200304302003 IRP.pdf

2003
Integrated
Resource
Plan

Table of Contents

Section 1. Introduction & Summary ..........................................................................................1
Overview.......................................................................................................................................1
Public Process...............................................................................................................................1
IRP Outline ...................................................................................................................................2
Summary.......................................................................................................................................2
Section 2. Loads & Resources .....................................................................................................4
Overview.......................................................................................................................................4
Resources and Contracts...............................................................................................................4
Load Forecast................................................................................................................................5
Energy Position.............................................................................................................................6
Capacity Position..........................................................................................................................7
Planning Reserves.........................................................................................................................8
Summary.....................................................................................................................................10
Section 3. Demand-Side Management......................................................................................11
Historic DSM Activities .............................................................................................................11
Future DSM Activities................................................................................................................13
Conservation Voltage Reduction................................................................................................14
Automated Meter Reading/Time of Use Metering.....................................................................14
DSM in AURORA......................................................................................................................15
Section 4. New Resource Alternatives ......................................................................................19
Overview.....................................................................................................................................19
General Approach.......................................................................................................................19
New Resource Alternatives Considered .....................................................................................19
Resources Not Evaluated............................................................................................................21
Section 5. Modeling....................................................................................................................22
Overview.....................................................................................................................................22
Modeling Process........................................................................................................................22
Assumptions and Inputs..............................................................................................................23
Analysis of Strategies .................................................................................................................25
Analysis of Scenarios..................................................................................................................26
Summary.....................................................................................................................................26
Section 6. Risk Analysis.............................................................................................................27
Overview.....................................................................................................................................27
Stochastic Risk Analysis.............................................................................................................27
Benefits and Risks of Resource Options.....................................................................................30
Section 7. Results........................................................................................................................32
Overview.....................................................................................................................................32
WECC Market Prices and Volatility ..........................................................................................32
The Preferred Resource Strategy ...............................................................................................36
Comparison of Strategies............................................................................................................47
Comparison of Scenarios............................................................................................................47
Summary and Conclusions .........................................................................................................48
Section 8. Action Plans & Avoided Costs.................................................................................49
Overview.....................................................................................................................................49
Summary Report for 2001 Action Plan ......................................................................................49
2003 Action Plan ........................................................................................................................52
Avoided Costs.............................................................................................................................54
Production Credits......................................................................................................................55
Table of Appendices

Appendix A. Resource & Contract Details............................................................................A-1
Utility-Owned Resources .......................................................................................................A-1
Power Purchase and Sale Contracts .......................................................................................A-3
Appendix B. Retail Load Forecast..........................................................................................B-1
Economic Growth...................................................................................................................B-1
Electric Retail Sales................................................................................................................B-4
Energy Load and Peak Load Forecasts ..................................................................................B-6
Enhancements to Forecasting Process....................................................................................B-6
Appendix C. Modeling Details ................................................................................................C-1
Selection of the AURORA Model..........................................................................................C-1
Cost of Capital for New Resources ........................................................................................C-2
Portfolio Optimization Using Linear Programming (LP) Module.........................................C-2
Capacity Expansion................................................................................................................C-5
Modeling Process Diagram ....................................................................................................C-6
Appendix D. Risk Details.........................................................................................................D-1
Resource Risk Profiles ...........................................................................................................D-1
Resource Characteristics ........................................................................................................D-2
Load Correlations...................................................................................................................D-3
Market Uncertainty.................................................................................................................D-4
Industry Restructuring............................................................................................................D-5
Appendix E. Detailed Results..................................................................................................E-1
Details of Preferred Resource Strategy.................................................................................. E-1
Details of Strategy Results ..................................................................................................... E-2
Details of Scenario Results..................................................................................................... E-7
Appendix F. Load and Resource Tables ................................................................................ F-1
Appendix G. TAC Meeting Agendas......................................................................................G-1
Appendix H. Wind Studies......................................................................................................H-1
Wind Energy...........................................................................................................................H-1
Appendix I. Capacity Expansion Process Details...................................................................I-1
Appendix J. Results of Capacity Expansion...........................................................................J-1
Appendix K. Spokane River Relicensing ...............................................................................K-1
Appendix L. Transmission Planning......................................................................................L-1
Relationship to Resource Planning......................................................................................... L-1
Current Issues......................................................................................................................... L-1
Expansion Possibilities & System Reconfiguration............................................................... L-2
Reliability............................................................................................................................... L-2
Appendix M. Distributed Generation....................................................................................M-1
Appendix N. Historic Data ......................................................................................................N-1
Hydroelectric Plants ...............................................................................................................N-1
Coal-Fired Plants....................................................................................................................N-4
Other Resources .....................................................................................................................N-6
PURPA Hydroelectric Plants .................................................................................................N-7
PURPA Thermal Plants........................................................................................................N-12
Appendix O. Avoided Cost Details .........................................................................................O-1
Appendix P. NWPPC Assumptions........................................................................................ P-1
Natural Gas Simple-Cycle Gas Turbine Power Plants........................................................... P-1
Coal-Fired Power Plants......................................................................................................... P-5
Natural Gas Combined-Cycle Gas Turbine Power Plants...................................................... P-9
Wind Power Plants............................................................................................................... P-18
Appendix Q. DSM Modeling Details......................................................................................Q-1
Section 1 Page 1 Introduction & Summary
Section

Introduction & Summary

Overview

The Company submits an Integrated Resource Plan (IRP) to its public utility commissions in
Idaho and Washington. The 2003 IRP is the seventh such submittal since 1989. In Washington,
IRP requirements are outlined in WAC 480-100-251 entitled “Least Cost Planning.” In Idaho,
the IRP requirements are outlined in Case No. U-1500-165 Order No. 22299, Case No. GNR-E-
93-1 Order No. 24729, and Case No. GNR-E-93-3 Order No. 25260. The plan describes the mix
of generating resources and improvements in efficiency that will meet future needs at the lowest
cost to the Company and its customers.

The Company has a statutory obligation to meet the electricity needs of its customers. To do so
reliably and at reasonable cost, the Company develops resource acquisition strategies and
business plans to acquire resources when supplies are insufficient. The Company will continue
to invest in conservation and cost-effective upgrades to existing generating facilities.

The Company views this IRP as a resource evaluation process, rather than a specific resource
acquisition plan. Primarily this is because significant resource deficiencies are many years ahead
of today. The 2005 IRP will likely include more specific plans for addressing future needs. The
2003 IRP is focused on developing a set of tools and methods within which various potential
resource decisions may be evaluated in future IRPs, requests for proposals, and other resource
planning analyses.

The Company believes it is prepared, even under low water conditions, to sufficiently meet retail
loads through at least 2007. The Company will continue to work with state commissions and
other interested parties to ensure that our resource planning decisions are cost effective,
reasonable, and responsive to an evolving industry.

Public Process

The Company strives to reach balanced business decisions by working with customers,
Commission Staff, and other key constituencies. An effective public involvement process
affords the opportunity to receive input from stakeholders, and exchange information and
perspectives regarding the IRP. The Company expects that public participation will continue to
play an important role in resource planning.

Specific to the IRP, the Company sponsored four Technical Advisory Committee (TAC)
meetings beginning in May 2002. Each of the meetings was designed to discuss the process,
provide preliminary results, and to obtain feedback on the IRP. Information shared at the TAC
meetings may be found in Appendix G.
Section 1 Page 2 Introduction & Summary

IRP Outline

In addition to this Introduction, the 2003 IRP contains the following sections:

• Section 2 details current loads and resources, and provides tabulations of future energy
and capacity balances.

• Section 3 discusses the Company’s current and future efforts in demand-side
management.

• Section 4 discusses those resources the Company is considering to meet future load
requirements.

• Section 5 details the modeling process used for the IRP, including the AURORA market
price-forecasting model, the Monte Carlo models used for stochastic analyses, and the
Linear Programming Module used to optimize the selection of hypothetical resource
acquisitions.

• Section 6 discusses the consideration of risk within the IRP, and identifies risk factors
specific to each new resource alternative.

• Section 7 explains the results of the IRP analyses. It provides the Preferred Resource
Strategy and compares other strategies that the Company might pursue. Scenarios are
also presented to quantify the potential impacts of specific future marketplaces.

• Section 8 provides the 2003 Action Plan resulting from the IRP, as well as avoided costs
for the Company.

The 2003 IRP also includes numerous appendices to support the sections listed above and
provide additional details for the document’s key elements. The IRP document and Technical
Appendices are available for download at the Company’s web site – www.avistautilities.com.

Summary

At this time, the Company has no immediate need for additional long-term resources. In fact, the
Company does not anticipate a significant deficit in energy, on an annual average basis, until
2008. Furthermore, the Company does not anticipate a deficit in capacity until 2010.

For this IRP the Company undertook a significant effort in computer modeling. This effort was
initiated with the acquisition of AURORA, an hourly production-cost model that dispatches
resources and develops a set of forward market prices based on numerous conditions. This effort
was substantiated through the development of numerous spreadsheet-based models, and the
incorporation of a Linear Programming (LP) Module.
Section 1 Page 3 Introduction & Summary

For the first ten years of the IRP timeframe (2004-2013), the IRP modeling process selected a
combination of combined and simple cycle combustion turbines, wind, and coal resources.
During the second ten year period of the IRP planning horizon (2014-2023), the modeling
process pointed towards acquisition of coal generation due to improvements in technology and
its fuel costs relative to other resources. Given no need for immediate resources, the Company
will continue to evaluate available options for future generating requirements.

Section 2 Page 4 Loads & Resources
Section

Loads & Resources

Overview

An essential element in integrated resource planning is the long-term forecast of future loads and
resources. The difference between the two illustrates resource needs that the Company must
address through its action plan. This section details Company resources and load obligations
through the twenty-year timeframe of the IRP, as well as the Company’s utilization of planning
reserves.

Resources and Contracts

The Company meets its load requirements through various owned and contracted resources. The
following table contains a listing of Company-owned resources and major contracts, as well as
some important details. Additional details on Company resources and contracts are provided in
Appendix A. A summary of the Company’s demand-side management activities may be found in
Section 3.

Section 2 Page 5 Loads & Resources
Table 2.1
Resource and Major Contract Summaries

River Start Capacity Energy End
Name System Fuel Location Date1 (MW)2 (aMW)2 Date3
Monroe Street Spokane Water Spokane, WA 1890 15.0 13.2 07-31-07
Post Falls Spokane Water Post Falls, ID 1906 18.0 9.9 07-31-07
Nine Mile Spokane Water Nine Mile Falls, WA 1925 24.5 16.4 07-31-07
Little Falls Spokane Water Ford, WA 1910 32.0 22.8 N/A
Long Lake Spokane Water Ford, WA 1915 88.0 52.4 07-31-07
Upper Falls Spokane Water Spokane, WA 1922 10.2 8.8 07-31-07
Cabinet Gorge Clark Fork Water Clark Fork, ID 1952 246.0 122.2 03-01-46
Noxon Rapids Clark Fork Water Noxon, MT 1959 527.0 202.9 03-01-46
Colstrip 3 N/A Coal Colstrip, MT 1984 111.0 95.6 N/A
Colstrip 4 N/A Coal Colstrip, MT 1986 111.0 95.6 N/A
Rathdrum N/A Gas Rathdrum, ID 1995 176.0 167.2 N/A
Northest N/A Gas/Oil Spokane, WA 1978 66.8 63.5 N/A
Boulder Park N/A Gas Spokane Valley, WA 2002 24.6 23.4 N/A
Coyote Springs 2 N/A Gas Boardman, OR 2003 143.5 136.3 N/A
Kettle Falls N/A Wood Kettle Falls, WA 1983 50.0 48.9 N/A
Kettle Falls CT N/A Gas Kettle Falls, WA 2002 6.9 6.5 N/A
Rocky Reach Mid-C Contract N/A 1961 37.7 20.5 10-31-11
Wells Mid-C Contract N/A 1967 28.6 9.9 08-31-18
Priest Rapids Mid-C Contract N/A 1965 129.3 71.0 TBD
PacifiCorp Exchange N/A Contract N/A 1954 50.0 0.0 03-31-04
PGE Capacity Sale N/A Contract N/A 1992 150.0 0.0 12-31-16
Upriver Dam Spokane Contract Spokane, WA 1966 14.4 10.0 06-30-04
WNP-3 N/A Contract N/A 1987 82.0 48.0 06-30-19
Medium-Term
Purchases
N/A Contract N/A 2004 100.0 100.0 12-31-10

Load Forecast

The Company develops a 20-year load forecast for the IRP process. Loads from 1997 through
2002 have been relatively flat. This is the result of several factors. The energy crisis of 2001
included the implementation of widespread conservation efforts. In 2002, higher retail electric
prices reinforced customer conservation efforts modestly. Also, due to the economic slowdown
in recent years, several large industrial facilities served by the Company were permanently
closed.

The twenty-year forecast assumes no additional large customer closures, retail electric prices that
increase slightly below the prevailing rate of inflation, and a modestly healthy economy.
Conservation acquisitions are expected to continue throughout the forecast horizon and energy
efficient equipment will be installed in new construction and replace retired equipment in
residences and businesses. The overall growth rate of retail electricity sales averages 3.2% per

1 indicates when ownership/contract began 2 represents Company share of project in 2004; hydro generation assumes "average" water from NWPP 2000/01
3 Indicates when contract/license will expire
Section 2 Page 6 Loads & Resources
year over the planning period. Refer to the following chart for a forecast of annual system load,
including weather-adjusted actual load for 1997 to 2002. Additional information regarding the
Company’s load forecast may be found in Appendix B.

Chart 2.1
System Retail Load Forecast

Energy Position

Table 2.2 contains a summary of annual loads and resources for 2004-2008, as well as 2013,
2018, and 2023. The table shows that, on an annual basis, the Company is surplus through 2006.

Table 2.2
Loads & Resources Energy Forecast (aMW)

2004 2005 2006 2007 2008 2013 2018 2023
Obligations
System Retail Load 985 1,014 1,051 1,083 1,120 1,326 1,569 1,860
DSM Load 2 5 10 14 19 41 64 56
80% Conf. Interval 189 189 189 189 189 189 189 153
Total Obligations 1,176 1,208 1,250 1,286 1,328 1,556 1,822 2,069
Resources
Hydro 550 545 530 530 529 477 471 458
DSM Resource 2 5 10 14 19 41 64 56
Net Contracts 156 157 175 177 177 58 59 12
Base Thermal 223 230 223 223 230 230 230 230
Gas Dispatch 158 156 158 158 156 158 158 156
Gas Peaking Units 181 181 181 181 181 181 181 181
Total Resources 1,270 1,274 1,277 1,283 1,292 1,145 1,163 1,093
Net Position 94 66 27 -3 -36 -411 -659 -976

800
1,000
1,200
1,400
1,600
1,800
2,000
19
9
7
19
9
9
20
0
1
20
0
3
20
0
5
20
0
7
20
0
9
20
1
1
20
1
3
20
1
5
20
1
7
20
1
9
20
2
1
20
2
3
Av
e
r
a
g
e

M
e
g
a
w
a
t
t
s
Actuals
Forecast
Section 2 Page 7 Loads & Resources
As referenced in Section 3, demand-side management (DSM) acquisitions are prescriptive. In
other words, without “programmatic” DSM acquisitions, retail loads and supply-side resource
acquisitions would be higher. This is represented in the table above by including DSM as both
an obligation and a resource. Subsequent tables, for simplification, net DSM obligations and
resources to zero. For detailed information about interactions between DSM and the Company’s
retail load forecast, refer to Appendix B. The DSM projections, as represented in this table, are
cumulative beginning in 2004, and illustrate the Company’s commitment to future acquisitions
of cost-effective DSM.

On a monthly basis, the Company expects to encounter energy deficits during some months in all
years of the forecast. In 2004, for example, the Company position is deficit in March,
September, and October, even though the annual position is surplus by 94 aMW. In other
months, particularly during spring runoff, the Company is in a surplus position. The Company
may balance its monthly positions through short-term market purchases or sales, exchanges, or
other resource arrangements.

As a general guideline, the annual energy position is used to determine when the Company needs
to acquire additional base-load energy resources. The first significant annual energy deficit is
expected in 2008. This deficit is forecasted to grow to 411 aMW by 2013 and 976 aMW by
2023. Load growth and reduced Mid-Columbia generation account for the significant majority
of increasing deficits during this period. For further details, including tabulations of annual and
monthly energy positions for 2004 to 2023, refer to Appendix F.

Capacity Position

The Company develops a twenty-year tabulation of peak capacity loads and resources. Peak
load is defined as the maximum one-hour load obligation on the expected average coldest day in
January, plus operating reserves. Peak resource capability is defined as the maximum one-hour
generation capability of Company resources, plus the net contract contribution. This tabulation
shows whether the Company has sufficient resources to meet its maximum expected one-hour
obligation.

The Company is in a surplus capacity position through 2009. Annual capacity deficits begin in
2010, with winter peak loads exceeding peak resource capability by more than 100 MW. The
deficits continue to grow as peaking requirements increase with load growth, and the Company's
resource base declines due to the expiration of market purchases and reductions in power from
Mid-Columbia project contracts. Table 2.3 includes the annual capacity forecast for 2004-2008,
as well as 2013, 2018, and 2023. For further details, including tabulations of annual and monthly
capacity positions for 2004 to 2023, refer to Appendix F.

Section 2 Page 8 Loads & Resources
Table 2.3
Loads & Resources Capacity Forecast (MW)

2004 2005 2006 2007 2008 2013 2018 2023
Obligations
Retail Load 1,470 1,515 1,570 1,617 1,672 1,982 2,349 2,780
Operating Reserves 110 110 108 108 108 104 103 101
Total Obligations 1,580 1,625 1,678 1,725 1,780 2,086 2,452 2,881
Resources
Hydro 1,177 1,177 1,135 1,134 1,133 1,043 1,035 998
Net Contracts 70 19 43 45 45 -73 78 -2
Base Thermal 272 272 272 272 272 272 272 272
Gas Dispatch 176 176 176 176 176 176 176 176
Gas Peaking Units 236 236 236 236 236 236 236 236
Total Resources 1,931 1,880 1,862 1,863 1,862 1,654 1,797 1,680
Net Position 351 255 184 138 82 -432 -655 -1,201
Reserve Margin 23.8% 16.8% 11.7% 8.5% 4.9% -21.8% -27.9% -43.2%

The Company currently has sufficient capacity resources, due primarily to the relative large
amount of hydroelectric generation in its resource portfolio. Typically, hydroelectric resources
provide a large amount of capacity in relation to the amount of energy they produce. Additional
capacity resources will be acquired when new resources are secured to meet future energy
deficits. For the most part, future capacity requirements will be met through the acquisition of
new resources, which provide both energy and capacity. However, as new resources are added
the Company’s resource base will include a lower percentage of hydro, and may include
resources, such as wind, which do not provide capacity.

This IRP focuses on meeting the Company's energy requirements to the eighty percent
confidence level. The eighty percent confidence level generally meets capacity requirements for
planning purposes. As explained in Section 7, only after 2009 do reserve margins fall below
twelve percent where resources are built to meet the 80 percent confidence interval. The
Company will address capacity planning margins in more detail in its 2005 IRP.

Planning Reserves

Planning reserves include components for meeting higher than expected loads due to severe
weather, unplanned generator-forced outages, adverse hydrological conditions, and other
contingencies. Historically, the Company’s planning reserves have not been based on unit size
or resource type; planning reserves have been set at a level equal to ten percent of the one-hour
system peak load, plus 90 MW. Together, these have equated to approximately a fifteen-percent
planning reserve margin during the Company’s peak load hour. The Company planning reserve
level, while not explicitly considered in the calculation, meets its operating reserve requirement
levels of five and seven percent for hydroelectric and thermal generation, respectively.

Section 2 Page 9 Loads & Resources
Confidence Interval Planning

The Company has evaluated a planning reserve methodology that accounts for deviations caused
by abnormal monthly weather patterns and below-average monthly hydroelectric capability.
Extreme weather can change monthly obligations by as much as 30 percent. In the event the
Company does not have adequate generation capability to meet this load variation, it is exposed
to the volatile short-term electricity marketplace.

Potentially more significant is hydroelectric generation variability. During 2001 the Company’s
hydroelectric generation level was the lowest ever recorded. In total, hydroelectric generation
over the year was down 181 aMW, or 33 percent, from an average of 550. Monthly reductions
were even more pronounced, with generation down nearly 50 percent in both February and
August.

Evaluation of the historical data shows that a superior planning criterion is the use of a
“confidence interval” based on 80 percent of the monthly variability of load and hydroelectric
generation. This means that for each month there is only a ten percent chance that the
combination of load and hydro variability would exceed the planning criteria. In other words,
for a given month there is a ten percent chance the Company would need to purchase some
amount of energy from the market.

The Company has considered confidence intervals higher than 80 percent, such as 95 or 99
percent, but believes based on current analysis that the cost of constructing resources to cover
this level of variability exceeds the potential benefits. For example, while building to the 99
percent confidence interval would decrease the frequency of market purchases significantly, such
a criterion would require approximately 200 MW of additional generation capability. This
would result in potential rate pressure resulting from additional capital expenditures.

On a monthly basis, the 80 percent confidence level varies between 77 and 268 aMW. The
average of the 80 percent confidence interval across the twelve months of the year equals 153
aMW. This level is similar to critical water planning on an annual basis, but is more precise
since it is based on the chance of exceedance by month.

In addition to load and hydroelectric variability, the Company’s WNP-3 contract with BPA
includes a return energy provision that can equate to an annual obligation of 36 aMW. The
contract would be exercised under adverse conditions, such as low hydroelectric generation
and/or high loads—coincident to conditions where the Company would expect its own system to
require additional resources. As a result, requirements under the confidence interval are
increased by 36 aMW to account for the WNP-3 obligation through its expiration in 2019.
Section 2 Page 10 Loads & Resources
Summary

The Company has adequate resources to meet its future annual load obligations through 2007,
including accounting for reserve margins and hydro and load variability. On an annual average
energy basis, the Company's first significant deficit occurs in 2008. On a capacity basis, the first
deficit occurs in 2010. However, on a monthly basis, the Company has deficiencies and will
investigate various ways to manage them.
Section 3 Page 11 Demand-Side Management
Section

Demand-Side Management

Historic DSM Activities

Since 1995 the Company has funded the acquisition of demand-side management (DSM)
resources through a “tariff rider” mechanism levied upon retail electric rates (through Schedule
91). Currently, the electric tariff rider stands at an amount equal to 1.48% of retail rates in
Washington and 1.95% in Idaho. Tariff rider revenues, DSM expenditures and the tariff rider
balances are separately tracked by jurisdiction. The following chart represents annual and
cumulative energy savings resulting from the Company’s DSM activities.

Chart 3.1
Annual and Cumulative Energy Savings
1978-2002

During the summer of 2001 the Company launched a series of emergency programs and
incentive enhancements to existing programs in response to the regional energy crisis. Final
calculations of the January to August 2001 impact of DSM programs indicate that the Company
acquired 437 percent of our energy savings goal during the first eight months of 2001 at the cost
of expending 281 percent of incoming tariff rider revenues. By the close of calendar year 2001
these extraordinary programs had resulted in a negative tariff rider balance of $12.2 million
($11.6 million electric and $0.6 million natural gas).

In the fall of 2001 a four-year (2002-2005) business plan was developed to simultaneously move
the tariff rider balance back to zero while continuing to deliver energy savings that are at least
proportionate to the percentage of tariff rider funds being expended. Based upon tariff rider
0
20
40
60
80
100
120
140
160
19
7
8
19
8
0
19
8
2
19
8
4
19
8
6
19
8
8
19
9
0
19
9
2
19
9
4
19
9
6
19
9
8
20
0
0
20
0
2
An
n
u
a
l

S
a
v
i
n
g
s

(
G
W
h
)
0
100
200
300
400
500
600
700
Cu
m
u
l
a
t
i
v
e

S
a
v
i
n
g
s

(
G
W
h
)
Annual
Savin s
Cumulative
Savings
Section 3 Page 12 Demand-Side Management
revenue projections over that time period, only 62 percent of incoming revenue will be available
for expenditure. The remaining amount would be dedicated to reducing the tariff rider balance.

The DSM business plan developed for 2002 to 2005 does not include any reductions to the
incentives specified to retail electric and natural gas customers (through Schedules 90 and 190),
nor is there a significant reduction in the availability of residential programs.

In addition to revenues generated from the DSM tariff rider, the Company also receives
$394,200 annually in Conservation and Renewable Discount (C&RD) program benefits from
Bonneville Power Administration (BPA). Though the C&RD funds are entirely separate from
the Company’s DSM funding, the two are managed to maximize the collective impact upon
DSM resource acquisition.

C&RD funding extends from October 2001 to September 2006; the five-year BPA rate case
period. The first year funded a 2001 compact fluorescent program. The remaining four years of
funding have been reserved for limited income programs (up to 75% of the funds) and a
conservation voltage reduction (CVR) pilot project that the Company is investigating with the
Northwest Energy Efficiency Alliance (NEEA).

The Company plans to deliver more energy savings per dollar expended than stated in our
Schedule 90 goal. Toward that end, the Company will target low-cost and no-cost efficiency
measures, lost opportunities and proven cost-effective measures. The programs are expected to
continue to be cost-effective.

The Company currently acquires DSM resources from a number of energy-efficiency
technologies delivered through commercial/industrial, residential, and limited income portfolios.
Please refer to the following chart for a depiction of each technology’s contribution to the
Company’s total DSM savings, using 2002 as an example.

Chart 3.2
DSM Resource Acquisition by Technology
2002
Appliances
Compressed Air
Controls
HVAC
Industrial Process
Lighting Motors
Renewables
Resource
Management
Shell
Section 3 Page 13 Demand-Side Management
Intervener Involvement

Company DSM activities are under the continuous review of an oversight board known as the
External Energy-Efficiency (Triple-E) board. This board is convened semi-annually to review
the status of electric and natural gas DSM programs. Analysis of the cost-effectiveness, energy
savings and other descriptive statistics are incorporated into periodic reports to the Triple-E
board.

Future DSM Activities

Near-term DSM operations follow through on the existing 2002 to 2005 business plan. Though
the implementation details are updated on a monthly basis, the core business plan rests upon
three fundamental priorities. These priorities are, in descending order of priority:

1. Satisfy least-cost resource requirements and expectations.
2. Field an overall DSM portfolio that is cost-effective on a societal and utility basis.
3. Return the tariff rider balance to zero in a timely manner.

In order to meet these objectives the Company has targeted:

• low-cost and no-cost DSM measures;
• traditional efficiency measures which are commercially-available, reliable, and generate
predictable and cost-effective energy savings; and
• lost opportunity measures.

With the exception of lost opportunities, the DSM business plan also calls for a diminished
emphasis on energy-efficiency technologies in the early commercialization phase. Historically,
these measures have been granted ”new technology” status that, under the provisions of Schedule
90, allow for enhanced customer direct incentives.

Current policy requires a business plan to be structured around any measure granted new
technology status. New technology business plans require all avenues to be reviewed for
enhancing the penetration of cost-effective measures in the early commercialization phase,
including non-incentive as well as incentive approaches. Exit strategies are a required
component of each new technology business plan. Under these circumstances the new
technology measures are essentially local-area market transformation ventures.

Since 1997, regional market transformation beyond the scope of an individual utility has been
within the realm of the Northwest Energy Efficiency Alliance (NEEA). At present, the
Company is contractually committed to funding four percent of NEEA expenditures through the
end of 2004. This proportion is based upon the Company’s percentage of end-use energy sales
within NEEA’s four-state area. The Company will evaluate continued participation in NEEA in
2005 and beyond during the last year of the existing contract.

Section 3 Page 14 Demand-Side Management
The Company is active in NEEA governance and operations. In recent years there has been an
increased coordination with NEEA on joint regional and local utility DSM operations. This
trend is expected to continue as the overlap between regional and local programs increases. As a
result of the successful collaborative effort with NEEA the Company does not plan on
independently initiating new energy-efficiency research and development efforts.

Conservation Voltage Reduction

Conservation voltage reduction (CVR) is likely to be the most significant measure to be
coordinated across the region. The Company has experimented with various approaches to
voltage control for energy-efficiency purposes in the past. Although these experiments have
indicated that CVR may be a significant and cost-effective DSM resource, they were not
extensive enough to be statistically significant. They have also been too limited in scope to
establish the determinants of the energy savings and the non-energy impacts of various
approaches to CVR.

In January 2003, NEEA adopted a CVR venture intended to complete extensive testing of a
variety of approaches to CVR. The expected outcome of the venture will be a determination of
the energy savings available and the non-energy impact under a variety of circumstances. The
study will also develop recommended protocols for implementation of CVR measures by
utilities. NEEA will also work with regulatory agencies to address the financial issues involved
in the adoption of these measures.

The Company is discussing the potential for coordinating with NEEA on a CVR pilot on its
system. The Company intends to contract with NEEA to complete and fully evaluate the pilot
study. Such a cooperative effort would meet the requirements for use of C&RD funding. Based
upon the results of this pilot, the Company will evaluate the cost-effectiveness of expanding the
adoption of the measure on a wider scale.

Automated Meter Reading/Time of Use Metering

The Company continues to monitor residential and industrial time-of-use (TOU) programs, as
well as ways to utilize automated meter reading (AMR) technologies to facilitate these efforts.
Market conditions and the current disconnect between wholesale and retail power markets has
recently been the focus of intense discussion. Wholesale power costs can vary significantly
across hours, days, and seasons. However, most customers in the Northwest pay fixed retail
prices. As a result, these customers do not have a price signal to incent them to manage their
usage during periods of high wholesale prices.

Various demand response mechanisms have been suggested to remedy this problem. Time-of-
use pricing has been studied in some detail in a variety of pilot and permanent programs. A
recent study on TOU provides insight into the potential benefits of this program for the
Company’s customers. Approximately 100 utilities across the nation were surveyed and the
analysis found that nearly 85% had some form of TOU tariff filed with their Utility Commission.
Section 3 Page 15 Demand-Side Management
However, the research found that less than one percent of the served residential customers
participated in the programs.

The Company has concluded that it is not economically viable to implement a full scale
residential time-of-use program prior to the implementation of an AMR system that bears the
metering and other technology costs necessary to support TOU. While an AMR system would
provide certain benefits, its immediate implementation is not critical for reliability or for the
ongoing business operations of the Company. The Company will continue to evaluate the costs
and benefits of an AMR system.

DSM in AURORA

Historically the Company has integrated supply and demand-side resources by evaluating
supply-side resource options, determining the deferrable resource and consequential avoided
cost, and subsequently applying that price signal to the selection of demand-side resources.
Integration of the two components of the resource plan is achieved by ensuring that demand-side
resources available at or below the avoided cost of that deferrable resource are acquired. This
approach does assume that demand-side resources are not a sufficiently large component of the
resource plan to change the selected deferrable resource. In this plan, and in prior plans, this has
been a reasonable assumption.

In the current IRP process the Company has applied a more explicit integration of supply and
demand-side resources, through incorporation of Company-specific DSM programs into the
AURORA model. This allowed Company DSM programs to be evaluated against hourly market
prices in parallel with supply-side resources.

Model Inputs and Assumptions

Developing demand-side resources for incorporation into AURORA involved several steps.
First, the Company identified six individual components of DSM measures based on actual
conservation activities. Utility costs and acquisition levels were indexed based on historic data.
These six components account for the vast majority of the historic energy savings, and are as
follows:

1. commercial heating, ventilation, and air conditioning (HVAC)
2. commercial lighting
3. commercial domestic hot water (DHW)
4. residential HVAC
5. residential lighting
6. residential DHW

Based upon a review of current projects and project economics, it was possible to estimate the
additional acquisition achievable given additional utility expenditures within each of the six
DSM components. For each component, the actual and three incremental points trace out the
DSM supply curve that is achievable with each incremental increase in utility expenditure. The
Section 3 Page 16 Demand-Side Management
incremental utility costs tested were based upon 25 percent increases to the current level of DSM
funding and represent alternative points on the supply curve. The estimated DSM acquisition
resulting from additional utility expenditure was based upon the technical and economic
potential for the measures represented in each DSM component and the ability of utility DSM
programs to capture that potential.

It was assumed that the Company would be able to move from the current point on the supply
curve to any of the three incremental points instantaneously and at no additional cost per aMW.
This assumption is based upon actual experience in ramping DSM acquisition activities up and
down over time. However, in the event that very substantial increases in utility acquisition were
necessary within a very short timeframe, such as was the case in the summer of 2001, it would
have been wise to assume significantly higher utility costs per aMW. Graphically this would be
depicted by a supply curve asymptotically approaching the vertical line representing the service
territory’s short-term technical DSM potential. Refer to Appendix Q for additional information.

In order to test each of the six DSM components against alternative resources or against the
avoided cost established by the AURORA model, it was necessary to develop hourly load
shapes. These 24-hour load shapes were estimated for a typical week for each of the twelve
months. The result was a "24 x 7 x 12" load shape for use in AURORA. There was a certain
amount of replication when, for example, there was no reason to believe that an hourly Tuesday
load shape would differ from the corresponding Thursday load shape. Similarly, some monthly
load shapes were combined into summer, winter, and shoulder seasons if appropriate for that
particular set of DSM measures.

Specific load shapes were derived from various sources available to the Company. Actual
measurement and evaluation (M&E) data from performance contracts or projects that were
sampled as part of the Company's analytical process was used as much as possible. This was
augmented by BPA End Use Load and Consumer Assessment Program (ELCAP) data on
occasion. The results were also modified to include engineering estimates of new technologies
that may not be fully represented in the Company's historic M&E process. For more detail
regarding the load shapes utilized in this analysis, refer to Appendix Q.

DSM Modeling Results

The DSM measures listed above were incorporated into AURORA as 24 individual resources
(four economic tiers for each of six measures). Each resource was modeled as non-dispatchable
and forced to sell into the marketplace for every hour of the twenty-year study term. The profit
or loss the resource generated was recorded for each hour, effectively resulting in the hourly
market value. The following table includes the results of this exercise, summarized for 2004-
2008, 2013, 2018, and 2023. The table also includes the twenty-year present value for each
measure, based on a discount rate of 8.22 percent as determined in the Company’s most recent
Washington General Rate Case. Please refer to Appendix Q for a table including results for all
years of the study.

Section 3 Page 17 Demand-Side Management
Table 3.1
DSM Resource Net Market Value
2004-2008, 2013, 2018 & 2023 (in thousands of dollars)

2004 2005 2006 2007 2008 2013 2018 2023 NPV
Com HVAC 1 -467 -417 -417 -330 -221 87 192 117 -1,090
Com HVAC 2 -116 -112 -114 -107 -98 -76 -78 -102 -903
Com HVAC 3 -20 -20 -20 -20 -19 -18 -20 -24 -190
Com HVAC 4 -3 -3 -3 -3 -3 -3 -4 -4 -31

Com Ltg 1 360 383 392 427 469 610 729 818 5,248
Com Ltg 2 29 31 32 35 39 52 63 70 443
Com Ltg 3 2 2 2 3 3 4 5 6 34
Com Ltg 4 0 0 0 0 0 0 0 0 2

Com DHW 1 6 7 7 7 8 11 13 14 92
Com DHW 2 1 1 1 1 1 1 1 1 8
Com DHW 3 0 0 0 0 0 0 0 0 1
Com DHW 4 0 0 0 0 0 0 0 0 0

Res HVAC 1 64 68 70 78 87 116 140 156 984
Res HVAC 2 4 5 5 6 6 9 11 12 74
Res HVAC 3 0 0 0 0 0 1 1 1 4
Res HVAC 4 0 0 0 0 0 0 0 0 0

Res Ltg 1 516 561 578 651 740 1,037 1,254 1,392 8,563
Res Ltg 2 28 32 33 40 48 75 93 101 588
Res Ltg 3 0 0 0 1 2 4 5 5 25
Res Ltg 4 0 0 0 0 0 0 0 0 0

Res DHW 1 0 0 0 0 0 1 1 1 6
Res DHW 2 0 0 0 0 0 0 0 0 0
Res DHW 3 0 0 0 0 0 0 0 0 0
Res DHW 4 0 0 0 0 0 0 0 0 0

As shown in the table above, each of the six DSM components and each of the four price
alternatives within each component was evaluated against AURORA-defined market prices for
the twenty-year planning period. This resulted in 24 streams of annual mark-to-market results.

By calculating a present value of these annual streams it is possible to determine if a resource
installed in a particular year will generate future value (relative to market) sufficient to make that
stream cost-effective. The most significant question lies in the appropriate term to be used for
that present value calculation. At least two reasonable alternatives exist. The first would be to
calculate a twenty-year present value covering the entire forecast period. The alternative would
be to calculate a moving present value equal to the measure life specific to that DSM component.

For purposes of deriving actionable information out of the integrated resource planning process,
this was not a significant issue. Two of the DSM components (those related to HVAC measures)
have a measure life of twenty years, thus encompassing the entire forecast period. The other two
measures (domestic hot water and lighting) have been deemed to generate ten years of savings in
the Company’s current cost-effectiveness analysis. However, most of the twenty-four individual
Section 3 Page 18 Demand-Side Management
streams of savings do not cross the zero line, and for these streams the resource would be
selected or not selected regardless of the term of the present valuing methodology.

The Company intends to create an actionable plan from this AURORA analysis of DSM
alternatives. Schedule 90, under which the Company acquires DSM resources, has historically
been interpreted as applying to any electrical-efficiency device available in the
commercial/industrial sector. Under this precedent it is not possible to exclude particular
measures from inclusion in the DSM resource portfolio. The Company does, however, have the
ability to target specific technology applications that appear to be more cost-effective than
others. Within the commercial/industrial sector the AURORA results will be used to perform
this targeting.

The Company implements residential DSM programs differently. Within this customer segment
prescriptive programs are developed and made available to customers. Thus there is a greater
ability to add or remove technology applications from this portfolio. The AURORA results will
also be used to identify technologies to be targeted in limited income residential programs.

An additional consideration is one of the most efficient ways to acquire the resources identified
as being cost-effective. Several technology applications are better pursued through a mix of
regional and local programs. The Company is supportive of funding cost-effective regional
market transformation when it is the most efficient way to acquire targeted DSM resources.
Section 4 Page 19 New Resource Alternatives
Section

New Resource Alternatives

Overview

This section will discuss the resource alternatives considered by the Company to meet its future
retail load requirements. In previous IRPs the Company included analyses for a very wide range
of resource alternatives. The approach in this IRP is to focus analysis on technologies likely to
be part of a least-cost mix.

General Approach

This IRP considers generic resource alternatives, rather than specific projects that the Company
might choose. This approach was selected for three reasons. First, the Company wants to
consider the affect on its portfolio of differing resource types without project-specific economics
impacting the result. This provides a more consistent comparison of technologies than site-
specific economics.

Second, the approach provides greater transparency of resource alternatives and assumptions. To
this end, this IRP adopts resources and associated characteristics from the forthcoming
Northwest Power Planning Council (NWPPC) Fifth Power Plan. The NWPPC resource
alternatives were formulated over a period of months through a committee of regional experts
drawn from utilities, developers, regulators, and other interested parties.

Third, the Company does not have an immediate resource deficiency on an annual average basis.
Without an immediate need on the horizon, the Company has not recently studied site-specific
projects. Instead, this IRP provides a framework of analysis that the Company expects to revisit
at the time it procures additional resources. At that time, assumptions would be updated to
include site-specific resource alternatives. Specific resource alternatives drawn from, for
example, a Request for Proposals (RFP) would be evaluated in the same manner as the NWPPC
resources used in this study.

New Resource Alternatives Considered

Five new resource alternatives were incorporated into the AURORA model as part of the 2004-
2023 capacity expansion plan for the WECC. Underlying assumptions for each resource were
taken from recent work by the NWPPC for its forthcoming Fifth Power Plan. The assumptions
were derived from a working forum of utility experts, merchant plant developers, BPA, and other
interested parties. For a more detailed discussion of the assumptions behind new resource
alternatives, see Appendix P.

Section 4 Page 20 New Resource Alternatives
The following table provides a brief description of each technology and key underlying
assumptions. The resource assumptions, excluding levelized cost calculations, were taken from
the NWPPC except where noted. Refer to Section 5 for more information on the AURORA
model and capacity expansion.

Table 4.1
New Resource Alternatives
(in 2000 Dollars)

Installed Unit Heat Unit Fixed Variable Levelized Cost
Cost Capacity Rate Availability O&M O&M AURORA Max Gen
Resource ($/kW) (MW) (Btu/kWh) (percent) ($/kW/yr) ($/MWh) ($/MWh) ($/MWh)
CCCT 686 280 6,946 92 26 2.80 56.21 51.91
SCCT 730 92 9,486 94 8 3.70 93.53 60.05
Coal 1,230 400 9,550 84 55 1.75 58.05 57.09
Wind 679 100 N/A 30 35 0.50 52.64 52.64
Solar 6,000 20 N/A 22 30 0.00 N/A N/A
Cogen 1,000 25 5,500 85 26 2.00 74.71 57.37

Unit availability accounts for both maintenance and forced outage, and is based on assumptions
from the NWPPC. Wind plant availability varies by region, but on average wind plants are
modeled to generate at a thirty percent capacity factor. Solar is shaped by hour over the year
with an average availability of 22 percent.

Heat rates for CCCT, SCCT, and coal plants are expected to improve over time. The NWPPC
assumes that, for example, CCCT heat rates will improve from an average of 6,946 Btu/kWh
today to 6,195 in 2023, a reduction of thirteen percent. Coal plant heat rates are expected to
improve by 4.5 percent over the same timeframe.

Fixed operation and maintenance (O&M) figures include maintenance and transmission costs of
$15 per kW-year, except for SCCT plants, where non-firm transmission service is assumed.
These assumptions are based on NWPPC datasets.

The levelized cost calculations are based on a discount rate of 8.22 percent as determined in the
Company’s most recent Washington General Rate Case. This discount rate is used for all
levelized cost and present value calculations throughout the document.

Levelized costs are presented assuming two levels of generation: the average output levels as
modeled in AURORA and maximum generation levels where economic dispatch is ignored. The
AURORA generation levelized costs are higher, as the plants are operated only when they are
lower cost than the wholesale marketplace. The levelized costs at maximum generation levels
assume that, except for maintenance and forced outage, plants run during all hours. Even though
levelized costs are lower, calculations at maximum generation are unrealistic, as the marketplace
dictates that most plants will not be economic during all hours of their lifetimes.

The Company diverged modestly from NWPPC resource assumptions in three areas: CCCT
configuration, the federal production tax credit for wind, and transmission costs for new coal
plants. The NWPPC assumes a “two-on-one” configuration for CCCTs. Two-on-one
Section 4 Page 21 New Resource Alternatives
configurations consist of two gas turbines exhausting waste heat into a single heat recovery
steam generator (HRSG), rather than one gas turbine matched to the HRSG as in more traditional
one-on-one configuration. The NWPPC assumes that modest cost efficiencies are gained
through the two-on-one configuration. However, based on its own experience, the Company is
concerned that the NWPPC has assumed costs that are too low for CCCT technology. The
Company believes that the larger size of the two-on-one configuration may be beyond the
incremental load requirements of utility companies building them. The IRP instead uses
NWPPC assumptions for a one-on-one configuration.

The NWPPC models the federal production tax credit for wind as an offset to variable O&M
costs. For the IRP, the Company instead reduced capital costs by an amount equal to the present
value of the NWPPC-assumed ten-year credit. The ultimate impact of this change was
negligible, but it simplified modeling within the IRP process.

The Company also does not believe that the NWPPC adequately addresses the incremental cost
of new transmission facilities necessary to integrate coal plants into the Northwest. Existing
transmission lines out of eastern-WECC states, where coal plants likely will be built (e.g.,
Montana, Wyoming), into the Northwest do not have capacity adequate to integrate large coal
plant developments. Therefore new and upgraded transmission facilities will be required to
integrate the plants. The IRP assumes that an additional $333 per installed kW of coal-fired
generation is required to cover the cost for new transmission facilities. This adjustment amounts
to an incremental levelized cost of about six to seven dollars per MWh of coal-fired generation.

The Company also included a generic cogeneration plant. This resource was not explicitly
modeled in the AURORA capacity expansion plan, but was evaluated as a potential future
resource. In addition, to evaluate the impact of a fixed-price contract on the Company’s risk
profile, a 100 MW contract was modeled as a potential resource.

Resources Not Evaluated

Many resource alternatives are available to the Company, however, applying basic cost-
effectiveness screens greatly reduces the opportunities. In the Company’s 2001 IRP, 32 resource
options were depicted, using information gathered from the NWPPC. While this list was
extensive, it was mostly comprised of uneconomic alternatives. For example, various
geothermal projects were evaluated, and estimated to cost more than 100 dollars per MWh.
Evaluating such resources within the IRP models would clearly lead to their exclusion from
consideration in a least-cost mix due. Other resources not considered in this IRP include nuclear,
advanced coal, bio-gasification, new hydroelectric generation facilities, and various high-cost
solar projects.
Section 5 Page 22 Modeling
Section

Modeling

Overview

Integrated resource planning typically considers many alternative strategies to identify an
optimum portfolio of resources matching future loads. Historically, IRP analyses have relied on
straightforward comparisons of future loads and resources on the basis of capacity and energy.
Resources were selected to meet deficiencies in a “least-cost” manner on a twenty-year present
value basis. Today, planning analyses are more quantitatively detailed for several reasons,
including:

• greater computing capabilities
• a viable wholesale electricity marketplace
• more capable resource modeling tools
• higher expectations from customers, regulators, and management

The result is a greater understanding of the potential impacts of varying resource decisions, and
enhanced assessment of strategies to reduce portfolio power supply risks.

In this IRP, the Company has enhanced its modeling capabilities even further, by including an
hourly production-cost model that dispatches resources to a given set of market conditions and
also develops a set of market prices responsive to varying levels of regional load, natural gas
prices, and hydroelectric conditions.

Modeling Process

For the purposes of this IRP, the AURORA model was used to simulate the entire Western
Electricity Coordinating Council (WECC) marketplace. Refer to Appendix C for a discussion of
the selection process whereby the Company chose AURORA for its planning efforts. The
WECC, as defined by AURORA, is separated into sixteen “load areas” based on geographical
regions of load concentration. Refer to the following table for a listing of the load areas included
in AURORA as part of the WECC. This table also provides a reference to define the acronyms
utilized throughout this document to describe these load areas.

Section 5 Page 23 Modeling
Table 5.1
AURORA Load Areas

Load
Area
Region(s)
Included
Load
Area
Region(s)
Included
AB Alberta IDSo Southern Idaho
AVA Avista MT Montana
AZ Arizona NM New Mexico
BajaN Baja Mexico NVNo Northern Nevada
BC British Columbia NVSo Southern Nevada
CANo Northern California OWI
Oregon, Washington,
and Northern Idaho
CASo Southern California UT Utah
CO Colorado WY Wyoming

The AVA load area listed above was developed in order to represent Company loads and
resources separately from those of OWI. For each of the load areas, the AURORA database
contains all of the corresponding loads and resources, and is capable of simulating the entire
system on an hourly basis. This simulation is used to derive market prices for each area and the
WECC as a whole. It also allows AURORA to compute statistics specific to individual
generating resources (e.g., fuel costs, dispatch margins, etc.) and individual loads (e.g., cost to
serve).

For this IRP, the Company utilized AURORA to simulate the WECC marketplace for twenty
years (2004-2023). As part of this simulation, AURORA builds new generation from a pool of
hypothetical resources to meet future load growth. This process is referred to as “capacity
expansion.” For further details on capacity expansion, refer to Appendix C.

AURORA is also capable of incorporating market uncertainty based on such variables as load,
fuel price, and hydroelectric generation. The Company utilized this capability by generating 200
sets of unique inputs for 200 distinct iterations of AURORA. Refer to Section 6 for more
information on this process. The results of the 200 iterations of AURORA were then input into a
spreadsheet model that utilized a Linear Programming (LP) Module to derive an optimal
solution. Refer to Appendix C for further details on utilization of the LP Module and a
discussion of linear programming theory.

Assumptions and Inputs

AURORA contains a database with generic data sets that provide a reasonable approximation of
market conditions in the future. To obtain more robust results, the Company modified many of
the base data sets. The following section describes the changes made by the Company.
Section 5 Page 24 Modeling
Hydroelectric Generation

The AURORA model contains a hydrological data set for the WECC. Northwest data includes
average monthly generation levels taken from BPA 50-year hydrological studies. The Company,
for its planning purposes, uses hydrological data from the Northwest Power Pool (NWPP) rather
than that from BPA. Presently the NWPP performs 60-year headwater benefit studies annually
for the Northwest hydroelectric system.

Data from the 2000-2001 version of this study was converted into an AURORA format and
utilized in place of existing Northwest data sets for IRP modeling. Results for the Northwest are
similar – the average annual generation level from the 60-year study for generation in Oregon,
Washington and Northern Idaho is 1.7 percent higher than the AURORA default data set.
AURORA data sets for hydroelectric systems outside the Northwest (e.g., California) were not
modified.

AURORA models hydroelectric generation by load area. In other words, every hydroelectric
facility located within a load area utilizes the same shaping factors. The results for each
hydroelectric system are accurate, but individual projects are not necessarily represented
correctly. To track Company hydroelectric resources more accurately, each of the Company’s
river systems was algebraically separated from the base AURORA data sets and assigned a
unique set of shaping factors.

Natural Gas Prices

Natural gas is a key underlying assumption in the model because gas-fired resources presently
set the marginal price for WECC electricity in many hours. Therefore the Company used a
natural gas price forecast developed for its 2003 Natural Gas IRP. The forecast was developed
in early July 2002 using forward prices for approximately the first five years, and then a long-
term forecast purchased from DRI/WEFA.

For the 2003 Electric IRP, a forecast of Henry Hub natural gas prices was developed in addition
to the traditional price forecast used in the Natural Gas IRP. This was necessary for the
AURORA model, as it develops all of its natural gas prices using Henry Hub. For the
Company’s natural gas-fired plants, the Company developed basis differentials from Henry Hub
using available market-based information.

WECC Load

The Company made two key modifications to the AURORA regional load database. The first
was algebraically separating the Company’s retail load forecast from the AURORA OWI load
area forecast. Separating the Company’s retail load allowed it to be tracked separately for IRP
reporting.

The second modification was to the hourly shape of the loads in each AURORA region. The
AURORA data set was based on actual hourly load shapes from calendar year 2000. The
Company had already reviewed data sets from 1998 and 1999 to obtain data sets for Monte Carlo
Section 5 Page 25 Modeling
runs, and determined that this information would potentially provide a better hourly shape
because calendar year 2000 loads were affected by actions taken during the 2000-2001 energy
shortages. Refer to Section 6 for more detailed information.

Resources

A Company review of the WECC resources included in the AURORA database found it to be
both comprehensive and accurate for IRP purposes. The only substantial change to the
AURORA database was the addition of 295 MW of wind resources per year between 2003 and
2012. These quantities were adopted from the NWPPC Fifth Power Plan model and are intended
to represent the implementation of Renewable Portfolio Standards (RPS) in states presently
having such requirements. This addition is miniscule in comparison to the resource quantities
added by AURORA, but guaranteed that RPS requirements would be satisfied. The Company
also modified the cost of capital for new resources, which is detailed further in Appendix C.

Integration into AURORA

The Company’s departure from the AURORA risk input structure required the development of
an interface to automate the 200 iterations of Monte Carlo. The Company developed a
spreadsheet containing the statistical relationships for natural gas, hydroelectric generation, and
load variability. A Visual Basic program was developed to write the 200 individual data sets to a
database that AURORA could interface with.

A second set of Visual Basic code was then used to upload each of the iterations into AURORA
and write the results back to the database. The results from AURORA were then queried from
the database and input into the LP Module. For more information, including a graphical
representation of the entire modeling process, refer to Appendix C.

Analysis of Strategies

As discussed in Section 4, several potential new resources were included in AURORA to meet
the Company’s future resource deficiencies. Based on this pool of resources, several alternative
resource strategies (or “strategies”) were derived. While the number of strategies can be
virtually unlimited, the LP Module provided a means to evaluate portfolios the Company
believed were essential to understand. Strategies considered in the IRP included the following:

• No Additions – used to simulate what would happen if the Company made no resource
additions over the term of the IRP, and instead relied entirely on the short-term electric
marketplace to serve load requirements.
• Lowest Cost – designed to minimize the NPV of average net power supply expense.
• Lowest Risk – designed to minimize the average variance of net power supply expense.
• All CCCT – comprised entirely of natural gas-fired combined-cycle combustion turbines.
• All Coal – comprised entirely of coal-fired plants.
• Wind Strategy – comprised of wind turbines, supplemented with simple-cycle
combustion turbines (SCCTs).
Section 5 Page 26 Modeling
The Wind Strategy was selected to provide a renewable portfolio. Because wind is an energy-
only resource, integrating wind resources has proven to be a complicated analysis. The
Company has completed various wind studies since early 2002, as provided in Appendix H. Still,
the answer is not clear and the evaluation continues. In the absence of a definitive study, a level
of 50 MW of peaking plants to support 75 MW of wind generation has been selected for the IRP.
This amount might be modestly too high or too low; however, based on analyses to date, the
Company believes this level is appropriate for IRP planning.

As it turns out, the Lowest Cost strategy is constructed entirely of CCCTs, so the Lowest Cost
and All CCCT strategies are the same (referred to hereafter as Lowest Cost/CCCT). For more
information on these strategies, including analysis results, refer to Appendix E.

Analysis of Scenarios

Scenarios continue to play an important role in the Company’s IRP studies. Numerous scenarios
were evaluated for this report. Aside from the Base Case, which incorporates the results of all
200 iterations of Monte Carlo simulation, there are eight scenarios included in this IRP. These
scenarios were intended to represent distinct market conditions the Company may face in the
future, and include the following:

• Average – incorporates average hydroelectric generation, natural gas prices, and loads.
• Critical Water – hydroelectric levels for the Northwest are set to 1936-1937 (critical
year) levels.
• High Gas – assumes natural gas prices that are 200 percent of average.
• High Load – utilizes WECC loads that are 12.5 percent higher than average.
• Load Loss – incorporates a 300 aMW loss of the Company’s retail load.
• New Trans – incorporates an additional 12,000 MW of transmission from Montana and
Wyoming into the Northwest and Southern California.
• Coal Build – replaces the CCCTs built in capacity expansion with equivalent coal plants.
• Carbon Tax – includes carbon taxes on applicable generating resources.

For more information on these scenarios, including analysis results, refer to Appendix E.

Summary

Many enhancements to the modeling process were made for the 2003 IRP. The Company
acquired a new hourly market price forecasting tool capable of tracking and valuing specific
portfolios of resources. By developing 200 sets of potential market conditions, the Company
was able to evaluate not only the expected values of various resource decisions, but also the
potential risk inherent in those decisions. The LP Module provided an efficient means to select
least-cost resources, account for risk considerations, and compare alternative scenarios. Overall,
the Company believes that this combination of analytical tools provides an excellent framework
for this type of analysis.
Section 6 Page 27 Risk Analysis
Section

Risk Analysis

Overview

This section provides a discussion of the stochastic risk analyses performed in this IRP. Risk
factors include hydroelectric generation, natural gas price, and WECC load variability. This
section also describes the varying risks associated with resource alternatives available to the
Company.

Stochastic Risk Analysis

Stochastic risk analysis provides a method to evaluate how relationships among variables change
over time. The IRP model considers variability in hydroelectric generation, natural gas prices,
and WECC loads in developing a robust model that considers many possible futures. In this IRP,
stochastic risk analysis is achieved by applying statistical methods to AURORA model inputs,
generating numerous unique input data sets. AURORA then utilizes these input data sets in
numerous iterations to generate unique sets of results. The following section describes analyses
performed to obtain 200 unique iterations based on the risk components mentioned above, as
well as how they were integrated into AURORA.

Hydroelectric Generation

Possibly the greatest power supply risk the Company presently faces is variation in hydroelectric
generation. In 2001 the Company saw its annual generation fall to approximately 67 percent of
average. Monthly generation levels can vary even further. Planning for this amount of
variability has challenged Northwest utilities since the first dams were built.

The Northwest Power Pool (NWPP) provides an estimate of hydroelectric generation based on a
60-year record of stream flows. For the IRP, the Company evaluated the hydrological record
stochastically in an attempt to infer statistical relationships from the data set. Each month of the
year was evaluated, along with correlations between the hydroelectric plants residing in the
various AURORA load areas. Special attention was paid to the Northwest load areas modeled
by the NWPP, as shown in Table 6.1.

Table 6.1
WECC Hydroelectric Generation by AURORA Load Area

Area OWI BC CANo IDSo CASo MT Other Total
Capacity (MW) 30,790 10,473 7,928 2,497 2,433 1,851 6,038 62,010
Percent of Total 49.7 16.9 12.8 4.0 3.9 3.0 9.7 100.0

Section 6 Page 28 Risk Analysis
The Northwest areas, (indicated above in italics and gray shading), encompass nearly 75 percent
of hydroelectric generation in the WECC, with the AURORA load area Oregon/Washington/
North Idaho (OWI) accounting for approximately half of the total.

Since hydroelectric generation is not normally distributed, the ability to randomly generate
monthly hydroelectric generation levels is limited. As an alternative, specific water years were
drawn randomly from the NWPP data set. For example, if the 1945 water year was drawn,
hydroelectric generation levels for the Northwest load areas (OWI, BC, IDSo, and MT) would be
based on the 1945 data set from the NWPP. Hydroelectric generation levels for load areas not
modeled by the NWPP were assumed to remain constant at the levels provided in the base
AURORA dataset. The following chart presents the distribution of hydroelectric generation
modeled for the WECC.

Chart 6.1
Distribution of Hydroelectric Generation in WECC

Natural Gas Prices

Natural gas-fired resources have recently become the most common selection for meeting new
electric load requirements. Increased reliance on natural gas has made gas-fired turbines
marginal cost resources during many hours in the WECC. As more natural gas-fired plants are
built, the Company expects electricity prices to become even more correlated to natural gas
prices than they are today.

AURORA develops electricity prices by determining the marginal resource used to serve load in
each hour. The extent to which natural gas-fired resources are the marginal resource during a
given hour depends on the level of generation from other lower-cost resources. Chief among
these is hydroelectric generation. Reductions in hydroelectric generation will increase the
number of hours where natural gas-fired generators set the marginal price of electricity. This
relationship is modeled in the IRP by inversely correlating natural gas prices by 50 percent to
0
5
10
15
20
25
22
.
5
23
.
0
23
.
5
24
.
0
24
.
5
25
.
0
25
.
5
26
.
0
26
.
5
27
.
0
27
.
5
28
.
0
28
.
5
29
.
0
29
.
5
30
.
0
30
.
5
31
.
0
31
.
5
32
.
0
32
.
5
33
.
0
Average Gigawatts
Fr
e
q
u
e
n
c
y
Average = 27.7
Section 6 Page 29 Risk Analysis
hydroelectric generation levels. In other words, gas prices rise as hydroelectric generation
declines, and vise versa.

The IRP assumes that natural gas prices have a standard deviation of 50 percent where prices rise
above the average forecast and 25 percent where prices fall below the average forecast. Half of
the standard deviation is then allocated to the annual price, with the remainder applied to
represent monthly volatility. Annual prices are correlated to hydroelectric generation as
described above, while monthly volatility is randomized. The Company chose to reduce the
standard deviation when prices fall below the average value to reflect that, while prices are
effectively capped on the down side at zero, upward price movement is potentially unlimited.

The following table illustrates the natural gas prices modeled in 2005 over 200 iterations.
Annual average prices ranged between $1.82 and $6.75 per decatherm, a range of 130 percent.
The monthly range was 220 percent, varying between $1.14 and $9.67 per decatherm.

Table 6.2
2005 Henry Hub Natural Gas Prices Over 200 IRP Iterations
(in 2000 dollars per decatherm)

Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Ann
Avg 4.28 3.89 3.64 3.57 3.12 3.83 4.01 4.07 3.90 3.68 3.77 4.05 3.82
Min 1.48 1.59 1.59 1.37 1.36 1.14 1.53 1.77 1.58 1.78 1.79 1.78 1.82
Max 9.67 7.27 8.61 7.75 6.26 7.77 7.62 9.59 8.64 7.42 8.31 7.68 6.75

Load Variability

AURORA includes historical data for each load area, as well as a set of annual growth rates.
The historical data sets are specific to a recent calendar year. In the case of the current version of
AURORA, the default data sets are based on calendar year 2000.

Due to the significant impact of high prices during 2000, the data may not be representative of
future load variability. As such, the Company has used hourly load information for each utility
in the WECC during 1998 and 1999 obtained from FERC Form 714, and has determined the
statistical relationships between areas within the WECC.

Standard deviations for each load area were developed on a monthly basis, but the Company was
interested in modeling loads in a fashion that varies them on more than just a monthly basis.
This desire was based on the observation that during “average” months loads oftentimes are both
significantly higher and lower than the average would indicate.

Varying loads on a weekly basis better represents weather patterns and more realistically
represents WECC loads. Daily load shapes were based on actual daily loads for 1998 and 1999,
and were represented as a percentage of the average load for the week within which they reside.

Without correlating loads across the WECC, higher loads in one area would be inappropriately
offset by lower loads in another area during many hours of the study. To better model load
variability across the WECC, correlations were identified between all load areas and the OWI
Section 6 Page 30 Risk Analysis
load area. The FERC Form 714 data were separated into weekdays by month to remove the
inherent bias that otherwise would have resulted due to normal intra-week trends. The resultant
correlations were then tested for statistical significance, which eliminated approximately half of
the values.

The following table describes load and correlation statistics for the Northwest and California,
using January and August of 2004 as examples. “NotSig” indicates that the relationship was not
statistically significant. Additional details are available in Appendix D.

Table 6.3
2004 Load Statistics

OWI BC IDSo Montana CANo CASo
January
Load (aGW)
Standard Deviation 6.0% 4.9% 4.3% 3.4% 6.8% 7.7%
Correlation to OWI 100% 92% 67% 89% NotSig NotSig
August
Load (aGW)
Standard Deviation 6.0% 5.0% 5.1% 3.5% 11.0% 8.5%
Correlation to OWI 100% NotSig 79% 65% 76% 50%

Benefits and Risks of Resource Options

The Company’s current resource portfolio contains a significant level of cost variability, which is
largely due to its large reliance on hydroelectric generation. This risk is significant and will be
difficult to mitigate completely. By changing the mix of future resources, however, power
supply cost variability can be reduced. There are several important underlying assumptions with
regard to selecting a portfolio of future resources, including the following:

1. Owning resources in lieu of utilizing spot market purchases reduces risk. This is due to the
fact that roughly one-third of the total resource costs (in the case of gas turbines) to almost all
of the costs (in the case of wind) are fixed costs consisting of capital recovery and fixed
O&M. These costs do not vary, unlike short-term market prices.

2. Risk is reduced by capital intensive, low operating cost resources with stable fuel supplies.
Future resources that meet this criterion include coal and wind. Both coal and wind costs are
dominated by capital recovery and fixed O&M. Both have fuel supplies that aren’t
correlated with electricity prices and typically have operating costs low enough for the plant
to be dispatched (or “in the money”) when available.

3. Being close to load/resource balance generally reduces risk. Being either very short or very
long increases exposure to market prices, which causes power supply costs to vary. This is
even the case if the resources added are lower risk resources, such as coal or wind.
Individual resource alternatives have unique risk profiles. Refer to the following table for a
summary of these profiles for the resource alternatives considered in this IRP:
Section 6 Page 31 Risk Analysis

Table 6.4
Resource Profiles

Fuel
Resource Capital Operating Price Operating Other Other
Type Required Cost Risk Flexibility Advantages Disadvantages
CCCT Low High High Medium Daily dispatch Gas price correlated
/High with electric price
SCCT Low Very High High High Hourly dispatch Gas price correlated
with electric price
Coal High Low Low Limited Stable fuel price Environmental issues
Long transmission
or coal haul
Wind Very High Very Low None None No fuel cost System integration
Not correlated to market No capacity
Long-term supply reliable Fuel supply is unreliable
Renewable requirements
Cogen High Medium Varies Limited Overall high efficiency Need host sites
/None Can't add when needed
Contract issues
Purchases None NA None None No fuel price risk Credit issues
No forced outages Counter-party issues
Supply reliability issues
DSM High None None None Good customer relations Savings hard to verify
High efficiency

Further details regarding particular risk factors, as well as risk characteristics for specific
resources can be found in Appendix D.
Section 7 Page 32 Results
Section

Results

Overview

The following section details the modeling results of the IRP. It provides the stochastic values
for natural gas, hydroelectric generation, and WECC loads. It also provides details regarding the
Preferred Resource Strategy (PRS) developed for the IRP, and discusses the strategies and
scenarios that were considered.

WECC Market Prices and Volatility

As discussed in Section 5, the Company ran 200 iterations of hydroelectric generation, natural
gas prices, and load using the stochastic variables through the AURORA model. Resultant
natural gas and electric market prices for each of the 200 model runs are discussed below.

Wholesale Natural Gas Prices

The following chart provides projected wholesale natural gas prices over the twenty-year IRP
study. Natural gas prices begin in 2004 at $4.30 per decatherm and rise on average to $6.04 by
2023, for an annual increase of 1.7 percent. The larger dashes represent the lowest and highest
prices observed over the 200 iterations. The smaller dashes represent the range between which
80 percent of all iterations of natural gas fell.
Section 7 Page 33 Results
Chart 7.1
Annual Wholesale Natural Gas Prices
2004-2023

The following chart details monthly wholesale natural gas prices for 2004 over 200 iterations.
Natural gas prices in 2004 average $4.30 per decatherm. Annual 2004 prices vary over the 200
iterations from $0.99 to $7.01. Eighty percent of all iterations fall between $2.78 and $5.59 per
decatherm, on an annual average basis.

Chart 7.2
Monthly and Annual Wholesale Natural Gas Prices
2004

0
2
4
6
8
10
12
20
0
4
20
0
5
20
0
6
20
0
7
20
0
8
20
0
9
20
1
0
20
1
1
20
1
2
20
1
3
20
1
4
20
1
5
20
1
6
20
1
7
20
1
8
20
1
9
20
2
0
20
2
1
20
2
2
20
2
3
Do
l
l
a
r
s

p
e
r

D
e
c
a
t
h
e
r
m
Minimum 10%Average 90%Maximum
80%
C.I.
0
2
4
6
8
10
12
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Ann
Do
l
l
a
r
s

p
e
r

D
e
c
a
t
h
e
r
m
Minimum 10%Average 90%Maximum
80%
C.I.
Section 7 Page 34 Results
Northwest Wholesale Electricity Prices

Wholesale electricity market prices in the Northwest trend upward by an average rate of 4.1
percent over the IRP study horizon. The average price in 2004 is $33.76 per MWh. In 2023, the
price is $75.33 per MWh. The following chart presents annual average wholesale prices in the
Northwest over the IRP term, as well as minimum and maximum annual values and the band
within which 80 percent of all observations occur.

Chart 7.3
Annual Northwest Wholesale Electricity Prices
2004-2023

The following chart details average monthly and annual wholesale prices for 2004. Prices over
the year average $33.77 per MWh, and range from $5.81 to $80.31. Eighty percent of the
monthly iterations for 2004 fall between $16.38 and $54.88.

10
30
50
70
90
110
130
20
0
4
20
0
5
20
0
6
20
0
7
20
0
8
20
0
9
20
1
0
20
1
1
20
1
2
20
1
3
20
1
4
20
1
5
20
1
6
20
1
7
20
1
8
20
1
9
20
2
0
20
2
1
20
2
2
20
2
3
Do
l
l
a
r
s

p
e
r

M
e
g
a
w
a
t
t
-
H
o
u
r
Minimum 10%Average 90%Maximum
80%
C.I.
Section 7 Page 35 Results
Chart 7.4
Monthly and Annual Northwest Wholesale Electricity Prices
2004

WECC Regional Electricity Prices

AURORA forecasts wholesale electric market prices across the WECC. While the Company is
most impacted by Northwest prices, other areas can affect Northwest levels. The following table
provides average annual market prices by area and the twenty-year average escalation of prices.
Across the WECC average prices are forecast to rise by 3.9% annually, or 1.4 percent above the
assumed rate of general price inflation.

Table 7.1
Average Market Prices by WECC Load Area

WECC 2004 Annual Average Market Price ($/MWh) by Year
Load
Area
Load
AMW

2004

2005

2006

2007

2008

2013

2018

2023
Average
Annual
Growth
CASo 20,025 37.0 38.1 39.4 41.4 45.8 65.7 70.7 76.4 3.7%
OWI 19,381 33.8 34.9 36.5 38.3 42.5 57.6 64.9 75.5 4.1%
CANo 12,787 36.4 37.5 39.0 41.0 45.6 64.9 69.1 74.4 3.6%
AZ 8,267 31.2 31.8 32.2 33.7 37.0 57.7 63.8 69.7 4.1%
BC 7,074 34.6 36.2 38.4 39.7 44.5 62.2 70.6 76.4 4.0%
AB 6,401 30.6 31.4 32.5 34.1 37.2 58.8 69.9 76.2 4.7%
CO 5,368 30.1 31.1 32.0 34.0 37.9 54.3 61.2 67.7 4.1%
NM 3,518 31.4 32.1 32.8 34.6 38.0 58.8 64.7 70.1 4.1%
UT 2,824 33.1 34.0 35.3 36.4 40.6 58.0 60.5 64.9 3.4%
IDSo 2,618 33.5 34.7 36.3 38.3 42.9 63.2 65.3 70.3 3.8%
NVSo 2,441 36.5 37.6 38.8 40.8 45.3 60.8 66.6 70.7 3.4%
WY 2,301 29.0 30.1 31.2 33.1 37.4 50.2 56.7 63.2 4.0%
MT 1,768 32.2 33.4 34.9 36.8 41.1 59.6 63.6 69.1 3.9%
NVNo 1,435 35.6 36.8 38.1 40.1 44.8 64.6 66.4 70.1 3.4%
Total 96,209 34.0 35.1 36.4 38.2 42.5 60.9 66.9 73.5 3.9%
0
10
20
30
40
50
60
70
80
90
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Ann
Do
l
l
a
r
s

p
e
r

M
e
g
a
w
a
t
t
-
H
o
u
r
Minimum 10%Average 90%Maximum
80%
C.I.
Section 7 Page 36 Results
The Preferred Resource Strategy

The Company reviewed the modeling results and developed a preferred mix of resource
additions, referred to as the Preferred Resource Strategy (PRS). This decision was made based
on a number of factors which are described below.

Focus on First Ten Years of Study

The Linear Programming (LP) Module (described in Appendix C) utilized to optimize resource
portfolios is set to weigh the first ten years of the study more heavily than the last ten. The LP
Module optimizes 2014 through 2023, but does so only after providing a least-cost solution for
the 2004 through 2013 timeframe. As a result, emphasis is put on the first ten years of the study
period.

Risk and Cost Are Equally Weighted

The Company was asked by Commission Staff and the Technical Advisory Committee to look
not only at lowest cost when evaluating various resource portfolio decisions, but also at resource
risk profiles. This request recognizes that a resource portfolio should be evaluated based on low
costs over time, as well as a reasonable range of variation around the expected cost.

The Company evaluated varying cost/risk relationships (i.e., varying between 30%/70% cost/risk
and 70%/30% cost/risk) and found that the resource selection was not affected substantially
across this range. Therefore the LP Module was set up to evaluate an optimized portfolio by
weighting absolute lowest cost at 50 percent and the variation in cost over the study at 50
percent.

Lowering risk is beneficial to customers where the incremental cost of doing so is relatively low.
To this end, the Company evaluated the expected risk of the PRS by using both a stochastic
approach (utilizing 200 iterations of hydroelectric generation, natural gas prices, and loads) and
by utilizing scenarios. The result of including risk as part of the portfolio decision criteria was a
slight increase in the portfolio costs. The cost increase was small enough that results can be
considered statistically equivalent to utilizing only the lowest expected cost. For further details
on the use of scenarios, refer to Comparison of Scenarios later in this section.

Eighty Percent Confidence Interval Build Level

As described in Section 2, confidence interval planning has numerous advantages over critical
water planning. For IRP planning, one benefit is that building to this level generally provides
enough resource capacity to serve peak load conditions with Company resources.

The LP Module selects a preferred resource mix that meets the 80 percent confidence interval
criteria. The following chart represents Company average requirements over the IRP timeframe,
and the increased requirements resulting from the 80 percent confidence interval. The difference
is approximately 189 aMW through 2018 and includes 153 aMW for load and hydro variability
and 36 aMW for the WNP-3 return obligation. In 2019, the return obligation for WNP-3 drops
to 20 aMW. In 2020 the WNP-3 contract expires.
Section 7 Page 37 Results
Chart 7.5
Average and 80 Percent Confidence Interval Build Requirements
2004-2023

Limitations Placed on LP Module Resource Selection

Limitations on resource selections are necessary for both quantitative and qualitative reasons.
These limitations do not significantly impact the lowest-risk and lowest-cost results. Listed
below are specific resource types and the limitations that apply.

Long-Term Purchases

While always in the short-term marketplace to optimize its portfolio, the Company’s present
strategy is not to rely on long-term market purchases to serve future base-load requirements.
This decision is based on a number of factors. Long-term contracts of five years or more are
difficult to procure in today’s marketplace. After the events of 2000-2001, fewer companies are
willing to sell long-term contracts. Current liquidity and credit issues are a concern for
transactions extending beyond a few months. In addition, the Company is concerned over
potential margin calls and counter party risk. In the current marketplace there is an increased
risk that a counter party will not remain in business long enough to deliver on future
commitments.

This is not to say that the Company will not utilize the marketplace to serve some portion of
customer loads, or capitalize on market opportunities as they present themselves. The Company
will still consider short- to medium-term power supply contracts where they provide benefits to
our customers.

80% Confidence
Interval
(400)
(200)
0
200
400
600
800
1,000
20
0
4
20
0
5
20
0
6
20
0
7
20
0
8
20
0
9
20
1
0
20
1
1
20
1
2
20
1
3
20
1
4
20
1
5
20
1
6
20
1
7
20
1
8
20
1
9
20
2
0
20
2
1
20
2
2
20
2
3
Av
e
r
a
g
e

M
e
g
a
w
a
t
t
s
Surplus Position
Deficit Position
Average
Conditions
80% CI
Section 7 Page 38 Results
Cogeneration

Cogeneration was not included as a new resource alternative for the Company. Cogeneration
offers the potential to increase societal efficiencies by capturing waste heat from industrial
processes, and by capturing a substantial portion of the emissions that otherwise would be
released into the environment; but the Company presently is not aware of any new cogeneration
alternatives within its service territory that it can rely on to meet long-term load obligations.

The exclusion of cogeneration does not indicate a Company preference to exclude this resource
from its portfolio. To the contrary, the Company would welcome a cost-effective cogeneration
facility to meet future resource requirements and would adjust its resource plan accordingly.

Wind

The Company has monitored the changing economics for wind generation in the Northwest.
Construction costs have decreased significantly, and federal tax credits have brought wind
turbines more in line with traditional generation alternatives. To further investigate wind power,
the Company has completed a preliminary wind integration study to help identify the integration
costs of wind. The result depicts significant integration expenses stemming primarily from
increased regulating margin requirements and transmission.

The other challenge of wind is its apparent inability to provide peaking capacity. Not all
generation resources may be relied on to meet the capacity requirements of the Company.
Capacity resources must be available, or reasonably expected to be available, at the times where
load requirements approach overall generating capability. Some wind proponents postulate that
wind energy can be used to serve peak requirements. Based upon internal studies, which are
included in Appendix H, the conclusion has been drawn that wind cannot be relied on to meet
Company peak load obligations.

Since wind generation is highly correlated across the Northwest, it is not possible for the
Company to acquire a wind product with enough geographic diversity to provide significant
capacity. The result is that the Company would most likely need to invest in other capacity
resources (e.g., SCCTs) to meet peaking requirements if significant wind resources are acquired,
or purchase wind from other sources that already includes shaping services.

Given the uncertainty around wind, the Company has elected to limit the preferred strategy to 75
MW of this resource, or around 25 aMW of energy. The Company also proposes to continue the
study of wind to stay well informed on issues, potential declining costs, and any future
opportunities. Where the Company can purchase cost-effective wind generation that includes an
integration service, it will re-evaluate this amount. However, the Company is not aware of an
entity in the Northwest that is providing wind integration services at this time.

In combination with 75 MW of wind energy, the Company would consider the installation of a
peaking unit as a firming service component. A peaking unit would also have the potential to
provide a portion of the Company’s future peaking requirements.

Section 7 Page 39 Results
The Preferred Resource Mix

Based on the conditions and limitations listed above, the LP Module determined a preferred mix
of new resources to meet the Company’s future requirements. The Preferred Resource Strategy
includes the following mix of resources and quantities during the first ten years of the study
(2004-2013):

• 149 aMW of CCCT
• 25 aMW of wind
• 197 aMW coal
• 40 aMW of SCCT

By the end of the first ten years, a total of 411 aMW are developed. A depiction of the Preferred
Resource Strategy is included in the following graph. Significant annual deficiencies do not
develop until 2008, so the chart details only the years 2008 through 2013.

Chart 7.6
Preferred Resource Mix (in aMW)
2008-2013

After 2013, only coal is selected as a result of a change in the relationship between natural gas
and coal prices. Natural gas prices over the IRP term increase faster than coal, making coal
generation less costly in later years. In total, between 2014 and 2023, an additional 566 aMW of
coal resources are selected in the Preferred Resource Strategy.

Costs of Preferred Resource Strategy Versus “No Additions”

Expected cost over the IRP term has traditionally been the benchmark of least-cost planning; and
generally includes capital recovery, operation and maintenance, fuel, and transmission costs.
This IRP continues to focus on expected power supply cost on a net present value (NPV) basis.
Under No Additions, where no resource acquisitions are made, the ten-year NPV of the power
0
50
100
150
200
250
300
350
400
450
2008 2009 2010 2011 2012 2013
Av
e
r
a
g
e

M
e
g
a
w
a
t
t
s
CCCT Peakers
Wind Coal
Section 7 Page 40 Results
supply cost is $1.11 billion. Over twenty years, the NPV rises to $2.73 billion. The Preferred
Resource Strategy (PRS) has NPV values of $1.11 and $2.69 billion, respectively. Over twenty
years, the NPV for the PRS is 1.6 percent lower than No Additions. Refer to the following chart
for a depiction of the difference in power supply expense between the PRS and No Additions
strategy.

Chart 7.7
Annual Net Power Supply Expenses
Difference Between PRS and No Additions
2004-2023

On a cost basis, the Preferred Resource Strategy provided a similar result to No Additions, with a
modestly higher ten-year cost and a modestly lower twenty-year cost. The significant difference
appears when assessing the risk profiles, detailed next.

Risk Assessment of Preferred Resource Strategy

Portfolio risk is based on the annual variance from the average power supply expense over 200
iterations of Monte Carlo. Over time the Company has an opportunity to lower its expected
variance relative to No Additions. The variance in net power supply expenses for the PRS and
No Additions strategy is shown in the chart below.

(70)
(60)
(50)
(40)
(30)
(20)
(10)
0
10
20
20
0
4
20
0
5
20
0
6
20
0
7
20
0
8
20
0
9
20
1
0
20
1
1
20
1
2
20
1
3
20
1
4
20
1
5
20
1
6
20
1
7
20
1
8
20
1
9
20
2
0
20
2
1
20
2
2
20
2
3
Mi
l
l
i
o
n
s

o
f

D
o
l
l
a
r
s
Section 7 Page 41 Results
Chart 7.8
Variance in Net Power Supply Expenses
Preferred Resource Strategy vs. No Additions
2004-2023

As load grows, the No Additions strategy becomes more risky as an increasing portion of system
loads are met with volatile spot market purchases. The Preferred Resource Strategy, on the
other hand, produces a substantially lower risk profile. By the end of twenty years, volatility
under the PRS has fallen to 40 percent of the No Additions strategy. In nominal dollars,
variability of net power supply expense under the PRS is 100 million dollars lower than under
the No Additions strategy.

The following chart shows the distribution over 200 iterations of 2013 power supply expense for
the PRS and No Additions strategy. The range of net power supply expense for the PRS is $273
million, based on an average of $319 million. The range of net power supply expense for the No
Additions strategy is $412 million, based on an average of $324 million. In other words, the
variation in power supply expense (risk) for the PRS is roughly one-third lower than the No
Additions strategy.

10
20
30
40
50
60
70
80
90
100
110
120
130
140
150
160
170
20
0
4
20
0
5
20
0
6
20
0
7
20
0
8
20
0
9
20
1
0
20
1
1
20
1
2
20
1
3
20
1
4
20
1
5
20
1
6
20
1
7
20
1
8
20
1
9
20
2
0
20
2
1
20
2
2
20
2
3
Av
e
r
a
g
e

V
a
r
i
a
t
i
o
n

o
f

N
e
t

P
o
w
e
r
S
u
p
p
l
y

E
x
p
e
n
s
e

(
$

M
il
l
i
o
n
s
)
No Action
Preferred
Section 7 Page 42 Results
Chart 7.9
Distribution of Net Power Supply Expenses – 2013
Preferred Resource Strategy vs. No Additions
(in Millions of Dollars)

Capital Expenditure Requirements

The modeling of future resource acquisitions includes built-in assumptions regarding new
construction costs. The Preferred Resource Strategy requires a capital investment of $725
million from 2007 to 2013 (or $610 million in 2004 dollars). Over twenty years, that amount
increases to nearly $2.4 billion. Capital expenditures in nominal dollars over time are presented
in the following table.

Table 7.2
Annual Capital Expenditures of Preferred Resource Strategy
2004-2023 ($millions)

Year Capital Year Capital
2004 0.0 2014 105.3
2005 0.0 2015 127.8
2006 0.0 2016 139.3
2007 2.4 2017 153.4
2008 39.4 2018 146.1
2009 44.9 2019 181.2
2010 146.5 2020 187.2
2011 222.2 2021 176.7
2012 164.8 2022 198.4
2013 104.7 2023 225.4
Total 725.0 1,640.8
over 20 years 2,365.9

0
10
20
30
40
50
60
70
80
90
175 200 225 250 275 300 325 350 375 400 425 450 475 500 525 550 575 600
Fr
e
q
u
e
n
c
y
No Additions
Preferred
Resource Strateg
Section 7 Page 43 Results
Rate Impact of Preferred Resource Strategy

Rate impacts of future resource acquisition strategies are difficult to accurately quantify.
However, it is important to compare resource strategies in a manner that indicates their potential
impact on rates. To simulate the rate impacts of the Preferred Resource Strategy, the Company
has calculated a power supply expense equal to the twenty-year NPV of the strategy divided by
the sum of energy sales over the same time. While this method does not provide the revenue
requirement for power supply costs, it does explain how rates would generally be impacted. The
following chart displays the difference in rate impact between the PRS and No Additions strategy
over the IRP timeframe.

Chart 7.10
Estimated Rate Impact
Difference Between PRS and No Additions
2004-2023

Qualitative Benefits of Preferred Strategy

Diversity of fuel supply is an important qualitative issue. The Company relies heavily on
hydroelectric generation, and is thereby subject to varying hydrological conditions. The current
resource mix of coal, wood, and natural gas-fired plants has helped to diversify the mix of fuels,
as well as the relationship between high capital/low variable cost and low capital/high variable
cost plants. The following chart provides a picture of the Company’s resource mix in 2004,
2013, and 2023 under the PRS.

(4.0)
(3.5)
(3.0)
(2.5)
(2.0)
(1.5)
(1.0)
(0.5)
0.0
0.5
1.0
20
0
4
20
0
5
20
0
6
20
0
7
20
0
8
20
0
9
20
1
0
20
1
1
20
1
2
20
1
3
20
1
4
20
1
5
20
1
6
20
1
7
20
1
8
20
1
9
20
2
0
20
2
1
20
2
2
20
2
3
Do
l
l
a
r
s

p
e
r

M
W
h

o
f

R
e
t
a
i
l

L
o
a
d
Section 7 Page 44 Results
Chart 7.11
Utility Resource Mix (aMW)
2004, 2013, and 2023

As discussed earlier, the No Additions strategy relies entirely on market purchases to serve load
growth. As a result, the Company would rely on market power for nearly 30 percent of its
annual average load over twenty years. Under the Preferred Resource Strategy, market
purchases account for an average of four percent of retail load. Modest purchases should be
expected under all strategies as the Company optimizes the operation of its gas turbines, but
significant purchases are indicative of an overly short position. Refer to Chart 7.12 for a
depiction of market purchases under the two strategies.

Chart 7.12
Annual Net Market Purchases
Preferred Resource Strategy vs No Additions
43%
12%18%
31%
4%
28%
35%
23%
49%
26%
0%
27%
2%
1%
1%
Hydro Contracts Base Thermal Gas-Fired Wind
-10%
0%
10%
20%
30%
40%
50%
60%
20
0
4
20
0
5
20
0
6
20
0
7
20
0
8
20
0
9
20
1
0
20
1
1
20
1
2
20
1
3
20
1
4
20
1
5
20
1
6
20
1
7
20
1
8
20
1
9
20
2
0
20
2
1
20
2
2
20
2
3
Pe
r
c
e
n
t

o
f

R
e
t
a
i
l

L
o
a
d
No Additions
Preferred
Resource Strateg
20042013
2023
Section 7 Page 45 Results
Adjusted Load and Resource Balance From Preferred Resource Strategy

A discussion of the Company’s forecasted load and resource balances for both energy and
capacity may be found in Section 2. The two tables below provide adjusted balances with the
inclusion of those resources selected in the Preferred Resource Strategy.

Table 7.3
Adjusted Loads and Resources – Energy
2004-2008, 2013, 2018, 2023

2004 2005 2006 2007 2008 2013 2018 2023
Obligations
Retail Load 985 1,014 1,051 1,083 1,120 1,326 1,569 1,860
80% Conf. Interval 189 189 189 189 189 189 189 153
Total Obligations 1,174 1,203 1,240 1,272 1,309 1,515 1,758 2,013
Existing Resources
Hydro 550 545 530 530 529 477 471 458
Net Contracts 156 157 175 177 177 58 59 12
Base Thermal 223 230 223 223 230 230 230 230
Gas Dispatch 158 156 158 158 156 158 158 156
Gas Peaking Units 181 181 181 181 181 181 181 181
Total Existing Resources 1,268 1,269 1,267 1,269 1,273 1,104 1,099 1,037
PRS Resource Additions
Wind 0 0 0 0 6 25 25 25
Base Thermal 0 0 0 0 0 197 446 763
Gas Dispatch 0 0 0 3 30 149 149 149
Gas Peaking Units 0 0 0 0 0 40 40 40
Total PRS Resources 0 0 0 3 36 411 660 977
Net Position 94 66 27 0 0 0 1 1

Section 7 Page 46 Results
Table 7.4
Adjusted Loads and Resources – Capacity
2004-2008, 2013, 2018, 2023

2004 2005 2006 2007 2008 2013 2018 2023
Obligations
Retail Load 1,470 1,515 1,570 1,617 1,672 1,982 2,349 2,780
Operating Reserves 107 107 105 105 107 130 150 174
Total Obligations 1,577 1,622 1,675 1,722 1,779 2,112 2,499 2,954
Existing Resources
Hydro 1,177 1,177 1,135 1,134 1,133 1,043 1,035 998
Net Contracts 70 19 43 45 45 -73 78 -2
Base Thermal 272 272 272 272 272 272 272 272
Gas Dispatch 176 176 176 176 176 176 176 176
Gas Peaking Units 236 236 236 236 236 236 236 236
Total Existing Resources 1,931 1,880 1,862 1,863 1,862 1,654 1,797 1,680
PRS Resource Additions
Wind 0 0 0 0 0 0 0 0
Base Thermal 0 0 0 0 0 229 518 886
Gas Dispatch 0 0 0 3 32 156 156 156
Gas Peaking Units 0 0 0 0 0 42 42 42
Total PRS Resources 0 0 0 3 32 428 716 1,084
Net Position 354 258 187 144 115 -30 14 -189
Reserve Margin 31.4% 24.1% 18.6% 15.4% 13.3% 5.0% 7.0% -0.6%

Based on the Preferred Resource Strategy, the Company maintains an energy balance at the 80
percent confidence level through the end of the IRP timeframe. Building to the 80 percent level
generally provides an adequate capacity reserve margin. As a result, the Preferred Resource
Strategy maintains planning reserve margins in excess of twelve percent through 2009. Falling
reserve margins after 2009 are a reflection of the Company outgrowing its hydroelectric
resources, which tend to have higher capacity to energy ratios than other generating facilities.
The Company will need to address a reduced capacity surplus in a later study, as discussed in
Section 8.

Resource Acquisition Under Preferred Resource Strategy

The Preferred Resource Strategy is designed without limitations on the quantity of megawatts
purchased by the Company in any given year. This assumption, while significantly reducing the
complexity of the LP Module logic, is not possible in reality. Instead, the Company would likely
implement the PRS in a less smooth manner. For example, it is unlikely that in 2009 the
Company would be able to procure 10 aMW from a CCCT plant, as directed by the LP Module.
Instead it might enter into an agreement that would cover the 149 aMW needs of 2008 through
2011.

Section 7 Page 47 Results
Comparison of Strategies

Section 5 describes the resource portfolio strategies considered in addition to the Preferred
Resource Strategy (Lowest Cost/CCCT, Lowest Risk, All Coal, and Wind Strategy). Each of the
strategies was compared using the same measurements used to compare the PRS and No
Additions strategies. These measurements include cost, risk, capital expenditures, rate impacts,
and reliance on the wholesale marketplace. The result was that the PRS performed well across
the criteria when compared to other strategies.

The ability of the PRS to reduce risk at a small incremental cost was the largest impact witnessed
in the comparisons. The Lowest Risk strategy reduced risk by an additional one percent of
average power supply expense, but only through much greater capital expense and further
reliance on coal. The Lowest Risk strategy also relied heavily on wind plants, which do not
provide capacity.

The capital costs of the Preferred Resource Strategy fell in the middle of the range of strategies.
Portfolios relying more heavily on coal have costs as much as $500 million more over twenty
years (in 2004 dollars). The Lowest Cost/CCCT strategy, which relies exclusively on CCCTs has
a substantially smaller capital requirement, but suffers from significant fuel price risk.

Rate impacts during the first ten years of the study were lower in the Lowest Cost/CCCT
strategy. Costs were higher under each of the remaining strategies. Reliance on the marketplace
was small and similar for all strategies except for No Additions and the Wind Strategy. Each of
these relied more heavily on market purchases to meet load requirements.

For further results from the analysis of strategies in this IRP, refer to Appendix E.

Comparison of Scenarios

Section 5 also describes eight scenarios considered by the Company to capture specific
marketplace futures (Average, Critical Water, High Gas, High Load, Load Loss, New Trans,
Coal Build, and Carbon Tax). Each scenario was included to test the Preferred Resource
Strategy and other strategies in the face of greatly different future market conditions. The PRS
performed well across the scenarios, as compared to the other strategies. The following text will
briefly describe the scenario results.

Under the Critical Water scenario, results across the strategies are similar to the results of 200
iterations of Monte Carlo included in the Base Case. The largest impact of low water is that it
drives average wholesale market prices up. The High Gas scenario results in the largest impact
on wholesale market prices, primarily as a result of the WECC’s heavy reliance on gas turbines.
Under high gas prices, the PRS outperforms the Lowest Cost strategy due to its reduced reliance
on natural gas-fired resources. However, strategies with more coal-fired generation benefit even
more.

Section 7 Page 48 Results
The High Load scenario, with an increase in WECC loads, drove wholesale market prices up to
levels nearly as high as the High Gas scenario. The No Additions strategy was significantly
more expensive than the PRS and all other strategies under high loads. The greatest benefactor
of the High Load scenario was the Lowest Risk portfolio, with its heavy reliance on coal and
wind.

The Load Loss scenario, in which the Company would lose 300 aMW of retail load, reduced the
amount of future resource requirements by the same amount. This scenario disadvantaged both
the All Coal and No Additions strategies. The PRS had similar costs to the remaining strategies.

The New Trans scenario, in which extensive transmission is added between Montana, Wyoming,
and several other load areas, actually benefited the No Additions strategy. Since the additional
transmission resulted in extensive additions of coal-fired generation, spot market prices were
kept low. See Section 4 for a discussion of why additional investment in transmission facilities
is necessary to support coal plant development.

The All Coal scenario benefited strategies with low capital cost investments (CCCTs) due to
reduced market price volatility. The No Additions strategy is also an attractive option under an
All Coal scenario, since exposure to market prices is significantly less risky.

The Carbon Tax scenario disadvantages coal plants, and to a lesser extent gas-fired resources.
Under the Carbon Tax scenario, No Additions and Lowest Cost (with its focus on CCCTs)
outperformed the PRS. The All Coal strategy was the highest cost, due to the new emission
taxes.

For further results from the analysis of scenarios in this IRP, refer to Appendix E.

Summary and Conclusions

This study represents a considerable analytical effort and provides a means to evaluate the
Preferred Resource Strategy against several alternative strategies under varying scenarios.
Overall, the PRS fairs well, not only in the Base Case, but also under numerous scenarios. The
PRS will meet not only the Company’s load obligations over time, but will also provide for
reserve margins in excess of twelve percent through 2009.

The PRS provides for a significant reduction of risk. This reduction comes at a very modest
impact to expected costs. Under the PRS, the average variation from net power supply expenses
is forecast to fall from about eighteen percent in 2004 to eight percent in 2011. The reduction in
risk under the PRS comes despite significant future variation in hydroelectric generation, natural
gas prices, and regional demand. The Company believes that customers will benefit from the
focus on risk reduction through greater rate stability. The Preferred Resource Strategy will
require significant additional investments over time. In the first ten years of the study, the
Company will need to invest nearly $725 million in new capital beyond present forecasts. Over
twenty years, a total of $2.4 billion will be required, nearly twice the current utility plant in
service figure.
Section 8 Page 49 Action Plans & Avoided Costs
Section

Action Plans & Avoided Costs

Overview

This section provides a summary of the 2001 IRP Action Plan and how the Company addressed
each of the items. A 2003 Action Plan follows and details the studies and actions the Company
will take between now and the 2005 IRP. Finally, avoided costs are presented for the IRP
timeline.

Summary Report for 2001 Action Plan

In the 2001 IRP, the Company listed specific action plan activities, which were to be
accomplished during the past two-year planning cycle. Each 2001 Action Item is listed below,
immediately followed by an explanation of the Company’s response in italics:

Public Process

1. Continue free flowing exchange of information with TAC members.

The number of TAC meetings was increased from three to four. Efforts were also made to
increase attendance. The mailing list was expanded to include additional customers who
might have an interest in resource planning. The Company now has a mailing list of 53
individuals who receive IRP information and TAC meeting invitations.

2. Propose changes to the IRP process that will be useful in the competitive market era.

The IRP process has been modified to incorporate significant modeling of present and future
market conditions. Monte Carlo risk analysis has been incorporated to evaluate volatile
market conditions.

Demand-Side Management

1. Pursue energy savings for the next three years with funding from the tariff rider.

The Company has continued to operate demand-side management programs focused on
obtaining available cost-effective resources. During the summer of 2001, the Company
launched a series of extraordinary temporary programs intended to immediately impact
utility load during a period of extreme wholesale electric price volatility. As a result of these
programs the tariff rider presently has a negative balance. Tariff rider funding is continuing
and the Company anticipates this balance will return to zero by the close of 2005.

Section 8 Page 50 Action Plans & Avoided Costs
2. Consider the development of programs that will allow peak shaving.

Proposals for peak-shaving measures were submitted to the Company in 2000 as part of the
Company's All-Resource RFP. Additionally, other proposals have been evaluated. To date
none of these programs have proven cost-effective. One demand-response program
proposal, submitted by an external engineering firm, remains in the evaluation stage.

3. Determine the potential for time-of-use (TOU) rates.

The Company continues evaluated the cost-effectiveness of various hypothetical TOU rate
options, but has no specific plans for implementation at this time.

4. Execute and implement DSM contracts that were selected under the 2000 RFP.

The Company selected and completed contracts for two proposals submitted under the 2000
RFP. Two resulting programs are currently available to qualifying customers.

Supply-Side Resource Options

1. Pursue the base plan for Spokane River relicensing.

Spokane River hydroelectric relicensing is proceeding following the Alternative Licensing
Procedures (ALP) used in the successful Clark Fork effort. The current license expires July
31, 2007. The ALP is a collaborative approach to decision making for relicensing. Over
100 stakeholder groups are involved in this effort. Primary studies will be conducted in
2003. Additional studies will follow in 2004, as will development of proposals guiding a new
license application. The Company must file a new application by July 31, 2005.

2. Upgrade at least two units at the Cabinet Gorge hydro facility.

The Company is currently in the initial phases of the Unit 2 upgrade at Cabinet Gorge. The
construction for the replacement runner, stator rewind, rotor refurbishment, machine
monitoring equipment, and other refurbishment work is scheduled to start late summer of
2003 and be completed by spring of 2004. The Unit 2 upgrade will provide a fifteen MW
increase in capacity at Cabinet Gorge. This estimate is based on the actual performance
realized with the upgrade of Unit 3, completed three years ago.

The Company has also identified other hydroelectric upgrades at Cabinet Gorge and Noxon
Rapids. While these upgrades are economically viable and beneficial for maintenance
purposes, they have been pushed out due to capital budget restrictions.

3. Evaluate the effects of a micro turbine on the system.

A micro turbine was added to the downtown Spokane system. The various operating
characteristics, under different loadings, have been recorded. These included fixed and
variable operating costs. The unit is only operated when it is economically beneficial to do
so.
Section 8 Page 51 Action Plans & Avoided Costs

4. Install inlet coolers at Rathdrum Combustion turbines for additional summer peaking output.

This was completed in July of 2000. The data shows a five MW/unit increase on hot days.

5. Evaluate RFP bids, compare to Company options, and select options that are cost effective
and that best meet the Company’s long-term resource needs. Complete transfer agreements
for selected supply-side resource.

The best options under the RFP were selected in December of 2000. Selected were three
DSM bids and one supply-side bid (Coyote Springs 2). Transfer agreements for Coyote
Springs 2 have been completed. The generating facility is essentially complete, with the
exception of the transformer. The original transformer was energized on March 3, 2002. It
failed due to an internal explosion on May 6, 2002. The second transformer was ordered on
June 21, 2002. This transformer failed its acceptance test in the factory on August 30, 2002.
The transformer had to be repaired at the factory and passed testing on November 5, 2002.
It was prepared for shipment and placed on a dedicated shipping vessel, which came across
the Atlantic into Houston.

It arrived in Houston on December 4, 2002 and was immediately placed on a railcar and
delivered to Coyote Springs 2 in Boardman, Oregon. It was moved to its foundation and on
December 18 apparent internal damage was observed. Representatives from the
manufacturer traveled to the site and performed further internal inspections. The results of
the inspections were that the transformer did have internal damage, most probably caused by
shipping, and that it could not be repaired onsite.

Arrangements were made to have the transformer repaired and it was shipped to California.
The initial inspection of the damage at the factory took place on February 10, 2003. The
repair is now complete and the transformer has gone through a “dryout” phase. It has been
filled with oil and is now in a resting stage to insure that all of the transformer insulation is
saturated with oil.

The testing of the transformer will take place the week of April 21, 2003. The transformer
should be returned to Coyote Springs on May 18, 2003 and be energized by May 28, 2003.
The plan is to have the plant commercial by July 1, 2003.

6. Pursue re-negotiation efforts with Mid-Columbia PUDs.

Renegotiation of the contracts with Grant County PUD has been completed. These contracts
affect the output of the Priest Rapids and Wanapum dams to the Company. As the contracts
for the Rocky Reach and Wells Canyon dams come up for re-negociation, the Company will
be actively involved.

Section 8 Page 52 Action Plans & Avoided Costs
7. Evaluate the need for additional supply or generation units to handle variability in hydro,
retail loads, and potential generation outages under projected market conditions.

The evaluation of potential new supply or generation units related to various market
conditions is addressed through the utilization of significant computer modeling. In this IRP,
the entire WECC has been modeled under multiple scenarios incorporating Monte Carlo
simulation. Numerous factors in market volatility have been simulated, including
hydroelectric generation variability, load variability, and fuel price variability.

Resource Management Issues

1. Implement relicensing programs on the Clark Fork River hydro projects, as part of the
“Living License” commitment.

The Company is working with other stakeholders in fulfilling this commitment under the new
license. The Company will spend about five million dollars per year for the next 45 years.

2. Continue to examine and pursue cost-effective efficiency improvements at generation
facilities.

Because of financial conditions all capital improvements are placed on hold. Future
upgrades include Unit 4 at Cabinet Gorge, two units at Noxon Rapids, and a new control
system at Long Lake.

2003 Action Plan

The Company’s Preferred Resource Strategy provides direction for long-term activities. The
Company’s new near-term action plan outlines activities that will support this strategy and
improve the planning process during 2003 and 2004. Progress on these activities will be
monitored over the two-year planning cycle and reported in the Company’s next Integrated
Resource Plan. They are designed to improve the planning and resource acquisition processes.

Public Process

1. Propose changes to WUTC on the IRP/RFP process that will provide improvements.

2. Continue to manage the free flow of information with TAC participants.

Demand-Side Management

1. Evaluate the cost-effectiveness and resource potential of conservation voltage reduction on
the Company’s system.

2. Acquire electric resources that are at least proportionate to the percentage of DSM revenues
being expended.

Section 8 Page 53 Action Plans & Avoided Costs
3. Field a DSM portfolio that continues to be cost-effective on a societal and utility basis.

4. Prepare contingency plans for future emergency responses to unexpected fluctuations in
wholesale electric markets.

5. Prepare for a reevaluation of continued participation in the Northwest Energy Efficiency
Alliance upon expiration of the current contract period (expiring at the end of 2004).

6. Convene a TAC meeting in the fall of 2003 to discuss the various alternatives for integrating
DSM into the 2005 IRP process.

Supply-Side Resource Options

1. Pursue a new license for the Spokane River projects by filing a new license application by
July 31, 2005.

2. Continue to evaluate the effects and costs of integrating wind generation into the Company’s
electrical system.

3. Consider and evaluate the potential to add coal facilities to the Company’s mix of existing
generating resources.

4. Determine the feasibility of entering into a medium-term firm power sale during the
Company’s surplus years.

5. Initiate a study to determine the optimal reserve margin for the Company, including the
benefits of additional peaking capacity.

6. Continue to assess the cost-effectiveness of new resource additions.

7. Continue to work with Commission Staff on methods whereby the Company can acquire
resources with development timelines beyond one or two years and increase the probability
for full rate recovery.

Resource Management Issues

1. Analyze the uncertainty of decisions as the Company confronts risks and opportunities.

2. Continue to assess the electric marketplace and its effect on the Company.

Section 8 Page 54 Action Plans & Avoided Costs
Avoided Costs

The Company develops avoided costs as part of the IRP process. The Company believes that the
marketplace provides the truest estimate of avoided cost. Models such as AURORA provide
insight into long-term market prices and therefore can be used to develop avoided costs. Results
from the 200 iterations were averaged to develop the annual avoided cost schedule for this IRP,
as shown in Table 8.1. For background information on avoided costs, refer to Appendix O.

Table 8.1
Avoided Costs ($/MWh, Flat 7x24)
2004-2023

Year Price Year Price Year Price Year Price
2004 33.72 2009 45.98 2014 58.28 2019 67.28
2005 35.06 2010 50.10 2015 60.20 2020 69.19
2006 36.49 2011 52.97 2016 62.63 2021 70.32
2007 38.20 2012 55.35 2017 64.87 2022 71.28
2008 42.44 2013 57.39 2018 65.41 2023 75.71

Page 55 Production Credits
Production Credits

Given the breadth of an integrated resource plan, its development is dependent on the work and
expertise of a broad range of individuals. The 2003 IRP was developed and documented by a
small team of utility staff; but they relied on a larger body of skills within the Company.
Following are lists of those individuals who contributed to the product.

Primary 2003 IRP Team

Individual Contribution
Clint Kalich, Manager of
Resource Planning & Analysis
Project Manager/
Author
Jason Fletcher, Power Supply
Analyst
Lead Modeler/
Author/Editor
Doug Young, Senior Engineer
Power Resources

Author/Historian
Randy Barcus, Chief
Corporate Economist
Load Forecast Author/
Statistics Consultant
Jon Powell, Partnership
Solutions Manager
Conservation & DSM
Author

Other Contributors

Contributor Contributor
Bill Johnson, Senior Power
Supply Analyst
Bruce Folsom, Manager of
Regulatory Compliance
George Perks, Joint
Generation Manager
Curt Rettenmier,
RAD Analyst
Linda Gervais, Regulatory
Analyst
Ed Groce, Manager of
Transmission Planning
Dick Winters, Gas Analyst Bruce Howard, Spokane
River Licensing Manager
Steve Lester, Oracle Database
Administrator
Brad Simcox, Utility Intern
Ross Taylor,
Telecommunications Project
Engineer
Todd Bryan, Technology
Coordinator
Steve Wenke, Chief
Generation Engineer
Patrick Maher, Senior
Transmission Contracts
Analyst
Bob Anderson, Director of
Environmental Affairs
Dave Heyamoto, Market
Solutions Manager