
5
1
I
L
14
.
0
%
,
J
5
.
6
.
.
8
.
c
i
)
L

(
~
1
.
0
9
t

.
I
3
,
i
6
.
l
i
.
.
.
1
6
,
~
~
h
.
J
8
?
3
.
4
:
t
J
3
.
4
:
?
:
3
.
6
.
)
;
.
.
.

2
5
9
.
g
.
.

6
4
:
7
8
.
.

1
2
~
,
5
6
)
'

(
9
.
:
2
5
1
\
1.
3
.
4
%

.

(
5
4
:
2
.
3
.
1
1
.

(
2
2
,
6
9
r
.
J
3
~
~
2
)
i
.
,
1
6
.
,
~
3
.
U
.
J
S
.
.
6
.
:
4
:
7
1
¡
.
1
.
3
.
4
S
.
:
e
?
)
i

.
,
.
.

2
7
1
.
8
!
,

.
.
'

6
?
~
~
5
.
:
.
.

,
.
J
!
~
~
5
7
l
;

(
?
I
i
,
O
~
I
L
.
a8
%
:

.
J
g
:
6
~
t

J
2
4
.
8
3
.
L
(
3
.
:
f
~
I
J
_
J
E
i
:
4
È
)
i
.
J
!
6
.
:
~
4
:
)
;

J
3
.
~
:
~
I
i
L
.
,
.

3
0
1
.
.
l
l
,

7
5
.
2
.
5
.
:
.

;
J
l
0
.
~
?
t
(
4
3
.
~
t
.

I
-
12
.
2
%
,

.
.

(
4
9
:
!
6
.
1
\
.
.

1
2
6
:
S
.
?
a
J
3
.
:
3
5
1
¡
.
.
.
,
.
.
.
.
I
.
6
.
.
.
3
.
?
~
J
~
5
:
?
S
.
l
L
J
3
.
I
1
3
.
.
.
0
0
)
;
,
.

3
0
3
.
2
9
:

7
5
,
8
.
2

,
.

:

.
.
.
.

(
~
:
~
3
.
I
¡

.

(
3
9
.
m
;
.
,
1
1
.
.
5
"
i
.
1
4
6
.
:
t
)
~
J
¡

(
i
.
6
.
.
.
?
~
l
L
(
~
4
:
~
)
I
.
.
.
_
_
.
~
:
S
S
.
l
¡
.
(
~
,
4
:
1
)
~
j
~
3
3
.
:
6
.
S
)
¡
,
L

3
g
,
S
c
i
.
.
?
8
:
!
3
.
,
!

l
.
.
J
~
.
2
~
l
l
.
J
2
1
:
1
4
!
¡
.
~
.
.
.
.
.:

.
t
(
:
9
'
:
:
J
4
4
:
0
9
1
¡

(
2
5
,
8
3
L
J
3
.
,
4
:
9
.
1
1
.
.
.
J
6
.
.
.
R
l
)
i
.
.
I
!
l
Q
:
i
.
~
1
¡
J
3
2
(
)
.
S
.
2
I
j

L
.

.
.

3
2
1
.
1
.
4

;
.
.
~
O
,
~
¡

.
.

,
.
C
l
:
2
3
¡

.

o
.
?
i
.
¡
,
.

L
,
.
:
i
~
~
3
.
'
J
L
.
.
4
!
:
S
.
6
.
)
~
.
J
2
5
,
~
S
.
l
l
.
J
3
.
~
S
.
6
.
I
L
_
.
j
6
.
.
:
!
!
!
t
.
.
l
~
:
!
:
S
)
.
i
.
J
3
.
.
i
.
~
!
l
c
i
i
¡

.
t
.
3
.
~
~
,
4
:
~
.
.

.

?
?
,
.
8
7

-
f
.
.

+
.
.
)
,
.
~
.
i
.
,

.

7
:
E
!
l
.

:
;
~:
c
i
i
¡
-
i
~
~
:
~
r
i
~
I
J
K
i
~
)
t
'
~
i
c
i
~
f
.
-
"
~
:
~
i
i
i
.
i
j
~
~
j
~
~
H
i
i
~
~
~
i
~
+
i
~
~
~
:
~

¡

.
:
~

I
.

.

j

i
~
:
i
:

:
~
c
i
:
~
~
¡
!

.
..
.
.
.
,

.
.
.
.
_

L
.
~
I
.
'
J
.

.
.
.
.
l
~
.
.
.
.
.
.
.
.
.
.
.

,
.
.
.
.
.
J
t
-
_
.
_
_
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
1
,
.

.
¡

.
.
.

.
.
.
.
.
.
.
.

.
.
.
.
.
,
.

.
.
.

.
.
.
.
.
.
.
.
.
.
.
,
.

'
.

.

.
.

.
,
.

.

.
-
.
~
.

.
.
,
.
.
.
.
'.

.
.
.
8
.
:
~
¡
.
.
.

J
3
3
.
:
~
~
4
.

1
2
.
~
.
1
5
1
~
J
3
,
?
8
1
.
.
_
.
.
.
Q
:
3
J
l
t
l
?
1
i
,
7
S
.
l
f
.
J
~
~
?
.
!
i
.
a
.
I
.
.
.
E
3
~
6
5

:

.
.
.

?
1
:
"
'
1

,
¡

,

.

a
1
3
.
l
J
6
.
:
5
~
~
+
.
.
.
.
.
.
7
.
S
.
,
,
~

.

(
3
.
:
i
.
,
4
:
~
~
.
.
.
(
~
0
.
1
s
.
!
t
.
.
.
(
È
~
.
8
?
-
!
t
.
.
.
,
_
,
.
J
?
~
1
1
.
1
?
2
.
:
~
1
!
.
.
.
H
~
p
5
l
j

.
.
.
L
.
.

.
.
.
3
.
e
S
.
:
~
8

;
.
.
.

!
:
.
6
:
.
s
,
0
+
.
.
.
.
.
.
.
2
.
3
.
,
5
6
.
.
¡
.
.

.
.
9
.
4
.
.
2
.
3
.

;
.
1
..
.

?
:
!
,
,
¡
.

.
.
.

J
2
l
!
:
9
.
(
)
!
L
.

.
l
i
.
~
,
~
!
:
l
.
.
.
~
:
.
~
3
.
I
I
.
.

.
.
.
_
J
?
:
6
.
.
s
.
L
l
!
(
)
,
4
:
?
i
~
ß
!
l
~
s
.
?
4
.
.
.

;
.
.
.
.
.
.
3
~
2
.
.
l
'
,
.
.
.

.
?
~
:
(
)
S
.
.
L
.
¡
.
.
3
?
6
.
2
.
.

t
.
.

~
~
0
,
4
7
.
+
.
+
..
:
.
.
.
.

6
.
,
5
"
"
.

.

.
.
.

(
~
6
.
.
3
.
6
.
1
L
I
2
.
9
.
S
.
6
.
)
L
1
4
,
O
l
)
"
.
.
.
.
.
.

(
?
:
!
~
m
.
.
I
~
.
l
!
I
J
.
ß
?
2
:
4
.
~
!
.
L
.
_
L
.

.
.
4
c
i
?
,
2
2
.
:
L
0
!
,
a
Q
.
;
.
.
.
:
.

3
.
3
.
.
6
9
:

1
S
4
:
?
?
L
.
.
'5
.
9
%
1

(
2
3
.
8
3
)
¡
'

(
3
2
.
1
2
)
J
J
4
:
:
c
i
9
)
L
.
.
.

(
7
.
8
:
1
L
J
6
7
.
8
~
1
!

J
2
7
1
.
5
2
l
.

j

.
.
4
:
1
2
.
2
0
.

1
0
3
.
0
5
.

.

3
5
.
1
7
:
.

1
4
0
.
6
8

¡

,
..Ma
r
k
e
t
..
.
.
.
.
.
.
.
.
Va
l
u
e
.

V
a
l
u
e
(S
j
k
\
N
~
y
r
l
(
$
r
n
l
!
t
J2
5
2
.
4
(

c
.

$
6
3
.
1
(
)
;
tf
t
V
a
l
u
e
($
m
i
l
l

'
(
$
/
k
W
-
y
r
l
'
¡(
s
i
ï
.
.
o
.
9
)
.
,
;
t
$
8
â
;
:
i
Z
I
(

_
,
.
.
,
..
.
.
.
Ex
h
i
b
i
t

N
O
.
5
Ca
s
e

N
o
.

A
V
U
-
E
-
I
O
-
O
l
C.

K
a
l
i
c
h
,

A

v
i
s
t
a
Sc
h
e
d
l
e

2
,

p
.

l
o
f

i
CONFIDENTIAL
Dispatch Model Sumry output
Pages i through 3
THESE PAGES ALLEGEDLY CONTAIN TRAE SECRETS OR
CONFIDENTIAL MATERIALS AN AR SEPARTELY
FILED.
Exhbit No.5
Case No. AVU-E-IO-OI
C. Kalich, A vista
Schedule 3, p. lof 3